www.megadrum.info

Hi,

Working on my A2E conversion using cake pans, Quartztrigger cones, and DDT mesh heads.

My 13" Snare is finished and triggers great. Drum rolls and flams are no problem. So, off to the toms

So, just assembled my 8" Tom. Same method: cake pan + cone trigger + DDT mesh heads. Figured since my
Snare triggers great, so will my 8" Tom. Well...

Solved: I have to disable notes on Tom1R to avoid conflicts and disable rim triggering

Having trouble with false triggering. Often when I hit (medium hit) I get false triggering, and
MDM Raw Midi shows this:
90.248 10 0x99 0x30 0x55 C 3 Note On
0.000 10 0x99 0x30 0x11 C 3 Note On
0.198 10 0x99 0x30 0x00 C 3 Note Off
0.000 10 0x99 0x30 0x00 C 3 Note Off

Two "Note On" messages happening at the same time??? The problem is an unwanted note about 20 milliseconds
after the first one. I've posted the waveform below that corresponds to those notes triggered. There's no
"spike" 20 milliseconds later, the waveform has pretty much flattened out. Also posted a waveform of my Snare
just as a comparison (waveform takes longer to flatten out, but causes no problem).

My pad settings are:

Tom1:
------
Note: 48 (C 3)
Alt Note: Disabled
Pressroll Note: Disabled
Threshold: 8
Gain: 1
HiLevel: 412
Retrigger Mask: 4
Dyn Level: 15 + DynTime: 28
MinScan: 10 (1ms should be fine according to waveform, could even go lower than this)
PosLevel: Disabled
Type: Single Piezo

Snare:
------
Note: 38 (D 2)
Alt Note: 38
Pressroll Note: 38
Threshold: 5
Gain: 4 (yeah, could use a smaller Resistor, but it triggers great, so won't change it)
HighLevel: 600
RetriggerMask: 8
DynLevel: 12 and DynTime: 24
MinScan: 16
Pos Level: 3
Pos Low: 20
Pos High: 33
Type: Single Piezo

Again, my 13" Snare triggers great. No false triggering at all. Only a minor hot spot, but very small hot spot, but acceptable for me.

Back to the 8" Tom: The waveform seems good because:
(1) No "spikes" after the main wave
(2) Flattens out quickly

Only thing a bit surprising is that the waveform settles mostly in the upper half... but with the rectifier board, would that even matter?

Only my 13" Snare and my 8" Tom are connected to Megadrum. No piezos for rim triggering are used (yet).

I've tried the procedure to optimize DynLevel+DynTime, but have not gotten it to get rid of false triggering completely. So above DynLevel+DynTime
setting is not the only one I've tried.

Any ideas what to try next? Oh, I used a 27k resistor for the 13" snare, but needed no resistor for the 8" Tom. I assume that's because
the mesh pad is smaller, and therefore reduces the force on the cone when hit.

Thanks
Andrej

[Megadrum 32: STM32F103RCT6 + Rectifier board, Firmware 20130612, MDM 20130706]

And here's the full pad settings for Snare (no false trigger) and Tom1 (false trigger happening about 20ms after real hit)...

I'm using Addictive Drums, and looks like it ignores a second NoteOn, as I can't hear the false triggered note.
So, maybe I'm making too much of it

Next: Rim trigger on Snare, Tom2, Tom3, and Kick. Then Yamaha cymbals and hihat.

Andrej wrote:
Having trouble with false triggering. Often when I hit (medium hit) I get false triggering, and
MDM Raw Midi shows this:
90.248 10 0x99 0x30 0x55 C 3 Note On
0.000 10 0x99 0x30 0x11 C 3 Note On
0.198 10 0x99 0x30 0x00 C 3 Note Off
0.000 10 0x99 0x30 0x00 C 3 Note Off

Two "Note On" messages happening at the same time???

It means you have more than one input with the same note and you get crosstalk false triggering.
It seems you a have dual zone pad but judging from config screenshots you configured it as a two single zone pads with the same note on both of them and hence apparent double triggering.

dmitri wrote:It means you have more than one input with the same note and you get crosstalk false triggering.
It seems you a have dual zone pad but judging from config screenshots you configured it as a two single zone pads with the same note on both of them and hence apparent double triggering.

Nope, no dual zone pad. One snare with one piezo plugged into port 4, and one tom with one piezo plugged into port 7.

But I think I found out the problem: My assumption that choosing type "Single Piezo" would ignore Tom1R settings was wrong. I think setting
Note to 0 on Tom1R should do the trick... Guess with my snare I just got lucky that it did not trigger SnareR note 37, which I've not disabled.

Now I realize that Megadrum can generate two notes with single-zone pads depending on strike force. Nice.

My Tom1 should trigger now just as good as my snare. I think I'll have a perfectly working drum set very soon

Feature request: Support MinScan down to 6 (0.6 milliseconds). Lowest is 1 millisecond, but I noticed that with a single layer mesh head on
my 8" tom, that rise time is about 0.8 milliseconds (verified with Oscilloscope). Lowering MinScan might help me with minimizing hot spots (not sure though).
(Roland supports 0 to 4.0 ms ScanTime)

Thanks Dmitri,
Andrej

Andrej wrote:

dmitri wrote:It means you have more than one input with the same note and you get crosstalk false triggering.
It seems you a have dual zone pad but judging from config screenshots you configured it as a two single zone pads with the same note on both of them and hence apparent double triggering.

Nope, no dual zone pad. One snare with one piezo plugged into port 4, and one tom with one piezo plugged into port 7.

Then you have to disable the rim inputs otherwise you will be getting electromagnetic interference on them and as a result false crosstalk triggering.

Feature request: Support MinScan down to 6 (0.6 milliseconds). Lowest is 1 millisecond, but I noticed that with a single layer mesh head on
my 8" tom, that rise time is about 0.8 milliseconds (verified with Oscilloscope). Lowering MinScan might help me with minimizing hot spots (not sure though).

It's been tested and even MinScans below 15 don't make any noticeable difference in terms of latency. I don't see how it can help with hot spots.

dmitri wrote:It's been tested and even MinScans below 15 don't make any noticeable difference in terms of latency. I don't see how it can help with hot spots.

Not asking because of latency. Roland modules can go to 0ms, and from what I've read on vdrums.com, some have success in setting a very low ScanTime
to force the module to not capture the highest voltage for a hit directly on the piezo. That assumes that hitting directly center causes the piezo to respond a little
longer, which I don't know if that's valid. I'm aware this would interfere with positional sensing.

I think I've optimized my pad for hot spots: The cone is only slightly touching the mesh head, and the mesh head is very tight. So, I don't have a big hot spot,
but it's still there. Can live with it, but if I can get rid of that, why not try...

Looking at the waveform, when hitting directly on the center, the first wave is very sharp, but the second is not. Wondering if a module could take that into consideration
for active hot-spot removal... this would require an increase in scan time (maybe adding 3ms) to be able to look at two wave crests.

What else can you do in Megadrum now to reduce hot spots? I don't think it's always and only a pad problem.

If you don't care about positional sensing and a spike from a hot spot is only in the first half wave than I'd suggest you to connect a piezo with such polarity that the first half wave is negative and use an input without a Positional sensing addon.

That's true, I could do that. But, it's ok for now. Will hopefully get to test a Roland drum set soon to see how much hot spots it has. If it also has some hot spots, then I know I should not waste time on optimizing it.

Thanks Dmitri.