Renoise EQ crackling

Just found a strange thing with the Renoise EQ. In the attached files there’s only a kick on track 1 and one EQ and a dip at 100 Hz.
When you loop the beat for a while you will hear some crackling noise that comes and goes. I assume it’s caused by the EQ which phase shifts the signal which leads to some kind of beat frequency. I tried another EQ from Overtone and with this I can’t hear it. Anyone ever stumbled over this?

crackles_EQ.xrns (17.3 KB)

How long is “a while”? I looped it for about a minute or so but didn’t hear any crackling.

1 Like

In the file I uploaded it starts exactly on the third loop, i.e. at 15 seconds.

mark the last few bits of the kick sample in the sampler and use the fade out function or download the altered xrns attached. crackles won’t appear no longer - with or without EQ10

crackles_EQ_edit.xrns (17.2 KB) .

1 Like

Thanks! It’s still happening with your edited file, but it’s better. Do you know why this happens in general? I don’t see why it changed over time.

the sound of the crackles reminded of a sample not ending at zero crossing, hence the approach of fading out the last bits of the waveform, in order to make sure that does not happen.
i’m not certain why this becomes audible with EQ10 enabled only. maybe the DSP causes a slight DC offset for some reason.

3 Likes

Ok, thanks. I’ll give it another try tomorrow.

one min to appreciate the master is here :pray:
keith 303

1 Like

:joy: Amen

1 Like

Sorry to bump this old thread, but I ran into this same problem, and after a long experimentation session, I do think it may actually be considered a bug.

Here’s the thing; if you bump a low frequency, and look at the result using an oscilloscope, you’ll notice the sample gets a longer tail with the low frequency. If this tail is longer than the sample length, it just cuts off and you get the click/crackle. (so yes, it’s fixable by just adding silence to the end of the sample.)

My theory; if I understand correctly, an EQ converts the audio (using a Fourier transform) from the time domain into the frequency domain. It then alters the amplitudes, and then converts it back to the time domain. In this final conversion, it seems these low frequencies go on for too long.

After all, where there was silence in the original, even after boosting a frequency using an EQ, there shouldn’t be any sound in the result, no? Yet, there is!

Now I’m not a mathematics professional, so I hope someone can confirm whether this is in fact a bug. It does feel like a bug, since (a) other EQs don’t cause this problem and (b) Renoise ‘should’ at least add room to the sample internally when applying an EQ that lengthens the sound duration :slight_smile:

At least there is a workaround, just add some silence to the end of the sample. The lower & harder the frequency boost, the longer the silence you need to add. Cheers!

2 Likes

yeah, post-ringing is a real thing and will impact phase and duration of low frequencies the most. Filters and EQ will both introduce post-ringing, while linear phase EQs will introduce pre-ringing. If you must filter or eq foundational elements of your low end, it can sometimes become necessary to resample post eq, so your tails aren’t all fucky. This is a bigger deal in some styles of music (esp psytrance), where you want a really clean low end that doesn’t have post-ringing in your bass bleeding over into your kick (and bass) transients.
Kind of cool to see how the post-ringing gets cut off by the termination of the sample, and I think you’re right @bitplanes, it’s like the pop you get when a sample doesn’t terminate at a zero-crossing.
Not sure I would consider it a bug, maybe just something to recognize in terms of how renoise handles sample playback… the end is, in fact, the end

2 Likes

Thanks for your insight, so as I understand you this is a ‘normal’ effect of EQs, I did not know that! The more you know :slight_smile:

I guess I would still call it a bug that it cuts off at the sample end; not to be mean to the Renoise devs, but mostly because users who do not know this EQ side effect, will be VERY puzzled, until they understand what’s going on, if they ever will at all - I mean it took me about an hour of debugging until I had a theory of what was going on, and not to be arrogant but I think most people will have given up by then.

I probably would have too, if I didn’t have the visual insight through the oscilloscope VST I had just installed the day before, because I happened to see it on your youtube video @slujr - isn’t that a funny and beautiful coincidence, haha :star_struck:

1 Like

Ha! Yeah, a funny synchronicity :slight_smile:
Oscilloscopes are honestly essential for fine tuning and detail work. I wish we could change the time window on the track scopes. They would be much more useful if we could. Especially if they could be synced to tempo as well…

1 Like

I also agree that this is not a bug per se, and more like an edge case that should absolutely be addressed and improved.

1 Like