New false triggering suppression to test.

[dmitri]

Hi,

I'm working on a new version of false triggering suppression algorithm. If anybody is willing to test it here it is:

megadrum_20160418_test.zip

A few notes about the new algorithm:
1. It will work only with STM32 based MegaDrum.
2. To enable new algorithm set AltFalseTrSupp to Yes.
3. Set Retrigger to a very low value, best set it to 1.
4. With this algorithm if either DynLevel or DynTime is set to 0, false triggering suppression is disabled completely and you can see unfiltered signal from MegaDrum in MDM.
5. Start with minimal DynLeveL (1) and DynTime(4) when testing.

There is at least one bug in the new algorithm and I wonder if you will be able to spot it.
I'm very interested in feedback on the new algorithm so please post your results. In particular what results you get for fast/press rolls and single hits with various types of pads.

[ignotus]

Just tried it out with a 10" snare using a Roland mesh head. With retrigger at 1, DynLevel at 1 and DynTime at 4, single hits are ok, but pressrolls and fast rolls miss a ton of hits - it's as if it were less sensitive, even though Threshold is as far down as it will go without firing off by itself.

One bug I noticed is that the rim zone on the snare doesn't work.

[dmitri]

Can you post a screenshot of a MIDI Log hit bars with 3-4 average strength hits ~1 second apart when Retrigger is set to 1 and DynTime or DynLevel is set to 0, i.e without any suppression ? Set Bars Number to 32 if needed to fit all 3-4 hits.
Also, can you test it with the Rim disabled (Note set 0)?

[ignotus]

Here goes.

With rim; Retrigger=1; DynLevel=0; DynTime=0

With rim.png

Rim disactivated; Retrigger=1; DynLevel=0; DynTime=0

without rim.png

[dmitri]

1. Can you please update your MDM to the latest version so that it shows signals intervals in milliseconds properly?
2. Can you test the head for fast rolls with rim disabled and DynLevel=1 and DynTime=4?

[ignotus]

Here's with rim activated again, with the latest MDM:

with rim.png

It's practically identical with rim disactivated, like above.

After some further testing of rolls, it's actually not that bad. I can't notice any difference with rim on or off. I did find that the rim will work for a little while, but then stops working. It just seems to be ever so slightly too aggressive suppressing false hits. That said, I've always had problems with double triggering and with this algorithm I get none at all.

[dmitri]

Here's with rim activated again, with the latest MDM:

with rim.png

It's practically identical with rim disactivated, like above.

After some further testing of rolls, it's actually not that bad. I can't notice any difference with rim on or off. I did find that the rim will work for a little while, but then stops working.

Does it stop working completely or does it resume after awhile? How long while?

It just seems to be ever so slightly too aggressive suppressing false hits. That said, I've always had problems with double triggering and with this algorithm I get none at all.

So if I make it less aggressive it will work better for you then the standard algorithm?

[ignotus]

Does it stop working completely or does it resume after awhile? How long while?

Yes, it stops working completely, after about 10s. 5 minutes later it still won't work.

So if I make it less aggressive it will work better for you then the standard algorithm?

I reckon it might do a better job of getting rid of false hits. I did however have it working very well with the previous algorithm, with DynLevel at 4 and Dyntime at 4, flams and everything work great and I very rarely get double hits; what I mean is that it will be hard to improve on what I already have.

Edit: By previous algorithm I mean I had AltFalseTrSupp set to 'no'.

[dmitri]

The purpose of the new algorithm is to make initial configuration for fast rolls/false triggering suppression easier, i.e. suppressing false hits without suppressing fast rolls. It's strange that rim stops responding completely, I will look into it tonight. Btw, I'm doing my tests with two extremes - 10" Pintech mesh snare and Yamaha PCY135.

[dmitri]

Please test this firmware version:

megadrum_20160420_test.zip