Elektron-Users

Hey !

A former topic rose the sample chopping / slicing question with a MDUW. (btw, is there any difference between "chopping" and "slicing" ?...) As I think it deserves its own topic, here I am again.



I tried it with the parameter locks on the START and END paramaters : it works. Though, for my purpose, it was better to lock only the START parameter and to set the HOLD parameter to 8 (the duration of a step, actually). But this way is a bit too long. I can hear your "Well, once in a kit, it's ok"... but I also want to make random slice triggering, etc.

So my second idea was to program a G2 patch, MIDI sync both MDUW and G2 and allez hop. Kind of gas factory, as say french people : many cables for something not so complicated.

Then I remembered the MDUW LFOs. I found a way to achieve -almost- what I want. Compared to the G2 solution, I am not be able to trigger slices randomly beginning at 16th beats only. Every randomly triggered slice can start anywhere in the sample. Using the G2, one can solve that minor problem.

Anyway... just considering simple slicing (without moving or swapping slices), I encountered a problem. I could not get a perfectly continuous playing of the sample. There is always a kind of gap, not always at the beginning.


Here are the settings I remember (I have not my MDUW on right now) :

RAM RECORDING machine :
length : 64.

RAM PLAYING machine :
LFO 2 set to ascending saw (Mix at 128),
Rate = 2,
Depth = 64, or 63
Dest = Start parameter of the current track
Trig mode = Hold or Trigger

START = 64
END = 128
HOLD = 8

I think that is all the necesary.



Can anybody tell me where I am wrong ?...

i asked a variant of this in another thread, got no responses. i was trying to do something similar, by placing a trigger for a loop on every beat, then using the LFO to randomly have the sample start on a different beat each time triggered. but, as you noted, can't figure out how to get the LFO to only result in start times that are multiples of 8. seemingly should be doable with square wave LFOs.

as you noted, can't figure out how to get the LFO to only result in start times that are multiples of 8. seemingly should be doable with square wave LFOs
I wonder how. I mean I don't see any method for quantizing the value of a LFO.

[EDIT : maybe by combining several square LFOs, yes... their values would be added]


Concerning my own problem, I have a clue in the "Granular style of looping" section of Tarekith Tips and Tricks : maybe one should try routing a LFO form another track (thanls you Syncretism and thank you Tarekith !)

I've already experienced this kind of phenomenon, when the evaluation order of the different parameters/controllers is important. The last time it was in MDUW with the order between RAM RECORD and RAM PLAY machines ( hopla ! )
The first time, I just read it in the Kurzweil K2500 user's guide...

i was thinking this way -- set sample start time at mid-point, use a couple square LFOs that should always increase/decrease sample start by 1, 1/2, 1/4, so that you always land on a beat for the sample start. haven't been able to get it to work, but didn't try too hard, not sure how the depth parameter would play into this.

as for your question, as far as i can follow it, i don't the gaps that you're talking about when i just randomly drop trigs of a loop across beats and then have LFO change start time. as you note, results aren't particularly musical if it's a drum loop because you're not always on a beat. but i like what it does with guitar loops.

I haven't tried this but you could use multiple LFO's so that their depth multiplies by 2x for each of them. If the smallest granularity you need is 8, then depth is 8, next 16, then 32 and so on. That way you get random modulation of 0, 8, 16, 24, 32, 40, 48, 56 (also negative) to the start point. Probably start needs to be set to 64 if modulating start doesn't wrap (0 --> 128 and so on)...

Just my $0.02

thx, i will try this out, seems potentially promising (one issue is that i'm not sure whether i'm shooting for multiples of 8, or multiples of 8 minus 1 (i think it's the former, but can't recall whether it's LFO or delay or sample start that is different from what you'd expect)).

I will try it too ! But this way, you need multiple LFOs for only one track.

That's why I would prefer use a saw-up LFO.
I think my problem is that LFOs values are positive and negative. Consequently, if we say that at zero depth, the value is 64, you can increase by 63 (up to 127) but you can decrease it by 64 (down to 0). That's why -IMHO- there is such a gap.

Trying to route the LFO to a filter cutOff, I cannot note that slipping phenomenon...

Here is the Jon's answer :
Regarding chopping, the Userwave isn't really capable of this. There are workaround, like the one you've described but they don't work 100%.

Plus...

If a parameter is set to 40 and is being modulated with an LFO with a depth of 2, the parameter values will fluctuade between 38 and 42.

Consequently, if my start parameter is set to 64 and is being modulated with an LFO with a depth of 64, the parameter values will fluctuate between 0 and 128 -> Problem.
If my start parameter is set to 63 and is being modulated with an LFO with a depth of 63, the parameter values will fluctuate between 0 and 126 -> Problem anyway.

There is no way, I guess... "Alalaa..."

I'll try with saw up or line up LFO on a lower track. That wastes a track, what a pitty.


PS : I still don't understand why routeing the LFO on another parameter does not reveal such a problem.

Kerrstinn and Tuareg do what I want to do.

Isn't there any little hope my beloved MD'LFOs can achieve that ?...

If you are a savvy programmer type my guess would be that the R&W midicommand would be able to do that.

Seems like the problem is that it's not possible to multiply values together using LFO's alone. If you had a random LFO, at base value 4, that fluctuated +/-4, multiplying by another random LFO at base value 8, that fluctuated at depth +/-8, and assign that to STRT it would work. Does that make any sense?