View Issue Details

IDProjectCategoryView StatusLast Update
0000668Cinelerra-GG[All Projects] Bugpublic2025-01-25 17:19
ReporterNafnlaus Assigned ToPhyllisSmith  
PriorityurgentSeveritymajorReproducibilityalways
Status acknowledgedResolutionopen 
Product Version2020-08 
Target VersionFixed in Version 
Summary0000668: Random audio pops make using Cinelerra-GG for mastering a maddening task (0000002)
DescriptionProbably relates to: https://www.cinelerra-gg.org/bugtracker/view.php?id=667
Version: CinGG-20241031-x86_64.AppImage (Fedora 41)

This is possibly the same bug as the last one, only more egregious (though as it's more egregious, it was thankfully easier to track down and isolate). One track (no_vocals.flac) plays. It has two neutral equalizers. The second track (vocals_new.wav) *is muted* and should have zero impact; it shares one equalizer with the first track. Yet when the second (muted) track begins, there is a pop. If you do anything - shut off either or both equalizers, or disable play for the muted track, or outright delete it - the pop goes away.

Two rendered examples are also included in the .xml file, one with the bug and one without the bug.

Note that unlike the last bug, this one doesn't need to be rendered to appear - it occurs even when just playing the track in Cinelerra.

Like with the last bug, since your maximum upload size is a mere 4,77MB, tracks are split and need to be catted back together to be recombined.
Steps To ReproduceSee above.
Additional InformationNone
TagsNo tags attached.

Activities

PhyllisSmith

PhyllisSmith

2025-01-25 17:10

manager   ~0005748

Last edited: 2025-01-25 17:19

View 2 revisions

@Nafnlaus
@Andrew-R
There is a native audio filter called Spectrogram in CinGG, but it has a lot of variables so I do not know how to use it.
The FFmpeg showspectrum and showspectrumpic have been commented out in CinGG due to input/output error. I will see if I can find variables that make them work (the problem may be that it requires a file name to put the output on, like a png, and Cinelerra has no way to interact with that)
There is also commentary in 667, but I have tried the FFmpeg flac with the same results as the native flac for loading AND rendering.

Note to Andrew specifically -- I converted loop_contact.flac to loop_contact.wav via external ffmpeg, but have not been able to determine if bug still there by comparing the waves because the original audio wave is different (still working on this though). Actually in 667, Nafnlaus says "converting loop_concat in my main project to wav, and rendering to wav "does NOT have the problem. But if he uses the flac and renders to wav, it does have the bug.

Nafnlaus

Nafnlaus

2025-01-25 14:38

reporter   ~0005747

There is no big FFMpeg button in the main window, but I did find what you were referring to (a tiny button in the upper right of green, white and black diagonal lines). No impact.

I did try replacing no_vocals with the output of:

sox -n -r 44100 -c 2 sine.flac synth 480 sine 440 sine 445 vol 0.2

The pop persists, but it doesn't make the pop more visible. The easiest way to see the pop is on a spectrogram, such as from Audacity, but there's no option in Cinelerra for that.
Andrew-R

Andrew-R

2025-01-25 13:07

reporter   ~0005746

@Nafnlaus

by decoder preferences I mean big FFmpeg button in main window.
You can click it and Cingg will try to use built-in (old) flac decoder instead of ffmpeg's one.


Can you retry with simple artifically generated sine wave? I hope we can guess what going on by looking at how deformed this simpke waveform become.
Nafnlaus

Nafnlaus

2025-01-25 12:55

reporter   ~0005745

@Andrew-R Oh, sorry - I was just checking up these bugs, and I noticed that you replied weeks ago; I never saw the notification!

The bug not only appears at other sample rates, but the pop becomes more prominent at e.g. 48k than 44,1k.

I don't know what you mean by "change decoder preference". Are you talking about rendering? This bug (unlike the other one) occurs even during playback within the app, not just during rendering.
Andrew-R

Andrew-R

2025-01-06 22:13

reporter   ~0005733

@Nafnlaus, does bug apper at another project samplerates?
I think files will be smaller if it occur all way down to 8khz or so.

Also, does bug occur if you change decoder preference away from ffmpeg first?
Nafnlaus

Nafnlaus

2025-01-05 13:10

reporter  

no_vocals.flac.pt7 (738,410 bytes)
vocals_new.flac.pt5 (4,696,706 bytes)
vocals_new.flac.pt4 (4,608,000 bytes)
vocals_new.flac.pt3 (4,608,000 bytes)
vocals_new.flac.pt2 (4,608,000 bytes)
no_vocals.flac.pt4 (4,608,000 bytes)
no_vocals.flac.pt5 (4,608,000 bytes)
no_vocals.flac.pt6 (4,608,000 bytes)
altman_bug.xml (12,507 bytes)
no_vocals.flac.pt1 (4,608,000 bytes)
no_vocals.flac.pt2 (4,608,000 bytes)
no_vocals.flac.pt3 (4,608,000 bytes)
altman_withbug.wav (624,792 bytes)
altman_withoutbug.wav (624,792 bytes)

Issue History

Date Modified Username Field Change
2025-01-05 13:10 Nafnlaus New Issue
2025-01-05 13:10 Nafnlaus File Added: altman_withbug.wav
2025-01-05 13:10 Nafnlaus File Added: altman_withoutbug.wav
2025-01-05 13:10 Nafnlaus File Added: altman_bug.xml
2025-01-05 13:10 Nafnlaus File Added: no_vocals.flac.pt1
2025-01-05 13:10 Nafnlaus File Added: no_vocals.flac.pt2
2025-01-05 13:10 Nafnlaus File Added: no_vocals.flac.pt3
2025-01-05 13:10 Nafnlaus File Added: no_vocals.flac.pt4
2025-01-05 13:10 Nafnlaus File Added: no_vocals.flac.pt5
2025-01-05 13:10 Nafnlaus File Added: no_vocals.flac.pt6
2025-01-05 13:10 Nafnlaus File Added: no_vocals.flac.pt7
2025-01-05 13:10 Nafnlaus File Added: vocals_new.flac.pt5
2025-01-05 13:10 Nafnlaus File Added: vocals_new.flac.pt4
2025-01-05 13:10 Nafnlaus File Added: vocals_new.flac.pt3
2025-01-05 13:10 Nafnlaus File Added: vocals_new.flac.pt2
2025-01-06 19:42 PhyllisSmith Assigned To => PhyllisSmith
2025-01-06 19:42 PhyllisSmith Status new => acknowledged
2025-01-06 22:13 Andrew-R Note Added: 0005733
2025-01-25 12:55 Nafnlaus Note Added: 0005745
2025-01-25 13:07 Andrew-R Note Added: 0005746
2025-01-25 14:38 Nafnlaus Note Added: 0005747
2025-01-25 17:10 PhyllisSmith Note Added: 0005748
2025-01-25 17:19 PhyllisSmith Note Edited: 0005748 View Revisions