Subject: Re: Re[2]: [linux-audio-dev] peakfiles and EDL's
From: Richard C. Burnett (burnett_AT_tality.com)
Date: Wed Feb 28 2001 - 20:19:32 EET
Well I just may have to do that :) I have been wanting to play with some
ways of displaying the data. So I can easily see how to use peaks for say
bins of 4000 and more ( even 2000 ) and of course dispaying the raw
data. My new question is how would you do a 1:64 zoom. All the symetry
stuff I was saying was pertaining to large enough bins where the frequency
does not matter. I can't imagine they downsample because we already know
all the aliasing effects that causes, do they use some form of bin still
though that is now affected by frequencies, at 1:64 signals at or below
689 Hz will start showing oscilations is amplitude.
Tonight I will generate some waveforms in Samplitude with the wavegen
feature and see how frequencies look at different zoom levels. I know
that when I did a 1k signal for 1 minutes and had it full zoomed on the
screen (1024x768) it appeared as a rectangle.
Rick
On Wed, 28 Feb 2001, rob wrote:
> On Wed, 28 Feb 2001, Richard C. Burnett wrote:
> > I agree with robby in not understanding the benefit of computing min
> > values, The top bottom is probably going to be pretty symetric for all
> > purposes, so you could just mirror it.
>
> for low zoom values (looking at alot of data in a single window),
> the data is mostly symmetrical (1pixel = 1024 frames, no DC offset),
> however looking at values around 1:64 it won't be generally (or even
> mostly). If i'm looking at 16bit PCM (+/- 32768) frame which has the
> following values 59 134 1515 23135 ... 24, where all values represented by
> the ... are between the two adjacent values, your represented signal won't
> be symmetrical about 0 even if it again has no DC offset.
> This is pretty much the way a person using a tool like soundforge
> etc expects the wave data display to act, and i don't see much motivation
> to change that.
> Perhaps the best way to convince yourself of this is to implement
> a small viewing program, when i did that, this method is pretty much the
> only way i found of making it look like i thought it should.
>
> rob
>
> ----
> Rob Melby
> Georgia Institute of Technology, Atlanta Georgia, 30332
> uucp: ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!gt4255a
> Internet: rob_AT_kaybee.org, gt4255a_AT_prism.gatech.edu
>
>
+------------------------+-----------------------+
| T a l i t y | +------+ |
+------------------------+ +----+-+ | |
| Richard Burnett | +-+ | |
| Senior Design Engineer +---+ +----+ |
| burnett_AT_tality.com | | |
| | | |
| Phone: 919.380.3014 | |
| Fax: 919.380.3903 | | |
+------------------------------------------------+
This archive was generated by hypermail 2b28 : Wed Feb 28 2001 - 21:01:57 EET