MIDI Expression with DIY optical hi hat controller

[Chrisblob]

Hi there,

I'm 99% finished building my electronic drumkit (A2E conversion), and have had trouble getting variable hi hat working. I'm using an Alesis Trigger I|O for trigger input, but the Alesis is not well suited to variable hi hats such as used by BFD2 - the Trigger I|O hi hat configuration requires 3 MIDI notes to be set - one each for open, closed, and foot chik sounds - problem is that BFD2 variable hat is played on just one MIDI note (F#1/42), but if I set the 3 MIDI notes on the Trigger I|O to that same note, when I depress the pedal from open to closed Trigger I|O sends a new note-on to BFD2 at around the halfway point (must be point of change from open to closed according to Trigger I|O), which sounds a hit even though I've not played the trigger.

I saw the MIDI Expression device mentioned on a forum on the internet, and looking at it I'm certain that using this device to send MIDI CC 4 to BFD2 instead of using the Trigger I|O will prevent Trigger I|O from sending these unwanted note-on's to BFD2 while I'm moving the hi hat pedal.

My optical pedal sensor is just an LDR on a TS cable. An LED illuminates the LDR, and a strip of opaque material with a very carefully designed angled notch cut in it is attached to the pedal - as the pedal moves up and down the strip moves between the LED/LDR and the angled notch varies the level of light received by the LDR: it seems to work well with BFD2 via the Trigger I|O's pedal input jack - the LDR specification is as copied from the website...

Dark Resistance: min 10M - Light Resistance: 48k - 140k - Max Dissipation: 30mW - Rise Time: (0~63%) 40mS- Decay Time: (100~37%) 10mS.

As it is in essence just a variable resistor is this likely to work with the MIDI Expression? - I can see no reason why it would not work, but it would be nice to hear some feedback from people who may have tried using optical devices with MIDI Expression.

Best regards, Chris W, New Zealand.

[Rob]

While I'd love to say that MIDI Expression would solve all your problems, I think you should be able to get the trigger i/o doing the job for you.

While I'm not 100% sure, I suspect the note being generated as you close the hihat is the foot chick articulation. The algorithms used to detect this kind of action are surprisingly complex--I know because I myself have coded one. There are a lot of variables that need to be measured and if your pedal is not performing the way trigger i/o expects, you might get mistriggers. But I think you can still get things working a lot better.

1. Setup the open / close / foot chick values to match BFD
2. In BFD find the setting called 'all hit variable' and check it.

That'll improve things dramatically. And at this point you can verify which articulation is being triggered when you close the hihat. If the trigger i/o is indeed mistriggering, I would suggest you just filter it out or set the note in the trigger i/o to something that BFD will simply ignore. After that, you might try playing around with BFD's 'auto pedal' feature to get your foot chick articulation back.

Good luck. If it turns out that the trigger i/o is indeed mistriggering, if you are willing to go through the effort to make a line graph of the CC4 data, I could probably tell you why.

Regards,
Rob

[Chrisblob]

While I'd love to say that MIDI Expression would solve all your problems, I think you should be able to get the trigger i/o doing the job for you.

1. Setup the open / close / foot chick values to match BFD
2. In BFD find the setting called 'all hit variable' and check it.

That'll improve things dramatically. And at this point you can verify which articulation is being triggered when you close the hihat. If the trigger i/o is indeed mistriggering, I would suggest you just filter it out or set the note in the trigger i/o to something that BFD will simply ignore. After that, you might try playing around with BFD's 'auto pedal' feature to get your foot chick articulation back.

Thanks Rob, I had to search to find 'all hits variable' under 'Session', but checking that seems to have fixed the false triggering, and 'auto pedal event' under the 'MIDI' settings was already checked and seemingly working, although before I checked 'all hits variable' it was hard to tell with a half open note still ringing.

So apart from the fragile wires on the cheap piezo's breaking every once in a while, and a couple of dodgy cables (thanks to a dodgy soldering iron operator, ie. myself) its now working near perfectly.

Best regards,

Chris W, NZ.

[Rob]

Assuming your trigger i/o is producing the foot chick already, then you probably shouldn't enable 'auto pedal'.

[Rob]

Also, I would love to see some pics / schematic of your pedal design. I might be interested in making my own.

[Chrisblob]

Also, I would love to see some pics / schematic of your pedal design. I might be interested in making my own.

Sure thing. I have a PDF file saved from a drawing I did in AutoCAD, showing the basic design and details, although I cannot see anywhere to attach a PDF or image file to this forum post. Do you have another means I can send you this file?

Regards,

Chris W, NZ.

[Chrisblob]

While I'd love to say that MIDI Expression would solve all your problems, I think you should be able to get the trigger i/o doing the job for you.

1. Setup the open / close / foot chick values to match BFD
2. In BFD find the setting called 'all hit variable' and check it.

That'll improve things dramatically. And at this point you can verify which articulation is being triggered when you close the hihat. If the trigger i/o is indeed mistriggering, I would suggest you just filter it out or set the note in the trigger i/o to something that BFD will simply ignore. After that, you might try playing around with BFD's 'auto pedal' feature to get your foot chick articulation back.

Thanks Rob, I had to search to find 'all hits variable' under 'Session', but checking that seems to have fixed the false triggering, and 'auto pedal event' under the 'MIDI' settings was already checked and seemingly working, although before I checked 'all hits variable' it was hard to tell with a half open note still ringing.

So apart from the fragile wires on the cheap piezo's breaking every once in a while, and a couple of dodgy cables (thanks to a dodgy soldering iron operator, ie. myself) its now working near perfectly.

I spoke too soon - the false triggering is back - I had the Trigger I|O closed hat note set to an unused note, so when the hat was played in the closed position no note was sounding, so I had to change the note back again to the same as for open and for foot chik (since they're all contained in the same articulation on note F#1/42 in BFD2), and the unwanted hit is back. Those settings in BFD2 made no difference sadly.

Next I unplugged the trigger from the hat, and used MIDI-0X to record the CC4 values as I press the pedal - Trigger I|O is definately generating the hit as the CC4 value exceeds 63 or 64, which makes sense to me as that is half of 127, and obviously that is where Trigger I|O thinks the drum brain/software needs to receive a new note to play the closed hat sample.

So Alesis provide an input that can be calibrated to convert a range of resistances to CC4 continuous data, but then they do not allow to configure input 3 to be just one note for variable hi hats as a few drum softwares require? - I would call that a design flaw or "oversight" on the part of Alesis...

So you might get to sell me a MIDI Expression after all...

Chris W. NZ

[Rob]

I'll break it down for you again.

I had the Trigger I|O closed hat note set to an unused note, so when the hat was played in the closed position no note was sounding, so I had to change the note back again to the same as for open and for foot chik (since they're all contained in the same articulation on note F#1/42 in BFD2), and the unwanted hit is back.

I think you are confusing the difference between the 'closed' hihat and 'foot chick'. The closed sound it when the hihat is closed and you hit the hihat. The foot chick is generated when you quickly press the pedal all the way down. They are two different things.

I promise you there is nothing wrong with your hardware.

1. Read the relevant parts of the BFD manual (it's conplex software)
2. Setup your Tigger i/o note values to match BFD.
3. Turn 'all hats variable'
4. Turn off 'auto pedal' and 'auto splash' in BFD
5. Enjoy your drums.

Regards,
Rob

[Chrisblob]

I'll break it down for you again.

I had the Trigger I|O closed hat note set to an unused note, so when the hat was played in the closed position no note was sounding, so I had to change the note back again to the same as for open and for foot chik (since they're all contained in the same articulation on note F#1/42 in BFD2), and the unwanted hit is back.

I think you are confusing the difference between the 'closed' hihat and 'foot chick'. The closed sound it when the hihat is closed and you hit the hihat. The foot chick is generated when you quickly press the pedal all the way down. They are two different things.

I promise you there is nothing wrong with your hardware.

1. Read the relevant parts of the BFD manual (it's conplex software)
2. Setup your Tigger i/o note values to match BFD.
3. Turn 'all hats variable'
4. Turn off 'auto pedal' and 'auto splash' in BFD
5. Enjoy your drums.

Regards,
Rob

I'm well aware of the difference between a closed hat hit and pedal 'chick' sound. When I press the pedal down quickly (or slowly) without hitting the bow, a note-on "hit" is generated as the CC4 value approaches 63 or 64. I can confirm it is Trigger I|O alone generating this, as it is still generated when the hi hat trigger is disabled. I tested the CC4 values that my optical pedal sensor generates via Trigger I|O, and it ranges from 0 (fully open) to about 122 (fully closed), or up to 125 or 126 if I push down hard. So the Trigger I|O seems to have calibrated to my pedal very well. Next in BFD2 I have made sure "all hi hats variable" is checked on, and that "auto splash" and "auto pedal" are off, and that the articulation is adjusted nicely (CC4 zone change values are 28/57/84/110). And finally Trigger I|O I've set the 3 MIDI notes to F#1/42 for open, F#1/42 for closed, and G#1/44 for pedal, because BFD2 manual recommends the open/closed notes are set to the note for the variable articulation (which I have set to F#1/42), but the pedal note should be set to a pedal sound (G#1/44 is the note I have assigned for pedal sound).

At this point I have read the BFD2 manuals hi hat section (section 9) several times through, and followed your recommendations to the letter, setting up Trigger I|O and BFD2 exactly as you describe, but the unwanted hits still happen as the pedal moves across the half open position.

I'm still wondering if it is to do with the LDR presenting resistance values that are not 100% good with the Trigger I|O, but Trigger I|O seems happy to calibrate to the values it sees.

For my final experiment tonite I unplugged the optical LDR, and then remapped the CC input for the hi hat articulation to CC7 so I could use an M-Audio Volume/expression pedal I have - worked flawlessly, even though it feels really weird trying to articulat hi hat position with a rocker pedal...

I'm stumped, and it is after 1am in the morning - my alarm for work is in less than 6 hours.

Chris W, NZ

[Rob]

Some questions:

• How have you wired up your sensor to the TRS cable?
• If you connect your optical pedal to Trigger I/O without calibrating, what is the value range of CC4?
• Am I correct in assuming that the note being generated when the pedal passes half is not the 'foot chick' note?

Regards,
Rob