[opendtv] Re: Precision
- From: Tom Barry <trbarry@xxxxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Mon, 28 May 2007 16:14:19 -0400
Craig Birkmaier wrote:
> The whole point of using digital video compression is to:
>
> 1. Remove redundancy - this is why temporal coding typically contributes
> significantly more coding gain than spatial coding. If done well this
> can be virtually lossless.
>
> 2. Substitute correlated noise for picture information - this is what we
> do when we quantize away high frequency image details. This is ALWAYS
> lossy.
I'd prefer to think of practical (lossy) compression as removing:
1) Redundancy, as you stated
2) Information we don't care about enough to encode, say very high
frequency information, and
3) Related, but not the same as 2), information we don't know about or
don't trust. This is information that was captured, but not reliably
due to sampling error, noise, whatever. There is a point of diminishing
returns on how many bits we can afford to spend encoding unreliable
samples or extra bit depth once these things become lost in the noise.
For 2) and 3) however it may be best to not filter or discard them but
instead allow the encoder to opportunistically choose whichever values
happen to encode nicest. That's one of the reasons I think capturing at
much higher bit depths and allowing encoders to quantize them away
seems, in some tests, to work more efficiently than some might predict.
- Tom
At 12:41 PM -0400 5/28/07, Tom Barry wrote:
All this is good but it would seem there would also be a point where
you don't need any noise or dithering because you were storing and
displaying sufficient colors/bits such that the HVS could not see the
difference anyway. Just on principle I'd like to get to that point.
And I worry that once you add (or keep) noise it becomes information,
subtracting from the available bit budget. We already live with this
in the various film grain debates that pop up periodically.
In theory this sounds wonderful. But in practice it is difficult to
achieve.
The whole point of using digital video compression is to:
1. Remove redundancy - this is why temporal coding typically contributes
significantly more coding gain than spatial coding. If done well this
can be virtually lossless.
2. Substitute correlated noise for picture information - this is what we
do when we quantize away high frequency image details. This is ALWAYS
lossy.
Bottom line, the better we can do at predicting what adjacent frames
look like, the more bits we have to preserve the high frequency details.
Regards
Craig
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
--
Tom Barry trbarry@xxxxxxxxxxx
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
