View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000667 | Cinelerra-GG | [All Projects] Bug | public | 2025-01-04 21:47 | 2025-01-06 19:44 |
| Reporter | Nafnlaus | Assigned To | PhyllisSmith | ||
| Priority | urgent | Severity | major | Reproducibility | always |
| Status | acknowledged | Resolution | open | ||
| Product Version | 2020-08 | ||||
| Target Version | Fixed in Version | ||||
| Summary | 0000667: Random audio pops make using Cinelerra-GG for mastering a maddening task | ||||
| Description | Version: CinGG-20241031-x86_64.AppImage (Fedora 41) Certain things - mainly filters - and in particular the freeverb and graphical equalizer filters (two of the most important when mixing audio) can add subtle (or sometimes not so subtle) pops into the audio. They're incredibly difficult to track down, and it's so much work to isolate them that before now I've not spent the many hours to try to narrow down a fill project into a simple test case. It's to the point where I'm seriously thinking about just giving up on Cinelerra-GG for music; I can't keep having the program ruin my tracks by introducing random pops :( I also do not believe that they represent a single problem, but rather multiple problems. In the process of submitting this, due to your incredibly low file size limit (come on, this is a bug report page for a media tool!), I tried converting my samples to m4a encoded with ffmpeg, and the problem *didn't* occur, whereas on FLAC it did (then again, these things are so random, who knows?). But the behavior is different for different types of bugs - for example, with freeverb, the pops seem most common when it's shifting parameters from one keyframe to the next (for example, amping up the wetness). In none of these cases is there any clipping. | ||||
| Steps To Reproduce | 1. I had to split loop_concat.flac to up load it here, so first mv loop_concat.flac.pt1 loop_concat.flac, then cat loop_concat.flac.pt2 >> loop_concat.flac 2. Open the xml. Note the marker. There will either be a pop here or not. You will *never hear the pop* just from playing the loop concat track in the interface, only via rendering (and possibly only via rendering to flac?) (note: when I say flac, I mean the flac option, not the ffmpeg suboption) 3. If you render the in-out points with the equalizer on, it will create the pop. If you shut the equalizer off, the pop will go away. Note that the equalizer is set to 0 - it should do absolutely nothing. But that's clearly not the case! 4. Other things that can make it go away are changing the render positioning (such as rendering a shorter section around the pop point) or removing the first chunk of the loop concat (before the in point, e.g. what we're not even rendering!) 5. I have included examples of it rendered with and without the bug. On close inspection, the pop seems to occur when cinelerra is - for some reason - getting skew between the left and right channels. I'm not sure if that's the /only/ thing going on, but it's certainly /something/ going on. See the png. | ||||
| Additional Information | As noted earlier, I don't believe that this is the /only/ pop bug. For example, I think the freeverb pop bug is an entirely separate thing, and there might be even more. But these render Cinelerra practically unusable for music to anyone who cares about mastering quality :( Just now, I tried rendering my main project (from which this highly simplified example was derived) to wav. But the pop was still there. I then tried converting loop_concat in my main project to wav, and rendering to wav. That particular pop seems to have vanished. So this particular one might be a flac-source bug rather than a flac-output bug. Since flac and wav should be identical (since flac is lossless), to debug this issue, you might want to have two different versions, one flac and one wav, and (A) see if they load identically, and if they do, (B) see where the processing of them starts to differ, and why. | ||||
| Tags | No tags attached. | ||||
 |
@Nafnlaus : Acknowledging that we have seen this 667 and 668 and will take a look at the files/issue. |
 |
Probably relates to: https://www.cinelerra-gg.org/bugtracker/view.php?id=668 |
|
|
loop_concat.flac.pt2 (3,723,050 bytes) d3cpcpr_withbug.flac (3,429,313 bytes) d3cpcpr_bug.xml (17,145 bytes) d3cpcpr_withoutbug.flac (3,429,310 bytes) loop_concat.flac.pt1 (3,072,000 bytes) |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2025-01-04 21:47 | Nafnlaus | New Issue | |
| 2025-01-04 21:47 | Nafnlaus | File Added: d3cpcpr_bug.xml | |
| 2025-01-04 21:47 | Nafnlaus | File Added: d3cpcpr_bug.png | |
| 2025-01-04 21:47 | Nafnlaus | File Added: d3cpcpr_withoutbug.flac | |
| 2025-01-04 21:47 | Nafnlaus | File Added: loop_concat.flac.pt1 | |
| 2025-01-04 21:47 | Nafnlaus | File Added: loop_concat.flac.pt2 | |
| 2025-01-04 21:47 | Nafnlaus | File Added: d3cpcpr_withbug.flac | |
| 2025-01-05 20:31 | Nafnlaus | Note Added: 0005729 | |
| 2025-01-06 19:44 | PhyllisSmith | Assigned To | => PhyllisSmith |
| 2025-01-06 19:44 | PhyllisSmith | Status | new => acknowledged |
| 2025-01-06 19:44 | PhyllisSmith | Note Added: 0005730 |
