strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Feedback and bug reports for Audacity 3.0.3 alpha
Forum rules
This forum is only for feedback and bug reports for the specified version of Audacity.

All bug reports must include the following information:
  • Exact version number of Audacity
  • Operating system version
  • Brief description of the problem
  • Detailed, step by step instructions for how to reproduce the problem
Bug reports that do not include the above information will be deleted

Additional information and / or screenshots should be included as necessary to adequately describe the problem.
Post Reply
broussaye
Posts: 15
Joined: Sat May 23, 2020 7:35 pm
Operating System: macOS 10.14 Mojave

strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by broussaye » Thu Mar 11, 2021 5:52 pm

Hello everyone,
I observed a strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06, I don't know if it is a normal behaviour or a bug.

I'm using Macbook pro 15 inch (2018) Mojave (10.14.6) 2,6 GHz Intel Core i7, 32 GB 2400 MHz DDR4

1- Empty project, generate->Nyquist->risset drum, with default effects

2-Without deleting the first sound obtained, I redo the same effect by increasing the "delay" of a few seconds

3-I observe a break in the time of the first effect and a gap of a few milliseconds

4-I do the same operation again, increasing the delay again, I observe a second cut-off at the time of the second effect (and a gap) and so on...

5-I switch to spectrogram mode and this is what I get.

Is this a bug or a normal behaviour, thanks.
And thanks for the work of all, developers, moderators of the site, and those who help on the forum.
Attachments
1- Empty project, generate->Nyquist->risset drum, with default effects.jpeg
1- Empty project, generate->Nyquist->risset drum, with default effects.jpeg (216.48 KiB) Viewed 430 times
2-same effect by increasing the "delay" of a few seconds.jpeg
2-same effect by increasing the "delay" of a few seconds.jpeg (234.99 KiB) Viewed 430 times
3-4_a gap of a few milliseconds.jpeg
3-4_a gap of a few milliseconds.jpeg (155.21 KiB) Viewed 430 times
5-spectrogram mode.jpg
5-spectrogram mode.jpg (79.97 KiB) Viewed 430 times

steve
Site Admin
Posts: 81379
Joined: Sat Dec 01, 2007 11:43 am
Operating System: Linux *buntu

Re: strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by steve » Thu Mar 11, 2021 6:20 pm

The "gap" in this image:


Image


isn't actually a "gap".
The dots represent the temporal positions of the samples. The lines joining the dots are imaginary - just a graphic aid to help visualise the analog waveform. You've just zoomed in close on a "split" between two audio clips (not a bug, perfectly normal)
See: https://manual.audacityteam.org/man/aud ... clips.html
9/10 questions are answered in the FREQUENTLY ASKED QUESTIONS (FAQ)

broussaye
Posts: 15
Joined: Sat May 23, 2020 7:35 pm
Operating System: macOS 10.14 Mojave

Re: strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by broussaye » Thu Mar 11, 2021 6:35 pm

steve wrote:
Thu Mar 11, 2021 6:20 pm
The "gap" in this image:


Image


isn't actually a "gap".
The dots represent the temporal positions of the samples. The lines joining the dots are imaginary - just a graphic aid to help visualise the analog waveform. You've just zoomed in close on a "split" between two audio clips (not a bug, perfectly normal)
See: https://manual.audacityteam.org/man/aud ... clips.html
Hi Steve,
Thank you for your quick response,
Maybe I don't understand everything correctly, I'm just an amateur who likes to play with sounds, I have already observed this hole when I split an audio clip in two but in my example there are not two audio clips but only one, and the spectrogram view shows something (??)
thanks
Attachments
new spectogram.jpg
new spectogram.jpg (98.95 KiB) Viewed 413 times
Last edited by steve on Thu Mar 11, 2021 7:25 pm, edited 1 time in total.

broussaye
Posts: 15
Joined: Sat May 23, 2020 7:35 pm
Operating System: macOS 10.14 Mojave

Re: strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by broussaye » Thu Mar 11, 2021 7:06 pm

I'm sorry I'm not used to the forums (bad manipulation on my part), the last image I sent corresponds to the spectogram view after having clicked on the cut line, as if to join two audio clips (normal behaviour with audacity) (but in my example there is normally only one audio clip) but I observe that the spectrogram is different...???

steve
Site Admin
Posts: 81379
Joined: Sat Dec 01, 2007 11:43 am
Operating System: Linux *buntu

Re: strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by steve » Thu Mar 11, 2021 7:40 pm

The spectrogram is calculated separately for each audio clip, thus a split will appear as a discontinuity in the spectrogram, even if the waveform continue seamlessly from one clip to the next.

First Track000.png
First Track000.png (51.84 KiB) Viewed 387 times
9/10 questions are answered in the FREQUENTLY ASKED QUESTIONS (FAQ)

broussaye
Posts: 15
Joined: Sat May 23, 2020 7:35 pm
Operating System: macOS 10.14 Mojave

Re: strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by broussaye » Thu Mar 11, 2021 8:26 pm

Thank you Steve for your explanations,
If I understand your example correctly, if you join the two regions, the spectrogram will be transformed as in my example (the image "5-spectrogram mode" becomes "new spectrogram" when I have joined the "two" clips)

Leaving aside the "hole" that is not a hole, what about the cut-off lines that appear at the location of the previous effect?

Thanks

steve
Site Admin
Posts: 81379
Joined: Sat Dec 01, 2007 11:43 am
Operating System: Linux *buntu

Re: strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by steve » Thu Mar 11, 2021 9:30 pm

broussaye wrote:
Thu Mar 11, 2021 8:26 pm
what about the cut-off lines that appear at the location of the previous effect?
That's normal. Other generators do the same - try it with the "Chirp" generator.
9/10 questions are answered in the FREQUENTLY ASKED QUESTIONS (FAQ)

broussaye
Posts: 15
Joined: Sat May 23, 2020 7:35 pm
Operating System: macOS 10.14 Mojave

Re: strange behavior with generate->Nyquist->risset drum with audacity 2.4.2 and 3.0.0 rc06

Post by broussaye » Fri Mar 12, 2021 6:45 am

Hi Steve,
Thanks again for the time you spend helping people, and giving explanations to neophytes like me.

As you suggested I did the test with the "chirp" generator, on Audacity 2.4.2 I observe the same behaviour when I increase the time of the effect (cut-off line at the end time of the previous effect) whereas with 3.0.0 rc06 this cut-off does not appear.

In fact I tried all the built-in generators and I observe the same behaviour as with the "chirp" generator (2.4.2 vs 3.0.0), on the other hand with the Nyquist generator it is the same behaviour on both versions (cut-off at the end of the previous effect).

So I feel like it is an improvement...
Thanks

Post Reply