[Ffmpeg-devel-irc] ffmpeg-devel.log.20170115

burek burek021 at gmail.com
Mon Jan 16 03:05:03 EET 2017

[03:47:57 CET] <kierank> atomnuker: what is your view of codec2?
[03:48:46 CET] <kierank> http://www.rowetel.com/?p=5373
[04:22:59 CET] <atomnuker> kierank: pretty impressive, though I disagree with how he uses an fft
[04:23:01 CET] <atomnuker> http://www.rowetel.com/?p=5245
[04:23:47 CET] <atomnuker> he complains about how if you have an odd frequency not fitting directly in any bin it just spreads out and makes things hard to encode
[04:24:02 CET] <atomnuker> he should choose a correct window so that doesn't happen
[04:24:31 CET] <atomnuker> blackman-harris minimizes the sidelobes (the spreading) at the cost of reduced frequency resolution
[04:29:57 CET] <atomnuker> he's using hamming, no wonder he's complaining, that window works in the general case but not very well in this special case
[04:29:58 CET] <atomnuker> he's also using the FFT directly without it being a real-only DCT-IV
[04:33:19 CET] <atomnuker> DCT makes things easier to compress (since the coefficients represent half-cycle basis functions, while FFT has them full-cycle)
[05:00:34 CET] <TD-Linux> atomnuker, I think the reason for the FFT is because phase information is intentionally discarded
[05:03:53 CET] <atomnuker> meh, he's a fully licenced ham radio guy making a codec for only fullly licenced ham radio guys to use
[05:05:31 CET] <atomnuker> that has like a direct analog to one of the main causes of why the French revolution happened
[05:06:07 CET] <kierank> lol wut?
[05:06:14 CET] <kierank> let them eat ham...
[05:06:24 CET] <TD-Linux> atomnuker is oppressed by the fcc.
[05:07:44 CET] <kierank> he is like eminem
[05:08:06 CET] <atomnuker> no, by the special status nobility occupying that establishment, and the high cost of food... I mean ham... licence
[05:08:48 CET] <kierank> it's like 30 quid in the uk
[05:09:00 CET] <TD-Linux> it's really cheap in the US... $15 per 10 years. hough other countries might differ.
[05:09:25 CET] <kierank> and to be fair you do need the training because some of those frequencies are close to distress frequencies
[05:09:37 CET] <kierank> 2.4ghz you can do what you want, when I cook my dinner that's blown to hell anyway
[05:09:42 CET] <TD-Linux> on the other hand some of the surrounding equipment can be dumb and expensive. codec2's competitor, AMBE, being a notable example.
[05:11:43 CET] <TD-Linux> they only sell AMBE encoders as dedicated hardware chips. if you want to do it on your PC you can get the chip mounted on a USB dongle for $300
[06:16:26 CET] <atomnuker> TD-Linux: on second thought, I don't really mind needing a special licence
[06:16:45 CET] <atomnuker> I don't mind it costing lots more and I don't mind the tests and education needed
[06:16:55 CET] <atomnuker> I think it's a genuinely a good thing
[06:17:08 CET] <atomnuker> heck, let them have a lot more spectrum, I don't care
[06:17:22 CET] <atomnuker> on Eminem's song: "FCC claimed that the edit was still considered indecent under its guidelines, as it contained references to sexual activity, which violated U.S. law."
[06:17:42 CET] <atomnuker> ...anything, everything, just keep them off the internet
[11:55:35 CET] <cone-734> ffmpeg 03Georgi D. Sotirov 07release/3.2:41f8a8843d21: lavf/chromaprint: Update for version 1.4
[18:50:23 CET] <cone-137> ffmpeg 03Rostislav Pehlivanov 07master:e7dec52d4da7: matroskaenc: remove unofficial compliance on color information
[20:45:21 CET] <JEEB> &32
[20:45:32 CET] <Sesse> >>5
[20:48:26 CET] <JEEB> I did -32 #elsehwere already, and now the other way. always fun when using irssi with multiple keyboard layouts that you keep switching between
[20:55:41 CET] <Shiz> lol
[20:55:46 CET] <Shiz> i need to get used to alt-j
[23:07:22 CET] <faLUCE> (libavcodec question) Hello. I can grab frames from a v4l2 device, resample them (yuyv422 to yuv420p) , x264-encode them and save the encoded stream into a output.h264 file, which I can see with players. Unfortunately, It's all ok, but I can't mux with libavformat the encoded frames. I used:  "avformat_alloc_output_context2(&avFormatOutputContext, NULL, "mpeg", "foobar"); .....;  av_interleaved_write_frame(
[23:07:24 CET] <faLUCE> [23:01] <faLUCE> avFormatOutputContext, &encodedPkt);"   <----- but I obtain a segfault when I call av_interleaved_write_frame(). Is there something basically wrong with that or I have to valgrind the program?
[23:07:25 CET] <faLUCE> [23:02] <faLUCE> encodedPkt is not  null. And I can write it to the otuput file, and see the output file with a player
[23:07:27 CET] <faLUCE> [23:04] <faLUCE> In addition: I used "foobar" instead of filename. I don't know what to insert as filename, given that there's not a file, but a living source
[23:08:06 CET] <Sesse> 16:55 < Sesse> faLUCE: #ffmpeg is the right place for this (cf. topic)
[23:08:35 CET] <faLUCE> Sesse: already asked there
[23:08:42 CET] <Sesse> and?
[23:09:05 CET] <Sesse> questions should be asked in #ffmpeg doesn't mean ask first in #ffmpeg, then you're free to take it here if you didn't get an answer :-)
[00:00:00 CET] --- Mon Jan 16 2017

More information about the Ffmpeg-devel-irc mailing list