“Reassignment” spectrograms take massive amount of time. While zooming in frequency is easy, zooming in time is hassle. Which is bad, since playing naturally scrolls the sound when it reaches right. So all time should be loaded. It happens very often, that I fit in window with Waveform, then change to spectrogram, use Reassignment and because frequency labels make it occupy more space, it’s end of the world when the song reaches end.
An important design feature of Audacity is that the visual representation of the track audio represents the current state of the audio. This is distinctly different from most “real-time DAW” programs that usually do not update the waveform. In Audacity, the waveform is updated “in real-time” (subject to the speed of the computer), so that “what you see is what you get”.
Spectrogram reassignment is off by default, because reassignment is computationally heavy and too slow for real-time updating.
But if I want to view in high quality, I’m screwed if I want to change zoom or scroll. I can’t do anything at all until spectrogram is loaded.
Audacity does NOT play the spectrogram, otherwise it would sound as if it was Paulstretched with time factor of 1 and time resolution of 1÷(sample rate÷window size) (for default settings, 0.0058). So why would what I see is what I get with spectrogram?
I didn’t know that audacity actually loads lower latency when more zoomed in. Thanks.
Bump - When you switch to spectrogram and change to Reassignment, but Audacity temporarily crashes when the song ends…