Major change in firmware. Mesh pads testing needed.

[dmitri]

Hi all,

I've completely re-worked the alternative false triggering suppression (AltFalseTrSupp). The main goal was to improve false triggering suppression without killing fast/press rolls. It is especially tricky to achieve with mesh pads. With the new algorithm I have a feeling the false triggering suppression works very well (much better than the original and the old alternative algorithms) and easier to configure.

I have only tested it with Pintech 10" mesh snare so I'm asking everyone who have mesh pads to test this firmware:

megadrum_20110623.zip

A few things to note before testing it:
1. The mesh pad should not be too "hot". You still should get good results even with slightly "hot" mesh pad.
2. HighLevel should be set as high as possible but so that you still can reliably get highest velocity (127) MIDI signals. This is true for any pads, I'm just stressing it here.
3. For most mesh pads out there MinScan should be set above 30 and even higher. If memory serves me well, a half wave on Roland PD-125 is 4ms so MinScan should be set to 40.
4. Retrigger should be set to ~2-5.
5. AltFalseTrSupp must be set to Yes.

Set DynLevel and DynTime to 0. With this you should be seeing massive false triggering after each hit.
Now try to increase DynTime one step at a time. With DynTime at 8-16 you should already see most of false triggering being eliminated without noticeable effect on fast/press rolls. With my Pintech increasing DynTime even to 60 still allows me to get fast/press rolls.
DynLevel has (very little) effect and it is mostly with light fast/press rolls.

I also found this new algorithm has positive effect on all other types pads/cymbals.

Please, give me your feedback on the new false triggering suppression algorithm.

[ignotus]

Hi Dmitri,

Tested it with 10" and 8" pads - one 10" (snare) has a roland mesh head and the rest use the lower quality and springier T-drum heads. My snare was working very well already so I didn't notice much difference there but with the others the improvement is quite significant - they sometimes missed hits and double triggered whatever the settings, but now I can even get good rolls out of them, so it's definitely an improvement for me.

Another thing that's changed (I was going to start a new topic but now it seems to be solved) is that before, when starting MD with AltFalseTrSupp set to 'yes', pads wouldn't start sending signals when hit for about a minute or more. Only the HH pedal would respond. If I scrolled through to this setting and set it to 'no', then back to 'yes' again, pads would work normally. This happened with all firmwares up until now and I thought there must be something wrong with my MD, since no one else complained about this, and maybe there is, but now it's gone. Strange.

Thanks!

[codemaster]

..pads wouldn't start sending signals when hit for about a minute or more

all samples have been loaded at this time?

[ignotus]

all samples have been loaded at this time?

I'm referring to the signals sent from the pads to MD, prior to reaching the computer - MD wouldn't even register the hits, as if the pads were disconnected. After about a full minute or more (unless I scrolled to AltFalseTrSupp and changed 'yes' to 'no' and back to 'yes' again) it would suddenly 'wake up' and function normally.

[dmitri]

ignotus, thanks for testing! I believe your mesh pads are DIY, right?

[ignotus]

ignotus, thanks for testing! I believe your mesh pads are DIY, right?

Thanks to you for your constant improvements!

Yes, my pads are DIY - sawn-in-half toms with crossbar and DIY cones made with swimming aid foam.

[magpies]

Massive improvement with my PD125 snare. Nice one Dimitri

[CyberFly]

At the end of next week I'll have the time to test out my pads.
I do need to set all '''all gains low'' to yes. Does this mean my pads are too hot to test?

[dmitri]

At the end of next week I'll have the time to test out my pads.
I do need to set all '''all gains low'' to yes. Does this mean my pads are too hot to test?

I think you'll be fine.

[dmitri]

Loaned a Drumtec 14" mesh snare, which was very tricky to configure before, and as far as I can judge got perfect results almost in no time.

1. Set AllGainsLow to Yes.
2. With HiLvlAuto set to Yes got HighLevel just above 800 for both the head and the rim so after disabling HiLvlAuto set HighLevels to 800 on both.
3. Set DualHead on the head to Yes.
4. Set Threshold to 20 on the head and 30 on the rim.
5. Set MinScan to 40.
6. Set Retrigger to 4.
7. Set DynLevel to 0.

After raising DynTime to 12 almost all double (triple,quadruple,...) triggering has been eliminated. With DynTime raised to 24 I could not see any double/false triggering at all.
Even after raising DynTime to 48-52 still couldn't see any missed hits with pressed rolls or with a free bouncing stick even at the very end (hence very fast) of a bouncing.