[FFmpeg-user] Unsophisticated user has a question

[Henk D. Schoneveld]

On Aug 12, 2013, at 4:42 PM, Jeff Evarts <riventree at gmail.com> wrote:

>> 
>> From phil_rhodes at rocketmail.com  Sun Aug 11 18:56:23 2013
>> From: phil_rhodes at rocketmail.com (Phil Rhodes)
>> Date: Sun, 11 Aug 2013 09:56:23 -0700 (PDT)
>> Subject: [FFmpeg-user] Unsophisticated user has a question
>> In-Reply-To: <5207B9CB.3080206 at thelounge.net>
>> References: <CADaSsY2sQwNO1WPX40hjENxmJ4bqZD28UKUrxvAvnQB8JxrXTw at mail.gmail.com>
>> 	<5207B9CB.3080206 at thelounge.net>
>> Message-ID: <1376240183.9929.YahooMailNeo at web121102.mail.ne1.yahoo.com>
>> 
>>>> Can anyone explain the 4x size difference with the data there?
>> 
>> 
>>> it is logical from MPEG3 to H264
>> 
>> I suspect you meant MPEG2.
>> 
>> But is this a situation where ffmpeg is intelligently picking a bitrate based on the characteristics of the video, or is it defaulting to a low bitrate, given the user hasn't specified one? In the latter case it may make sense for the the user to specify a higher bitrate.
>> 
>> P
>> 
>> 
> It appears I should've been more clear. If I am losing a large amount of
> visual data, then this conversion is a bad plan. If I'm not, then the 4x
> storage decrease would be a boon. It looks like the original bitrate was
> 5.2kb/s and the output bitrate was merely 1.3kb/s.
> 
> Is there some way for me to understand whether this is "a lot of loss and I
> should keep the original" vs "no real loss and save the space".
Real loss means, you can visibly detect it. If you add -crf 22 and inspect both visibly and the size of the file, then if not satisfied with the result, try -crf 20 etc. With -crf 20 file size will be larger and quality will be better. Decide yourself if things are going in the desired direction.
> 
> -Jeff
> _______________________________________________
> ffmpeg-user mailing list
> ffmpeg-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user