Skip to content

Commit 1983077

Browse files
authored
Minor tweak to prores-ks removing the suggestion to use qscale. (#107)
Signed-off-by: [email protected] <[email protected]> Signed-off-by: Sam Richards <[email protected]>
1 parent 38195b7 commit 1983077

File tree

1 file changed

+3
-5
lines changed

1 file changed

+3
-5
lines changed

EncodeProres.md

Lines changed: 3 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,7 @@ parent: Codec Comparisons
88
# ProRes <a name="prores"></a>
99
There are four ProRes encoders, Prores, Prores_ks, Prores_aw and prores_videotoolbox, which is a hardware based OSX M1 encoder/decoder (ffmpeg version 5 or higher).
1010

11-
From [https://trac.ffmpeg.org/wiki/Encode/VFX](https://trac.ffmpeg.org/wiki/Encode/VFX) the recommendation is to use Prores_ks with -profile:v 3 and the qscale of 11 unless you are on OSX where we would recommend using prores_videotoolbox.
11+
If you are doing lighting or compositing reviews the recommendation is to use Prores_ks with -profile:v 3. If you are able we would recommend using [OpenAPV](https://academysoftwarefoundation.github.io/EncodingGuidelines/EncodeOpenAPV.html) or prores_videotoolbox over prores_ks, and look at a 444 10-bit or 12-bit encode, since they both give a superior result.
1212

1313
The two encoders we are reviewing are:
1414
* [Prores_ks](#Prores_ks)
@@ -20,8 +20,6 @@ While ProRes does support up to 12 bits of resolution, prores_ks can only encode
2020

2121
Supported pixel formats: yuv422p10le yuv444p10le yuva444p10le
2222

23-
24-
2523
Example encode would look like:
2624

2725
<!---
@@ -49,8 +47,8 @@ ffmpeg -r 24 -start_number 100 -i inputfile.%04d.png \
4947
Options that can be used include:
5048

5149
| --- | --- |
52-
| -profile:v 4444xq | Which ProRes encoding profile to use, see below |
53-
| -qscale:v 3 | between values of 1-6 give a good result, 0 being default, see below for some wedge tests. |
50+
| -profile:v 4444xq | Which prores encoding profile to use, see below |
51+
| -qscale:v 3 | between values of 1-4, 0 being the same as specifying no qscale |
5452
| -vendor apl0 | tricks the codec into believing its from an Apple codec. |
5553
| -alpha_bits 16 | Allows you to specify how many bits to use for the alpha channel (default 16) |
5654

0 commit comments

Comments
 (0)