[studiorecorder] Re: Normalization question
- From: "Neal Ewers" <neal.ewers@xxxxxxxxxxxxxx>
- To: <studiorecorder@xxxxxxxxxxxxx>
- Date: Fri, 23 Jun 2006 15:04:24 -0500
Rob, Sound forge calls these Minimum and maximum sample values. I am
not sure these are any more correct, and could be misleading, but I
think the word peak is what may be throwing people. I think many people
think of peak as a positive number and relate it to peak meters which
causes some to think of peak as being the highest possible level in the
file. Thus, minimum peak could be rather confusing.
Neal
-----Original Message-----
From: studiorecorder-bounce@xxxxxxxxxxxxx
[mailto:studiorecorder-bounce@xxxxxxxxxxxxx] On Behalf Of ROB MEREDITH
Sent: Friday, June 23, 2006 2:21 PM
To: studiorecorder@xxxxxxxxxxxxx
Subject: [studiorecorder] Re: Normalization question
Perhaps the terminology is confusing. The negative peak is the smallest
value in the document. The positive peak is the largest value in the
document. But, when converting amplitude to dB, they are considered
absolute values, where full scale is 0dB. So, in 16-bit land, +22387 and
-22387 are both -2.31dB, and neither one is more or less important than
the other.
By the way, there is nothing to say that the positive peak must be a
positive value, or the negative peak must be a negative value. It
generally is the case, however. Perhaps I should change the wording on
this at some point.
Rob Meredith
>>> neal.ewers@xxxxxxxxxxxxxx 06/23/06 02:56PM >>>
So, guess what my negative peak was. 0 DB. You're right again.
Neal
-----Original Message-----
From: studiorecorder-bounce@xxxxxxxxxxxxx
[mailto:studiorecorder-bounce@xxxxxxxxxxxxx] On Behalf Of ROB MEREDITH
Sent: Friday, June 23, 2006 1:45 PM
To: studiorecorder@xxxxxxxxxxxxx
Subject: [studiorecorder] Re: Normalization question
Neal:
The answer is that you need to look at the positive and the negative
peak. They normally don't match, and the one with the largest value is
used to determine the normalize offset. I'll bet your negative peak in
this case was very close to 0 to begin with. That would explain the
clipping when forcing the positive value to 0dB.
Note: this is normal; all audio programs work this way.
Rob Meredith
>>> neal.ewers@xxxxxxxxxxxxxx 06/23/06 02:31PM >>>
Rob, On several occasions when using SR, I normalize a file to 0 DB and
when I check the measurements, the positive peak is something like minus
3.2 or something like that. When I raise the volume to 0 with the
volume module, the file is distorted. If I undo the volume change and
try normalizing again, I am told that the file is already at the level I
have set. It is often a mono file. I know there is an answer here, I
just don't know what it is. I bet, however, that you do. So, may we
have your wisdom on this one?
Neal
Other related posts:
