Bazille delay section generates some clicks on tempo change

[Tchandra]

Here is an exemple with the preset "ST distorted reality" (in the "keys" preset section), you can here the clicks sent in the feedback loop :

http://pat.prajna.gouriet.free.fr/demo/ ... eClick.mp3

with Bazille delay turned off :

http://pat.prajna.gouriet.free.fr/demo/ ... oDelay.mp3

With the delay turned on, but without project tempo change :

http://pat.prajna.gouriet.free.fr/demo/ ... Change.mp3


I use reaper v5 and the last version of Bazille (rev 3898), no external effect applied. The tempo change is very small (125 to 128 on both cases) - I use this kind of little tempo change because I like to record something first, and if I like the groove, I use reaper tempo mapping tools to have the metronome grid matching the rhythm of what I have just recorded. But here, it causes some glitches... not everytime though.

I have slightly changed the preset : the "left" button on the delay section is set to 7 instead of 4, to make the effect more noticeable. I have tried the multicore and HQ buttons and it has no effect.

Unfortunately, there is no way to set the delay time to a fixed value in ms, it is always synchronized to the host tempo. I have found a way to avoid these glitches with 2 automation curves, I set to O the Center and Side volumes of the delay section just around the tempo change, that way I get a clean sound, so it won't ruin my project ^^ but I wanted to report this issue

[Tchandra]

I use the 64 bits vst version of Bazille, win7, an intel i5 proc with 8go of ram and a SSD.

[Urs]

We've discussed this internally, but we have no immediate fix for this I'm afraid

It's a tempo synced delay. Tempo synced delays glitch on tempo changes.

While it doesn't necessarily have to be clicks, there certainly would have to be some sort of glitch. We'll evaluate some other ways of dealing with this, and try to figure out the lesser evil.

[EvilDragon]

Isn't that a matter of interpolating delay buffers somehow when delay time (ergo, tempo) changes? I'm talking from my ass though... I remember some talks back at NI forums about glitching vs non-glitching delay building in Reaktor...

[Tchandra]

Personally, an option to set the delay time in ms would be fine...

In my project, I have set the volume of Bazille delay at a lower value than in the examples above, and blended it with another delay plugin and a reverb, so the echos are more subtle and fluid, and the glitches are more difficult to notice, but if I add a vari-mu type compressor to keep the level between some boudaries, these klicks can become really louder and easily noticeable, so it can be a problem in my case...

Well I might disable the Bazille delay and use an external plug-in for that...

Thanks for your reply.

[Urs]

Isn't that a matter of interpolating delay buffers somehow when delay time (ergo, tempo) changes? I'm talking from my ass though... I remember some talks back at NI forums about glitching vs non-glitching delay building in Reaktor...

Yes, one can smooth out tempo changes but then the delay changes pitch for a while. An alternative might be crossfading between two sets of taps...

[Chris-S]

Or one clears the delay buffer on each tempo change. Then you have no glitches (but also no echos for a while).

Or you simply ignore the tempo change.

So many options....

[hakey]

Real world analogue delays change pitch with tempo changes. It's an interesting effect in itself.

[Tchandra]

Real world analogue delays change pitch with tempo changes. It's an interesting effect in itself.

Well, a kind of "glide" sound, or any sounds produced by modulating the delay parameters can be interesting, but clicks just sound bad imo.

[hakey]

I was replying to Urs' comment that "one can smooth out tempo changes but then the delay changes pitch for a while".

I agree, clicks are bad.

[aaron aardvark]

I just happen to be working on a song in Cubase LE4 that changes in tempo a few times. The verse is at 105 beats per minute, and the chorus is at 115 BPM (instant tempo change going from verse to the chorus). I have lead guitar at the end of the chorus and I wanted to transition back into the 105 BPM verse without it sounding terribly slow. I have a few seconds of silence except for a long echo delay (tempo synced) coming off the lead guitar. It is a (from Cubase) digital delay, though it has a slight analog quality to it, I assume caused by the ramped down tempo change (that takes places over those few seconds). There are no clicks though. Perhaps a gradual tempo change would cause less problems.

[exmatproton]

Real world analogue delays change pitch with tempo changes. It's an interesting effect in itself.

Well, a kind of "glide" sound, or any sounds produced by modulating the delay parameters can be interesting, but clicks just sound bad imo.

I use clicks all the time. Love to glitch with digital delays. For me, the best solution would be a switch to be able to use time-based/tempo-sync delay and ms.

[exmatproton]

I just happen to be working on a song in Cubase LE4 that changes in tempo a few times. The verse is at 105 beats per minute, and the chorus is at 115 BPM (instant tempo change going from verse to the chorus). I have lead guitar at the end of the chorus and I wanted to transition back into the 105 BPM verse without it sounding terribly slow. I have a few seconds of silence except for a long echo delay (tempo synced) coming off the lead guitar. It is a (from Cubase) digital delay, though it has a slight analog quality to it, I assume caused by the ramped down tempo change (that takes places over those few seconds). There are no clicks though. Perhaps a gradual tempo change would cause less problems.

No. Slow tempo progression would stretch and shrink the audio. It is sample based, so maybe the effect seemes less clicky, it is still there. (And i love that )

[aaron aardvark]

exmatproton,
If it is there, I can't hear any clicking. If I can't hear it, it might as well not exist as far as I am concerned.