Recent changes to format_changes

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Sat, 13 Jan 2018 17:09:35 -0000

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Sun, 17 Sep 2017 21:04:48 -0000

--- v14
+++ v15
@@ -64,6 +64,10 @@
 ---------- | ---------- 
 **b64a**             | Utvideo, MagicYUV, VapourSynth, AviSynth+
 **BGRA64**        | MagicYUV, AviSynth+
+**r210**        | MagicYUV, VapourSynth
+**R10k**        | MagicYUV
+**P010**             | VapourSynth, AviSynth+  
+**P016**             | VapourSynth, AviSynth+  
 **v210**              | Utvideo, MagicYUV, CineformHD, VapourSynth, AviSynth+
 **P210**             | MagicYUV, VapourSynth, AviSynth+  
 **YUV422P10**   | MagicYUV, AviSynth+

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Wed, 11 Jan 2017 23:45:25 -0000

--- v13
+++ v14
@@ -62,15 +62,16 @@

 format  | sources
 ---------- | ---------- 
-**b64a**             | Utvideo, MagicYUV, VapourSynth
-**BGRA64**        | MagicYUV
-**v210**              | Utvideo, MagicYUV, CineformHD, VapourSynth
-**P210**             | MagicYUV, VapourSynth  
-**YUV422P10**   | MagicYUV
-**P216**             | VapourSynth  
-**Y416**             | VapourSynth  
+**b64a**             | Utvideo, MagicYUV, VapourSynth, AviSynth+
+**BGRA64**        | MagicYUV, AviSynth+
+**v210**              | Utvideo, MagicYUV, CineformHD, VapourSynth, AviSynth+
+**P210**             | MagicYUV, VapourSynth, AviSynth+  
+**YUV422P10**   | MagicYUV, AviSynth+
+**P216**             | VapourSynth, AviSynth+  
+**Y416**             | VapourSynth, AviSynth+  

-Note that when input source can decode to multiple formats, more precise format is not selected automatically. Check with "Video->Decode format..." dialog:
+Note that when input source can decode to multiple formats, more precise format is selected automatically. 
+The actual format in use is displayed in "Video->Decode format..." dialog at the top:
    [[img src=decode_format.png]]

 Deep color output options

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Sun, 08 Jan 2017 21:13:20 -0000

--- v12
+++ v13
@@ -65,8 +65,10 @@
 **b64a**             | Utvideo, MagicYUV, VapourSynth
 **BGRA64**        | MagicYUV
 **v210**              | Utvideo, MagicYUV, CineformHD, VapourSynth
-**P210**             | MagicYUV  
+**P210**             | MagicYUV, VapourSynth  
 **YUV422P10**   | MagicYUV
+**P216**             | VapourSynth  
+**Y416**             | VapourSynth

 Note that when input source can decode to multiple formats, more precise format is not selected automatically. Check with "Video->Decode format..." dialog:
    [[img src=decode_format.png]]

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Fri, 30 Dec 2016 19:46:33 -0000

--- v11
+++ v12
@@ -68,6 +68,9 @@
 **P210**             | MagicYUV  
 **YUV422P10**   | MagicYUV

+Note that when input source can decode to multiple formats, more precise format is not selected automatically. Check with "Video->Decode format..." dialog:
+   [[img src=decode_format.png]]
+
 Deep color output options
 -------------

@@ -79,6 +82,8 @@
 * Image sequence: TIFF sequence will store RGB64 at full precision (either 64 or 48 bits depending on alpha).
 * External encoder: FFMPEG is easy to send RGB64 to, and it then can encode to any supported format.
   The key is to specify  **-pix_fmt %(pix_fmt)** for input and also  **-pix_fmt desired_format** for output.
+  Example profiles: [ffv1-yuva64, ffv1-rgb14](https://sourceforge.net/p/vdfiltermod/wiki/format_changes/attachment/example_vdprof.txt)
+

 Some known VFW codecs accepting bitmap formats:

@@ -88,62 +93,3 @@
 **BGRA64**        | MagicYUV
 **v210**              | Utvideo, MagicYUV, CineformHD
 **YUV422P10**   | MagicYUV
-
-Example profiles for FFMPEG encoding:
-~~~text
-{
-  "description": "VirtualDub external encoder profile collection",
-  "externalEncoders": {
-    "sets": {
-      "ffv1-yuva64": {
-        "videoEncoder": "ffv1-yuva64",
-        "audioEncoder": "",
-        "multiplexer": "",
-        "description": "",
-        "extension": "mkv",
-        "processPartial": false,
-        "useOutputAsTemp": true
-      },
-      "ffv1-rgb14": {
-        "videoEncoder": "ffv1-rgb14",
-        "audioEncoder": "",
-        "multiplexer": "",
-        "description": "",
-        "extension": "mkv",
-        "processPartial": false,
-        "useOutputAsTemp": true
-      },
-    },
-    "profiles": {
-      "ffv1-rgb14": {
-        "name": "ffv1-rgb14",
-        "program": "E:\\download\\vd\\ffmpeg1\\ffmpeg.exe",
-        "commandArguments": "-f rawvideo -pix_fmt %(pix_fmt) -s %(width)x%(height) -r %(fpsnum)/%(fpsden) -i - -c:v ffv1 -pix_fmt gbrp14 %(tempvideofile)",
-        "outputFilename": "",
-        "type": 0,
-        "pixelFormat": "bgra64le",
-        "inputFormat": 0,
-        "checkReturnCode": true,
-        "logStdout": true,
-        "logStderr": true,
-        "bypassCompression": false,
-        "predeleteOutputFile": true
-      },
-      "ffv1-yuva64": {
-        "name": "ffv1-yuva64",
-        "program": "E:\\download\\vd\\ffmpeg1\\ffmpeg.exe",
-        "commandArguments": "-f rawvideo -pix_fmt %(pix_fmt) -s %(width)x%(height) -r %(fpsnum)/%(fpsden) -i - -c:v ffv1 %(tempvideofile)",
-        "outputFilename": "",
-        "type": 0,
-        "pixelFormat": "bgra64le",
-        "inputFormat": 0,
-        "checkReturnCode": true,
-        "logStdout": true,
-        "logStderr": true,
-        "bypassCompression": false,
-        "predeleteOutputFile": true
-      },
-    }
-  }
-}
-~~~

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Fri, 30 Dec 2016 17:50:42 -0000

--- v10
+++ v11
@@ -1,45 +1,52 @@
 Enable alpha
 ------------

-Can read Alpha with image sequence: tiff, tga, png, bmp.
-Can write Alpha with image sequence: tiff, tga.
-There is difference how Alpha is handled on output:
+Possible inputs with Alpha:

-*  Tiff writes image without alpha at all when there is no alpha
-*  Tga always writes 32bpp image, if there is no alpha it writes completely white alpha
+*  image sequences: tiff, tga, png, bmp (supported internally)
+*  gif, apng
+*  raw rgb32, bgra64
+*  certain VFW codecs (Utvideo, MagicYUV)
+*  all video/image/sequence formats supported by FFMPEG

-I think Tiff way is more valid, maybe I will change Tga someday.
+Possible outputs with Alpha:

-Fill filter can use alpha:
-This was needed to bring at least anything that can display alpha.
+*  image sequences: tiff, tga, png (supported internally)
+*  gif, apng
+*  raw rgb32, bgra64
+*  certain VFW codecs (Utvideo, MagicYUV)
+*  external encoder (for example ffmpeg)
+*  FFV1, Huffyuv encoders

-Every image in sequence can either have or not have alpha.
-The filter like fill, when enabled to use alpha, is supposed to change only areas flagged by alpha or do nothing at all if the frame has no alpha.
+Filters using alpha:
+
+* Fill (internal): when enabled, fills only areas marked with alpha
+* fflayer: uses source alpha for compositing

 Alpha can be displayed in viewports if dx9 display is enabled.
    [[img src=alpha_display.png]]

-Enable RGB64 mode
+New 16-bit processing formats
 ------------

-RGB64 is mode similar to XRGB but with 16 bit per component precision.
+**XRGB64** is format similar to XRGB but with 16 bit per component precision.  
+**YUV420P16**, **YUV422P16**, **YUV444P16** are planar YUV formats similar to 8-bit YUV formats.

 Supported features:

 *  Input and Output in a variety of ways (see below).
 *  Display (downconverted to 8 bit)
-*  Color picker can show fractional components and also real R/G/B/A codes as stored in buffer:
+*  Color picker can show fractional components and also raw R/G/B/A or Y/U/V values as stored in buffer:
    [[img src=picker.png]]

 Supported internal filters:

-*  cropping (also known as null transform)
+*  crop / null transform
 *  flip vertical
 *  flip horizontal
 *  fill
-*  convert format - handles conversions XRGB<->RGB64
-Other conversions work too, but are not optimized now.
-In particular V210->RGB64 and RGB64->V210 conversions are useful but are several times slower than more common UYVY->RGB32. 
+*  resize
+*  convert format

 Not supported now (for example):

@@ -47,27 +54,40 @@

 Deep color input options
 -------------
-*  Caching input driver can read all formats supported by FFMPEG, this can be FFV1 video or DPX sequence or anything else  
-   rgb formats are passed to VD automatically as RGB64 (when bit depth is above 8), actual number of used bits it preserved  
-   yuv formats can be expanded to 16bit rgb if Video->Decode format option is set to RGB64  
+*  Caching input driver can read all formats supported by FFMPEG, this can be FFV1 video or DPX sequence or anything else (most useful formats translate directly to RGB64 or YUV###P16 without precision changes)
 *  Tiff image sequence is also supported internally and can read 16bit RGB/RGBA
-*  internal AVI input driver can read V210 and b64a from uncompressed file or vfw decoder
-*  also internal input driver can read b64a from **VapourSynth**
+*  internal AVI input driver can read some bitmap formats from uncompressed file or vfw decoder (b64a, v210, P210, P216, ...)

+Some known sources by bitmap format:
+
+format  | sources
+---------- | ---------- 
+**b64a**             | Utvideo, MagicYUV, VapourSynth
+**BGRA64**        | MagicYUV
+**v210**              | Utvideo, MagicYUV, CineformHD, VapourSynth
+**P210**             | MagicYUV  
+**YUV422P10**   | MagicYUV

 Deep color output options
 -------------

 * integrated encoders: 
   **x264** supports 10 bit yuv (4:2:0, 4:2:2, 4:4:4) and 10 bit rgb
-  **FFV1** supports rgb from 9 bit to 16 bit  
-  **FFVHuff** supports rgb from 9 bit to 14 bit
-* AVI: output as v210. Many vfw codecs can encode from v210. Specifically I tested it with Cineform 9.0.5 codec.
-  Select output format "v210", also there are options for color space and luma range, all combinations work with Cineform (you also have to change codec options accordingly).
-* AVI: output as RGB64. The output will be presented to codec as b64a. Tested with **Utvideo** 10 bit rgb.
+  **FFV1** supports rgb/yuv from 9 bit to 16 bit  
+  **FFVHuff** supports rgb/yuv from 9 bit to 14 bit
+* AVI: set output format to v210, YUV422P16 or XRGB64.
 * Image sequence: TIFF sequence will store RGB64 at full precision (either 64 or 48 bits depending on alpha).
 * External encoder: FFMPEG is easy to send RGB64 to, and it then can encode to any supported format.
   The key is to specify  **-pix_fmt %(pix_fmt)** for input and also  **-pix_fmt desired_format** for output.
+
+Some known VFW codecs accepting bitmap formats:
+
+format  | codecs
+---------- | ---------- 
+**b64a**             | Utvideo, MagicYUV
+**BGRA64**        | MagicYUV
+**v210**              | Utvideo, MagicYUV, CineformHD
+**YUV422P10**   | MagicYUV

 Example profiles for FFMPEG encoding:
 ~~~text

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Sun, 25 Sep 2016 19:13:17 -0000

--- v9
+++ v10
@@ -26,28 +26,45 @@

 Supported features:

-*  Read and export tiff sequences in 16 bit per channel format.
-*  Caching input driver delivers all rgb formats which are more than 8 bit precision as RGB64.
-Also it is possible to request conversion to RGB64 for all formats (Video->Color depth). In this case FFMPEG (swscale) will be responsible for the conversion quality/performance.
+*  Input and Output in a variety of ways (see below).
 *  Display (downconverted to 8 bit)
 *  Color picker can show fractional components and also real R/G/B/A codes as stored in buffer:
    [[img src=picker.png]]

 Supported internal filters:

-*  null transform
+*  cropping (also known as null transform)
 *  flip vertical
 *  flip horizontal
+*  fill
 *  convert format - handles conversions XRGB<->RGB64
 Other conversions work too, but are not optimized now.
 In particular V210->RGB64 and RGB64->V210 conversions are useful but are several times slower than more common UYVY->RGB32. 

+Not supported now (for example):
+
+* blend curve (only fully opaque or fully transparent)  
+
+Deep color input options
+-------------
+*  Caching input driver can read all formats supported by FFMPEG, this can be FFV1 video or DPX sequence or anything else  
+   rgb formats are passed to VD automatically as RGB64 (when bit depth is above 8), actual number of used bits it preserved  
+   yuv formats can be expanded to 16bit rgb if Video->Decode format option is set to RGB64  
+*  Tiff image sequence is also supported internally and can read 16bit RGB/RGBA
+*  internal AVI input driver can read V210 and b64a from uncompressed file or vfw decoder
+*  also internal input driver can read b64a from **VapourSynth**
+
+
 Deep color output options
 -------------

+* integrated encoders: 
+  **x264** supports 10 bit yuv (4:2:0, 4:2:2, 4:4:4) and 10 bit rgb
+  **FFV1** supports rgb from 9 bit to 16 bit  
+  **FFVHuff** supports rgb from 9 bit to 14 bit
 * AVI: output as v210. Many vfw codecs can encode from v210. Specifically I tested it with Cineform 9.0.5 codec.
   Select output format "v210", also there are options for color space and luma range, all combinations work with Cineform (you also have to change codec options accordingly).
-* AVI: output as RGB64. The output will be presented to codec as b64a. Tested with Utvideo 10 bit rgb.
+* AVI: output as RGB64. The output will be presented to codec as b64a. Tested with **Utvideo** 10 bit rgb.
 * Image sequence: TIFF sequence will store RGB64 at full precision (either 64 or 48 bits depending on alpha).
 * External encoder: FFMPEG is easy to send RGB64 to, and it then can encode to any supported format.
   The key is to specify  **-pix_fmt %(pix_fmt)** for input and also  **-pix_fmt desired_format** for output.

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Mon, 25 Jul 2016 19:47:52 -0000

--- v8
+++ v9
@@ -15,6 +15,9 @@

 Every image in sequence can either have or not have alpha.
 The filter like fill, when enabled to use alpha, is supposed to change only areas flagged by alpha or do nothing at all if the frame has no alpha.
+
+Alpha can be displayed in viewports if dx9 display is enabled.
+   [[img src=alpha_display.png]]

 Enable RGB64 mode
 ------------

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Tue, 31 May 2016 16:36:34 -0000

--- v7
+++ v8
@@ -26,7 +26,7 @@
 *  Read and export tiff sequences in 16 bit per channel format.
 *  Caching input driver delivers all rgb formats which are more than 8 bit precision as RGB64.
 Also it is possible to request conversion to RGB64 for all formats (Video->Color depth). In this case FFMPEG (swscale) will be responsible for the conversion quality/performance.
-*  Display in GDI/DirectDraw mode
+*  Display (downconverted to 8 bit)
 *  Color picker can show fractional components and also real R/G/B/A codes as stored in buffer:
    [[img src=picker.png]]

format_changes modified by Anton Shekhovtsov

Anton Shekhovtsov — Thu, 05 May 2016 09:42:46 -0000

--- v6
+++ v7
@@ -27,6 +27,8 @@
 *  Caching input driver delivers all rgb formats which are more than 8 bit precision as RGB64.
 Also it is possible to request conversion to RGB64 for all formats (Video->Color depth). In this case FFMPEG (swscale) will be responsible for the conversion quality/performance.
 *  Display in GDI/DirectDraw mode
+*  Color picker can show fractional components and also real R/G/B/A codes as stored in buffer:
+   [[img src=picker.png]]

 Supported internal filters: