#1

I’m thinking out loud here because I can’t quite get these to match up nicely and I think they should for some reason, perhaps I am wrong and someone can explain why? Anyway, bear with me…

The output range of the ER-101 is 0 to 8.196V and the input range of the ER-301 is -10.24 to 10.24V.

There are 100 values in each voltage table in the ER-101, I’m using the 12TET table.

If I want to map the first 88 voltages from the ER-101 to 88 slices in the ER-301 I thought if I set the slice select input gain to 1.0 it would just work.

It almost does, but occasionally skips slices… what am I doing wrong?

I know I can adjust the voltage table for the ER-101, and have done to get this to work, but this seems like a mistake for some reason.

Anyone got any clues? Thoughtful intelligent insights? :smiley:

1 Like
Mellotron Samples
How do you map the triggering of sample slices to a range of voltages?
Custom Unit - Bokeh 4-Step Melody Generator/Sequencer
#2

I don’t have the answer but Im very curious to hear the answer

#3

So the value of the 88th entry in the ER-101 12ET table (which is index 87 because the table starts on zero) is 7.250 volts. You can see this value by momentarily switching the view over to Nr in the track configuration screen. Or you can just calculate it 87/12 = 7.25.

The ER-301 is distributing the slices evenly across the positive half of its input range (0 to 10.24) so the last slice will always map to 10.24V.

You want the 7.25V coming from the ER-101 to map to the top range of 10.24V in the ER-301. Therefore the gain should be 10.24/7.25 = 1.41.

By the way, Neil has shown how to do this without math:

  • Arrange for your source to swing between its lowest and highest value (for example punch a two-step sequence in to the ER-101, one with CV index 0 and the other with CV index 87, turn on smoothing, set both gate lengths to zero).
  • On the ER-301 increase/decrease the modulation gain until you see the post-modulation cursor scanning exactly the entire range.
  • If your source is not unipolar, then you would first adjust the bias so your source sweep is centered (in the positive upper half) and then adjust the gain.
1 Like
#4

Oh and if you are having trouble activating the last slice then just add a tiny bit of bias. Half a semitone would be perfect: 1.41 * (0.5/12) = 0.06.

However, at least for me the last slice is usually just there to indicate the end of the previous slice. So I would not be interested in activating it.

1 Like
#5

Got it - thank you!!

I don’t think I’d have worked that out, but now you mention it I do remember the demo from Neil, I just didn’t make the connection but now it makes perfect sense :slight_smile:

I think the thing that threw me off was that it almost worked, i.e. at 1.0 gain there was only a few errors.

Anyway, I’ll report back when I’ve done the tests :slight_smile:

#6

Redundant but a picture is worth a thousand words right?

slice-mapping.png600×580 41.1 KB

1 Like
#7

awesome, thanks Brian

#8

Oh yes, nice graphic, cheers :smile:

I have just recreated the entire thing from scratch and reloaded the 12TET voltage table and it’s still not quite right! It’s definitely better though :slight_smile:

I’m getting the first slice triggered twice and slice 10 is not played, apart from that it seems perfect!

I need to sleep on it and try again with a clear head, it’s late here!

I also think I need to update my 101/102 firmware.

#9

I bet you need to just bias everything up by half a semitone (i.e. 1.41 * (0.5/12) = 0.06) on the ER-301 so that the boundaries between slices fall halfway between notes on the ER-101.

#10

And the picture for that:

slice-mapping-2.png799×552 68.4 KB

This is exactly why two quantizers in a row hardly ever work as you would want.

1 Like
#11

I would like to add some encouragement here. If you completely master the concepts of gain and bias (also known as attenuation and offset) then your modular-fu will be unstoppable.

7 Likes
#12

Thanks, I already tried that and it makes a difference, but I am going round in circles with the adjustment, too low and I still get the double slice at the beginning, too high and an odd slice here and there will start being skipped or repeat.

#13

You are using one of the ABCD inputs right?

#14

Haha, a while back no, I wasn’t, I was using one of the ins and that cause all kinds of mayhem, but yes I am now :smiley:

#15

And are you delaying the gate with respect to the CV?

#16

I tried that too and it don’t seem to make any difference, but I’ll try again now…

5ms seems to sort it out, I think I might have got it!!!

Thank you :smile:

#17

And thank you for your perserverance!

#18

A pleasure, and I’m not giving up any time soon hehe, nearly 30 years and counting, I’m a lifer :smiley:

Hopefully this will help others, I have a really nice custom unit coming from this I think… will need another day or so to finish it off, but bank holiday weekend so plenty of time to get it right :slight_smile:

So to recap, sat the min/max levels, add a delay to the trigger and offset the input voltage!

Nice, I feel like I have complete control of this now!! Weird to do it with numbers though :wink:

1 Like
#19

Interesting thread - this brings me back to my request for having a slice address mode that uses a fixed semi-tone 1V/Oct option so that you don’t have to tweak ER101 voltage table if you add or remove N slices. Thinking about it this request would break down if you had more than 100 slices but if that happens (I can’t imagine having 100 slices but I’m sure someone could/will) you’d have to revert to the existing method.

#20

I totally understand why you would want this, it was exactly how I was thinking, but having finally totally got my head around this it seems superfluous because it really is a very elegant and flexible solution.

I think you would still have a problem removing slices anyway, would it not mess up which slice you were triggering if you removed one, i.e. say index 34 triggers slice 34, nice and simple, then you remove slice 30, would slice 34 not end up triggering slice 33?