Random Layers?

[nonoduweb]

Hi

I try to understand how layers work in the kit, and I come to a conclusion that the samples are sometimes played in a random way.

It happens when I use the sound CH Bell1 (for example) of Cliff Hewitt: samples are played randomly, no matter how I bang the pad.
When I use the file "count" in the examples folder of Louis (with dsndtool), it seems to be the same thing: disorder.

Is it a way to avoid machine gunning?

Am I the only one in that case?

[nonoduweb]

There are maybe too many questions in my thread.

The main one is: are Cliff Hewitt's CH Bell1 samples played randomly for everyone or have I got a problem in the brain? (no remarks unfriendly on this point  )

Thanks

[puttenvr]

Is it a way to avoid machine gunning?

Yes, sell your Roland kit

Sometimes, I have crash and bell samples randomly on the ride, too. Perhaps I don't hit the small (even if it's 14 inch, it's still smaller than 20 inch) pad correctly

[nonoduweb]

Thanks for your answer.

Yes, the pad bell on the cymbals is sometimes difficult to reach.

My question is: Does the Hewitt CHBell1 sound play a different sample each time we hit a cymbal (or a pad) with the same strengh?

[Louis]

My hypothesis is that the module registers the strength with which you hit a pad, and then randomly picks the sound to play from a fixed amount of samples (perhaps something like 5), that is irrespective of the total amount of layers. This works well with sounds with many layers (since all 5 layers would sound quite like each other), but not with sounds with few layers (where the first sound would be very different from the 5th sound).

This would explain why the "count 10" example in dsoundtool gives very unpredictable results. The CH1 bell has only 8 layers, so if I am right, there will be big variation in the sound you get, even if you repeatedly hit with exactly the same force.

If this hypothesis is right, if you want consistent results, you should select sounds with many sound layers.

A way to test is is to take the CH1 bell into DSoundTool, split the file, duplicate each wave file and reconstruct it (using the "many wave files in 1 directory) and check if this gives better results.

If this would give better results, I think it could be considered as a bug in the module: if a sound contains, say only 8 layers, then it should narrow down the number of sounds to pick from to perhaps 2.

[nonoduweb]

A lot of thanks for your answer, Louis.

I do not know if this answer addresses my personal case or if it describes the usual functioning of the module.

Effectively when I use a sound with a lot of samples (a tom for example) I have no impression that they are randomly played.
The random effect appears only with sounds with few layers (as you say),like the sound CHBell1 where at least four differents samples are played for the same strike force.
When I make a file CHBell1 with a lot of layers, this unpredictable effect is less visible.
On the other hand, the volume follows well my strength, as well as the indication bars  in the "TRIG" menu. So I do not see how there could be a bug of the module because the choice of the samples to play is made (certainly) by software.

[nonoduweb]

Ok

The random effect in Cliff's sounds is normal, other users of the forum tested it and confirmed me that. (Thanks to patbat, mandubien and alainV for their help  ).

I do not still understand why the file "count" of the software dsndtool gives me so unpredictable results; if somebody tests this file in the future, he can inform me about that.

Thank you