Page 2 of 3

Re: Discontinuous, and spectral selections

Posted: Mon Jun 30, 2014 2:58 pm
by Paul L
steve, I figured out procedures like what you described, yes.

I thought there is little difference between effects and generators and analyzers, when it comes to programming them in Nyquist. You can put your plug in in one or another menu with header comments and little differs.

The punch paste you wrote for me -- which should we call that? We call it an effect. Well, it does have that cross fading at the boundaries.

Letting Nyquist read labels would be interesting. Still I suppose the effect would have to produce unchanged sound for what is between labels, but I think that might be needlessly slow. After all, the trivial effect that returns s takes time to evaluate on a long selection, right?

Whereas applying effects repeatedly to discontinuous pieces of the selection would, I imagine, not suffer that and generalize to other than Nyquist effects.

Re: Discontinuous, and spectral selections

Posted: Mon Jun 30, 2014 4:00 pm
by steve
Paul L wrote:I thought there is little difference between effects and generators and analyzers,
There are some significant differences.

For example, in a "generate" type plug-in, the value of the global "s" is 0.25, whereas for analyze and process type plug-ins, the value of "s" is the (mono or stereo) sound passed from Audacity.

Another difference concerns *warp*. For generate type plug-ins, (osc 60 1) will generate a sine wave of 1 second duration, whereas for a process type plug-in the environment is stretched so that it produces a sine wave of the same duration as the selection.
Paul L wrote:Still I suppose the effect would have to produce unchanged sound for what is between labels,
In (experimental) version 4 plug-ins, Audacity can pass the start/end times of audio clips, Nyquist can access specified track times, and Nyquist can make multiple passes over the selected audio. This overcomes several of the current limitations of Nyquist plug-ins (including the age-old problems of: (a) not being able to normalize very long tracks. (b) not being able to differentiate between white-space and absolute silence. (c) not being able to access audio outside of the current selection).

Re: Discontinuous, and spectral selections

Posted: Mon Jun 30, 2014 6:58 pm
by Paul L
Is this sort of thing protected by software patent or not?

Have patents ever stopped anything else in Audacity development?

Re: Discontinuous, and spectral selections

Posted: Mon Jun 30, 2014 7:20 pm
by steve
No doubt that making a direct copy of features from Adobe Audition or Izotope RX would breach IP rights (intellectual property) in one way or another, but I doubt that the general idea of selecting audio in spectrogram track view would be patentable.
Paul L wrote:Have patents ever stopped anything else in Audacity development?
There are a number of features that cannot be included with Audacity due to licensing/patent restrictions, (such as ASIO, ReWire, built-in MP3 export, built-in WMA import/export, Steinberg VST GUI support ..) though workarounds have been developed for some of these (such as using LAME/FFMpeg external libraries to support restricted formats, and using an open source VST header rather than the Steinberg VST header).

Disclaimer: I am not a lawyer ;)

Re: Discontinuous, and spectral selections

Posted: Mon Jun 30, 2014 8:10 pm
by Paul L
I worked in commercial software, and it is surprising what trivial GUI things they attempt to patent. Patent may have become a tool of mutual legal harassment of software companies.

The USPO sometimes patents stupid things.

Re: Discontinuous, and spectral selections

Posted: Mon Jun 30, 2014 8:30 pm
by steve
The fact that two different large companies (Adobe and iZotope) both have a very similar feature is probably a reasonable indication that such IP claims will not stand up.

Re: Discontinuous, and spectral selections

Posted: Tue Jul 01, 2014 1:24 am
by Paul L
Great! Now which of us will add it to Audacity? One of these days I will go beyond Nyquist... but I will start smaller.

The hints about greater Nyquist capabilities are pleasing.

Re: Discontinuous, and spectral selections

Posted: Wed Sep 24, 2014 11:51 am
by waxcylinder
Do I assume correctly that this relates to the "Experimental Spectral Patch" that Paul has been posting about in -devel?

And if so I'm assuming that I can just delete this thread - unless, that is, anyone wants me to archive it?

Peter

Re: Discontinuous, and spectral selections

Posted: Wed Sep 24, 2014 1:06 pm
by Paul L
steve wrote:
Paul L wrote:
Gale Andrews wrote: People have asked to select arbitrary frequencies such as 1000 Hz to 2000 Hz then apply an effect only to those frequencies.

Gale
Can you point me to a detailed discussion of that? I am not exactly sure what is meant by the brief description.
Not a detailed discussion, but this is the most recent request that I recall seeing for this type of feature: http://forum.audacityteam.org/viewtopic ... 15#p245215
I noticed this link is no longer good.

Re: Discontinuous, and spectral selections

Posted: Wed Sep 24, 2014 1:11 pm
by Paul L
waxcylinder wrote:Do I assume correctly that this relates to the "Experimental Spectral Patch" that Paul has been posting about in -devel?

And if so I'm assuming that I can just delete this thread - unless, that is, anyone wants me to archive it?

Peter
Yes, relevant, I wanted the feature so much, I did it! Might as well archive this.

I noticed this page
http://wiki.audacityteam.org/wiki/Sugge ... pabilities

contains

Selection Tool for spectrum: with upper and lower frequency thresholds to apply effects to selected frequencies. I'd like to be able to draw a box on the spectrum and know my effect would only be applied in there. Ideally we could draw any shaped selection we wanted, not just a rectangle.
This is misplaced on that page which is about the Plot Spectrum window. And it doesn't exactly describe what I did. If what it implied is splitting the selection into bands, then doing ANY effect in the selected band, and then mixing again -- no.