- From: Max Romantschuk <max@romantschuk.fi>
- Date: Mon, 17 Aug 2009 13:22:50 +0300
Dr. Markus Walther wrote:
> The much weaker goal I would propose is to support at least one simple
> lossless audio format in this regard (I am not qualified to comment on
> the video case). Simple means 'simple to generate, simple to decode',
> and PCM WAVE meets these requirements, so would be an obvious candidate.
>
> For that candidate at least I think one could give sample-accurate
> implementations of subinterval selection - tons of audio applications
> demonstrate this is possible.
That is a good point, and having done sample editing it's true that for
applications like speech sub-millisecond resolution makes sense.
The problem becomes what makes sense as the chosen unit. Time is the
intuitive option, but with samples we have aliasing issues when trying
to match a particle time offset to a sample offset. Allowing
microseconds would be a possible solution, but I doubt any authors would
like to think in the terms of microseconds.
But on the other hand if samples are the unit of choice that instantly
ties the interval scale to the file. If an author wants to halve his
bandwidth usage by halving the sample rate for PCM or FLAC (comparable
to increasing the compression for a lossy format) all the interval
information has to be redone. (The offsets will point to the wrong parts
of the file at a different sample rate.) Time based intervals would not
have this issue.
The problem becomes even more convoluted when taking video into account.
Then there are multiple permutations of sample rates and frame rates to
take into account.
It would of course be an option to have a sample-accurate option for
audio only, but would this really be implemented in user agents? Authors
will likely not adopt any solution poorly supported in user agents and
stick to existing plugin solutions like Flash.
Regards,
Max
--
Max Romantschuk
max at romantschuk.fi
http://max.romantschuk.fi/
Received on Monday, 17 August 2009 03:22:50 UTC