[Ffmpeg-devel-irc] ffmpeg-devel.log.20160715

[burek]

[00:51:19 CEST] <ubitux> https://blogs.dropbox.com/tech/2016/07/lepton-image-compression-saving-22-losslessly-from-images-at-15mbs/ not pasted yet?
[00:51:42 CEST] <jkqxz> nevcairiel:  The code is just wrong - the allocation has to be inside the loop, because the pointers are written to the fifo and freed after reading.  <http://sprunge.us/bKVj> makes it not-entirely-broken, I think?
[02:25:26 CEST] <cone-908> ffmpeg 03Burt P 07master:33fc4cc7c6bd: af_hdcd: fewer false positives by ignoring code_counterC in HDCD detection
[02:25:27 CEST] <cone-908> ffmpeg 03Burt P 07master:5c7457c02856: af_hdcd: only hdcd_update_info() when something changes
[02:25:28 CEST] <cone-908> ffmpeg 03Burt P 07master:c41718c0d0e6: af_hdcd: detect and report encoding errors and oddities
[04:41:07 CEST] <bp0> michaelni, maybe not
[05:50:40 CEST] <bp0> michaelni, I submitted some trimmed versions, now about 2.3MB. Hope that is alright
[07:45:52 CEST] <thardin> I am noticing I have no motivation/time to fix that mxf regression
[07:48:36 CEST] <thardin> do any of the other people I've noticed contributing mxf patches work with the format in their dayjob?
[09:10:04 CEST] <kode54> what the hell happened to ffurl_register_protocol?
[09:10:24 CEST] <kode54> I used to register my own protocol handlers instead of compiling the stock ones in, so I could do my own file reading
[09:11:29 CEST] <nevcairiel> ff* api is privaet and you should have never used it
[09:12:01 CEST] <kode54> I compile and bundle my own ffmpeg anyway
[09:12:02 CEST] <nevcairiel> you can implement your own reading callbacks using the avio api
[09:12:06 CEST] <kode54> oh
[09:12:12 CEST] <kode54> I didn't know there was an API for that
[09:12:13 CEST] <nevcairiel> in a public and proper way
[09:12:20 CEST] <nevcairiel> it gives you read write and seek callbacks
[10:28:41 CEST] <cone-246> ffmpeg 03Carl Eugen Hoyos 07master:040b4e139b94: lavf/img2: Remove ffv1-img.
[10:28:41 CEST] <cone-246> ffmpeg 03Carl Eugen Hoyos 07master:e1023aa1dd0c: lavf/img2dec: Also auto-detect (unusual) uncompressed pcx.
[10:36:39 CEST] <cone-246> ffmpeg 03Clément BSsch 07master:77726d32a872: lavf/vplayerdec: Improve auto-detection.
[10:38:02 CEST] <cone-246> ffmpeg 03Clément BSsch 07release/3.1:7da59005bec1: lavf/vplayerdec: Improve auto-detection.
[10:39:13 CEST] <michaelni> bp0, where did you submit the trimmed versions ? 
[10:39:38 CEST] <michaelni> dont see any so obviously i look at the wrong plae
[10:39:40 CEST] <michaelni> place
[11:50:07 CEST] <marcin__> re all
[11:50:44 CEST] <marcin__> i'm facing this issue (memory leak) -> http://ffmpeg.org/pipermail/libav-user/2016-July/009343.html
[11:51:01 CEST] <marcin__> has anybody similar leaking problems?
[13:08:56 CEST] <Rathann> marcin__: please check trac if there's a ticket open and if not, please open one
[13:30:24 CEST] <marcin__> Rathann: ok, I will check then, thanks for responding 
[14:02:00 CEST] <cone-246> ffmpeg 03Pavel Koshevoy 07master:25ca74d314c3: Restore compatibility with powerpc-apple-darwin9-gcc-4.2.1
[14:57:25 CEST] <bp0> michaelni, I sent them as an email to ffmpeg-devel as a reply to the patch that adds the tests
[14:58:08 CEST] <bp0> I received a response that they are awaiting moderation
[14:59:03 CEST] <nevcairiel> sending samples directly to the ML is not really a good idea, which is probably why the filter flagged it for moderation
[15:01:32 CEST] <bp0> what is the preferred way to submit samples?
[15:16:57 CEST] <michaelni> bp0, upload somewhere and post url, (somewhere can be our ftp if you dont know any place)
[15:36:42 CEST] <bp0> upload.fffmpeg.org?
[15:38:30 CEST] <bp0> no route to host
[15:38:34 CEST] <bp0> I'll figure something out
[15:41:23 CEST] <Compn> you had one too many f's
[15:41:31 CEST] <Compn> bp0
[15:42:27 CEST] <bp0> :) It's true, but not when I actually tried to ftp
[15:43:51 CEST] <funman> bp0: the machine is being replaced
[15:44:42 CEST] <bp0> ah, I see then
[15:59:01 CEST] <michaelni> bp0, ive seen people use www.wetransfer.com recently for sharing some files but you can use anything (services with less bloat & spam preferred)
[16:01:27 CEST] <Dev^Null> I am sure i am in the wrong channel but there is no one in the other ones. I am trying to use ffmpeg to connect to a ELP 1280*720p camera I got the url from Ispy and I keep getting this error "method DESCRIBE failed: 451 ERROR
[16:01:27 CEST] <Dev^Null> rtsp://admin@10.10.10.231/user=admin_password=_channel1=1_stream=1.sdp: Server returned 4XX Client Error, but not one of 40{0,1,3,4}" this is the syntax I am using ffmpeg -i rtsp://admin@10.10.10.231/user=admin_password=_channel1=1_stream=1.sdp -r 15 C:/DB_Videos/2013-04-30 17_18_34.703.mp4 can someone help me figure out what is wrong I am running ubuntu 16.04 and ffmpeg ver 3.1.1
[16:11:18 CEST] <bp0> https://www.dropbox.com/s/fy4r5nsr5ysweg8/hdcd-false-positive.flac?dl=1
[16:11:20 CEST] <bp0> https://www.dropbox.com/s/ifkl0x7mqhiqc4t/hdcd-encoding-errors.flac?dl=1
[16:11:23 CEST] <bp0> works?
[16:33:12 CEST] <bp0> michaelni, are those dropbox links going to work for you?
[16:38:30 CEST] <michaelni> bp0, yes
[16:40:03 CEST] <cone-246> ffmpeg 03Steven Liu 07master:b701ec4152b8: refine the method option describe of hlsenc doc
[16:41:38 CEST] <bp0> ok
[17:20:42 CEST] <michaelni> bp0, hdcd files uploaded to the fate samples
[17:21:40 CEST] <bp0> right on, thank you
[17:22:17 CEST] <bp0> how do you feel about the added compare "grep" added to fate-run.sh?
[17:23:11 CEST] <bp0> if there was another way to look for a result in the log, I didn't find it
[17:26:57 CEST] <nevcairiel> testing log output is not very reliable as we generally do not guarantee it to remain stable
[17:27:20 CEST] <nevcairiel> and its also not a good way for user apps to be notified
[17:27:37 CEST] <nevcairiel> so maybe you identified a flaw and should add proper metadata to show the presence :)
[17:29:16 CEST] <bp0> So for the HDCD filter, perhaps that summary line could be encoded into metadata?
[20:35:43 CEST] <KGB> [13FFV1] 15retokromer opened pull request #21: escaping not required for markdown (06master...06patch-1) 02https://git.io/vKuwJ
[21:57:58 CEST] <cone-124> ffmpeg 03Josh de Kock 07master:d52dd768a322: lavf: add libopenmpt demuxer
[22:43:35 CEST] <ubitux> > NVIDIA 367.35 Released, Supports 8K H.265 Video Decoding
[22:43:40 CEST] <ubitux> what an era to live in...
[22:43:47 CEST] <JEEB> :D
[22:47:57 CEST] <jamrial> ubitux: what gpus? and that's probably hybrid decoding
[22:49:06 CEST] <ubitux> > Added support for VDPAU Feature Set H to the NVIDIA VDPAU driver. GPUs with VDPAU Feature Set H are capable of hardware-accelerated decoding of 8192x8192 (8k) H.265/HEVC video streams.
[22:49:08 CEST] <ubitux> http://www.nvidia.com/download/driverResults.aspx/105343/en-us
[22:49:19 CEST] <ubitux> pascal gpus probably
[23:11:16 CEST] <Chloe> does ffmpeg support demuxing multiple subsongs from a single file?
[23:14:47 CEST] <Rathann> Chloe: what are "subsongs"? what kind of file are you talking about?
[23:18:59 CEST] <Chloe> As far as I can tell, they're just sections of an audio track which are songs, I guess it'd be something like a CD where the audio file is the CD and the songs on it are the subsongs, but I'm not entirely sure. The kind of file I'm talking about is an audio file in which there are multiple songs (note: not tracks)--in this case, a module.
[00:00:00 CEST] --- Sat Jul 16 2016