Hi,
I am using Theora codec w/ farsigh2 (no Telepathy) for video conferencing and I am facing distorted video on the receiving end when using camera with relatively high resolution on the sending end. Low resolution streams are decoded alright by the receiver. I ran receiver/sender on the same host so the bandwidth should not be a problem. CPU consumption was low too. The artifacts in the video I see most often are small squares(up to 16x16 or so pixels) with 2 white and 1 black stripes in it. Here is versions of the gstreamer/farsight2 I am using: set VER_GST=0.10.25 set VER_GST_PLUGINS_GL=0.10.1 set VER_GST_PLUGINS_BASE=0.10.25 set VER_GST_PLUGINS_GOOD=0.10.17 set VER_GST_PLUGINS_UGLY=0.10.13 set VER_GST_PLUGINS_BAD=0.10.17 Farsight2 0.0.16 I read on this list about incompatibilities between certain versions of the Fs and gstreamer. I wonder if I am running in it here or it is something else. Thanks, Miron ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ gstreamer-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gstreamer-devel |
Hi, I updated to Farsight2, 0.0.20 set VER_GST=0.10.29 set VER_GST_PLUGINS_BASE=0.10.29 set VER_GST_PLUGINS_GOOD=0.10.22 But I am still getting distorted video when Theora codec is used. Looks like frames between I-frames are not properly decoded. Anything I can check? Thanks, Mironoz > Hi, > > I am using Theora codec w/ farsigh2 (no Telepathy) for video conferencing and I am facing distorted video on the receiving end when using camera with relatively high resolution on the sending end. Low resolution streams are decoded alright by the receiver. I ran receiver/sender on the same host so the bandwidth should not be a problem. CPU consumption was low too. > The artifacts in the video I see most often are small squares(up to 16x16 or so pixels) with 2 white and 1 black stripes in it. > > Here is versions of the gstreamer/farsight2 I am using: > > set VER_GST=0.10.25 > set VER_GST_PLUGINS_GL=0.10.1 > set VER_GST_PLUGINS_BASE=0.10.25 > set VER_GST_PLUGINS_GOOD=0.10.17 > set VER_GST_PLUGINS_UGLY=0.10.13 > set VER_GST_PLUGINS_BAD=0.10.17 > Farsight2 0.0.16 > > I read on this list about incompatibilities between certain versions of the Fs and gstreamer. I wonder if I am running in it here or it is something else. > > Thanks, > Miron Найди все ролики Интернета в поиске по видео http://r.mail.ru/cln5070/go.mail.ru/ ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ gstreamer-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gstreamer-devel |
Hi,
My problem seems to be related to Theora encoder. With the number of discontinued buffers increasing I am likely to get distorted image on the receiving end: 0:01:55.343750000 4492 00DF2698 DEBUG theoraenc gsttheoraenc.c:829:theora_enc_is_discontinuous:<theoraen c4> Incoming TS 0:01:50.345187500 exceeds expected value 0:01:48.955312500 by too much, marking discontinuity If I increase threshold for late/discontinued buffers - the receiving end shows video w/o artifacts. I am not sure yet how to interpret the observation because I see discontinued buffers in this pipeline which still produces the fine video: gst-launch --gst-debug="theoraenc:5" dshowvideosrc!ffmpegcolorspace!theoraenc!theoradec!ffmpegcolorspace!directdrawsink One more thing - on every discontinued packet, the theora codec is deallocated and reincarnated again by theora_enc_reset(). For performance sake it looks suspicious to me. Thanks, Miron > > Hi, > > I updated to Farsight2, 0.0.20 > set VER_GST=0.10.29 > set VER_GST_PLUGINS_BASE=0.10.29 > set VER_GST_PLUGINS_GOOD=0.10.22 > > > But I am still getting distorted video when Theora codec is used. > > Looks like frames between I-frames are not properly decoded. > > Anything I can check? > > Thanks, > Mironoz > > > > > Hi, > > > > I am using Theora codec w/ farsigh2 (no Telepathy) for video conferencing and I am facing distorted video on the receiving end when using camera with relatively high resolution on the sending end. Low resolution streams are decoded alright by the receiver. I ran receiver/sender on the same host so the bandwidth should not be a problem. CPU consumption was low too. > > The artifacts in the video I see most often are small squares(up to 16x16 or so pixels) with 2 white and 1 black stripes in it. > > > > Here is versions of the gstreamer/farsight2 I am using: > > > > set VER_GST=0.10.25 > > set VER_GST_PLUGINS_GL=0.10.1 > > set VER_GST_PLUGINS_BASE=0.10.25 > > set VER_GST_PLUGINS_GOOD=0.10.17 > > set VER_GST_PLUGINS_UGLY=0.10.13 > > set VER_GST_PLUGINS_BAD=0.10.17 > > Farsight2 0.0.16 > > > > I read on this list about incompatibilities between certain versions of the Fs and gstreamer. I wonder if I am running in it here or it is something else. > > > > Thanks, > > Miron > > > Найди все ролики Интернета в поиске по видео > http://r.mail.ru/cln5070/go.mail.ru/ > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Sprint > What will you do first with EVO, the first 4G phone? > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > _______________________________________________ > gstreamer-devel mailing list > [hidden email] > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel Тесты новейших гаджетов на Hi-tech.Mail.Ru http://r.mail.ru/cln5823/hi-tech.mail.ru/ ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ gstreamer-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gstreamer-devel |
Hi ,
Use "sync=false" property with both the sinks at sending and receiving end , and then see the behaviour .
2010/7/14 Miron Kunz <[hidden email]> Hi, -- Regards, Sudarshan Bisht ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ gstreamer-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gstreamer-devel |
Hi, All my video/audio sinks on receiving end have sync=false. The sending part is handled by Farsight2 and it looks like sync=false is set there too. Thanks, Mironoz > > Hi , > > Use "sync=false" property with both the sinks at sending and receiving end , and then see the behaviour . > > > > 2010/7/14 Miron Kunz <[hidden email]> > Hi, > > My problem seems to be related to Theora encoder. > > With the number of discontinued buffers increasing I am likely to get distorted image on the receiving end: > > 0:01:55.343750000 4492 00DF2698 DEBUG theoraenc gsttheoraenc.c:829:theora_enc_is_discontinuous:<theoraen > c4> Incoming TS 0:01:50.345187500 exceeds expected value 0:01:48.955312500 by too much, marking discontinuity > > If I increase threshold for late/discontinued buffers - the receiving end shows video w/o artifacts. > > I am not sure yet how to interpret the observation because I see discontinued buffers in this pipeline which still produces the fine video: > gst-launch --gst-debug="theoraenc:5" dshowvideosrc!ffmpegcolorspace!theoraenc!theoradec!ffmpegcolorspace!directdrawsink > > One more thing - on every discontinued packet, the theora codec is deallocated and reincarnated again by theora_enc_reset(). For performance sake it looks suspicious to me. > > Thanks, > Miron > > > > > > > > > > Hi, > > > > I updated to Farsight2, 0.0.20 > > set VER_GST=0.10.29 > > set VER_GST_PLUGINS_BASE=0.10.29 > > set VER_GST_PLUGINS_GOOD=0.10.22 > > > > > > But I am still getting distorted video when Theora codec is used. > > > > Looks like frames between I-frames are not properly decoded. > > > > Anything I can check? > > > > Thanks, > > Mironoz > > > > > > > > > Hi, > > > > > > I am using Theora codec w/ farsigh2 (no Telepathy) for video conferencing and I am facing distorted video on the receiving end when using camera with relatively high resolution on the sending end. Low resolution streams are decoded alright by the receiver. I ran receiver/sender on the same host so the bandwidth should not be a problem. CPU consumption was low too. > > > The artifacts in the video I see most often are small squares(up to 16x16 or so pixels) with 2 white and 1 black stripes in it. > > > > > > Here is versions of the gstreamer/farsight2 I am using: > > > > > > set VER_GST=0.10.25 > > > set VER_GST_PLUGINS_GL=0.10.1 > > > set VER_GST_PLUGINS_BASE=0.10.25 > > > set VER_GST_PLUGINS_GOOD=0.10.17 > > > set VER_GST_PLUGINS_UGLY=0.10.13 > > > set VER_GST_PLUGINS_BAD=0.10.17 > > > Farsight2 0.0.16 > > > > > > I read on this list about incompatibilities between certain versions of the Fs and gstreamer. I wonder if I am running in it here or it is something else. > > > > > > Thanks, > > > Miron > > > > > > Найди все ролики Интернета в поиске по видео > > http://r.mail.ru/cln5070/go.mail.ru/ > > > > ------------------------------------------------------------------------------ > > This SF.net email is sponsored by Sprint > > What will you do first with EVO, the first 4G phone? > > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > > _______________________________________________ > > gstreamer-devel mailing list > > [hidden email] > > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel > > > > > Тесты новейших гаджетов на Hi-tech.Mail.Ru > http://r.mail.ru/cln5823/hi-tech.mail.ru/ > > > > > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Sprint > What will you do first with EVO, the first 4G phone? > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > _______________________________________________ > gstreamer-devel mailing list > [hidden email] > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel > > > > > > > -- > Regards, > > Sudarshan Bisht "Барахолка" на Товары@Mail.Ru. Покупай дешево, продавай выгодно. http://r.mail.ru/cln4270/torg.mail.ru/used/ ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ gstreamer-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gstreamer-devel |
Hi,
So I came to the conclusion that latency caused by encoder and network are destroying my Fs2/Theora video on the receiving end. With high (for ex 1/15) fps encoder drops many buffers and does not try to lower fps to adjust to CPU capabilities. Thanks, Mironoz > > Hi, > All my video/audio sinks on receiving end have sync=false. The sending part is handled by Farsight2 and it looks like sync=false is set there too. > > Thanks, > Mironoz > > > > > Hi , > > > > Use "sync=false" property with both the sinks at sending and receiving end , and then see the behaviour . > > > > > > > > 2010/7/14 Miron Kunz <[hidden email]> > > Hi, > > > > My problem seems to be related to Theora encoder. > > > > With the number of discontinued buffers increasing I am likely to get distorted image on the receiving end: > > > > 0:01:55.343750000 4492 00DF2698 DEBUG theoraenc gsttheoraenc.c:829:theora_enc_is_discontinuous:<theoraen > > c4> Incoming TS 0:01:50.345187500 exceeds expected value 0:01:48.955312500 by too much, marking discontinuity > > > > If I increase threshold for late/discontinued buffers - the receiving end shows video w/o artifacts. > > > > I am not sure yet how to interpret the observation because I see discontinued buffers in this pipeline which still produces the fine video: > > gst-launch --gst-debug="theoraenc:5" dshowvideosrc!ffmpegcolorspace!theoraenc!theoradec!ffmpegcolorspace!directdrawsink > > > > One more thing - on every discontinued packet, the theora codec is deallocated and reincarnated again by theora_enc_reset(). For performance sake it looks suspicious to me. > > > > Thanks, > > Miron > > > > > > > > > > > > > > > > > > Hi, > > > > > > I updated to Farsight2, 0.0.20 > > > set VER_GST=0.10.29 > > > set VER_GST_PLUGINS_BASE=0.10.29 > > > set VER_GST_PLUGINS_GOOD=0.10.22 > > > > > > > > > But I am still getting distorted video when Theora codec is used. > > > > > > Looks like frames between I-frames are not properly decoded. > > > > > > Anything I can check? > > > > > > Thanks, > > > Mironoz > > > > > > > > > > > > > Hi, > > > > > > > > I am using Theora codec w/ farsigh2 (no Telepathy) for video conferencing and I am facing distorted video on the receiving end when using camera with relatively high resolution on the sending end. Low resolution streams are decoded alright by the receiver. I ran receiver/sender on the same host so the bandwidth should not be a problem. CPU consumption was low too. > > > > The artifacts in the video I see most often are small squares(up to 16x16 or so pixels) with 2 white and 1 black stripes in it. > > > > > > > > Here is versions of the gstreamer/farsight2 I am using: > > > > > > > > set VER_GST=0.10.25 > > > > set VER_GST_PLUGINS_GL=0.10.1 > > > > set VER_GST_PLUGINS_BASE=0.10.25 > > > > set VER_GST_PLUGINS_GOOD=0.10.17 > > > > set VER_GST_PLUGINS_UGLY=0.10.13 > > > > set VER_GST_PLUGINS_BAD=0.10.17 > > > > Farsight2 0.0.16 > > > > > > > > I read on this list about incompatibilities between certain versions of the Fs and gstreamer. I wonder if I am running in it here or it is something else. > > > > > > > > Thanks, > > > > Miron > > > > > > > > > Найди все ролики Интернета в поиске по видео > > > http://r.mail.ru/cln5070/go.mail.ru/ > > > > > > ------------------------------------------------------------------------------ > > > This SF.net email is sponsored by Sprint > > > What will you do first with EVO, the first 4G phone? > > > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > > > _______________________________________________ > > > gstreamer-devel mailing list > > > [hidden email] > > > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel > > > > > > > > > > Тесты новейших гаджетов на Hi-tech.Mail.Ru > > http://r.mail.ru/cln5823/hi-tech.mail.ru/ > > > > > > > > > > > > ------------------------------------------------------------------------------ > > This SF.net email is sponsored by Sprint > > What will you do first with EVO, the first 4G phone? > > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > > _______________________________________________ > > gstreamer-devel mailing list > > [hidden email] > > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel > > > > > > > > > > > > > > -- > > Regards, > > > > Sudarshan Bisht > > > "Барахолка" на Товары@Mail.Ru. Покупай дешево, продавай выгодно. http://r.mail.ru/cln4270/torg.mail.ru/used/ > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Sprint > What will you do first with EVO, the first 4G phone? > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > _______________________________________________ > gstreamer-devel mailing list > [hidden email] > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel Тесты новейших гаджетов на Hi-tech.Mail.Ru http://r.mail.ru/cln5823/hi-tech.mail.ru/ ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ gstreamer-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gstreamer-devel |
Hi,
can you monitor a session with wireshark on the receiving side and check the lost packets? When bandwidth usage increases you're more likely to loose packets or receive them too in late. Regards 2010/7/15 Miron Kunz <[hidden email]>: > Hi, > > So I came to the conclusion that latency caused by encoder and network are destroying my Fs2/Theora video on the receiving end. > > With high (for ex 1/15) fps encoder drops many buffers and does not try to lower fps to adjust to CPU capabilities. > > Thanks, > Mironoz > >> >> Hi, >> All my video/audio sinks on receiving end have sync=false. The sending part is handled by Farsight2 and it looks like sync=false is set there too. >> >> Thanks, >> Mironoz >> >> > >> > Hi , >> > >> > Use "sync=false" property with both the sinks at sending and receiving end , and then see the behaviour . >> > >> > >> > >> > 2010/7/14 Miron Kunz <[hidden email]> >> > Hi, >> > >> > My problem seems to be related to Theora encoder. >> > >> > With the number of discontinued buffers increasing I am likely to get distorted image on the receiving end: >> > >> > 0:01:55.343750000 4492 00DF2698 DEBUG theoraenc gsttheoraenc.c:829:theora_enc_is_discontinuous:<theoraen >> > c4> Incoming TS 0:01:50.345187500 exceeds expected value 0:01:48.955312500 by too much, marking discontinuity >> > >> > If I increase threshold for late/discontinued buffers - the receiving end shows video w/o artifacts. >> > >> > I am not sure yet how to interpret the observation because I see discontinued buffers in this pipeline which still produces the fine video: >> > gst-launch --gst-debug="theoraenc:5" dshowvideosrc!ffmpegcolorspace!theoraenc!theoradec!ffmpegcolorspace!directdrawsink >> > >> > One more thing - on every discontinued packet, the theora codec is deallocated and reincarnated again by theora_enc_reset(). For performance sake it looks suspicious to me. >> > >> > Thanks, >> > Miron >> > >> > >> > >> > >> > >> > >> > > >> > > Hi, >> > > >> > > I updated to Farsight2, 0.0.20 >> > > set VER_GST=0.10.29 >> > > set VER_GST_PLUGINS_BASE=0.10.29 >> > > set VER_GST_PLUGINS_GOOD=0.10.22 >> > > >> > > >> > > But I am still getting distorted video when Theora codec is used. >> > > >> > > Looks like frames between I-frames are not properly decoded. >> > > >> > > Anything I can check? >> > > >> > > Thanks, >> > > Mironoz >> > > >> > > >> > > >> > > > Hi, >> > > > >> > > > I am using Theora codec w/ farsigh2 (no Telepathy) for video conferencing and I am facing distorted video on the receiving end when using camera with relatively high resolution on the sending end. Low resolution streams are decoded alright by the receiver. I ran receiver/sender on the same host so the bandwidth should not be a problem. CPU consumption was low too. >> > > > The artifacts in the video I see most often are small squares(up to 16x16 or so pixels) with 2 white and 1 black stripes in it. >> > > > >> > > > Here is versions of the gstreamer/farsight2 I am using: >> > > > >> > > > set VER_GST=0.10.25 >> > > > set VER_GST_PLUGINS_GL=0.10.1 >> > > > set VER_GST_PLUGINS_BASE=0.10.25 >> > > > set VER_GST_PLUGINS_GOOD=0.10.17 >> > > > set VER_GST_PLUGINS_UGLY=0.10.13 >> > > > set VER_GST_PLUGINS_BAD=0.10.17 >> > > > Farsight2 0.0.16 >> > > > >> > > > I read on this list about incompatibilities between certain versions of the Fs and gstreamer. I wonder if I am running in it here or it is something else. >> > > > >> > > > Thanks, >> > > > Miron >> > > >> > > >> > > Найди все ролики Интернета в поиске по видео >> > > http://r.mail.ru/cln5070/go.mail.ru/ >> > > >> > > ------------------------------------------------------------------------------ >> > > This SF.net email is sponsored by Sprint >> > > What will you do first with EVO, the first 4G phone? >> > > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> > > _______________________________________________ >> > > gstreamer-devel mailing list >> > > [hidden email] >> > > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel >> > >> > >> > >> > >> > Тесты новейших гаджетов на Hi-tech.Mail.Ru >> > http://r.mail.ru/cln5823/hi-tech.mail.ru/ >> > >> > >> > >> > >> > >> > ------------------------------------------------------------------------------ >> > This SF.net email is sponsored by Sprint >> > What will you do first with EVO, the first 4G phone? >> > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> > _______________________________________________ >> > gstreamer-devel mailing list >> > [hidden email] >> > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel >> > >> > >> > >> > >> > >> > >> > -- >> > Regards, >> > >> > Sudarshan Bisht >> >> >> "Барахолка" на Товары@Mail.Ru. Покупай дешево, продавай выгодно. http://r.mail.ru/cln4270/torg.mail.ru/used/ >> >> ------------------------------------------------------------------------------ >> This SF.net email is sponsored by Sprint >> What will you do first with EVO, the first 4G phone? >> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> _______________________________________________ >> gstreamer-devel mailing list >> [hidden email] >> https://lists.sourceforge.net/lists/listinfo/gstreamer-devel > > > Тесты новейших гаджетов на Hi-tech.Mail.Ru > http://r.mail.ru/cln5823/hi-tech.mail.ru/ > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Sprint > What will you do first with EVO, the first 4G phone? > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > _______________________________________________ > gstreamer-devel mailing list > [hidden email] > https://lists.sourceforge.net/lists/listinfo/gstreamer-devel > ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ gstreamer-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gstreamer-devel |
Free forum by Nabble | Edit this page |