DSP trigger, Midi Expression and HiHat

[Alteregoxxx]

Hello,
I've to convert an acoustic drum kit (HiHat,Snare,Tom,Floor Tom, Ride, Crash) in an electronic "silent" one. I want a pc/mac soundcard based solution, no "drum brain" for me! My plan is to have the utter awesome DSPtrigger as the core of the system.
hence, my intention is to buy DSPtrigger and also Midi Expression but, before I buy them, I would need to have 2 questions answered, to be sure the entire thing is doable.

For the question 1, I need to do a preamble, so I apologize for the long post. Let's try....


The system I've in mind requires a soundcard with lots of line inputs:

-SNARE (need rimshots)-->2 mono/1 stereo
-RIDE (need chocking)-->2 mono/1 stereo
-CRASH (need chocking)-->2 mono/1 stereo
-TOM-->1 mono
-FLOOR TOM-->1 mono
-HI-HAT-->1 mono

So, a total of 9 mono line inputs = Big problem, because almost no cheap/medium price audio card come with such a feature but maybe there is one that could be the right ticket, the Tascam US-16x08. Even if It's "only" an 8 line input (mono) audiocard it has also 8 mic preamp input, for a total of 8 + 8 "potentials" triggers inputs.

I called mic preamps "potential" triggers inputs because of this: piezoelectric sensors has an high output impedance, especially at low frequency, in other words they have lot of capacitive reactance (500pF to 15nF) and that, coupled with the low impedance of mic preamplifier (1-3 Kohm usually) forms a "nasty" hi-pass filter, with the cutoff frequency around 1Khz, generally speaking. Obviously also line inputs suffer of this kind of problems, but having line inputs an impedance in the order of 10-22 Kohm, the Hi-pass cut off occurs in a lower freq region, let's say 100Hz.

Now, as every electronic engineer, I didn't resist the temptation of performing a bit of "reverse engineering" on DSPtrigger, in the meaning of analyzing its behaviour under the stimulus of various signal, such as sinusoids of various frequency with flipped/unflipped phase. I guess I understood part of the working principle of the underlying algorythms for positional sensing and sidestick triggering. Let's say, in a simplicistic way, that it seems that the first one, positional sensing, seems to look somewhat also at the frequency content of the signal (It seems there is also a smart autolearning algo on this point...) and the other one looks only to the phase of the signal.

So, question number one: Is the marked hi-pass filtering effect of mic preamp going to disturb the correct functioning of the positional sensing algorythm?

Then, the simpler question 2: Could someone explain how do they use, if possible, HiHat pedal, let's say Roland FD-8 for example, to control the opening closing etc etc of the HiHat? Obviously MidiExpression is needed in addition to DSPtrigger, but is it a fully satisfying solution for this scope?

Thank you!!

[Rob]

The first thing I'll say about DSP TRIGGER and preamps, is don't use them. Preamps are very sensitive and the relatively high voltage spikes from the drum triggers. You are warned.

As for the the MIIDI Expression for your hihat, that should work fine. I actually originally thought up the idea for MIDI Expression devices simply as a way to support hihat pedals inside DSP TRIGGER. I have a Roland FD-8 that I used when I first started developing MIDI Expression, but unfortunately the rubbed arm inside is now broken, so I can't really comment on how it performs now.

Regards,
Rob

[Alteregoxxx]

The first thing I'll say about DSP TRIGGER and preamps, is don't use them. Preamps are very sensitive and the relatively high voltage spikes from the drum triggers. You are warned.

As for the the MIIDI Expression for your hihat, that should work fine. I actually originally thought up the idea for MIDI Expression devices simply as a way to support hihat pedals inside DSP TRIGGER. I have a Roland FD-8 that I used when I first started developing MIDI Expression, but unfortunately the rubbed arm inside is now broken, so I can't really comment on how it performs now.

Regards,
Rob

Hello Rob,
thanks for the feedback.

Well, I'm experienced in electronic design and analysis, not a newbie, in other words. I know preamp expects lower voltage levels than line inputs and slightly lower levels than instruments input too.
Nevertheless, the Tascam i mentioned before, at minimum gain, specifies the following max voltage input values:

LINE LEVEL: +24dBu -->12.182 Vrms -->17.22 Vpeak

INSTRUMENT LEVEL: +8dBV--> 2.512 Vrms --> 3.55 Vpeak (the Tascam site wrongly reports 0.2512Vrms, obviously a typo.... )

MIC PREAMP LEVEL: +8dBu --> 1.947 Vrms --> 2.75 Vpeak

So between the max input for line level and mic preamp there is almost a factor of 6 (without taking in account the loading effect of the mic preamp itself): it will be sufficient a voltage divider, in the meaning of two simple resistors in series, to attenuate the trigger signal to the "right" level for the mic preamp. Moreover you have to take in account that, still the mic preamp having lower max input level capability compared to the line level, it has lower input impedance too, and this will cause the signal from the trigger, that has an high output impedance, to arrive attenuated at the mic preamp input, in comparison to the same trigger connected to a line input (universe behaviour is, partially, wise! )

However, I was more interested in the answer to the hi-pass effect question; Would you mind to reply to that too?

Thanks!

[Alteregoxxx]

Ok,
I'll report my findings here, just in case someone is planning to make a system similar to the one I described at the beginning of this topic....

DSP trigger works really good with both, LINE inputs and MIC PREAMP inputs with my converted acoustic snare.
I'm using for the snare a diy mesh and 2 piezoelectric for the triggering; the first is in the center (27mm), mounted below the mesh with the classic "cone" between the mesh and the top of the piezo and a soft rubber decoupler between the bottom of the piezo and the wood bat (that serves as sustain for the 2 piezos) tightened to the shell of the snare. The second piezo (35mm) is mounted directly on the wood bat, as near as possible to the center piezo.
I've checked multiple times, using an oscilloscope, that the levels coming out from both piezos, hitting hard on the mesh head and the rim, were never out of the maximum range allowed by the MIC PREAMP of my audio card (MotuMicrobook II). By the way, with the center piezo connected to the Motu mic preamp, voltage levels never went over 2-2.5V (and this was hitting the mesh really really hard), a value that is completely super safe for the 99.99% of audio cards mic preamps (obviously triple check, before hitting the head/rim, to have setted the gain of your mic preamp at minimum). If your mic preamp is not able to manage neither such a low voltage levels, you can still use the magic 20dB (10x attenuation) pad attenuator present almost on all the serious audiocards on the market. With the 20dB pad attenuator engaged, paradoxically, at this point you'll probably need to rise the gain of the mic preamp a little bit to have a decent signal to feed DSP trigger.

P.S. (discovered, probably a DSP Trigger BUG)

At the moment I'm using the demo version of DSP Trigger 1.6.1.0, the up to date version, and probably I've found a bug: sometimes, even after performing an accurate calibration as suggested by the DSP Trigger manual (so, also tweaking the envelope curve and, eventually, the threshold) very light hits are not detected. BUT, and this is the strange thing, with that calibration just done, if I slightly tweak the default velocity curve of the center articulation, then bring this curve back again to its default state and then I repeat the calibration procedure, very lights hits are now perfectly detected, without the needs of tweaking the envelope curve nor the treshold!! When I discovered this strange behaviour, I tried to save a preset of a working calibration (one the recognizes the very light hits), then calibrating again to induce the wrong calibration phenomenon /that regularly happens...) and then loaded the preset I had made: the preset recall correctly the perfectly working calibration.
So at least there is a way to overcome this annoying bug (?).
I want to clarify that for the above procedure I'm using regular LINE INPUTS as inputs for the trigger, so the MIC PREAMP has nothing to do with all this.

[Rob]

I'll look into this. Thanks for reporting all your fundings.

Rob

[Alteregoxxx]

I'll look into this. Thanks for reporting all your fundings.

Rob

Hi Rob, any news?

[Rob]

I haven't had time to look into this yet. I'll try to look into it this weekend or early next week.

Rob