[MXXX 7.01] Granular Freeze etc !

[farfadetfarfelu]

"Restore when disabled" seems to work well in the modulator window, very useful when you want to use multiparameters to activate modulators for very weird sounds, and then you want everything to go back to normal... perfect.

I love the freeze function in the granular, i had asked for it with the delay but it's certainly more powerful in the granular with the reverse and other transformations than can be applied on the freezed audio !


Now here are a few strange things i noticed with the freeze :
- In MXXX, if you freeze something in the granular and then with the freezed sound playing, you change anything in the routing of the fx blocks : the freezed audio stops playing back, it's lost.
Edit : well it seems the same with the reverb and delay... so i guess it's normal.


I'm still interested to see the freeze implemented in the delay because the delay does different sounds, like you can wait for a feedback buildup and then freeze the wall of sound, or you can modulate the delay and then freeze it... but plenty of stuff is covered with the freeze in the granular, really cool, thank you so much.

[farfadetfarfelu]

Okay i have another request for the delay, the addition of a "Left Offset" parameter similar to the right offset.
It might seem strange but let me try to explain.
I'm trying to build a modulation delay.

It sounds very different when modulating the 10 sec delay time or when modulating the 100 ms "Right offset" parameter.

Modulating the Right offset parameter is easier because the timeframe is smaller and it sounds better.
But the problem : if i'm correct, it works only on one side of the stereo field + Each time is passes through zero it makes a clicking sound.
So i modulate from 0.1 to 100, but then there is only modulation on the right side of the stereo field...
So i'm pretty sure having a seperate left offset and right offset will allow for huge stereo modulated delays with the possibility to have different modulation on left and right....!!

[MeldaProduction]

Thanks!

It actually is normal. When you change the layout, all processors must be reinitialized and therefore you start over. I'll add the delay freeze to to-do list, though I'm not sure how much useful this will be .

[farfadetfarfelu]

Oops, we posted at about the same time last night so you probably missed my post above.

I'll do more testing with the delay tonight.

[MeldaProduction]

Ojojoj, ok, I'll check

[farfadetfarfelu]

Okay here is a video to illustrate what i'm saying : HERE

Modulating the Right offset is more practical than the bigger delay time, and it sounds cool BUT it works only on one side of the stereo, so you have to use two delay modules...

Suggestions for the delay :
1- Add a left offset parameter to allow stereo modulation (possibility to modulate left and right differently)
2- It might still be cool to have the option of letting the non-modulated delayed sound go through as shown in the video after 1"42 min.

3- Freeze button !
4- Saturation (most of your other effects have it)
5- Maybe a Higher Q on the filter to do more radical filtered delay sounds, but how does this filter work, how is the frequency set ?
6- There is a huge volume difference between default and classic delay modes (classic is louder), this is annoying when comparing.

MmultibandDelay could benefit from this also.

[MeldaProduction]

Ok, it's interesting indeed .

So,
left offset - ok,
nonmod sounds - rather not, the plugins is getting too big...
freeze - ok,
saturation - ok,
higher Q - no because of backward compatibility and actually it was not possible, because it was getting unstable
volume differences - not possible, that's just the way these 2 modes work, also backwards compatibility

[farfadetfarfelu]

Ok, it's interesting indeed .

So,
left offset - ok,
nonmod sounds - rather not, the plugins is getting too big...
freeze - ok,
saturation - ok,
higher Q - no because of backward compatibility and actually it was not possible, because it was getting unstable
volume differences - not possible, that's just the way these 2 modes work, also backwards compatibility

Cool, should be great !

I have yet again more stuff for you :
- i think there is something strange going on with the "even harmonics" in the saturator module, it is very VERY loud to the point of being impossible to use. It's a lot louder than the same parameter in the Mmultibandsaturator and it sounds different too.
Maybe that's a result of the change of the dry/wet.

- Strange BUG, when using a bandpass in a feedback loop at zero delay, clicking on the bandpass processor in the grid can make it behave strangely, it's even more obvious with several band pass processors... video HERE

[MeldaProduction]

Ok, to the "bugs":

1) saturator - I checked and found no problems. Any video?

2) bandpass with feedback - well, what you did is actually you created a very unstable filter . You know, feedback is a VERY dangerous thing. Basically any filter (IR at least, like the bandpass for example) is essentially a set of "1 sample delayed" feedbacks. You destroyed the whole filter structure by making it a feedback in a feedback. Making such a thing unstable is a piece of cake really . So, it's not a bug, it's an inevitable result of a really hardcore processing .

[farfadetfarfelu]

Ok, to the "bugs":

1) saturator - I checked and found no problems. Any video?

Okay then, video of the Even HArmonics thing : HERE

I'm pretty sure there's something wrong.
Sorry i'm still using the same sound sample lol.

2) bandpass with feedback - well, what you did is actually you created a very unstable filter . You know, feedback is a VERY dangerous thing. Basically any filter (IR at least, like the bandpass for example) is essentially a set of "1 sample delayed" feedbacks. You destroyed the whole filter structure by making it a feedback in a feedback. Making such a thing unstable is a piece of cake really . So, it's not a bug, it's an inevitable result of a really hardcore processing .

I tried using filer processors with HP selected instead of the "bandpass" processors, and it does not have that strange behaviour of cuting of or changing settings, it's strange it only happens with the LP/HP in the bandpass and not the LP/HP in the filter ?

[MeldaProduction]

1) Aaaaha, the difference is, that the saturator in MXXX doesn't clip it, so if you want to sound the same, you need to put some clipper afterwards - e.g. waveshaper. Maybe the optional clipper could be added, I'll add it to to-do list.

2) Well, it may sound different as the algorithms are different, but trust me, feedback with one sample delay is a bad idea, especially with any filter in it.

[farfadetfarfelu]

Well, imo it's more usable with the clipping.
And from now on i'll avoid using bandpass with less than 2 sample feedback, no big deal.

I've been doing more testing of the "restore when disabled", sometimes it works very well, sometimes it has more trouble.
For example i have mod 1 (follower) controling the freq of a filter (in the eq), it returns to the start value maybe 50% of the time.

VIDEO

That's all for today, now i'm off to try stuff with the vocoder...

[MeldaProduction]

Yej I'll check it out...

About the feedback - it's not just 1 sample, it is just getting close to 0. For example, 120dB/oct LP takes I think 20 taps, so having feedback below 20 samples is a problem.

[MeldaProduction]

Weird thing here - I couldn't verify the "restore when disable"...

[MeldaProduction]

Aaaaha, it indeed does that, but only with the equalizer...