traverso-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Traverso-devel] Some confirmations for the manual


From: Jonatan Liljedahl
Subject: Re: [Traverso-devel] Some confirmations for the manual
Date: Mon, 12 Feb 2007 22:12:25 +0100
User-agent: Thunderbird 1.5 (X11/20051025)

Remon wrote:
>> No, I'm talking of offline transformations... Say for example that you
>> normalize the audio after recording, with a integer format this means
>> there will be rounding errors which will be interpolated. But with a
>> float format this doesn't happen, the resolution isn't changed when the
>> sample values are scaled.
> 
> I'm not completely sure about this. In respect to this particular case, 
> Traverso calculates the normalization factor during import/record, and at 
> that time it's all in the float domain.
> 
> Applying normalization is thus instantanious in traverso ;-)

Yes, this particular example was not very good in this case... =)

Anyhow, I was only trying to point out that it's a good idea to save and
work with 32bit floats even when the recorded input (and eventually the
output, when burned to CD) is only 16bit int.
Like in a sample editor like Audacity, where you can do lots of
processing "offline" (not realtime), the sample data is actually changed
and would loose information if it was 16bit ints! It's just like when
editing photo's, you don't save a work-in-progress photo in JPEG, but in
a non-destructive format like XCF or TIFF.

-- 
/Jonatan    -=( http://kymatica.com )=-




reply via email to

[Prev in Thread] Current Thread [Next in Thread]