Composers Desktop Project Forum

CDP Software Maintenance engineer Richard Dobson writes regarding the question about PVOC frequency & amplitude v. phase shift & amplitude:

The short answer is, yes, we really do mean frequencies. Techncially, "frequency" (as a nominally continuous artefact) is defined as the rate of change of phase. This is why sometimes the phase vocoder has been referred to as a "tracking" vocoder, simnply because it tracks the (rate of) change of phase over time. A zero rate of change would equate to a constant DC level.

In the original pvoc, understood very much as an accumulating (sliding/overlapping frames) FFT-based process, data could be output in all the possible formats: raw "complex" values as delivered by the FFT, or either of amplitude/phase or all the way to ampltude/frequency. For pure analysis tasks, amp/phase is still often used. But in CDP (and likewise in Csound etc), we want to make weird transformations of the source and resynthesise it - this is very much easier and more "intuitive" with the classic amp/frequency format. It means, for example, that resynthesis can be done (from the same data) either by pvoc or by direct oscillator bank.

Some transformations (such as zeroing the amplitudes in some analysis bins) are "incorrect" mathematically (in terms of strict band-limited "linear" DSP), but composers rarely care about that. So long as the results are both interesting and mostly correlate with the intention, we just get on with it!

The "pvocex" program now included in the CDP system is a straight port of the origial pvoc (but creating pvx files), with (almost) all the original flag options.

Cheers,
Richard

Hello,  Thank you for your query about PVOC usage.  The references to pre-Release 4 usage flags were put there at the time of Release 4.5 in order to alert existing users at the time to the changes.  These references are now completely redundant and the current usage and descriptions are the ones to be followed.  I will recommend that the yes, confusing, references to pre-Release 4 usage be removed from the Manual.  I myself just use the default settings, so the command lines are very simple, e.g., pvoc anal 1 infile.wav outfile.ana and pvoc synth infile.ana outfile.wav.

As to the query about frequencies and amplitudes v. phase shifts and amplitudes, I will have to check with the experts on this before replying.  To be continued.

We think the AV software is probably reacting to the self-extracting exe used for setup. I looked at the individual exe files inside the distribution and a random selection of the CDP ones scored completely "clean", as did Soundloom. Soundshaper,exe upset just one of the AV programs, which saw it as a PUP - Potentially Unwanted Program (same as PUA).

We'll probably switch to a disk-image ISO or a plain zip file, but this will mean revising the installation notes, which can't be done until after next week.  I'll post again when completed.

To add a little more to this important investigation:

The reference to "PUA..." in the Sophos instance is explained by them here:

https://support.sophos.com/support/s/article/KBA-000004926?language=en_US

I find it interesting (not least as I used Sophos for all AV checking when I had a free account through Bath University; I now use McAfee) that they refer to their "Deep Learning" system. This is of course the modern AI approach, which is by its nature statistical, rather than the "conventional" AV paradigm of comparing against a database of known threats. On that page "PUA" means "Potentially Unwanted Application" (especially in the context of a "business netowrk") - which looks to me remarkably like the Gatekeeper system on the Mac for checking uncertificated downloads - and for which we document a standard workaround.

We are of course continuing to look at this, and will report here as and when.

Thanks for this analysis. We'll try to follow these up. Two thoughts though: (1) I've not heard of any of the programs that found fault with the files - all the well-known anti-virus packages seem to pass them. (2) it's hard to believe that the Lite zip would be seen as more malicious than the full one, though, given that it has far fewer files - it's exactly the same as Full less the CDP Documentation, so that doesn't add up somehow.