Subject: Re: Fwd: Re: Re[2]: [linux-audio-dev] peakfiles and EDL's
From: Robert Schrem (Robert.Schrem_AT_WiredMinds.de)
Date: Thu Mar 01 2001 - 22:00:20 EET
On Thursday 01 March 2001 18:44, Paul Davis wrote:
> In message <01030118195401.06627_AT_pc-robert>you write:
> >On Wednesday 28 February 2001 19:57, Paul Davis wrote:
> >> >At least I never felt any neccesarity for two seperate peak meters
> >> >showing the positive and negative peak values of the same signal
> >> >seperatly. Maybe you can you convince me, why I would need this an
> >> >for what purpose.
> >>
> >> i think thats because your peak meters are almost certainly computing
> >> RMS levels, even if they say "peak meter".
> >
> >It depends. The Yamaha 02/R digital mixer, Alesis ADAT recorders,
> >usual DAT recorders and the DR-8 HardDisk recorder display
> >like most digital equipment 'digital' peak data on the peak
> >meters, that's for sure (I'm using these things for years now).
>
> as the author of a few software digital meters, i can explain to you
> that the answer is very simple: peak meters are designed to show
> clipping. it doesn't matter whether the clip is positive or
> negative. hence, zero is at the bottom of the meter, and the clip
> value is somewhere toward the top. and yes, you use abs() to figure
> out the value to display.
>
> but with a "waveform" display, this is rather different. for some
> reason (and who can really tell why?) people expect to see a zero axis
> that is roughly in the middle of the data, and to see amplitude values
> for the wave above and below the line.
>
> i don't disagree that for the purposes of spotting clipping, it would
> work to use abs(), and display the entire curve using positive values
> above an axis. but it wouldn't fit in very well with what most people
> think of when they see a waveform display, and it misses the critical
> distinction between a peak meter and a waveform display.
... and has only the half resolution (in pixels) for the whole
dynamic range :)
Wouldn't it be great to have your waveform view be configurable
like this:
1.) The 'classic' view using positive and negative peak values
and faking some sort of 'waveform-like' display as seen in
many sound editors. If you zoom out the 'waveform-fake' will
not neccessarily look like the detail view of the waveform.
If you zoom out very far, it will usually look like a rectangle.
2.) The more sofisticated view mode:
a) For the close up's - where you would have a sample accurate
display just like in 1.): Display a positive/negativ
y axis with y=0 in the middle. In this mode the view will
be computed by the raw sample data directly. You would
use this view for cutting a signal at a zero crossing.
b) If you zoom out -looking at the bigger context of your
editing on the time axis- the editor would automatically
switch to the envelope display mode, showing the abs()
peak values on an positive only y-axis (no need for negative
peak values and using all of the y-axis pixels for maximum
dynamic resolution that is possible with the aviable pixels).
Additionaly to the peak envolope we could optionaly display
a 'loudness' envelope _curve_ computed by some cooseable
algorithm in another color as an overlay. Computationally
more extensive RMS values and pre-filtered RMS would be
candidates for this.
In the envolope display mode the y-axis should be switchable
to a logarithmic scale (in db?) to enhance the editing of
signals with a big dynamic ranges. In mode a) we would always
display in a linear y-axis scale to avoid unnatural
distortion - elsewise a sine wouldn't look like a sine at
all...).
If you zoom closer into the time axis it would then switch
automatically back into view mode a). This would clarify to
the user what he really sees on his display. Eg: No reasonable
operator would think that he would be able to do zero crossing
cuts using a far to unpercise time scale... It would be pretty
obvious that he sees only peak data and not waveform data.
Finally:
Depending on your computer power, the number of tracks and
visulation demands you could witch between mode 1 and 2
to use the view that fits best to your needs (and hardware).
Of course these are just my sweet dreams... :)
robby.
This archive was generated by hypermail 2b28 : Thu Mar 01 2001 - 22:23:45 EET