Crackles in the reverb

So. I tend to get crackles in the reverb from LX-24. Seems to get worse with transient sounds, but even slow changes in amplitude seem to trigger it. CPU is at below 10%, so unlikely to be CPU-related. Doesn’t happen with other reverbs, so it’s not a “click” in the sound that gets smeared by the delay lines. This is on VST3, Ableton Live on M1 Air. Anybody else experiencing this?

Hi @pinkyfluff welcome to our Sound Explorers community!

Sorry to read this, in order to help you better please log in to your account to contact our support team that will help you to fix this situation.

Best :zap:

Ps
If you haven’t already, then please make sure to subscribe to our Newsletter to be the first to know about any new and exciting news from Arturia.

1 Like

Hi @pinkyfluff, Can you say whether the same crackling sounds are heard if you choose another algorithm (1) or another converter (2) in the Rev-LX?

Cheers,

2 Likes

Have emailed support.

But yes. It changes with different algos and converters, I think it’s the worst with LargeConcertHallB and Modern combination, but as far as I can tell it happens in all combinations, it’s just less obvious and less frequent in some.

Hi all,

@pinkyfluff - Does it mainly have to do with the PITCH SHIFT and the MODE ENH settings?

No, they might change things, but not even sure they do that. Happens regardless of settings…

I see.
By me the Pitch Shift can add crackles/ noise. The amount depends on the settings.

A sound example perhaps could help.

Are the 10% CPU usage you mention for the whole CPU or for the single core that have the highest load?

If the crackles is subtle, then perhaps it’s about the emulation also have the original reverbs imperfections.
A way to choose to avoid this would in that case be great to have.
But i really cab’t tell.

Just what i can think of right now. Perhaps Arturia find something.

Measuring CPU load correctly is always tricky. The number I referred to was the highest core load, but even that is an average at best. The exact number isn’t that important, just that the CPU isn’t under heavy load in the test project (so, if the problem is in fact due to intermittent CPU spikes that doesn’t show up on meter/as overloads, it’s still the LX-24 causing them).

There’s an audio example (and screen recording of the session recording the audio) here:

I’m on Windows.

If you for example have a 12-16 core/ thread CPU, then one core can be overloaded, if the full CPU load is 10%. Each core can’t run 100% of the full CPU capacity each. So it’s important to know if the CPU usage is for the full CPU capacity or for the CPU core that have the highest load.

Thanks for your audio example and Screen recording. Both are helpful.

I understand your screen recording show the audio recording you have posted, so no other parameters are changed during the audio recording.

What happens if you set the DECAY OPT to a high level or turn it off completely by clicking on the parameter text?

Yeah. CPU-load is reported differently on OSX. The “full” load is the sum of all cores, so you can have 358% cpu load and still be completely fine, which is confusing in other ways (or you can have 101% load overall and have one core at overload). But, yes, I’m aware that it’s possible to overload a single core (especially in audio, where each linear chain of plugins, on a channel for instance, generally needs to be calculated on a single core).

Correct. The audio is what’s being recorded as a clip on channel 2, so nothing going on other than what’s visible.

Anyway. Thanx for the suggestion! Turning DECAY OPT off, as opposed to just turning it up/down, actually seems to solve most/all of the problems. They tend to come back in another shape with high pitch shift, but that’s easy to just not use. =) (But it’s also pretty strange, I don’t think it should sound like a blender filled with gravel even at high values)

By me 100% CPU usage is max. A core can have a 100% CPU usage in it self, but it will not appear as 100% in the full CPU usage in Windows. Perhaps it’s a Windows VS MAC thing.
DAW’s show this in different ways. My DAW show the usage the most loaded core use. Some hosts i believe show the load of the full CPU. I prefere the first.

The sound you use has an impact as you say. The DECAY OPT act depending on the sounds. But i also don’t understand why it clicks.
I also don’t understand the Pitch Shift clicks at least allready at so low values. I assume it’s about “Grains”.
The manual unfortunately does’nt explain much aboout this.

Hopefully @niks.arturia are on the case and will come with explanations.