audacity VS audacity

"For what it's worth, I forked Audacity yesterday and removed all telemetry that could potentially spy on users from the codebase. We're currently deciding on a nice rebranded name (as Audacity is trademarked) and are in the process of founding a GitHub organization afterwards. We're also looking for maintainers that could help with Windows and MacOS builds. The repo's link is (for now):" https://github.com/cookiengineer/audacity

See also, https://www.darkaudacity.com/

Title feels misleading - skip to the end:

> As for me, I still believe in open source, and even in the commercial potential of open source. It requires creativity and a clever business acumen to identify and exploit market opportunities within this collaborative framework. To win in open source you must embrace this collaboration and embrace the fact that you will share the commercial market for the software with other entities. If you’re up to that challenge, then let’s keep beating the open source drum together. If not, these new movements(source-available software) may be a home for you – but know that a lot of hard work still lies ahead of you in that path.

This article feels more like a critique of the new "source-available" trend in tech whereby existing OSS projects sell out and try to retroactively re-license themselves to be more closed-source friendly. I've been sort of aware of this phenomenon since there was a small upset when Muse Group bought Audacity, though it appears it is still GPL[1].

[1] https://github.com/audacity/audacity/blob/master/LICENSE.txt

#5467 Fix 24-bit recording.

The time stretch algorithm is implemented in https://github.com/audacity/audacity/blob/master/libraries/l... particularly functions _time_stretch and _process_hop. It looks to me like a classic phase vocoder with vertical phase coherence (c.f. https://en.wikipedia.org/wiki/Phase_vocoder).

The basic idea is this. For a time-stretch factor of, say, 2x, the frequency spectrum of the stretched output at 2 sec should be the same as the frequency spectrum of the unstretched input at 1 sec. The naive algorithm therefore takes a short section of signal at 1s, translates it to 2s and adds it to the result. Unfortunately, this method generates all sorts of unwanted artifacts.

Imagine a pure sine wave. Now take 2 short sections of the wave from 2 random times, overlap them, and add them together. What happens? Well, it depends on the phase of each section. If the sections are out of phase, they cancel on the overlap; if in phase, they constructively interfere.

The phase vocoder is all about overlapping and adding sections together so that the phases of all the different sine waves in the sections line up. Thus, in any phase vocoder algorithm, you will see code that searches for peaks in the spectrum (see _time_stretch code). Each peak is an assumed sine wave, and corresponding peaks in adjacent frames should have their phases match.

https://github.com/audacity/audacity/releases

If you are able to build from source, you could download the latest source code, or the latest release version source code, and manually apply the patch.

FWIW: We'll be ditching the atlases for 3.4 most likely, in favor of having themes work like they do in source anyway: With individual PNGs for the icons and a Colors.txt containing all the color definitions. With that, custom themes would be much easier to author and also would stop breaking every time we introduce a new icon.

https://github.com/audacity/audacity/issues/4639 this bug mentions everything, I guess it's Audacity after all. To be fixed in 3.3.2