gosu 0.12.1 → 0.13.0

data/src/Utility.cpp

@@ -8,6 +8,7 @@
 #include <algorithm>
 #include <stdexcept>
 #include <vector>
+using namespace std;
 
 #ifndef GOSU_IS_IPHONE
 
@@ -31,7 +32,7 @@ wstring Gosu::utf8_to_wstring(const string& s)
 {
     return iconvert<wstring, UCS_4_INTERNAL, UTF_8>(s);
 }
-string Gosu::wstring_to_utf8(const std::wstring& ws)
+string Gosu::wstring_to_utf8(const wstring& ws)
 {
     return iconvert<string, UTF_8, UCS_4_INTERNAL>(ws);
 }
@@ -57,7 +58,7 @@ string Gosu::wstring_to_utf8(const wstring& ws)
 #endif
 #endif
 
-bool Gosu::has_extension(const std::string& filename, const char* extension)
+bool Gosu::has_extension(const string& filename, const char* extension)
 {
     size_t ext_len = strlen(extension);
     if (ext_len > filename.length()) {

data/src/UtilityApple.cpp

@@ -5,16 +5,15 @@
 #import <Foundation/Foundation.h>
 #import <stdexcept>
 #import <vector>
+using namespace std;
 
 #ifdef GOSU_IS_IPHONE
-std::wstring Gosu::utf8_to_wstring(const std::string& s)
+wstring Gosu::utf8_to_wstring(const string& s)
 {
-    if (s.empty()) {
-        return std::wstring();
-    }
+    if (s.empty()) return wstring();
     
     NSString* string = [NSString stringWithUTF8String:s.c_str()];
-    std::vector<wchar_t> buffer(s.size());
+    vector<wchar_t> buffer(s.size());
     NSUInteger buffer_size;
     if (![string getBytes:&buffer[0]
                 maxLength:buffer.size() * sizeof(wchar_t)
@@ -23,25 +22,25 @@ std::wstring Gosu::utf8_to_wstring(const std::string& s)
                   options:0
                     range:NSMakeRange(0, string.length)
            remainingRange:nullptr]) {
-        throw std::runtime_error("String " + s + " could not be converted to UTF-32");
+        throw runtime_error("String " + s + " could not be converted to UTF-32");
     }
-    return std::wstring(&buffer[0], &buffer[0] + buffer_size / sizeof(wchar_t));
+    return wstring(&buffer[0], &buffer[0] + buffer_size / sizeof(wchar_t));
 }
 
-std::string Gosu::wstring_to_utf8(const std::wstring& ws)
+string Gosu::wstring_to_utf8(const wstring& ws)
 {
-    if (ws.empty()) return std::string();
+    if (ws.empty()) return string();
 
     @autoreleasepool {
         NSString* string = [[NSString alloc] initWithBytes:ws.data()
                                                     length:ws.size() * sizeof(wchar_t)
                                                   encoding:NSUTF32LittleEndianStringEncoding];
-        return string.UTF8String ?: std::string();
+        return string.UTF8String ?: "";
     }
 }
 #endif
 
-std::string Gosu::language()
+string Gosu::language()
 {
     @autoreleasepool {
         NSString* language = [NSLocale preferredLanguages][0];

data/src/UtilityWin.cpp

@@ -4,8 +4,9 @@
 #include <Gosu/Utility.hpp>
 #include "WinUtility.hpp"
 #include <windows.h>
+using namespace std;
 
-std::string Gosu::language()
+string Gosu::language()
 {
     LCID lcid = GetUserDefaultLCID();
     char buffer[9];

data/src/Version.cpp

@@ -1,12 +1,13 @@
 #include <Gosu/Version.hpp>
 #include <Gosu/Platform.hpp>
 #include <string>
+using namespace std;
 
-const std::string Gosu::VERSION = std::to_string(GOSU_MAJOR_VERSION) + '.' +
-                                  std::to_string(GOSU_MINOR_VERSION) + '.' +
-                                  std::to_string(GOSU_POINT_VERSION);
+const string Gosu::VERSION = to_string(GOSU_MAJOR_VERSION) + '.' +
+                             to_string(GOSU_MINOR_VERSION) + '.' +
+                             to_string(GOSU_POINT_VERSION);
 
-const std::string Gosu::LICENSES =
+const string Gosu::LICENSES =
     "This software may utilize code from the following third-party libraries:\n"
     "\n"
     "Gosu, https://www.libgosu.org, MIT License, https://opensource.org/licenses/MIT\n"

data/src/WinMain.cpp

@@ -23,7 +23,7 @@ vector<string> split_cmd_line()
                 is_quoted_arg = true;
             }
             else if (is_quoted_arg) {
-                result.push_back(std::string(arg_begin, cmd_line));
+                result.push_back(string(arg_begin, cmd_line));
                 arg_begin = nullptr;
             }
         }
@@ -32,7 +32,7 @@ vector<string> split_cmd_line()
             is_quoted_arg = false;
         }
         else if (isspace((unsigned char)*cmd_line) && arg_begin != nullptr && !is_quoted_arg) {
-            result.push_back(std::string(arg_begin, cmd_line + 1));
+            result.push_back(string(arg_begin, cmd_line + 1));
             arg_begin = nullptr;
         }
         ++cmd_line;
@@ -55,7 +55,7 @@ int WINAPI WinMain(HINSTANCE, HINSTANCE, LPSTR, int)
             argv[i] = const_cast<char*>(arguments[i].c_str());
         return main(argv.size(), &argv[0]);
     }
-    catch (const std::exception& e) {
+    catch (const exception& e) {
         ::MessageBoxA(0, e.what(), "Uncaught Exception", MB_OK | MB_ICONERROR);
         return EXIT_FAILURE;
     }

data/src/WinUtility.cpp

@@ -4,9 +4,10 @@
 #include "WinUtility.hpp"
 #include <Gosu/Utility.hpp>
 #include <stdexcept>
-#include <windows.h>
+#include <windows.h>
+using namespace std;
 
-void Gosu::throw_last_winapi_error(const std::string& action)
+void Gosu::throw_last_winapi_error(const string& action)
 {
     // Obtain error message from Windows.
     wchar_t* buffer;
@@ -16,11 +17,11 @@ void Gosu::throw_last_winapi_error(const std::string& action)
                         FORMAT_MESSAGE_IGNORE_INSERTS, 0, GetLastError(),
                         MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT), (LPWSTR) &buffer, 0, 0)
             || buffer == nullptr) {
-        throw std::runtime_error("Unknown error");
+        throw runtime_error("Unknown error");
     }
 
-    // Safely move the message into a std::string.
-    std::string message;
+    // Safely move the message into a string.
+    string message;
     try {
         message = wstring_to_utf8(buffer);
     }
@@ -35,7 +36,7 @@ void Gosu::throw_last_winapi_error(const std::string& action)
         message = "While " + action + ", the following error occured: " + message;
     }
 
-    throw std::runtime_error(message);
+    throw runtime_error(message);
 }
 
 #endif

data/src/Window.cpp

@@ -6,6 +6,7 @@
 #include <cstdlib>
 #include <algorithm>
 #include <stdexcept>
+using namespace std;
 
 namespace Gosu
 {
@@ -14,10 +15,10 @@ namespace Gosu
         void register_frame();
     }
 
-    static void throw_sdl_error(const std::string& operation)
+    static void throw_sdl_error(const string& operation)
     {
         const char* error = SDL_GetError();
-        throw std::runtime_error(operation + ": " + (error ? error : "(unknown error)"));
+        throw runtime_error(operation + ": " + (error ? error : "(unknown error)"));
     }
 
     static void cleanup();
@@ -30,7 +31,7 @@ namespace Gosu
                 throw_sdl_error("Could not initialize SDL Video");
             }
 
-            std::atexit(cleanup);
+            atexit(cleanup);
 
             Uint32 flags = SDL_WINDOW_OPENGL | SDL_WINDOW_HIDDEN;
             
@@ -90,8 +91,8 @@ struct Gosu::Window::Impl
     // (transition from CLOSED to OPEN).
     enum { CLOSED, OPEN, CLOSING } state = CLOSED;
     
-    std::unique_ptr<Graphics> graphics;
-    std::unique_ptr<Input> input;
+    unique_ptr<Graphics> graphics;
+    unique_ptr<Input> input;
 };
 
 Gosu::Window::Window(unsigned width, unsigned height, bool fullscreen, double update_interval)
@@ -151,7 +152,7 @@ void Gosu::Window::resize(unsigned width, unsigned height, bool fullscreen)
 
         double scale_x = 1.0 * actual_width / width;
         double scale_y = 1.0 * actual_height / height;
-        scale_factor = std::min(scale_x, scale_y);
+        scale_factor = min(scale_x, scale_y);
 
         if (scale_x < scale_y) {
             black_bar_height = (actual_height / scale_x - height) / 2;
@@ -165,7 +166,7 @@ void Gosu::Window::resize(unsigned width, unsigned height, bool fullscreen)
         double max_height = Gosu::available_height();
         
         if (width > max_width || height > max_height) {
-            scale_factor = std::min(max_width / width, max_height / height);
+            scale_factor = min(max_width / width, max_height / height);
             actual_width = width * scale_factor;
             actual_height = height * scale_factor;
         }
@@ -205,13 +206,13 @@ void Gosu::Window::set_update_interval(double update_interval)
     pimpl->update_interval = update_interval;
 }
 
-std::string Gosu::Window::caption() const
+string Gosu::Window::caption() const
 {
     const char* title = SDL_GetWindowTitle(shared_window());
     return title ? title : "";
 }
 
-void Gosu::Window::set_caption(const std::string& caption)
+void Gosu::Window::set_caption(const string& caption)
 {
     SDL_SetWindowTitle(shared_window(), caption.c_str());
 }
@@ -262,7 +263,7 @@ bool Gosu::Window::tick()
             case (SDL_DROPFILE): {
                 char* dropped_filedir = e.drop.file;
                 if (dropped_filedir == nullptr) break;
-                drop(std::string(dropped_filedir));
+                drop(string(dropped_filedir));
                 SDL_free(dropped_filedir);
                 break;
             }

data/src/WindowUIKit.cpp

@@ -3,6 +3,7 @@
 
 #include "GosuViewController.h"
 #include <Gosu/Gosu.hpp>
+using namespace std;
 
 namespace Gosu
 {
@@ -35,12 +36,12 @@ struct Gosu::Window::Impl
 {
     ::UIWindow* window;
     GosuViewController* controller;
-    std::unique_ptr<Graphics> graphics;
-    std::unique_ptr<Input> input;
+    unique_ptr<Graphics> graphics;
+    unique_ptr<Input> input;
     
     bool fullscreen;
     double update_interval;
-    std::string caption;
+    string caption;
 };
 
 Gosu::Window::Window(unsigned width, unsigned height, bool fullscreen, double update_interval)
@@ -90,7 +91,7 @@ bool Gosu::Window::fullscreen() const
 
 void Gosu::Window::resize(unsigned width, unsigned height, bool fullscreen)
 {
-    throw std::logic_error("Cannot resize windows on iOS");
+    throw logic_error("Cannot resize windows on iOS");
 }
 
 double Gosu::Window::update_interval() const
@@ -100,15 +101,15 @@ double Gosu::Window::update_interval() const
 
 void Gosu::Window::set_update_interval(double update_interval)
 {
-    throw std::logic_error("Cannot change the update interval on iOS");
+    throw logic_error("Cannot change the update interval on iOS");
 }
 
-std::string Gosu::Window::caption() const
+string Gosu::Window::caption() const
 {
     return pimpl->caption;
 }
 
-void Gosu::Window::set_caption(const std::string& caption)
+void Gosu::Window::set_caption(const string& caption)
 {
     pimpl->caption = caption;
 }
@@ -144,7 +145,7 @@ bool Gosu::Window::tick()
 
 void Gosu::Window::close()
 {
-    throw std::logic_error("Cannot close windows manually on iOS");
+    throw logic_error("Cannot close windows manually on iOS");
 }
 
 void Gosu::Window::button_down(Button button)

data/src/stb_image.h

@@ -1,4 +1,4 @@
-/* stb_image - v2.12 with fixes from https://github.com/nothings/stb/pull/343 - public domain image loader - http://nothings.org/stb_image.h
+/* stb_image - v2.16 - public domain image loader - http://nothings.org/stb_image.h
                                      no warranty implied; use at your own risk
 
    Do this:
@@ -21,7 +21,7 @@
           avoid problematic images and only need the trivial interface
 
       JPEG baseline & progressive (12 bpc/arithmetic not supported, same as stock IJG lib)
-      PNG 1/2/4/8-bit-per-channel (16 bpc not supported)
+      PNG 1/2/4/8/16-bit-per-channel
 
       TGA (not sure what subset, if a subset)
       BMP non-1bpp, non-RLE
@@ -42,136 +42,23 @@
    Full documentation under "DOCUMENTATION" below.
 
 
-   Revision 2.00 release notes:
-
-      - Progressive JPEG is now supported.
-
-      - PPM and PGM binary formats are now supported, thanks to Ken Miller.
-
-      - x86 platforms now make use of SSE2 SIMD instructions for
-        JPEG decoding, and ARM platforms can use NEON SIMD if requested.
-        This work was done by Fabian "ryg" Giesen. SSE2 is used by
-        default, but NEON must be enabled explicitly; see docs.
-
-        With other JPEG optimizations included in this version, we see
-        2x speedup on a JPEG on an x86 machine, and a 1.5x speedup
-        on a JPEG on an ARM machine, relative to previous versions of this
-        library. The same results will not obtain for all JPGs and for all
-        x86/ARM machines. (Note that progressive JPEGs are significantly
-        slower to decode than regular JPEGs.) This doesn't mean that this
-        is the fastest JPEG decoder in the land; rather, it brings it
-        closer to parity with standard libraries. If you want the fastest
-        decode, look elsewhere. (See "Philosophy" section of docs below.)
-
-        See final bullet items below for more info on SIMD.
-
-      - Added STBI_MALLOC, STBI_REALLOC, and STBI_FREE macros for replacing
-        the memory allocator. Unlike other STBI libraries, these macros don't
-        support a context parameter, so if you need to pass a context in to
-        the allocator, you'll have to store it in a global or a thread-local
-        variable.
-
-      - Split existing STBI_NO_HDR flag into two flags, STBI_NO_HDR and
-        STBI_NO_LINEAR.
-            STBI_NO_HDR:     suppress implementation of .hdr reader format
-            STBI_NO_LINEAR:  suppress high-dynamic-range light-linear float API
-
-      - You can suppress implementation of any of the decoders to reduce
-        your code footprint by #defining one or more of the following
-        symbols before creating the implementation.
-
-            STBI_NO_JPEG
-            STBI_NO_PNG
-            STBI_NO_BMP
-            STBI_NO_PSD
-            STBI_NO_TGA
-            STBI_NO_GIF
-            STBI_NO_HDR
-            STBI_NO_PIC
-            STBI_NO_PNM   (.ppm and .pgm)
-
-      - You can request *only* certain decoders and suppress all other ones
-        (this will be more forward-compatible, as addition of new decoders
-        doesn't require you to disable them explicitly):
-
-            STBI_ONLY_JPEG
-            STBI_ONLY_PNG
-            STBI_ONLY_BMP
-            STBI_ONLY_PSD
-            STBI_ONLY_TGA
-            STBI_ONLY_GIF
-            STBI_ONLY_HDR
-            STBI_ONLY_PIC
-            STBI_ONLY_PNM   (.ppm and .pgm)
-
-         Note that you can define multiples of these, and you will get all
-         of them ("only x" and "only y" is interpreted to mean "only x&y").
-
-       - If you use STBI_NO_PNG (or _ONLY_ without PNG), and you still
-         want the zlib decoder to be available, #define STBI_SUPPORT_ZLIB
-
-      - Compilation of all SIMD code can be suppressed with
-            #define STBI_NO_SIMD
-        It should not be necessary to disable SIMD unless you have issues
-        compiling (e.g. using an x86 compiler which doesn't support SSE
-        intrinsics or that doesn't support the method used to detect
-        SSE2 support at run-time), and even those can be reported as
-        bugs so I can refine the built-in compile-time checking to be
-        smarter.
-
-      - The old STBI_SIMD system which allowed installing a user-defined
-        IDCT etc. has been removed. If you need this, don't upgrade. My
-        assumption is that almost nobody was doing this, and those who
-        were will find the built-in SIMD more satisfactory anyway.
-
-      - RGB values computed for JPEG images are slightly different from
-        previous versions of stb_image. (This is due to using less
-        integer precision in SIMD.) The C code has been adjusted so
-        that the same RGB values will be computed regardless of whether
-        SIMD support is available, so your app should always produce
-        consistent results. But these results are slightly different from
-        previous versions. (Specifically, about 3% of available YCbCr values
-        will compute different RGB results from pre-1.49 versions by +-1;
-        most of the deviating values are one smaller in the G channel.)
-
-      - If you must produce consistent results with previous versions of
-        stb_image, #define STBI_JPEG_OLD and you will get the same results
-        you used to; however, you will not get the SIMD speedups for
-        the YCbCr-to-RGB conversion step (although you should still see
-        significant JPEG speedup from the other changes).
-
-        Please note that STBI_JPEG_OLD is a temporary feature; it will be
-        removed in future versions of the library. It is only intended for
-        near-term back-compatibility use.
-
-
-   Latest revision history:
+LICENSE
+
+  See end of file for license information.
+
+RECENT REVISION HISTORY:
+
+      2.16  (2017-07-23) all functions have 16-bit variants; optimizations; bugfixes
+      2.15  (2017-03-18) fix png-1,2,4; all Imagenet JPGs; no runtime SSE detection on GCC
+      2.14  (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs
+      2.13  (2016-12-04) experimental 16-bit API, only for PNG so far; fixes
       2.12  (2016-04-02) fix typo in 2.11 PSD fix that caused crashes
       2.11  (2016-04-02) 16-bit PNGS; enable SSE2 in non-gcc x64
                          RGB-format JPEG; remove white matting in PSD;
-                         allocate large structures on the stack; 
+                         allocate large structures on the stack;
                          correct channel count for PNG & BMP
       2.10  (2016-01-22) avoid warning introduced in 2.09
       2.09  (2016-01-16) 16-bit TGA; comments in PNM files; STBI_REALLOC_SIZED
-      2.08  (2015-09-13) fix to 2.07 cleanup, reading RGB PSD as RGBA
-      2.07  (2015-09-13) partial animated GIF support
-                         limited 16-bit PSD support
-                         minor bugs, code cleanup, and compiler warnings
-      2.06  (2015-04-19) fix bug where PSD returns wrong '*comp' value
-      2.05  (2015-04-19) fix bug in progressive JPEG handling, fix warning
-      2.04  (2015-04-15) try to re-enable SIMD on MinGW 64-bit
-      2.03  (2015-04-12) additional corruption checking
-                         stbi_set_flip_vertically_on_load
-                         fix NEON support; fix mingw support
-      2.02  (2015-01-19) fix incorrect assert, fix warning
-      2.01  (2015-01-17) fix various warnings
-      2.00b (2014-12-25) fix STBI_MALLOC in progressive JPEG
-      2.00  (2014-12-25) optimize JPEG, including x86 SSE2 & ARM NEON SIMD
-                         progressive JPEG
-                         PGM/PPM support
-                         STBI_MALLOC,STBI_REALLOC,STBI_FREE
-                         STBI_NO_*, STBI_ONLY_*
-                         GIF bugfix
 
    See end of file for full revision history.
 
@@ -186,34 +73,30 @@
     Tom Seddon (pic)                       Omar Cornut (1/2/4-bit PNG)
     Thatcher Ulrich (psd)                  Nicolas Guillemot (vertical flip)
     Ken Miller (pgm, ppm)                  Richard Mitton (16-bit PSD)
-    urraka@github (animated gif)           Junggon Kim (PNM comments)
+    github:urraka (animated gif)           Junggon Kim (PNM comments)
                                            Daniel Gibson (16-bit TGA)
-
+                                           socks-the-fox (16-bit PNG)
+                                           Jeremy Sawicki (handle all ImageNet JPGs)
  Optimizations & bugfixes
     Fabian "ryg" Giesen
     Arseny Kapoulkine
+    John-Mark Allen
 
  Bug & warning fixes
     Marc LeBlanc            David Woo          Guillaume George   Martins Mozeiko
-    Christpher Lloyd        Martin Golini      Jerry Jansson      Joseph Thomson
-    Dave Moore              Roy Eltham         Hayaki Saito       Phil Jordan
-    Won Chun                Luke Graham        Johan Duparc       Nathan Reed
-    the Horde3D community   Thomas Ruf         Ronny Chevalier    Nick Verigakis
-    Janez Zemva             John Bartholomew   Michal Cichon      svdijk@github
-    Jonathan Blow           Ken Hamada         Tero Hanninen      Baldur Karlsson
-    Laurent Gomila          Cort Stratton      Sergio Gonzalez    romigrou@github
-    Aruelien Pocheville     Thibault Reuille   Cass Everitt       Matthew Gregan
-    Ryamond Barbiero        Paul Du Bois       Engin Manap        snagar@github
-    Michaelangel007@github  Oriol Ferrer Mesia socks-the-fox
-    Blazej Dariusz Roszkowski
-
-
-LICENSE
-
-This software is dual-licensed to the public domain and under the following
-license: you are granted a perpetual, irrevocable license to copy, modify,
-publish, and distribute this file as you see fit.
-
+    Christpher Lloyd        Jerry Jansson      Joseph Thomson     Phil Jordan
+    Dave Moore              Roy Eltham         Hayaki Saito       Nathan Reed
+    Won Chun                Luke Graham        Johan Duparc       Nick Verigakis
+    the Horde3D community   Thomas Ruf         Ronny Chevalier    Baldur Karlsson
+    Janez Zemva             John Bartholomew   Michal Cichon      github:rlyeh
+    Jonathan Blow           Ken Hamada         Tero Hanninen      github:romigrou
+    Laurent Gomila          Cort Stratton      Sergio Gonzalez    github:svdijk
+    Aruelien Pocheville     Thibault Reuille   Cass Everitt       github:snagar
+    Ryamond Barbiero        Paul Du Bois       Engin Manap        github:Zelex
+    Michaelangel007@github  Philipp Wiesemann  Dale Weiler        github:grim210
+    Oriol Ferrer Mesia      Josh Tobin         Matthew Gregan     github:sammyhw
+    Blazej Dariusz Roszkowski                  Gregory Mullen     github:phprus
+    Christian Floisand      Kevin Schmidt                         github:poppolopoppo
 */
 
 #ifndef STBI_INCLUDE_STB_IMAGE_H
@@ -238,10 +121,10 @@ publish, and distribute this file as you see fit.
 //    stbi_image_free(data)
 //
 // Standard parameters:
-//    int *x       -- outputs image width in pixels
-//    int *y       -- outputs image height in pixels
-//    int *comp    -- outputs # of image components in image file
-//    int req_comp -- if non-zero, # of image components requested in result
+//    int *x                 -- outputs image width in pixels
+//    int *y                 -- outputs image height in pixels
+//    int *channels_in_file  -- outputs # of image components in image file
+//    int desired_channels   -- if non-zero, # of image components requested in result
 //
 // The return value from an image loader is an 'unsigned char *' which points
 // to the pixel data, or NULL on an allocation failure or if the image is
@@ -249,11 +132,12 @@ publish, and distribute this file as you see fit.
 // with each pixel consisting of N interleaved 8-bit components; the first
 // pixel pointed to is top-left-most in the image. There is no padding between
 // image scanlines or between pixels, regardless of format. The number of
-// components N is 'req_comp' if req_comp is non-zero, or *comp otherwise.
-// If req_comp is non-zero, *comp has the number of components that _would_
-// have been output otherwise. E.g. if you set req_comp to 4, you will always
-// get RGBA output, but you can check *comp to see if it's trivially opaque
-// because e.g. there were only 3 channels in the source image.
+// components N is 'desired_channels' if desired_channels is non-zero, or
+// *channels_in_file otherwise. If desired_channels is non-zero,
+// *channels_in_file has the number of components that _would_ have been
+// output otherwise. E.g. if you set desired_channels to 4, you will always
+// get RGBA output, but you can check *channels_in_file to see if it's trivially
+// opaque because e.g. there were only 3 channels in the source image.
 //
 // An output image with N components has the following components interleaved
 // in this order in each pixel:
@@ -265,10 +149,10 @@ publish, and distribute this file as you see fit.
 //       4           red, green, blue, alpha
 //
 // If image loading fails for any reason, the return value will be NULL,
-// and *x, *y, *comp will be unchanged. The function stbi_failure_reason()
-// can be queried for an extremely brief, end-user unfriendly explanation
-// of why the load failed. Define STBI_NO_FAILURE_STRINGS to avoid
-// compiling these strings at all, and STBI_FAILURE_USERMSG to get slightly
+// and *x, *y, *channels_in_file will be unchanged. The function
+// stbi_failure_reason() can be queried for an extremely brief, end-user
+// unfriendly explanation of why the load failed. Define STBI_NO_FAILURE_STRINGS
+// to avoid compiling these strings at all, and STBI_FAILURE_USERMSG to get slightly
 // more user-friendly ones.
 //
 // Paletted PNG, BMP, GIF, and PIC images are automatically depalettized.
@@ -287,13 +171,13 @@ publish, and distribute this file as you see fit.
 // and for best performance I may provide less-easy-to-use APIs that give higher
 // performance, in addition to the easy to use ones. Nevertheless, it's important
 // to keep in mind that from the standpoint of you, a client of this library,
-// all you care about is #1 and #3, and stb libraries do not emphasize #3 above all.
+// all you care about is #1 and #3, and stb libraries DO NOT emphasize #3 above all.
 //
 // Some secondary priorities arise directly from the first two, some of which
 // make more explicit reasons why performance can't be emphasized.
 //
 //    - Portable ("ease of use")
-//    - Small footprint ("easy to maintain")
+//    - Small source code footprint ("easy to maintain")
 //    - No dependencies ("ease of use")
 //
 // ===========================================================================
@@ -325,13 +209,6 @@ publish, and distribute this file as you see fit.
 // (at least this is true for iOS and Android). Therefore, the NEON support is
 // toggled by a build flag: define STBI_NEON to get NEON loops.
 //
-// The output of the JPEG decoder is slightly different from versions where
-// SIMD support was introduced (that is, for versions before 1.49). The
-// difference is only +-1 in the 8-bit RGB channels, and only on a small
-// fraction of pixels. You can force the pre-1.49 behavior by defining
-// STBI_JPEG_OLD, but this will disable some of the SIMD decoding path
-// and hence cost some performance.
-//
 // If for some reason you do not want to use any of SIMD code, or if
 // you have issues compiling it, you can disable it entirely by
 // defining STBI_NO_SIMD.
@@ -387,6 +264,41 @@ publish, and distribute this file as you see fit.
 // says there's premultiplied data (currently only happens in iPhone images,
 // and only if iPhone convert-to-rgb processing is on).
 //
+// ===========================================================================
+//
+// ADDITIONAL CONFIGURATION
+//
+//  - You can suppress implementation of any of the decoders to reduce
+//    your code footprint by #defining one or more of the following
+//    symbols before creating the implementation.
+//
+//        STBI_NO_JPEG
+//        STBI_NO_PNG
+//        STBI_NO_BMP
+//        STBI_NO_PSD
+//        STBI_NO_TGA
+//        STBI_NO_GIF
+//        STBI_NO_HDR
+//        STBI_NO_PIC
+//        STBI_NO_PNM   (.ppm and .pgm)
+//
+//  - You can request *only* certain decoders and suppress all other ones
+//    (this will be more forward-compatible, as addition of new decoders
+//    doesn't require you to disable them explicitly):
+//
+//        STBI_ONLY_JPEG
+//        STBI_ONLY_PNG
+//        STBI_ONLY_BMP
+//        STBI_ONLY_PSD
+//        STBI_ONLY_TGA
+//        STBI_ONLY_GIF
+//        STBI_ONLY_HDR
+//        STBI_ONLY_PIC
+//        STBI_ONLY_PNM   (.ppm and .pgm)
+//
+//   - If you use STBI_NO_PNG (or _ONLY_ without PNG), and you still
+//     want the zlib decoder to be available, #define STBI_SUPPORT_ZLIB
+//
 
 
 #ifndef STBI_NO_STDIO
@@ -397,7 +309,7 @@ publish, and distribute this file as you see fit.
 
 enum
 {
-   STBI_default = 0, // only used for req_comp
+   STBI_default = 0, // only used for desired_channels
 
    STBI_grey       = 1,
    STBI_grey_alpha = 2,
@@ -406,6 +318,7 @@ enum
 };
 
 typedef unsigned char stbi_uc;
+typedef unsigned short stbi_us;
 
 #ifdef __cplusplus
 extern "C" {
@@ -433,22 +346,44 @@ typedef struct
    int      (*eof)   (void *user);                       // returns nonzero if we are at end of file/data
 } stbi_io_callbacks;
 
-STBIDEF stbi_uc *stbi_load               (char              const *filename,           int *x, int *y, int *comp, int req_comp);
-STBIDEF stbi_uc *stbi_load_from_memory   (stbi_uc           const *buffer, int len   , int *x, int *y, int *comp, int req_comp);
-STBIDEF stbi_uc *stbi_load_from_callbacks(stbi_io_callbacks const *clbk  , void *user, int *x, int *y, int *comp, int req_comp);
+////////////////////////////////////
+//
+// 8-bits-per-channel interface
+//
+
+STBIDEF stbi_uc *stbi_load_from_memory   (stbi_uc           const *buffer, int len   , int *x, int *y, int *channels_in_file, int desired_channels);
+STBIDEF stbi_uc *stbi_load_from_callbacks(stbi_io_callbacks const *clbk  , void *user, int *x, int *y, int *channels_in_file, int desired_channels);
 
 #ifndef STBI_NO_STDIO
-STBIDEF stbi_uc *stbi_load_from_file  (FILE *f,                  int *x, int *y, int *comp, int req_comp);
+STBIDEF stbi_uc *stbi_load            (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels);
+STBIDEF stbi_uc *stbi_load_from_file  (FILE *f, int *x, int *y, int *channels_in_file, int desired_channels);
 // for stbi_load_from_file, file pointer is left pointing immediately after image
 #endif
 
+////////////////////////////////////
+//
+// 16-bits-per-channel interface
+//
+
+STBIDEF stbi_us *stbi_load_16_from_memory   (stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels);
+STBIDEF stbi_us *stbi_load_16_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels);
+
+#ifndef STBI_NO_STDIO
+STBIDEF stbi_us *stbi_load_16          (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels);
+STBIDEF stbi_us *stbi_load_from_file_16(FILE *f, int *x, int *y, int *channels_in_file, int desired_channels);
+#endif
+
+////////////////////////////////////
+//
+// float-per-channel interface
+//
 #ifndef STBI_NO_LINEAR
-   STBIDEF float *stbi_loadf                 (char const *filename,           int *x, int *y, int *comp, int req_comp);
-   STBIDEF float *stbi_loadf_from_memory     (stbi_uc const *buffer, int len, int *x, int *y, int *comp, int req_comp);
-   STBIDEF float *stbi_loadf_from_callbacks  (stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp, int req_comp);
+   STBIDEF float *stbi_loadf_from_memory     (stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels);
+   STBIDEF float *stbi_loadf_from_callbacks  (stbi_io_callbacks const *clbk, void *user, int *x, int *y,  int *channels_in_file, int desired_channels);
 
    #ifndef STBI_NO_STDIO
-   STBIDEF float *stbi_loadf_from_file  (FILE *f,                int *x, int *y, int *comp, int req_comp);
+   STBIDEF float *stbi_loadf            (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels);
+   STBIDEF float *stbi_loadf_from_file  (FILE *f, int *x, int *y, int *channels_in_file, int desired_channels);
    #endif
 #endif
 
@@ -650,12 +585,14 @@ typedef unsigned char validate_uint32[sizeof(stbi__uint32)==4 ? 1 : -1];
 #define STBI__X86_TARGET
 #endif
 
-#if defined(__GNUC__) && (defined(STBI__X86_TARGET) || defined(STBI__X64_TARGET)) && !defined(__SSE2__) && !defined(STBI_NO_SIMD)
-// NOTE: not clear do we actually need this for the 64-bit path?
+#if defined(__GNUC__) && defined(STBI__X86_TARGET) && !defined(__SSE2__) && !defined(STBI_NO_SIMD)
 // gcc doesn't support sse2 intrinsics unless you compile with -msse2,
-// (but compiling with -msse2 allows the compiler to use SSE2 everywhere;
-// this is just broken and gcc are jerks for not fixing it properly
-// http://www.virtualdub.org/blog/pivot/entry.php?id=363 )
+// which in turn means it gets to use SSE2 everywhere. This is unfortunate,
+// but previous attempts to provide the SSE2 functions with runtime
+// detection caused numerous issues. The way architecture extensions are
+// exposed in GCC/Clang is, sadly, not really suited for one-file libs.
+// New behavior: if compiled with -msse2, we use SSE2 without any
+// detection; if not, we don't use it at all.
 #define STBI_NO_SIMD
 #endif
 
@@ -703,7 +640,7 @@ static int stbi__cpuid3(void)
 
 #define STBI_SIMD_ALIGN(type, name) __declspec(align(16)) type name
 
-static int stbi__sse2_available()
+static int stbi__sse2_available(void)
 {
    int info3 = stbi__cpuid3();
    return ((info3 >> 26) & 1) != 0;
@@ -711,16 +648,12 @@ static int stbi__sse2_available()
 #else // assume GCC-style if not VC++
 #define STBI_SIMD_ALIGN(type, name) type name __attribute__((aligned(16)))
 
-static int stbi__sse2_available()
+static int stbi__sse2_available(void)
 {
-#if defined(__GNUC__) && (__GNUC__ * 100 + __GNUC_MINOR__) >= 408 // GCC 4.8 or later
-   // GCC 4.8+ has a nice way to do this
-   return __builtin_cpu_supports("sse2");
-#else
-   // portable way to do this, preferably without using GCC inline ASM?
-   // just bail for now.
-   return 0;
-#endif
+   // If we're even attempting to compile this on GCC/Clang, that means
+   // -msse2 is on, which means the compiler is allowed to use SSE2
+   // instructions at will, and so are we.
+   return 1;
 }
 #endif
 #endif
@@ -828,57 +761,70 @@ static void stbi__rewind(stbi__context *s)
    s->img_buffer_end = s->img_buffer_original_end;
 }
 
+enum
+{
+   STBI_ORDER_RGB,
+   STBI_ORDER_BGR
+};
+
+typedef struct
+{
+   int bits_per_channel;
+   int num_channels;
+   int channel_order;
+} stbi__result_info;
+
 #ifndef STBI_NO_JPEG
 static int      stbi__jpeg_test(stbi__context *s);
-static stbi_uc *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__jpeg_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_PNG
 static int      stbi__png_test(stbi__context *s);
-static stbi_uc *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__png_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_BMP
 static int      stbi__bmp_test(stbi__context *s);
-static stbi_uc *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_TGA
 static int      stbi__tga_test(stbi__context *s);
-static stbi_uc *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__tga_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_PSD
 static int      stbi__psd_test(stbi__context *s);
-static stbi_uc *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc);
 static int      stbi__psd_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_HDR
 static int      stbi__hdr_test(stbi__context *s);
-static float   *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static float   *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__hdr_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_PIC
 static int      stbi__pic_test(stbi__context *s);
-static stbi_uc *stbi__pic_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__pic_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__pic_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_GIF
 static int      stbi__gif_test(stbi__context *s);
-static stbi_uc *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__gif_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
 #ifndef STBI_NO_PNM
 static int      stbi__pnm_test(stbi__context *s);
-static stbi_uc *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp);
+static void    *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri);
 static int      stbi__pnm_info(stbi__context *s, int *x, int *y, int *comp);
 #endif
 
@@ -1007,33 +953,38 @@ STBIDEF void stbi_set_flip_vertically_on_load(int flag_true_if_should_flip)
     stbi__vertically_flip_on_load = flag_true_if_should_flip;
 }
 
-static unsigned char *stbi__load_main(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__load_main(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc)
 {
+   memset(ri, 0, sizeof(*ri)); // make sure it's initialized if we add new fields
+   ri->bits_per_channel = 8; // default is 8 so most paths don't have to be changed
+   ri->channel_order = STBI_ORDER_RGB; // all current input & output are this, but this is here so we can add BGR order
+   ri->num_channels = 0;
+
    #ifndef STBI_NO_JPEG
-   if (stbi__jpeg_test(s)) return stbi__jpeg_load(s,x,y,comp,req_comp);
+   if (stbi__jpeg_test(s)) return stbi__jpeg_load(s,x,y,comp,req_comp, ri);
    #endif
    #ifndef STBI_NO_PNG
-   if (stbi__png_test(s))  return stbi__png_load(s,x,y,comp,req_comp);
+   if (stbi__png_test(s))  return stbi__png_load(s,x,y,comp,req_comp, ri);
    #endif
    #ifndef STBI_NO_BMP
-   if (stbi__bmp_test(s))  return stbi__bmp_load(s,x,y,comp,req_comp);
+   if (stbi__bmp_test(s))  return stbi__bmp_load(s,x,y,comp,req_comp, ri);
    #endif
    #ifndef STBI_NO_GIF
-   if (stbi__gif_test(s))  return stbi__gif_load(s,x,y,comp,req_comp);
+   if (stbi__gif_test(s))  return stbi__gif_load(s,x,y,comp,req_comp, ri);
    #endif
    #ifndef STBI_NO_PSD
-   if (stbi__psd_test(s))  return stbi__psd_load(s,x,y,comp,req_comp);
+   if (stbi__psd_test(s))  return stbi__psd_load(s,x,y,comp,req_comp, ri, bpc);
    #endif
    #ifndef STBI_NO_PIC
-   if (stbi__pic_test(s))  return stbi__pic_load(s,x,y,comp,req_comp);
+   if (stbi__pic_test(s))  return stbi__pic_load(s,x,y,comp,req_comp, ri);
    #endif
    #ifndef STBI_NO_PNM
-   if (stbi__pnm_test(s))  return stbi__pnm_load(s,x,y,comp,req_comp);
+   if (stbi__pnm_test(s))  return stbi__pnm_load(s,x,y,comp,req_comp, ri);
    #endif
 
    #ifndef STBI_NO_HDR
    if (stbi__hdr_test(s)) {
-      float *hdr = stbi__hdr_load(s, x,y,comp,req_comp);
+      float *hdr = stbi__hdr_load(s, x,y,comp,req_comp, ri);
       return stbi__hdr_to_ldr(hdr, *x, *y, req_comp ? req_comp : *comp);
    }
    #endif
@@ -1041,56 +992,123 @@ static unsigned char *stbi__load_main(stbi__context *s, int *x, int *y, int *com
    #ifndef STBI_NO_TGA
    // test tga last because it's a crappy test!
    if (stbi__tga_test(s))
-      return stbi__tga_load(s,x,y,comp,req_comp);
+      return stbi__tga_load(s,x,y,comp,req_comp, ri);
    #endif
 
    return stbi__errpuc("unknown image type", "Image not of any known type, or corrupt");
 }
 
-static unsigned char *stbi__load_flip(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static stbi_uc *stbi__convert_16_to_8(stbi__uint16 *orig, int w, int h, int channels)
 {
-   unsigned char *result = stbi__load_main(s, x, y, comp, req_comp);
+   int i;
+   int img_len = w * h * channels;
+   stbi_uc *reduced;
 
-   if (stbi__vertically_flip_on_load && result != NULL) {
-      int w = *x, h = *y;
-      int depth = req_comp ? req_comp : *comp;
-      int row,col,z;
-      stbi_uc temp;
-
-      // @OPTIMIZE: use a bigger temp buffer and memcpy multiple pixels at once
-      for (row = 0; row < (h>>1); row++) {
-         for (col = 0; col < w; col++) {
-            for (z = 0; z < depth; z++) {
-               temp = result[(row * w + col) * depth + z];
-               result[(row * w + col) * depth + z] = result[((h - row - 1) * w + col) * depth + z];
-               result[((h - row - 1) * w + col) * depth + z] = temp;
-            }
-         }
+   reduced = (stbi_uc *) stbi__malloc(img_len);
+   if (reduced == NULL) return stbi__errpuc("outofmem", "Out of memory");
+
+   for (i = 0; i < img_len; ++i)
+      reduced[i] = (stbi_uc)((orig[i] >> 8) & 0xFF); // top half of each byte is sufficient approx of 16->8 bit scaling
+
+   STBI_FREE(orig);
+   return reduced;
+}
+
+static stbi__uint16 *stbi__convert_8_to_16(stbi_uc *orig, int w, int h, int channels)
+{
+   int i;
+   int img_len = w * h * channels;
+   stbi__uint16 *enlarged;
+
+   enlarged = (stbi__uint16 *) stbi__malloc(img_len*2);
+   if (enlarged == NULL) return (stbi__uint16 *) stbi__errpuc("outofmem", "Out of memory");
+
+   for (i = 0; i < img_len; ++i)
+      enlarged[i] = (stbi__uint16)((orig[i] << 8) + orig[i]); // replicate to high and low byte, maps 0->0, 255->0xffff
+
+   STBI_FREE(orig);
+   return enlarged;
+}
+
+static void stbi__vertical_flip(void *image, int w, int h, int bytes_per_pixel)
+{
+   int row;
+   size_t bytes_per_row = (size_t)w * bytes_per_pixel;
+   stbi_uc temp[2048];
+   stbi_uc *bytes = (stbi_uc *)image;
+
+   for (row = 0; row < (h>>1); row++) {
+      stbi_uc *row0 = bytes + row*bytes_per_row;
+      stbi_uc *row1 = bytes + (h - row - 1)*bytes_per_row;
+      // swap row0 with row1
+      size_t bytes_left = bytes_per_row;
+      while (bytes_left) {
+         size_t bytes_copy = (bytes_left < sizeof(temp)) ? bytes_left : sizeof(temp);
+         memcpy(temp, row0, bytes_copy);
+         memcpy(row0, row1, bytes_copy);
+         memcpy(row1, temp, bytes_copy);
+         row0 += bytes_copy;
+         row1 += bytes_copy;
+         bytes_left -= bytes_copy;
       }
    }
+}
 
-   return result;
+static unsigned char *stbi__load_and_postprocess_8bit(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+{
+   stbi__result_info ri;
+   void *result = stbi__load_main(s, x, y, comp, req_comp, &ri, 8);
+
+   if (result == NULL)
+      return NULL;
+
+   if (ri.bits_per_channel != 8) {
+      STBI_ASSERT(ri.bits_per_channel == 16);
+      result = stbi__convert_16_to_8((stbi__uint16 *) result, *x, *y, req_comp == 0 ? *comp : req_comp);
+      ri.bits_per_channel = 8;
+   }
+
+   // @TODO: move stbi__convert_format to here
+
+   if (stbi__vertically_flip_on_load) {
+      int channels = req_comp ? req_comp : *comp;
+      stbi__vertical_flip(result, *x, *y, channels * sizeof(stbi_uc));
+   }
+
+   return (unsigned char *) result;
+}
+
+static stbi__uint16 *stbi__load_and_postprocess_16bit(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+{
+   stbi__result_info ri;
+   void *result = stbi__load_main(s, x, y, comp, req_comp, &ri, 16);
+
+   if (result == NULL)
+      return NULL;
+
+   if (ri.bits_per_channel != 16) {
+      STBI_ASSERT(ri.bits_per_channel == 8);
+      result = stbi__convert_8_to_16((stbi_uc *) result, *x, *y, req_comp == 0 ? *comp : req_comp);
+      ri.bits_per_channel = 16;
+   }
+
+   // @TODO: move stbi__convert_format16 to here
+   // @TODO: special case RGB-to-Y (and RGBA-to-YA) for 8-bit-to-16-bit case to keep more precision
+
+   if (stbi__vertically_flip_on_load) {
+      int channels = req_comp ? req_comp : *comp;
+      stbi__vertical_flip(result, *x, *y, channels * sizeof(stbi__uint16));
+   }
+
+   return (stbi__uint16 *) result;
 }
 
 #ifndef STBI_NO_HDR
 static void stbi__float_postprocess(float *result, int *x, int *y, int *comp, int req_comp)
 {
    if (stbi__vertically_flip_on_load && result != NULL) {
-      int w = *x, h = *y;
-      int depth = req_comp ? req_comp : *comp;
-      int row,col,z;
-      float temp;
-
-      // @OPTIMIZE: use a bigger temp buffer and memcpy multiple pixels at once
-      for (row = 0; row < (h>>1); row++) {
-         for (col = 0; col < w; col++) {
-            for (z = 0; z < depth; z++) {
-               temp = result[(row * w + col) * depth + z];
-               result[(row * w + col) * depth + z] = result[((h - row - 1) * w + col) * depth + z];
-               result[((h - row - 1) * w + col) * depth + z] = temp;
-            }
-         }
-      }
+      int channels = req_comp ? req_comp : *comp;
+      stbi__vertical_flip(result, *x, *y, channels * sizeof(float));
    }
 }
 #endif
@@ -1125,27 +1143,66 @@ STBIDEF stbi_uc *stbi_load_from_file(FILE *f, int *x, int *y, int *comp, int req
    unsigned char *result;
    stbi__context s;
    stbi__start_file(&s,f);
-   result = stbi__load_flip(&s,x,y,comp,req_comp);
+   result = stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp);
    if (result) {
       // need to 'unget' all the characters in the IO buffer
       fseek(f, - (int) (s.img_buffer_end - s.img_buffer), SEEK_CUR);
    }
    return result;
 }
+
+STBIDEF stbi__uint16 *stbi_load_from_file_16(FILE *f, int *x, int *y, int *comp, int req_comp)
+{
+   stbi__uint16 *result;
+   stbi__context s;
+   stbi__start_file(&s,f);
+   result = stbi__load_and_postprocess_16bit(&s,x,y,comp,req_comp);
+   if (result) {
+      // need to 'unget' all the characters in the IO buffer
+      fseek(f, - (int) (s.img_buffer_end - s.img_buffer), SEEK_CUR);
+   }
+   return result;
+}
+
+STBIDEF stbi_us *stbi_load_16(char const *filename, int *x, int *y, int *comp, int req_comp)
+{
+   FILE *f = stbi__fopen(filename, "rb");
+   stbi__uint16 *result;
+   if (!f) return (stbi_us *) stbi__errpuc("can't fopen", "Unable to open file");
+   result = stbi_load_from_file_16(f,x,y,comp,req_comp);
+   fclose(f);
+   return result;
+}
+
+
 #endif //!STBI_NO_STDIO
 
+STBIDEF stbi_us *stbi_load_16_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels)
+{
+   stbi__context s;
+   stbi__start_mem(&s,buffer,len);
+   return stbi__load_and_postprocess_16bit(&s,x,y,channels_in_file,desired_channels);
+}
+
+STBIDEF stbi_us *stbi_load_16_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels)
+{
+   stbi__context s;
+   stbi__start_callbacks(&s, (stbi_io_callbacks *)clbk, user);
+   return stbi__load_and_postprocess_16bit(&s,x,y,channels_in_file,desired_channels);
+}
+
 STBIDEF stbi_uc *stbi_load_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp, int req_comp)
 {
    stbi__context s;
    stbi__start_mem(&s,buffer,len);
-   return stbi__load_flip(&s,x,y,comp,req_comp);
+   return stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp);
 }
 
 STBIDEF stbi_uc *stbi_load_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp, int req_comp)
 {
    stbi__context s;
    stbi__start_callbacks(&s, (stbi_io_callbacks *) clbk, user);
-   return stbi__load_flip(&s,x,y,comp,req_comp);
+   return stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp);
 }
 
 #ifndef STBI_NO_LINEAR
@@ -1154,13 +1211,14 @@ static float *stbi__loadf_main(stbi__context *s, int *x, int *y, int *comp, int
    unsigned char *data;
    #ifndef STBI_NO_HDR
    if (stbi__hdr_test(s)) {
-      float *hdr_data = stbi__hdr_load(s,x,y,comp,req_comp);
+      stbi__result_info ri;
+      float *hdr_data = stbi__hdr_load(s,x,y,comp,req_comp, &ri);
       if (hdr_data)
          stbi__float_postprocess(hdr_data,x,y,comp,req_comp);
       return hdr_data;
    }
    #endif
-   data = stbi__load_flip(s, x, y, comp, req_comp);
+   data = stbi__load_and_postprocess_8bit(s, x, y, comp, req_comp);
    if (data)
       return stbi__ldr_to_hdr(data, *x, *y, req_comp ? req_comp : *comp);
    return stbi__errpf("unknown image type", "Image not of any known type, or corrupt");
@@ -1428,26 +1486,75 @@ static unsigned char *stbi__convert_format(unsigned char *data, int img_n, int r
       unsigned char *src  = data + j * x * img_n   ;
       unsigned char *dest = good + j * x * req_comp;
 
-      #define COMBO(a,b)  ((a)*8+(b))
-      #define CASE(a,b)   case COMBO(a,b): for(i=x-1; i >= 0; --i, src += a, dest += b)
+      #define STBI__COMBO(a,b)  ((a)*8+(b))
+      #define STBI__CASE(a,b)   case STBI__COMBO(a,b): for(i=x-1; i >= 0; --i, src += a, dest += b)
+      // convert source image with img_n components to one with req_comp components;
+      // avoid switch per pixel, so use switch per scanline and massive macros
+      switch (STBI__COMBO(img_n, req_comp)) {
+         STBI__CASE(1,2) { dest[0]=src[0], dest[1]=255;                                     } break;
+         STBI__CASE(1,3) { dest[0]=dest[1]=dest[2]=src[0];                                  } break;
+         STBI__CASE(1,4) { dest[0]=dest[1]=dest[2]=src[0], dest[3]=255;                     } break;
+         STBI__CASE(2,1) { dest[0]=src[0];                                                  } break;
+         STBI__CASE(2,3) { dest[0]=dest[1]=dest[2]=src[0];                                  } break;
+         STBI__CASE(2,4) { dest[0]=dest[1]=dest[2]=src[0], dest[3]=src[1];                  } break;
+         STBI__CASE(3,4) { dest[0]=src[0],dest[1]=src[1],dest[2]=src[2],dest[3]=255;        } break;
+         STBI__CASE(3,1) { dest[0]=stbi__compute_y(src[0],src[1],src[2]);                   } break;
+         STBI__CASE(3,2) { dest[0]=stbi__compute_y(src[0],src[1],src[2]), dest[1] = 255;    } break;
+         STBI__CASE(4,1) { dest[0]=stbi__compute_y(src[0],src[1],src[2]);                   } break;
+         STBI__CASE(4,2) { dest[0]=stbi__compute_y(src[0],src[1],src[2]), dest[1] = src[3]; } break;
+         STBI__CASE(4,3) { dest[0]=src[0],dest[1]=src[1],dest[2]=src[2];                    } break;
+         default: STBI_ASSERT(0);
+      }
+      #undef STBI__CASE
+   }
+
+   STBI_FREE(data);
+   return good;
+}
+
+static stbi__uint16 stbi__compute_y_16(int r, int g, int b)
+{
+   return (stbi__uint16) (((r*77) + (g*150) +  (29*b)) >> 8);
+}
+
+static stbi__uint16 *stbi__convert_format16(stbi__uint16 *data, int img_n, int req_comp, unsigned int x, unsigned int y)
+{
+   int i,j;
+   stbi__uint16 *good;
+
+   if (req_comp == img_n) return data;
+   STBI_ASSERT(req_comp >= 1 && req_comp <= 4);
+
+   good = (stbi__uint16 *) stbi__malloc(req_comp * x * y * 2);
+   if (good == NULL) {
+      STBI_FREE(data);
+      return (stbi__uint16 *) stbi__errpuc("outofmem", "Out of memory");
+   }
+
+   for (j=0; j < (int) y; ++j) {
+      stbi__uint16 *src  = data + j * x * img_n   ;
+      stbi__uint16 *dest = good + j * x * req_comp;
+
+      #define STBI__COMBO(a,b)  ((a)*8+(b))
+      #define STBI__CASE(a,b)   case STBI__COMBO(a,b): for(i=x-1; i >= 0; --i, src += a, dest += b)
       // convert source image with img_n components to one with req_comp components;
       // avoid switch per pixel, so use switch per scanline and massive macros
-      switch (COMBO(img_n, req_comp)) {
-         CASE(1,2) dest[0]=src[0], dest[1]=255; break;
-         CASE(1,3) dest[0]=dest[1]=dest[2]=src[0]; break;
-         CASE(1,4) dest[0]=dest[1]=dest[2]=src[0], dest[3]=255; break;
-         CASE(2,1) dest[0]=src[0]; break;
-         CASE(2,3) dest[0]=dest[1]=dest[2]=src[0]; break;
-         CASE(2,4) dest[0]=dest[1]=dest[2]=src[0], dest[3]=src[1]; break;
-         CASE(3,4) dest[0]=src[0],dest[1]=src[1],dest[2]=src[2],dest[3]=255; break;
-         CASE(3,1) dest[0]=stbi__compute_y(src[0],src[1],src[2]); break;
-         CASE(3,2) dest[0]=stbi__compute_y(src[0],src[1],src[2]), dest[1] = 255; break;
-         CASE(4,1) dest[0]=stbi__compute_y(src[0],src[1],src[2]); break;
-         CASE(4,2) dest[0]=stbi__compute_y(src[0],src[1],src[2]), dest[1] = src[3]; break;
-         CASE(4,3) dest[0]=src[0],dest[1]=src[1],dest[2]=src[2]; break;
+      switch (STBI__COMBO(img_n, req_comp)) {
+         STBI__CASE(1,2) { dest[0]=src[0], dest[1]=0xffff;                                     } break;
+         STBI__CASE(1,3) { dest[0]=dest[1]=dest[2]=src[0];                                     } break;
+         STBI__CASE(1,4) { dest[0]=dest[1]=dest[2]=src[0], dest[3]=0xffff;                     } break;
+         STBI__CASE(2,1) { dest[0]=src[0];                                                     } break;
+         STBI__CASE(2,3) { dest[0]=dest[1]=dest[2]=src[0];                                     } break;
+         STBI__CASE(2,4) { dest[0]=dest[1]=dest[2]=src[0], dest[3]=src[1];                     } break;
+         STBI__CASE(3,4) { dest[0]=src[0],dest[1]=src[1],dest[2]=src[2],dest[3]=0xffff;        } break;
+         STBI__CASE(3,1) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]);                   } break;
+         STBI__CASE(3,2) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]), dest[1] = 0xffff; } break;
+         STBI__CASE(4,1) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]);                   } break;
+         STBI__CASE(4,2) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]), dest[1] = src[3]; } break;
+         STBI__CASE(4,3) { dest[0]=src[0],dest[1]=src[1],dest[2]=src[2];                       } break;
          default: STBI_ASSERT(0);
       }
-      #undef CASE
+      #undef STBI__CASE
    }
 
    STBI_FREE(data);
@@ -1547,7 +1654,7 @@ typedef struct
    stbi__context *s;
    stbi__huffman huff_dc[4];
    stbi__huffman huff_ac[4];
-   stbi_uc dequant[4][64];
+   stbi__uint16 dequant[4][64];
    stbi__int16 fast_ac[4][1 << FAST_BITS];
 
 // sizes for components, interleaved MCUs
@@ -1583,6 +1690,8 @@ typedef struct
    int            succ_high;
    int            succ_low;
    int            eob_run;
+   int            jfif;
+   int            app14_color_transform; // Adobe APP14 tag
    int            rgb;
 
    int scan_n, order[4];
@@ -1653,7 +1762,7 @@ static void stbi__build_fast_ac(stbi__int16 *fast_ac, stbi__huffman *h)
             // magnitude code followed by receive_extend code
             int k = ((i << len) & ((1 << FAST_BITS) - 1)) >> (FAST_BITS - magbits);
             int m = 1 << (magbits - 1);
-            if (k < m) k += (-1 << magbits) + 1;
+            if (k < m) k += (~0U << magbits) + 1;
             // if the result is small enough, we can fit it in fast_ac table
             if (k >= -128 && k <= 127)
                fast_ac[i] = (stbi__int16) ((k << 8) + (run << 4) + (len + magbits));
@@ -1668,6 +1777,7 @@ static void stbi__grow_buffer_unsafe(stbi__jpeg *j)
       int b = j->nomore ? 0 : stbi__get8(j->s);
       if (b == 0xff) {
          int c = stbi__get8(j->s);
+         while (c == 0xff) c = stbi__get8(j->s); // consume fill bytes
          if (c != 0) {
             j->marker = (unsigned char) c;
             j->nomore = 1;
@@ -1792,7 +1902,7 @@ static stbi_uc stbi__jpeg_dezigzag[64+15] =
 };
 
 // decode one 64-entry block--
-static int stbi__jpeg_decode_block(stbi__jpeg *j, short data[64], stbi__huffman *hdc, stbi__huffman *hac, stbi__int16 *fac, int b, stbi_uc *dequant)
+static int stbi__jpeg_decode_block(stbi__jpeg *j, short data[64], stbi__huffman *hdc, stbi__huffman *hac, stbi__int16 *fac, int b, stbi__uint16 *dequant)
 {
    int diff,dc,k;
    int t;
@@ -2501,7 +2611,7 @@ static stbi_uc stbi__get_marker(stbi__jpeg *j)
    x = stbi__get8(j->s);
    if (x != 0xff) return STBI__MARKER_none;
    while (x == 0xff)
-      x = stbi__get8(j->s);
+      x = stbi__get8(j->s); // consume repeated 0xff fill bytes
    return x;
 }
 
@@ -2516,7 +2626,7 @@ static void stbi__jpeg_reset(stbi__jpeg *j)
    j->code_bits = 0;
    j->code_buffer = 0;
    j->nomore = 0;
-   j->img_comp[0].dc_pred = j->img_comp[1].dc_pred = j->img_comp[2].dc_pred = 0;
+   j->img_comp[0].dc_pred = j->img_comp[1].dc_pred = j->img_comp[2].dc_pred = j->img_comp[3].dc_pred = 0;
    j->marker = STBI__MARKER_none;
    j->todo = j->restart_interval ? j->restart_interval : 0x7fffffff;
    j->eob_run = 0;
@@ -2648,7 +2758,7 @@ static int stbi__parse_entropy_coded_data(stbi__jpeg *z)
    }
 }
 
-static void stbi__jpeg_dequantize(short *data, stbi_uc *dequant)
+static void stbi__jpeg_dequantize(short *data, stbi__uint16 *dequant)
 {
    int i;
    for (i=0; i < 64; ++i)
@@ -2690,13 +2800,14 @@ static int stbi__process_marker(stbi__jpeg *z, int m)
          L = stbi__get16be(z->s)-2;
          while (L > 0) {
             int q = stbi__get8(z->s);
-            int p = q >> 4;
+            int p = q >> 4, sixteen = (p != 0);
             int t = q & 15,i;
-            if (p != 0) return stbi__err("bad DQT type","Corrupt JPEG");
+            if (p != 0 && p != 1) return stbi__err("bad DQT type","Corrupt JPEG");
             if (t > 3) return stbi__err("bad DQT table","Corrupt JPEG");
+
             for (i=0; i < 64; ++i)
-               z->dequant[t][stbi__jpeg_dezigzag[i]] = stbi__get8(z->s);
-            L -= 65;
+               z->dequant[t][stbi__jpeg_dezigzag[i]] = (stbi__uint16)(sixteen ? stbi__get16be(z->s) : stbi__get8(z->s));
+            L -= (sixteen ? 129 : 65);
          }
          return L==0;
 
@@ -2729,12 +2840,50 @@ static int stbi__process_marker(stbi__jpeg *z, int m)
          }
          return L==0;
    }
+
    // check for comment block or APP blocks
    if ((m >= 0xE0 && m <= 0xEF) || m == 0xFE) {
-      stbi__skip(z->s, stbi__get16be(z->s)-2);
+      L = stbi__get16be(z->s);
+      if (L < 2) {
+         if (m == 0xFE)
+            return stbi__err("bad COM len","Corrupt JPEG");
+         else
+            return stbi__err("bad APP len","Corrupt JPEG");
+      }
+      L -= 2;
+
+      if (m == 0xE0 && L >= 5) { // JFIF APP0 segment
+         static const unsigned char tag[5] = {'J','F','I','F','\0'};
+         int ok = 1;
+         int i;
+         for (i=0; i < 5; ++i)
+            if (stbi__get8(z->s) != tag[i])
+               ok = 0;
+         L -= 5;
+         if (ok)
+            z->jfif = 1;
+      } else if (m == 0xEE && L >= 12) { // Adobe APP14 segment
+         static const unsigned char tag[6] = {'A','d','o','b','e','\0'};
+         int ok = 1;
+         int i;
+         for (i=0; i < 6; ++i)
+            if (stbi__get8(z->s) != tag[i])
+               ok = 0;
+         L -= 6;
+         if (ok) {
+            stbi__get8(z->s); // version
+            stbi__get16be(z->s); // flags0
+            stbi__get16be(z->s); // flags1
+            z->app14_color_transform = stbi__get8(z->s); // color transform
+            L -= 6;
+         }
+      }
+
+      stbi__skip(z->s, L);
       return 1;
    }
-   return 0;
+
+   return stbi__err("unknown marker","Corrupt JPEG");
 }
 
 // after we see SOS
@@ -2808,7 +2957,7 @@ static int stbi__process_frame_header(stbi__jpeg *z, int scan)
    s->img_y = stbi__get16be(s);   if (s->img_y == 0) return stbi__err("no header height", "JPEG format not supported: delayed height"); // Legal, but we don't handle it--but neither does IJG
    s->img_x = stbi__get16be(s);   if (s->img_x == 0) return stbi__err("0 width","Corrupt JPEG"); // JPEG requires
    c = stbi__get8(s);
-   if (c != 3 && c != 1) return stbi__err("bad component count","Corrupt JPEG");    // JFIF requires
+   if (c != 3 && c != 1 && c != 4) return stbi__err("bad component count","Corrupt JPEG");
    s->img_n = c;
    for (i=0; i < c; ++i) {
       z->img_comp[i].data = NULL;
@@ -2821,13 +2970,8 @@ static int stbi__process_frame_header(stbi__jpeg *z, int scan)
    for (i=0; i < s->img_n; ++i) {
       static unsigned char rgb[3] = { 'R', 'G', 'B' };
       z->img_comp[i].id = stbi__get8(s);
-      if (z->img_comp[i].id != i+1)   // JFIF requires
-         if (z->img_comp[i].id != i) {  // some version of jpegtran outputs non-JFIF-compliant files!
-            // somethings output this (see http://fileformats.archiveteam.org/wiki/JPEG#Color_format)
-            if (z->img_comp[i].id != rgb[i])
-               return stbi__err("bad component ID","Corrupt JPEG");
-            ++z->rgb;
-         }
+      if (s->img_n == 3 && z->img_comp[i].id == rgb[i])
+         ++z->rgb;
       q = stbi__get8(s);
       z->img_comp[i].h = (q >> 4);  if (!z->img_comp[i].h || z->img_comp[i].h > 4) return stbi__err("bad H","Corrupt JPEG");
       z->img_comp[i].v = q & 15;    if (!z->img_comp[i].v || z->img_comp[i].v > 4) return stbi__err("bad V","Corrupt JPEG");
@@ -2899,6 +3043,8 @@ static int stbi__process_frame_header(stbi__jpeg *z, int scan)
 static int stbi__decode_jpeg_header(stbi__jpeg *z, int scan)
 {
    int m;
+   z->jfif = 0;
+   z->app14_color_transform = -1; // valid values are 0,1,2
    z->marker = STBI__MARKER_none; // initialize cached marker to empty
    m = stbi__get_marker(z);
    if (!stbi__SOI(m)) return stbi__err("no SOI","Corrupt JPEG");
@@ -2940,12 +3086,15 @@ static int stbi__decode_jpeg_image(stbi__jpeg *j)
                if (x == 255) {
                   j->marker = stbi__get8(j->s);
                   break;
-               } else if (x != 0) {
-                  return stbi__err("junk before marker", "Corrupt JPEG");
                }
             }
             // if we reach eof without hitting a marker, stbi__get_marker() below will fail and we'll eventually return 0
          }
+      } else if (stbi__DNL(m)) {
+         int Ld = stbi__get16be(j->s);
+         stbi__uint32 NL = stbi__get16be(j->s);
+         if (Ld != 4) stbi__err("bad DNL len", "Corrupt JPEG");
+         if (NL != j->s->img_y) stbi__err("bad DNL height", "Corrupt JPEG");
       } else {
          if (!stbi__process_marker(j, m)) return 0;
       }
@@ -3164,38 +3313,9 @@ static stbi_uc *stbi__resample_row_generic(stbi_uc *out, stbi_uc *in_near, stbi_
    return out;
 }
 
-#ifdef STBI_JPEG_OLD
-// this is the same YCbCr-to-RGB calculation that stb_image has used
-// historically before the algorithm changes in 1.49
-#define float2fixed(x)  ((int) ((x) * 65536 + 0.5))
-static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step)
-{
-   int i;
-   for (i=0; i < count; ++i) {
-      int y_fixed = (y[i] << 16) + 32768; // rounding
-      int r,g,b;
-      int cr = pcr[i] - 128;
-      int cb = pcb[i] - 128;
-      r = y_fixed + cr*float2fixed(1.40200f);
-      g = y_fixed - cr*float2fixed(0.71414f) - cb*float2fixed(0.34414f);
-      b = y_fixed                            + cb*float2fixed(1.77200f);
-      r >>= 16;
-      g >>= 16;
-      b >>= 16;
-      if ((unsigned) r > 255) { if (r < 0) r = 0; else r = 255; }
-      if ((unsigned) g > 255) { if (g < 0) g = 0; else g = 255; }
-      if ((unsigned) b > 255) { if (b < 0) b = 0; else b = 255; }
-      out[0] = (stbi_uc)r;
-      out[1] = (stbi_uc)g;
-      out[2] = (stbi_uc)b;
-      out[3] = 255;
-      out += step;
-   }
-}
-#else
 // this is a reduced-precision calculation of YCbCr-to-RGB introduced
 // to make sure the code produces the same results in both SIMD and scalar
-#define float2fixed(x)  (((int) ((x) * 4096.0f + 0.5f)) << 8)
+#define stbi__float2fixed(x)  (((int) ((x) * 4096.0f + 0.5f)) << 8)
 static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step)
 {
    int i;
@@ -3204,9 +3324,9 @@ static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc
       int r,g,b;
       int cr = pcr[i] - 128;
       int cb = pcb[i] - 128;
-      r = y_fixed +  cr* float2fixed(1.40200f);
-      g = y_fixed + (cr*-float2fixed(0.71414f)) + ((cb*-float2fixed(0.34414f)) & 0xffff0000);
-      b = y_fixed                               +   cb* float2fixed(1.77200f);
+      r = y_fixed +  cr* stbi__float2fixed(1.40200f);
+      g = y_fixed + (cr*-stbi__float2fixed(0.71414f)) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000);
+      b = y_fixed                                     +   cb* stbi__float2fixed(1.77200f);
       r >>= 20;
       g >>= 20;
       b >>= 20;
@@ -3220,7 +3340,6 @@ static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc
       out += step;
    }
 }
-#endif
 
 #if defined(STBI_SSE2) || defined(STBI_NEON)
 static void stbi__YCbCr_to_RGB_simd(stbi_uc *out, stbi_uc const *y, stbi_uc const *pcb, stbi_uc const *pcr, int count, int step)
@@ -3339,9 +3458,9 @@ static void stbi__YCbCr_to_RGB_simd(stbi_uc *out, stbi_uc const *y, stbi_uc cons
       int r,g,b;
       int cr = pcr[i] - 128;
       int cb = pcb[i] - 128;
-      r = y_fixed + cr* float2fixed(1.40200f);
-      g = y_fixed + cr*-float2fixed(0.71414f) + ((cb*-float2fixed(0.34414f)) & 0xffff0000);
-      b = y_fixed                             +   cb* float2fixed(1.77200f);
+      r = y_fixed + cr* stbi__float2fixed(1.40200f);
+      g = y_fixed + cr*-stbi__float2fixed(0.71414f) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000);
+      b = y_fixed                                   +   cb* stbi__float2fixed(1.77200f);
       r >>= 20;
       g >>= 20;
       b >>= 20;
@@ -3367,18 +3486,14 @@ static void stbi__setup_jpeg(stbi__jpeg *j)
 #ifdef STBI_SSE2
    if (stbi__sse2_available()) {
       j->idct_block_kernel = stbi__idct_simd;
-      #ifndef STBI_JPEG_OLD
       j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd;
-      #endif
       j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd;
    }
 #endif
 
 #ifdef STBI_NEON
    j->idct_block_kernel = stbi__idct_simd;
-   #ifndef STBI_JPEG_OLD
    j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd;
-   #endif
    j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd;
 #endif
 }
@@ -3399,9 +3514,16 @@ typedef struct
    int ypos;    // which pre-expansion row we're on
 } stbi__resample;
 
+// fast 0..255 * 0..255 => 0..255 rounded multiplication
+static stbi_uc stbi__blinn_8x8(stbi_uc x, stbi_uc y)
+{
+   unsigned int t = x*y + 128;
+   return (stbi_uc) ((t + (t >>8)) >> 8);
+}
+
 static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp, int req_comp)
 {
-   int n, decode_n;
+   int n, decode_n, is_rgb;
    z->s->img_n = 0; // make stbi__cleanup_jpeg safe
 
    // validate req_comp
@@ -3411,9 +3533,11 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp
    if (!stbi__decode_jpeg_image(z)) { stbi__cleanup_jpeg(z); return NULL; }
 
    // determine actual number of components to generate
-   n = req_comp ? req_comp : z->s->img_n;
+   n = req_comp ? req_comp : z->s->img_n >= 3 ? 3 : 1;
 
-   if (z->s->img_n == 3 && n < 3)
+   is_rgb = z->s->img_n == 3 && (z->rgb == 3 || (z->app14_color_transform == 0 && !z->jfif));
+
+   if (z->s->img_n == 3 && n < 3 && !is_rgb)
       decode_n = 1;
    else
       decode_n = z->s->img_n;
@@ -3473,7 +3597,7 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp
          if (n >= 3) {
             stbi_uc *y = coutput[0];
             if (z->s->img_n == 3) {
-               if (z->rgb == 3) {
+               if (is_rgb) {
                   for (i=0; i < z->s->img_x; ++i) {
                      out[0] = y[i];
                      out[1] = coutput[1][i];
@@ -3484,6 +3608,28 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp
                } else {
                   z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n);
                }
+            } else if (z->s->img_n == 4) {
+               if (z->app14_color_transform == 0) { // CMYK
+                  for (i=0; i < z->s->img_x; ++i) {
+                     stbi_uc m = coutput[3][i];
+                     out[0] = stbi__blinn_8x8(coutput[0][i], m);
+                     out[1] = stbi__blinn_8x8(coutput[1][i], m);
+                     out[2] = stbi__blinn_8x8(coutput[2][i], m);
+                     out[3] = 255;
+                     out += n;
+                  }
+               } else if (z->app14_color_transform == 2) { // YCCK
+                  z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n);
+                  for (i=0; i < z->s->img_x; ++i) {
+                     stbi_uc m = coutput[3][i];
+                     out[0] = stbi__blinn_8x8(255 - out[0], m);
+                     out[1] = stbi__blinn_8x8(255 - out[1], m);
+                     out[2] = stbi__blinn_8x8(255 - out[2], m);
+                     out += n;
+                  }
+               } else { // YCbCr + alpha?  Ignore the fourth channel for now
+                  z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n);
+               }
             } else
                for (i=0; i < z->s->img_x; ++i) {
                   out[0] = out[1] = out[2] = y[i];
@@ -3491,25 +3637,54 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp
                   out += n;
                }
          } else {
-            stbi_uc *y = coutput[0];
-            if (n == 1)
-               for (i=0; i < z->s->img_x; ++i) out[i] = y[i];
-            else
-               for (i=0; i < z->s->img_x; ++i) *out++ = y[i], *out++ = 255;
+            if (is_rgb) {
+               if (n == 1)
+                  for (i=0; i < z->s->img_x; ++i)
+                     *out++ = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]);
+               else {
+                  for (i=0; i < z->s->img_x; ++i, out += 2) {
+                     out[0] = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]);
+                     out[1] = 255;
+                  }
+               }
+            } else if (z->s->img_n == 4 && z->app14_color_transform == 0) {
+               for (i=0; i < z->s->img_x; ++i) {
+                  stbi_uc m = coutput[3][i];
+                  stbi_uc r = stbi__blinn_8x8(coutput[0][i], m);
+                  stbi_uc g = stbi__blinn_8x8(coutput[1][i], m);
+                  stbi_uc b = stbi__blinn_8x8(coutput[2][i], m);
+                  out[0] = stbi__compute_y(r, g, b);
+                  out[1] = 255;
+                  out += n;
+               }
+            } else if (z->s->img_n == 4 && z->app14_color_transform == 2) {
+               for (i=0; i < z->s->img_x; ++i) {
+                  out[0] = stbi__blinn_8x8(255 - coutput[0][i], coutput[3][i]);
+                  out[1] = 255;
+                  out += n;
+               }
+            } else {
+               stbi_uc *y = coutput[0];
+               if (n == 1)
+                  for (i=0; i < z->s->img_x; ++i) out[i] = y[i];
+               else
+                  for (i=0; i < z->s->img_x; ++i) *out++ = y[i], *out++ = 255;
+            }
          }
       }
       stbi__cleanup_jpeg(z);
       *out_x = z->s->img_x;
       *out_y = z->s->img_y;
-      if (comp) *comp  = z->s->img_n; // report original components, not output
+      if (comp) *comp = z->s->img_n >= 3 ? 3 : 1; // report original components, not output
       return output;
    }
 }
 
-static unsigned char *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri)
 {
    unsigned char* result;
    stbi__jpeg* j = (stbi__jpeg*) stbi__malloc(sizeof(stbi__jpeg));
+   STBI_NOTUSED(ri);
    j->s = s;
    stbi__setup_jpeg(j);
    result = load_jpeg_image(j, x,y,comp,req_comp);
@@ -3520,11 +3695,12 @@ static unsigned char *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *com
 static int stbi__jpeg_test(stbi__context *s)
 {
    int r;
-   stbi__jpeg j;
-   j.s = s;
-   stbi__setup_jpeg(&j);
-   r = stbi__decode_jpeg_header(&j, STBI__SCAN_type);
+   stbi__jpeg* j = (stbi__jpeg*)stbi__malloc(sizeof(stbi__jpeg));
+   j->s = s;
+   stbi__setup_jpeg(j);
+   r = stbi__decode_jpeg_header(j, STBI__SCAN_type);
    stbi__rewind(s);
+   STBI_FREE(j);
    return r;
 }
 
@@ -3536,7 +3712,7 @@ static int stbi__jpeg_info_raw(stbi__jpeg *j, int *x, int *y, int *comp)
    }
    if (x) *x = j->s->img_x;
    if (y) *y = j->s->img_y;
-   if (comp) *comp = j->s->img_n;
+   if (comp) *comp = j->s->img_n >= 3 ? 3 : 1;
    return 1;
 }
 
@@ -3593,7 +3769,7 @@ stbi_inline static int stbi__bit_reverse(int v, int bits)
    return stbi__bitreverse16(v) >> (16-bits);
 }
 
-static int stbi__zbuild_huffman(stbi__zhuffman *z, stbi_uc *sizelist, int num)
+static int stbi__zbuild_huffman(stbi__zhuffman *z, const stbi_uc *sizelist, int num)
 {
    int i,k=0;
    int code, next_code[16], sizes[17];
@@ -3883,9 +4059,24 @@ static int stbi__parse_zlib_header(stbi__zbuf *a)
    return 1;
 }
 
-// @TODO: should statically initialize these for optimal thread safety
-static stbi_uc stbi__zdefault_length[288], stbi__zdefault_distance[32];
-static void stbi__init_zdefaults(void)
+static const stbi_uc stbi__zdefault_length[288] =
+{
+   8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,
+   8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,
+   8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,
+   8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,
+   8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,
+   9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,
+   9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,
+   9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,
+   7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, 7,7,7,7,7,7,7,7,8,8,8,8,8,8,8,8
+};
+static const stbi_uc stbi__zdefault_distance[32] =
+{
+   5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5
+};
+/*
+Init algorithm:
 {
    int i;   // use <= to match clearly with spec
    for (i=0; i <= 143; ++i)     stbi__zdefault_length[i]   = 8;
@@ -3895,6 +4086,7 @@ static void stbi__init_zdefaults(void)
 
    for (i=0; i <=  31; ++i)     stbi__zdefault_distance[i] = 5;
 }
+*/
 
 static int stbi__parse_zlib(stbi__zbuf *a, int parse_header)
 {
@@ -3913,7 +4105,6 @@ static int stbi__parse_zlib(stbi__zbuf *a, int parse_header)
       } else {
          if (type == 1) {
             // use fixed code lengths
-            if (!stbi__zdefault_distance[31]) stbi__init_zdefaults();
             if (!stbi__zbuild_huffman(&a->z_length  , stbi__zdefault_length  , 288)) return 0;
             if (!stbi__zbuild_huffman(&a->z_distance, stbi__zdefault_distance,  32)) return 0;
          } else {
@@ -4106,15 +4297,14 @@ static int stbi__create_png_image_raw(stbi__png *a, stbi_uc *raw, stbi__uint32 r
 
    img_width_bytes = (((img_n * x * depth) + 7) >> 3);
    img_len = (img_width_bytes + 1) * y;
-   if (s->img_x == x && s->img_y == y) {
-      if (raw_len != img_len) return stbi__err("not enough pixels","Corrupt PNG");
-   } else { // interlaced:
-      if (raw_len < img_len) return stbi__err("not enough pixels","Corrupt PNG");
-   }
+   // we used to check for exact match between raw_len and img_len on non-interlaced PNGs,
+   // but issue #276 reported a PNG in the wild that had extra data at the end (all zeros),
+   // so just check for raw_len < img_len always.
+   if (raw_len < img_len) return stbi__err("not enough pixels","Corrupt PNG");
 
    for (j=0; j < y; ++j) {
       stbi_uc *cur = a->out + stride*j;
-      stbi_uc *prior = cur - stride;
+      stbi_uc *prior;
       int filter = *raw++;
 
       if (filter > 4)
@@ -4126,6 +4316,7 @@ static int stbi__create_png_image_raw(stbi__png *a, stbi_uc *raw, stbi__uint32 r
          filter_bytes = 1;
          width = img_width_bytes;
       }
+      prior = cur - stride; // bugfix: need to compute this after 'cur +=' computation above
 
       // if first row, use special filter that doesn't sample previous row
       if (j == 0) filter = first_row_filter[filter];
@@ -4166,37 +4357,37 @@ static int stbi__create_png_image_raw(stbi__png *a, stbi_uc *raw, stbi__uint32 r
       // this is a little gross, so that we don't switch per-pixel or per-component
       if (depth < 8 || img_n == out_n) {
          int nk = (width - 1)*filter_bytes;
-         #define CASE(f) \
+         #define STBI__CASE(f) \
              case f:     \
                 for (k=0; k < nk; ++k)
          switch (filter) {
             // "none" filter turns into a memcpy here; make that explicit.
             case STBI__F_none:         memcpy(cur, raw, nk); break;
-            CASE(STBI__F_sub)          cur[k] = STBI__BYTECAST(raw[k] + cur[k-filter_bytes]); break;
-            CASE(STBI__F_up)           cur[k] = STBI__BYTECAST(raw[k] + prior[k]); break;
-            CASE(STBI__F_avg)          cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k-filter_bytes])>>1)); break;
-            CASE(STBI__F_paeth)        cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],prior[k],prior[k-filter_bytes])); break;
-            CASE(STBI__F_avg_first)    cur[k] = STBI__BYTECAST(raw[k] + (cur[k-filter_bytes] >> 1)); break;
-            CASE(STBI__F_paeth_first)  cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],0,0)); break;
+            STBI__CASE(STBI__F_sub)          { cur[k] = STBI__BYTECAST(raw[k] + cur[k-filter_bytes]); } break;
+            STBI__CASE(STBI__F_up)           { cur[k] = STBI__BYTECAST(raw[k] + prior[k]); } break;
+            STBI__CASE(STBI__F_avg)          { cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k-filter_bytes])>>1)); } break;
+            STBI__CASE(STBI__F_paeth)        { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],prior[k],prior[k-filter_bytes])); } break;
+            STBI__CASE(STBI__F_avg_first)    { cur[k] = STBI__BYTECAST(raw[k] + (cur[k-filter_bytes] >> 1)); } break;
+            STBI__CASE(STBI__F_paeth_first)  { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],0,0)); } break;
          }
-         #undef CASE
+         #undef STBI__CASE
          raw += nk;
       } else {
          STBI_ASSERT(img_n+1 == out_n);
-         #define CASE(f) \
+         #define STBI__CASE(f) \
              case f:     \
                 for (i=x-1; i >= 1; --i, cur[filter_bytes]=255,raw+=filter_bytes,cur+=output_bytes,prior+=output_bytes) \
                    for (k=0; k < filter_bytes; ++k)
          switch (filter) {
-            CASE(STBI__F_none)         cur[k] = raw[k]; break;
-            CASE(STBI__F_sub)          cur[k] = STBI__BYTECAST(raw[k] + cur[k- output_bytes]); break;
-            CASE(STBI__F_up)           cur[k] = STBI__BYTECAST(raw[k] + prior[k]); break;
-            CASE(STBI__F_avg)          cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k- output_bytes])>>1)); break;
-            CASE(STBI__F_paeth)        cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],prior[k],prior[k- output_bytes])); break;
-            CASE(STBI__F_avg_first)    cur[k] = STBI__BYTECAST(raw[k] + (cur[k- output_bytes] >> 1)); break;
-            CASE(STBI__F_paeth_first)  cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],0,0)); break;
+            STBI__CASE(STBI__F_none)         { cur[k] = raw[k]; } break;
+            STBI__CASE(STBI__F_sub)          { cur[k] = STBI__BYTECAST(raw[k] + cur[k- output_bytes]); } break;
+            STBI__CASE(STBI__F_up)           { cur[k] = STBI__BYTECAST(raw[k] + prior[k]); } break;
+            STBI__CASE(STBI__F_avg)          { cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k- output_bytes])>>1)); } break;
+            STBI__CASE(STBI__F_paeth)        { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],prior[k],prior[k- output_bytes])); } break;
+            STBI__CASE(STBI__F_avg_first)    { cur[k] = STBI__BYTECAST(raw[k] + (cur[k- output_bytes] >> 1)); } break;
+            STBI__CASE(STBI__F_paeth_first)  { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],0,0)); } break;
          }
-         #undef CASE
+         #undef STBI__CASE
 
          // the loop above sets the high byte of the pixels' alpha, but for
          // 16 bit png files we also need the low byte set. we'll do that here.
@@ -4428,27 +4619,6 @@ static int stbi__expand_png_palette(stbi__png *a, stbi_uc *palette, int len, int
    return 1;
 }
 
-static int stbi__reduce_png(stbi__png *p)
-{
-   int i;
-   int img_len = p->s->img_x * p->s->img_y * p->s->img_out_n;
-   stbi_uc *reduced;
-   stbi__uint16 *orig = (stbi__uint16*)p->out;
-
-   if (p->depth != 16) return 1; // don't need to do anything if not 16-bit data
-
-   reduced = (stbi_uc *)stbi__malloc(img_len);
-   if (reduced == NULL) return stbi__err("outofmem", "Out of memory");
-
-   for (i = 0; i < img_len; ++i)
-      reduced[i] = (stbi_uc)((orig[i] >> 8) & 0xFF); // top half of each byte is a decent approx of 16->8 bit scaling
-
-   p->out = reduced;
-   STBI_FREE(orig);
-
-   return 1;
-}
-
 static int stbi__unpremultiply_on_load = 0;
 static int stbi__de_iphone_flag = 0;
 
@@ -4483,9 +4653,10 @@ static void stbi__de_iphone(stbi__png *z)
             stbi_uc a = p[3];
             stbi_uc t = p[0];
             if (a) {
-               p[0] = p[2] * 255 / a;
-               p[1] = p[1] * 255 / a;
-               p[2] =  t   * 255 / a;
+               stbi_uc half = a / 2;
+               p[0] = (p[2] * 255 + half) / a;
+               p[1] = (p[1] * 255 + half) / a;
+               p[2] = ( t   * 255 + half) / a;
             } else {
                p[0] = p[2];
                p[2] = t;
@@ -4539,7 +4710,7 @@ static int stbi__parse_png_file(stbi__png *z, int scan, int req_comp)
             s->img_y = stbi__get32be(s); if (s->img_y > (1 << 24)) return stbi__err("too large","Very large image (corrupt?)");
             z->depth = stbi__get8(s);  if (z->depth != 1 && z->depth != 2 && z->depth != 4 && z->depth != 8 && z->depth != 16)  return stbi__err("1/2/4/8/16-bit only","PNG not supported: 1/2/4/8/16-bit only");
             color = stbi__get8(s);  if (color > 6)         return stbi__err("bad ctype","Corrupt PNG");
-			if (color == 3 && z->depth == 16)                  return stbi__err("bad ctype","Corrupt PNG");
+            if (color == 3 && z->depth == 16)                  return stbi__err("bad ctype","Corrupt PNG");
             if (color == 3) pal_img_n = 3; else if (color & 1) return stbi__err("bad ctype","Corrupt PNG");
             comp  = stbi__get8(s);  if (comp) return stbi__err("bad comp method","Corrupt PNG");
             filter= stbi__get8(s);  if (filter) return stbi__err("bad filter method","Corrupt PNG");
@@ -4588,7 +4759,7 @@ static int stbi__parse_png_file(stbi__png *z, int scan, int req_comp)
                if (c.length != (stbi__uint32) s->img_n*2) return stbi__err("bad tRNS len","Corrupt PNG");
                has_trans = 1;
                if (z->depth == 16) {
-                  for (k = 0; k < s->img_n; ++k) tc16[k] = stbi__get16be(s); // copy the values as-is
+                  for (k = 0; k < s->img_n; ++k) tc16[k] = (stbi__uint16)stbi__get16be(s); // copy the values as-is
                } else {
                   for (k = 0; k < s->img_n; ++k) tc[k] = (stbi_uc)(stbi__get16be(s) & 255) * stbi__depth_scale_table[z->depth]; // non 8-bit images will be larger
                }
@@ -4648,6 +4819,9 @@ static int stbi__parse_png_file(stbi__png *z, int scan, int req_comp)
                if (req_comp >= 3) s->img_out_n = req_comp;
                if (!stbi__expand_png_palette(z, palette, pal_len, s->img_out_n))
                   return 0;
+            } else if (has_trans) {
+               // non-paletted image with tRNS -> source image has (constant) alpha
+               ++s->img_n;
             }
             STBI_FREE(z->expanded); z->expanded = NULL;
             return 1;
@@ -4675,20 +4849,22 @@ static int stbi__parse_png_file(stbi__png *z, int scan, int req_comp)
    }
 }
 
-static unsigned char *stbi__do_png(stbi__png *p, int *x, int *y, int *n, int req_comp)
+static void *stbi__do_png(stbi__png *p, int *x, int *y, int *n, int req_comp, stbi__result_info *ri)
 {
-   unsigned char *result=NULL;
+   void *result=NULL;
    if (req_comp < 0 || req_comp > 4) return stbi__errpuc("bad req_comp", "Internal error");
    if (stbi__parse_png_file(p, STBI__SCAN_load, req_comp)) {
-      if (p->depth == 16) {
-         if (!stbi__reduce_png(p)) {
-            return result;
-         }
-      }
+      if (p->depth < 8)
+         ri->bits_per_channel = 8;
+      else
+         ri->bits_per_channel = p->depth;
       result = p->out;
       p->out = NULL;
       if (req_comp && req_comp != p->s->img_out_n) {
-         result = stbi__convert_format(result, p->s->img_out_n, req_comp, p->s->img_x, p->s->img_y);
+         if (ri->bits_per_channel == 8)
+            result = stbi__convert_format((unsigned char *) result, p->s->img_out_n, req_comp, p->s->img_x, p->s->img_y);
+         else
+            result = stbi__convert_format16((stbi__uint16 *) result, p->s->img_out_n, req_comp, p->s->img_x, p->s->img_y);
          p->s->img_out_n = req_comp;
          if (result == NULL) return result;
       }
@@ -4703,11 +4879,11 @@ static unsigned char *stbi__do_png(stbi__png *p, int *x, int *y, int *n, int req
    return result;
 }
 
-static unsigned char *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri)
 {
    stbi__png p;
    p.s = s;
-   return stbi__do_png(&p, x,y,comp,req_comp);
+   return stbi__do_png(&p, x,y,comp,req_comp, ri);
 }
 
 static int stbi__png_test(stbi__context *s)
@@ -4789,19 +4965,22 @@ static int stbi__bitcount(unsigned int a)
 
 static int stbi__shiftsigned(int v, int shift, int bits)
 {
-   int result;
-   int z=0;
-
-   if (shift < 0) v <<= -shift;
-   else v >>= shift;
-   result = v;
-
-   z = bits;
-   while (z < 8) {
-      result += v >> z;
-      z += bits;
-   }
-   return result;
+    static unsigned int mul_table[9] = {
+        0,
+        0xff/*0b11111111*/, 0x55/*0b01010101*/, 0x49/*0b01001001*/, 0x11/*0b00010001*/,
+        0x21/*0b00100001*/, 0x41/*0b01000001*/, 0x81/*0b10000001*/, 0x01/*0b00000001*/,
+    };
+    static unsigned int shift_table[9] = {
+        0, 0,0,1,0,2,4,6,0,
+    };
+    if (shift < 0)
+        v <<= -shift;
+    else
+        v >>= shift;
+    assert(v >= 0 && v < 256);
+    v >>= (8-bits);
+    assert(bits >= 0 && bits <= 8);
+    return (int) ((unsigned) v * mul_table[bits]) >> shift_table[bits];
 }
 
 typedef struct
@@ -4820,7 +4999,7 @@ static void *stbi__bmp_parse_header(stbi__context *s, stbi__bmp_data *info)
    info->offset = stbi__get32le(s);
    info->hsz = hsz = stbi__get32le(s);
    info->mr = info->mg = info->mb = info->ma = 0;
-   
+
    if (hsz != 12 && hsz != 40 && hsz != 56 && hsz != 108 && hsz != 124) return stbi__errpuc("unknown BMP", "BMP type not supported: unknown");
    if (hsz == 12) {
       s->img_x = stbi__get16le(s);
@@ -4895,7 +5074,7 @@ static void *stbi__bmp_parse_header(stbi__context *s, stbi__bmp_data *info)
 }
 
 
-static stbi_uc *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri)
 {
    stbi_uc *out;
    unsigned int mr=0,mg=0,mb=0,ma=0, all_a;
@@ -4903,8 +5082,9 @@ static stbi_uc *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int
    int psize=0,i,j,width;
    int flip_vertically, pad, target;
    stbi__bmp_data info;
+   STBI_NOTUSED(ri);
 
-   info.all_a = 255;   
+   info.all_a = 255;
    if (stbi__bmp_parse_header(s, &info) == NULL)
       return NULL; // error code already set
 
@@ -5023,7 +5203,7 @@ static stbi_uc *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int
          stbi__skip(s, pad);
       }
    }
-   
+
    // if alpha channel is all 0s, replace with all 255s
    if (target == 4 && all_a == 0)
       for (i=4*s->img_x*s->img_y-1; i >= 0; i -= 4)
@@ -5169,18 +5349,18 @@ errorEnd:
 }
 
 // read 16bit value and convert to 24bit RGB
-void stbi__tga_read_rgb16(stbi__context *s, stbi_uc* out)
+static void stbi__tga_read_rgb16(stbi__context *s, stbi_uc* out)
 {
-   stbi__uint16 px = stbi__get16le(s);
+   stbi__uint16 px = (stbi__uint16)stbi__get16le(s);
    stbi__uint16 fiveBitMask = 31;
    // we have 3 channels with 5bits each
    int r = (px >> 10) & fiveBitMask;
    int g = (px >> 5) & fiveBitMask;
    int b = px & fiveBitMask;
    // Note that this saves the data in RGB(A) order, so it doesn't need to be swapped later
-   out[0] = (r * 255)/31;
-   out[1] = (g * 255)/31;
-   out[2] = (b * 255)/31;
+   out[0] = (stbi_uc)((r * 255)/31);
+   out[1] = (stbi_uc)((g * 255)/31);
+   out[2] = (stbi_uc)((b * 255)/31);
 
    // some people claim that the most significant bit might be used for alpha
    // (possibly if an alpha-bit is set in the "image descriptor byte")
@@ -5188,7 +5368,7 @@ void stbi__tga_read_rgb16(stbi__context *s, stbi_uc* out)
    // so let's treat all 15 and 16bit TGAs as RGB with no alpha.
 }
 
-static stbi_uc *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri)
 {
    //   read in the TGA header stuff
    int tga_offset = stbi__get8(s);
@@ -5210,10 +5390,11 @@ static stbi_uc *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int
    unsigned char *tga_data;
    unsigned char *tga_palette = NULL;
    int i, j;
-   unsigned char raw_data[4];
+   unsigned char raw_data[4] = {0};
    int RLE_count = 0;
    int RLE_repeating = 0;
    int read_next_pixel = 1;
+   STBI_NOTUSED(ri);
 
    //   do a tiny bit of precessing
    if ( tga_image_type >= 8 )
@@ -5431,7 +5612,7 @@ static int stbi__psd_decode_rle(stbi__context *s, stbi_uc *p, int pixelCount)
    return 1;
 }
 
-static stbi_uc *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc)
 {
    int pixelCount;
    int channelCount, compression;
@@ -5439,6 +5620,7 @@ static stbi_uc *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int
    int bitdepth;
    int w,h;
    stbi_uc *out;
+   STBI_NOTUSED(ri);
 
    // Check identifier
    if (stbi__get32be(s) != 0x38425053)   // "8BPS"
@@ -5500,7 +5682,13 @@ static stbi_uc *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int
       return stbi__errpuc("too large", "Corrupt PSD");
 
    // Create the destination image.
-   out = (stbi_uc *) stbi__malloc_mad3(4, w, h, 0);
+
+   if (!compression && bitdepth == 16 && bpc == 16) {
+      out = (stbi_uc *) stbi__malloc_mad3(8, w, h, 0);
+      ri->bits_per_channel = 16;
+   } else
+      out = (stbi_uc *) stbi__malloc(4 * w*h);
+
    if (!out) return stbi__errpuc("outofmem", "Out of memory");
    pixelCount = w*h;
 
@@ -5541,48 +5729,77 @@ static stbi_uc *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int
 
    } else {
       // We're at the raw image data.  It's each channel in order (Red, Green, Blue, Alpha, ...)
-      // where each channel consists of an 8-bit value for each pixel in the image.
+      // where each channel consists of an 8-bit (or 16-bit) value for each pixel in the image.
 
       // Read the data by channel.
       for (channel = 0; channel < 4; channel++) {
-         stbi_uc *p;
-
-         p = out + channel;
          if (channel >= channelCount) {
             // Fill this channel with default data.
-            stbi_uc val = channel == 3 ? 255 : 0;
-            for (i = 0; i < pixelCount; i++, p += 4)
-               *p = val;
-         } else {
-            // Read the data.
-            if (bitdepth == 16) {
-               for (i = 0; i < pixelCount; i++, p += 4)
-                  *p = (stbi_uc) (stbi__get16be(s) >> 8);
+            if (bitdepth == 16 && bpc == 16) {
+               stbi__uint16 *q = ((stbi__uint16 *) out) + channel;
+               stbi__uint16 val = channel == 3 ? 65535 : 0;
+               for (i = 0; i < pixelCount; i++, q += 4)
+                  *q = val;
             } else {
+               stbi_uc *p = out+channel;
+               stbi_uc val = channel == 3 ? 255 : 0;
                for (i = 0; i < pixelCount; i++, p += 4)
-                  *p = stbi__get8(s);
+                  *p = val;
+            }
+         } else {
+            if (ri->bits_per_channel == 16) {    // output bpc
+               stbi__uint16 *q = ((stbi__uint16 *) out) + channel;
+               for (i = 0; i < pixelCount; i++, q += 4)
+                  *q = (stbi__uint16) stbi__get16be(s);
+            } else {
+               stbi_uc *p = out+channel;
+               if (bitdepth == 16) {  // input bpc
+                  for (i = 0; i < pixelCount; i++, p += 4)
+                     *p = (stbi_uc) (stbi__get16be(s) >> 8);
+               } else {
+                  for (i = 0; i < pixelCount; i++, p += 4)
+                     *p = stbi__get8(s);
+               }
             }
          }
       }
    }
 
+   // remove weird white matte from PSD
    if (channelCount >= 4) {
-      for (i=0; i < w*h; ++i) {
-         unsigned char *pixel = out + 4*i;
-         if (pixel[3] != 0 && pixel[3] != 255) {
-            // remove weird white matte from PSD
-            float a = pixel[3] / 255.0f;
-            float ra = 1.0f / a;
-            float inv_a = 255.0f * (1 - ra);
-            pixel[0] = (unsigned char) (pixel[0]*ra + inv_a);
-            pixel[1] = (unsigned char) (pixel[1]*ra + inv_a);
-            pixel[2] = (unsigned char) (pixel[2]*ra + inv_a);
+      if (ri->bits_per_channel == 16) {
+         for (i=0; i < w*h; ++i) {
+            stbi__uint16 *pixel = (stbi__uint16 *) out + 4*i;
+            if (pixel[3] != 0 && pixel[3] != 65535) {
+               float a = pixel[3] / 65535.0f;
+               float ra = 1.0f / a;
+               float inv_a = 65535.0f * (1 - ra);
+               pixel[0] = (stbi__uint16) (pixel[0]*ra + inv_a);
+               pixel[1] = (stbi__uint16) (pixel[1]*ra + inv_a);
+               pixel[2] = (stbi__uint16) (pixel[2]*ra + inv_a);
+            }
+         }
+      } else {
+         for (i=0; i < w*h; ++i) {
+            unsigned char *pixel = out + 4*i;
+            if (pixel[3] != 0 && pixel[3] != 255) {
+               float a = pixel[3] / 255.0f;
+               float ra = 1.0f / a;
+               float inv_a = 255.0f * (1 - ra);
+               pixel[0] = (unsigned char) (pixel[0]*ra + inv_a);
+               pixel[1] = (unsigned char) (pixel[1]*ra + inv_a);
+               pixel[2] = (unsigned char) (pixel[2]*ra + inv_a);
+            }
          }
       }
    }
 
+   // convert to desired output format
    if (req_comp && req_comp != 4) {
-      out = stbi__convert_format(out, 4, req_comp, w, h);
+      if (ri->bits_per_channel == 16)
+         out = (stbi_uc *) stbi__convert_format16((stbi__uint16 *) out, 4, req_comp, w, h);
+      else
+         out = stbi__convert_format(out, 4, req_comp, w, h);
       if (out == NULL) return out; // stbi__convert_format frees input on failure
    }
 
@@ -5766,10 +5983,13 @@ static stbi_uc *stbi__pic_load_core(stbi__context *s,int width,int height,int *c
    return result;
 }
 
-static stbi_uc *stbi__pic_load(stbi__context *s,int *px,int *py,int *comp,int req_comp)
+static void *stbi__pic_load(stbi__context *s,int *px,int *py,int *comp,int req_comp, stbi__result_info *ri)
 {
    stbi_uc *result;
-   int i, x,y;
+   int i, x,y, internal_comp;
+   STBI_NOTUSED(ri);
+
+   if (!comp) comp = &internal_comp;
 
    for (i=0; i<92; ++i)
       stbi__get8(s);
@@ -6154,11 +6374,12 @@ static stbi_uc *stbi__gif_load_next(stbi__context *s, stbi__gif *g, int *comp, i
    STBI_NOTUSED(req_comp);
 }
 
-static stbi_uc *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri)
 {
    stbi_uc *u = 0;
    stbi__gif* g = (stbi__gif*) stbi__malloc(sizeof(stbi__gif));
    memset(g, 0, sizeof(*g));
+   STBI_NOTUSED(ri);
 
    u = stbi__gif_load_next(s, g, comp, req_comp);
    if (u == (stbi_uc *) s) u = 0;  // end of animated gif marker
@@ -6184,20 +6405,24 @@ static int stbi__gif_info(stbi__context *s, int *x, int *y, int *comp)
 // Radiance RGBE HDR loader
 // originally by Nicolas Schulz
 #ifndef STBI_NO_HDR
-static int stbi__hdr_test_core(stbi__context *s)
+static int stbi__hdr_test_core(stbi__context *s, const char *signature)
 {
-   const char *signature = "#?RADIANCE\n";
    int i;
    for (i=0; signature[i]; ++i)
       if (stbi__get8(s) != signature[i])
-         return 0;
+          return 0;
+   stbi__rewind(s);
    return 1;
 }
 
 static int stbi__hdr_test(stbi__context* s)
 {
-   int r = stbi__hdr_test_core(s);
+   int r = stbi__hdr_test_core(s, "#?RADIANCE\n");
    stbi__rewind(s);
+   if(!r) {
+       r = stbi__hdr_test_core(s, "#?RGBE\n");
+       stbi__rewind(s);
+   }
    return r;
 }
 
@@ -6251,7 +6476,7 @@ static void stbi__hdr_convert(float *output, stbi_uc *input, int req_comp)
    }
 }
 
-static float *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static float *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri)
 {
    char buffer[STBI__HDR_BUFLEN];
    char *token;
@@ -6262,10 +6487,12 @@ static float *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int re
    int len;
    unsigned char count, value;
    int i, j, k, c1,c2, z;
-
+   const char *headerToken;
+   STBI_NOTUSED(ri);
 
    // Check identifier
-   if (strcmp(stbi__hdr_gettoken(s,buffer), "#?RADIANCE") != 0)
+   headerToken = stbi__hdr_gettoken(s,buffer);
+   if (strcmp(headerToken, "#?RADIANCE") != 0 && strcmp(headerToken, "#?RGBE") != 0)
       return stbi__errpf("not HDR", "Corrupt HDR image");
 
    // Parse header
@@ -6382,6 +6609,11 @@ static int stbi__hdr_info(stbi__context *s, int *x, int *y, int *comp)
    char buffer[STBI__HDR_BUFLEN];
    char *token;
    int valid = 0;
+   int dummy;
+
+   if (!x) x = &dummy;
+   if (!y) y = &dummy;
+   if (!comp) comp = &dummy;
 
    if (stbi__hdr_test(s) == 0) {
        stbi__rewind( s );
@@ -6423,14 +6655,14 @@ static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp)
    void *p;
    stbi__bmp_data info;
 
-   info.all_a = 255;   
+   info.all_a = 255;
    p = stbi__bmp_parse_header(s, &info);
    stbi__rewind( s );
    if (p == NULL)
       return 0;
-   *x = s->img_x;
-   *y = s->img_y;
-   *comp = info.ma ? 4 : 3;
+   if (x) *x = s->img_x;
+   if (y) *y = s->img_y;
+   if (comp) *comp = info.ma ? 4 : 3;
    return 1;
 }
 #endif
@@ -6438,7 +6670,10 @@ static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp)
 #ifndef STBI_NO_PSD
 static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp)
 {
-   int channelCount;
+   int channelCount, dummy;
+   if (!x) x = &dummy;
+   if (!y) y = &dummy;
+   if (!comp) comp = &dummy;
    if (stbi__get32be(s) != 0x38425053) {
        stbi__rewind( s );
        return 0;
@@ -6471,9 +6706,13 @@ static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp)
 #ifndef STBI_NO_PIC
 static int stbi__pic_info(stbi__context *s, int *x, int *y, int *comp)
 {
-   int act_comp=0,num_packets=0,chained;
+   int act_comp=0,num_packets=0,chained,dummy;
    stbi__pic_packet packets[10];
 
+   if (!x) x = &dummy;
+   if (!y) y = &dummy;
+   if (!comp) comp = &dummy;
+
    if (!stbi__pic_is4(s,"\x53\x80\xF6\x34")) {
       stbi__rewind(s);
       return 0;
@@ -6549,14 +6788,17 @@ static int      stbi__pnm_test(stbi__context *s)
    return 1;
 }
 
-static stbi_uc *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp)
+static void *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri)
 {
    stbi_uc *out;
+   STBI_NOTUSED(ri);
+
    if (!stbi__pnm_info(s, (int *)&s->img_x, (int *)&s->img_y, (int *)&s->img_n))
       return 0;
+
    *x = s->img_x;
    *y = s->img_y;
-   *comp = s->img_n;
+   if (comp) *comp = s->img_n;
 
    if (!stbi__mad3sizes_valid(s->img_n, s->img_x, s->img_y, 0))
       return stbi__errpuc("too large", "PNM too large");
@@ -6610,16 +6852,20 @@ static int      stbi__pnm_getinteger(stbi__context *s, char *c)
 
 static int      stbi__pnm_info(stbi__context *s, int *x, int *y, int *comp)
 {
-   int maxv;
+   int maxv, dummy;
    char c, p, t;
 
-   stbi__rewind( s );
+   if (!x) x = &dummy;
+   if (!y) y = &dummy;
+   if (!comp) comp = &dummy;
+
+   stbi__rewind(s);
 
    // Get identifier
    p = (char) stbi__get8(s);
    t = (char) stbi__get8(s);
    if (p != 'P' || (t != '5' && t != '6')) {
-       stbi__rewind( s );
+       stbi__rewind(s);
        return 0;
    }
 
@@ -6726,6 +6972,19 @@ STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *c, void *user, int
 
 /*
    revision history:
+      2.16  (2017-07-23) all functions have 16-bit variants;
+                         STBI_NO_STDIO works again;
+                         compilation fixes;
+                         fix rounding in unpremultiply;
+                         optimize vertical flip;
+                         disable raw_len validation;
+                         documentation fixes
+      2.15  (2017-03-18) fix png-1,2,4 bug; now all Imagenet JPGs decode;
+                         warning fixes; disable run-time SSE detection on gcc;
+                         uniform handling of optional "return" values;
+                         thread-safe initialization of zlib tables
+      2.14  (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs
+      2.13  (2016-11-29) add 16-bit API, only supported for PNG right now
       2.12  (2016-04-02) fix typo in 2.11 PSD fix that caused crashes
       2.11  (2016-04-02) allocate large structures on the stack
                          remove white matting for transparent PSD
@@ -6886,3 +7145,46 @@ STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *c, void *user, int
       0.50  (2006-11-19)
               first released version
 */
+
+
+/*
+------------------------------------------------------------------------------
+This software is available under 2 licenses -- choose whichever you prefer.
+------------------------------------------------------------------------------
+ALTERNATIVE A - MIT License
+Copyright (c) 2017 Sean Barrett
+Permission is hereby granted, free of charge, to any person obtaining a copy of
+this software and associated documentation files (the "Software"), to deal in
+the Software without restriction, including without limitation the rights to
+use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
+of the Software, and to permit persons to whom the Software is furnished to do
+so, subject to the following conditions:
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+------------------------------------------------------------------------------
+ALTERNATIVE B - Public Domain (www.unlicense.org)
+This is free and unencumbered software released into the public domain.
+Anyone is free to copy, modify, publish, use, compile, sell, or distribute this
+software, either in source code form or as a compiled binary, for any purpose,
+commercial or non-commercial, and by any means.
+In jurisdictions that recognize copyright laws, the author or authors of this
+software dedicate any and all copyright interest in the software to the public
+domain. We make this dedication for the benefit of the public at large and to
+the detriment of our heirs and successors. We intend this dedication to be an
+overt act of relinquishment in perpetuity of all present and future rights to
+this software under copyright law.
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+------------------------------------------------------------------------------
+*/