[MEncoder-users] new doom9 codec comparission (submission)

Jan Knutar jknutar at nic.fi
Thu Dec 15 08:07:49 CET 2005

On Thursday 15 December 2005 05:29, Corey Hickey wrote:

> Nobody should be using snow at this point for anything but testing
> purposes. It could change incompatibly in the future. What makes snow
> remarkable is that it can compete with mature, highly-tweaked and
> -optimized codecs despite not being finished. If Michael ever starts
> working on it again, I wouldn't be surprised if it won your comparison
> someday. For now, though, I can definitely understand why you have
> problems with it.

What's even more remarkable, is that Snow, which seemingly appeared
overnight into the lavc cvs repository, and having none of the MMX and
SSE speedup work done by yartrebo applied yet, can still compete with
BBC's dirac. A codec which has the support of a very large company,
but still fails to meet quality and performance expectations.

Mind, both snow and dirac aren't exactly usable on current hardware, but
atleast snow will be usable in the near future. Plus, it reportedly has better
quality vs bitrate than Dirac.

It amazes me that this experimental, essentially one-man work, can
compete so well with a codec from a large company. Although, in other
ways, I'm not surprised. 

The dirac developers seem unaware of the rest of the multimedia community,
and suggest a strange container format nobody has ever heard of before.
Upon quick browsing, I discover, that to implementthis supposedly "Open"
media format, you need to buy the spec, and that is for the
"single user license". I'm not sure what that means in this context. Can only
one person at a time read the specs, can only one implementation be made
out of the spec, can only one person ever work on the implementation, can
only one user ever use the implementation? Can you at all implement this in
free software? Beats me. The cost for this spec seems to be, to put in relative
terms, a mere two months of rent. For a container format. I think I'd rather have
a place to live in, really. Cardboard box, laptop and wifi hotspots isn't exactly my
idea of place to live in. Mind, inaccessible specs has never prevented ffmpeg
before, and if MXF actually catches on (which I doubt), there'll be people
reverse engineering it for ffmpeg.

But I digress. What really makes snow remarkable is its performance
and its applicability that is miles ahead of the so called competition in
the same field.

It might also stand as a great testament to Mr Newton's profound words.

(I was going to CC everyone on this, but I decided to spare them the
probably to them seemingly philosophical drivel)

More information about the MEncoder-users mailing list