headers
Basil Mohamed Gohar | 1 Feb 01:05
Gravatar

Re: converting a mov/h264 screencast to ogg/theora keeping same quality

On 01/31/2010 06:34 PM, Fabio Varesano wrote:
> Hi all,
>
> I would like to convert the following screencast to an Ogg/Theora
> video while keeping the same quality.
>   
Your best bet for this kind of thing is to use the most recent version
of ffmpeg2theora, which you can get here:

http://v2v.cc/~j/ffmpeg2theora/download.html

It is optimized for this kind of thing.  ffmpeg2theora can convert
almost anything that ffmpeg supports into Theora video & Vorbis audio. 
However, it should be kept in mind that there is no way to transcode
from one codec to another and preserve the same quality at the same
rate, unless you are doing lossless to lossless conversions.  By
definition, a lossy codec (H.264) transcoded to another lossy codec
(Theora) will incur yet more loss.  However, with ffmpeg2theora and its
two-pass encoding, you'll likely be able to get acceptable results.

--

-- 
      Basil Mohamed Gohar
abu_hurayrah <at> hidayahonline.org
http://www.basilgohar.com/blog
basilgohar on irc.freenode.net
GPG Key Fingerprint:  5AF4B362
Permalink | Reply | 0 comments |
headers
Jan Stary | 1 Feb 01:17
Picon

understanding the sync options

This must have been asked before, but I couldn't find
a difinitive answer anywhere.

What exactly does -isync do? man says

	-isync	Synchronize read on input.

which confuses me a bit: the video and audio input
should always be synced (give me a counterexample please).

Also, 

	-vsync parameter

                   Video sync method. Video will be stretched/squeezed
                   to match the timestamps, it is done by duplicating
                   and dropping frames. With -map you can select
                   from which stream the timestamps should be taken.
                   You can leave either video or audio unchanged
                   and sync the remaining stream(s) to the unchanged
                   one.

What are the possible values of "parameter"?

	Thanks

		Jan
Permalink | Reply | 0 comments |
headers
James | 1 Feb 10:12
Picon

Error when lame encoding

Hi,

Has this error been resolved?

lame: output buffer too small

The mp3 has been encoded correctly, but this error is still occurring.
Permalink | Reply | 1 comment |
headers
Benjamin Larsson | 1 Feb 11:43
Picon

Re: Wrong Audio when extracting (sometimes silences where there is none!)

Hi mr Wagner.

Sebastian Wagner wrote:
> hi,
>
> I had a previous posting complaining that FFMPEG does ignore gaps in the FLV
> when extracting Audio out of it. (
> http://lists.mplayerhq.hu/pipermail/ffmpeg-user/2010-January/023582.html)
>   

What you have is this issue:

https://roundup.ffmpeg.org/roundup/ffmpeg/issue271

> I now encounter the opposite:
> It does also ADD silence at some places where there is actually none!
>
> Sample Case:
> Input FLV:
> http://www.openmeetings.de/rec/rec_10_seconds_skits.flv (around 2:50)
> Output Wave:
> http://www.openmeetings.de/rec/rec_10_seconds_skits.wav (around 1:40)
>
> The FLV file is correct and it has quite exactly 10 seconds of silence + 10
> seconds of sound bit by bit/rotating till the end.
> The output wave is produced with ffmpeg -i input.flv output.wav (I have
> compiled the latest FFmpeg version SVN-r21544).
>
> The output Wave adds at:
> 0:09 - 0:11 2 seconds of silence.
(Continue reading)

Permalink | Reply | 8 comments |
headers
Sebastian Wagner | 1 Feb 12:12
Picon
Gravatar

Re: Wrong Audio when extracting (sometimes silences where there is none!)

hi,

thanks for your answers.

*This is likely the cause of the flashplayer having a 2 second window before
it cuts the stream.*
=> You are talking about the way the Decoder/Demuxer reads in the stream and
writes the Audio out yes?
Cause in a scenario where there are no Gaps/Silence in the FLV Audio Track
those 2 second window Issue is not available.

More Samples are no problem I can produce you as many as you need.

Budget for fixing is another topic with the sponsors ^^.
We will see what is the most feasable.

anyway thanks a lot for your time and effort.

Sebastian

2010/2/1 Benjamin Larsson <banan <at> ludd.ltu.se>

> Hi mr Wagner.
>
>
> Sebastian Wagner wrote:
>
>> hi,
>>
>> I had a previous posting complaining that FFMPEG does ignore gaps in the
(Continue reading)

Permalink | Reply | 7 comments |
headers
Benjamin Larsson | 1 Feb 12:24
Picon

Re: Wrong Audio when extracting (sometimes silences where there is none!)

Sebastian Wagner wrote:
> hi,
>
> thanks for your answers.
>
> *This is likely the cause of the flashplayer having a 2 second window before
> it cuts the stream.*
> => You are talking about the way the Decoder/Demuxer reads in the stream and
> writes the Audio out yes?
>   

I'm not following you.

> Cause in a scenario where there are no Gaps/Silence in the FLV Audio Track
> those 2 second window Issue is not available.
>   

Ok one more try. When the Flashplayer records from some input it has 
some kind of sound level detector, when there is sound it will send 
nellymoser frames. When there is no sound for 2 seconds it will stop 
sending nellymoser frames and just send null frames instead. This is how 
I assume it could work. Anyway how this works is irrelevant in the case 
of fixing this issue.

> More Samples are no problem I can produce you as many as you need.
>   

5 stream with different flashplayer versions and sample rates should be 
fine.
(Continue reading)

Permalink | Reply | 6 comments |
headers
Sebastian Wagner | 1 Feb 12:50
Picon
Gravatar

Re: Wrong Audio when extracting (sometimes silences where there is none!)

hi,

2010/2/1 Benjamin Larsson <banan <at> ludd.ltu.se>

> Sebastian Wagner wrote:
>
>> hi,
>>
>> thanks for your answers.
>>
>> *This is likely the cause of the flashplayer having a 2 second window
>> before
>> it cuts the stream.*
>> => You are talking about the way the Decoder/Demuxer reads in the stream
>> and
>> writes the Audio out yes?
>>
>>
>
> I'm not following you.
>
>
>  Cause in a scenario where there are no Gaps/Silence in the FLV Audio Track
>> those 2 second window Issue is not available.
>>
>>
>
> Ok one more try. When the Flashplayer records from some input it has some
> kind of sound level detector, when there is sound it will send nellymoser
> frames. When there is no sound for 2 seconds it will stop sending nellymoser
(Continue reading)

Permalink | Reply | 5 comments |
headers
Benjamin Larsson | 1 Feb 12:59
Picon

Re: Wrong Audio when extracting (sometimes silences where there is none!)

Sebastian Wagner wrote:
> hi,
>
> 2010/2/1 Benjamin Larsson <banan <at> ludd.ltu.se>
>
>   
>> Sebastian Wagner wrote:
>>
>>     
>>> hi,
>>>
>>> thanks for your answers.
>>>
>>> *This is likely the cause of the flashplayer having a 2 second window
>>> before
>>> it cuts the stream.*
>>> => You are talking about the way the Decoder/Demuxer reads in the stream
>>> and
>>> writes the Audio out yes?
>>>
>>>
>>>       
>> I'm not following you.
>>
>>
>>  Cause in a scenario where there are no Gaps/Silence in the FLV Audio Track
>>     
>>> those 2 second window Issue is not available.
>>>
>>>
(Continue reading)

Permalink | Reply | 4 comments |
headers
Charles Buysschaert | 1 Feb 15:15

Creating V210 videos from tif 16 bits RGB content

Hello all,

I am trying to create V210 "avi" from a tif 16 bits RGB content
(1920x1080). 

I use this command

ffmpeg -f image2 -i ./tif16/out_%06d.tif -vcodec v210 BigBuckBunny.avi

I get a 1920x1080 video. 

However, the second half of the video width is completely green, while
the first half contains my shrunk original content.

Is this a bug? Or am I supposed to specify that ffmpeg has to make a
4:4:4 to 4:2:2 downsampling?

Thank you in advance for your help. 

Charles

P.S: my ffmpeg version is SVN-r19283

Charles Buysschaert

Senior Software Engineer,

IntoPIX S.A
Permalink | Reply | 1 comment |
headers
Ra Phair | 1 Feb 16:42
Picon

How to enable the JFIF marker in JPEG video thumbnails ?

Hi All,

The generation of a jpeg thumbnail give a standard jpeg file without
the (non mandatory) JFIF marker :

ffmpeg -y -i /tmp/test.flv -f mjpeg -ss 6 -vframes 1 -s 100x74 -an
/tmp/thumb.jpg

Editing the file, we get a Start Of Image marker (FF D8), directly
followed by the comment marker (FF FE).

For compatibility with some external services, I would like to have a
JFIF marker (FF E0) following the Start Of Image marker.

Looking into the thumbnail generator code in libavcodec/mjpegenc.c
(jpeg_put_comments function) shows that this is possible :

-----------------------------------------------------------------------------------------------------
   if (s->aspect_ratio_info /* && !lossless */)
   {
   /* JFIF header */
   put_marker(p, APP0);
   put_bits(p, 16, 16);
   ff_put_string(p, "JFIF", 1); /* this puts the trailing zero-byte too */
-----------------------------------------------------------------------------------------------------

But it depends on aspect_ratio_info, which is always 0 in my case, and
I can't see how to set it from the command line :
ffmpeg -y -i /tmp/test.flv -f mjpeg -ss 6 -vframes 1 -s 100x74 -aspect
[whatever] -an /tmp/thumb.jpg
(Continue reading)

Permalink | Reply | 0 comments |

Gmane