Re: [FFmpeg-devel] [PATCH 1/4] avcodec/proresdec2: change profile only if it is unknown

2018-12-05 Thread Gregor Riepl
There seems to be an indentation issue here: > +if (avctx->profile == FF_PROFILE_UNKNOWN) { > switch (avctx->codec_tag) { > case MKTAG('a','p','c','o'): and here: > av_log(avctx, AV_LOG_WARNING, "Unknown prores profile %d\n", > avctx->codec_tag); > } > +} ___

Re: [FFmpeg-devel] [PATCH] cmdutils: replace strncpy() with direct assignment

2021-03-09 Thread Gregor Riepl
> // Change all the ' --' strings to '~--' so that > // they can be identified as tokens. > while ((conflist = strstr(str, " --")) != NULL) { > -strncpy(conflist, "~--", 3); > +conflist[0] = '~'; > } Doesn't this simply replace -- with ~- ? The comment seems wro

Re: [FFmpeg-devel] [PATCH 11/11] avformat/wavdecuse FFABS to instead of fabs

2017-11-20 Thread Gregor Riepl
> FFABS is slower than fabs. Not generally. That is highly dependent on compiler/flags/architecture. In some cases, FFABS may actually be faster, in others maybe not. The expected result would be on-par performance. ___ ffmpeg-devel mailing list ffmpeg

[FFmpeg-devel] [PATCH] ffserver: fixed deallocation bug in build_feed_streams

2016-05-26 Thread Gregor Riepl
rom 00c9203f0349dbae6e701104d5a7f6c4e6fa0159 Mon Sep 17 00:00:00 2001 From: Gregor Riepl Date: Tue, 24 May 2016 15:17:22 +0200 Subject: [PATCH] ffserver: fixed deallocation bug in build_feed_streams Signed-off-by: Gregor Riepl --- ffserver.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/ffserver.c b/ffserve

[FFmpeg-devel] [PATCH] avformat/dashdec: Differentiate unassigned and zero attributes

2022-10-07 Thread Gregor Riepl
the issue tracker was incorrect and changed other parts of the segment calculation logic. With this patch, only the interpretation of the attributes is changed. Some warnings are added to account for potential errors in manifests. Fixes: #8522 Signed-off-by: Gregor Riepl --- libavformat

Re: [FFmpeg-devel] [PATCH] avformat/dashdec: Differentiate unassigned and zero attributes

2022-10-17 Thread Gregor Riepl
+#include Not sure this header can be used every enveriment. Or what about use #define TRUE 1 #define FALSE 0 > ? You are right, this is C99. I'd rather not define a custom boolean type just for this, it would be better if FFmpeg had something like FF_TRUE/FF_FALSE. fftools/ffmpeg_opt.c se

[FFmpeg-devel] [PATCH v2] avformat/dashdec: Differentiate unassigned and zero attributes

2022-10-18 Thread Gregor Riepl
what's done in fftools/ffmpeg_opt.c. Fixes: #8522 Signed-off-by: Gregor Riepl --- libavformat/dashdec.c | 209 -- 1 file changed, 161 insertions(+), 48 deletions(-) diff --git a/libavformat/dashdec.c b/libavformat/dashdec.c index 29d4680c68..df5d4

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-22 Thread Gregor Riepl
The result should be no need to provide "crypto://". The ffmpeg file format detection should detect that ".crypto" should be handled by the crypto plugin. Instead of a custom descriptor file format that is only used for this particular special case, you could also define a custom URI, such as:

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-23 Thread Gregor Riepl
I can't find a single thing about this in the ffmpeg documentation. How is this called, where can I read more about it and - most importantly - does it work out of the box? No, it does not. This was a suggestion that sounds like a better option than a custom metadata descriptor file that is onl

Re: [FFmpeg-devel] seeking consultancy around FFmpeg

2021-01-24 Thread Gregor Riepl
> What we are looking for is pretty straight forward. I like to capture > the HDMI video/audio output of a Chromecast dongle and distribute it > into a LAN network via UDP multicast. Something like this: > > Chromecast -> HDMI -> HDMI-USB Capture -> RPi4/8G (with FFmpeg) -> > UDP/Multicast -> LAN