[MEncoder-users] Encoding interlaced content: what am I doing wrong?

[Jarred Nicholls]

>
>
>> In short, going from one codec to another will lose more quality than
>> not, and the only benefit to newer codecs like x264 is in squeezing
>> slightly higher quality out of slightly smaller files, at the expense of
>> lots of CPU time.  At high bitrates, H.264 has nothing special to offer.
>>
>
> But going from MPEG-2 to MPEG-2 is losing quality too, not? Going in two
> steps is AFAIK the only way to use an EDL. And I thought because x.264 has
> the possibility to encode on quality instead of bitrate (I'm using --crf 15)
> that would result in big but quite good files.
>
> Re-encoding video will always lose quality, but if you're encoding from a
high quality source, then the result will look better than if you encoded
from a low quality source - obviously.  Going from a high quality x264 to a
high quality mpeg-2 will probably visually look fine.  Technically though,
quality loss occurs.  But just understand that choosing a codec isn't about
quality, it should be about whether a certain codec is required (i.e. mpeg-2
for DVD, h.264 for blu-ray), getting the most bang (quality) for the buck
(bitrate), decoding requirements (i.e. cpu), and encoding requirements (cpu,
time, etc.).  My point is, you can have an mpeg-2 file be just as high of
quality as a "-crf 15" x264 file...it will just be a larger bitrate
generally.  So your codec choice should be based on what you want/need to do
given everything you know about your codec candidates.  If you need an
mpeg-2 file in the end, then just capture an mpeg-2 file from the get go :-)