Ride cymbal with positional sensing

[Brian Jackson]

Greetings.
I'm new to this forum but very enthused about my new A2E / DIY MegaDrum build. Drums arrived today to replace an Alesis kit that was pure frustration to play, and am ordering the MD next week with positional sensing board.

I'm cool with or without PS on the snare, but for the ride cymbal I'm willing to go to extreme measures to have this feature. I'm a licensed user of AD2 and am happy to know they support PS on rides, though was disappointed to learn (from their tech support person) that it's for stroke type (shank>tip) and not for actual position (bell>bow) as one would expect. But I do like the way they handle the smooth gradations from one sound to another by adjusting the balance between the 2 sounds playing together.

My question is this: Has anyone here managed to either buy or create a true positional sensing ride that works with popular VSTs? Really I'm probably just looking for a 2-zone ride (piezo/piezo acoustic conversion I think) that plays 2 sounds together at varying mixtures... I imagine it basically being 2 separate single-zone circuits that are physically attached to the same cymbal (bell and edge) so the bell gets louder than the bow the closer to the bell you move, but both sounds still trigger with each hit.

Make sense? I just never liked "zones" in the e-rides I've played because of the sharp cut-off from one sound to another rather than the gradual variation you get in a real ride. Thanks for reading, and all opinions and suggestions are welcome. Cheers.

Respectfully,
Brian Jackson

[dmitri]

Really I'm probably just looking for a 2-zone ride (piezo/piezo acoustic conversion I think) that plays 2 sounds together at varying mixtures... I imagine it basically being 2 separate single-zone circuits that are physically attached to the same cymbal (bell and edge) so the bell gets louder than the bow the closer to the bell you move, but both sounds still trigger with each hit.

I guess you know that the positional sensing works from a signal of single piezo, based on the form of the first half wave. Calculating position from two piezos based the level of signal is very unreliable for several reasons, e.g. due to unreliable timing/level difference of incoming signals.

[ignotus]

Hi Brian,

I think I see what you want to do: you don't want complete separation between zones, but rather want the two zones to play simultaneously, with differing amounts of each depending on how close you play to one or the other. That should be pretty easy: you just have to use a stereo input (2 mono inputs, one per piezo), and set DualHead to 'no'. You could then fine tune how they interact with the settings, by playing around with xtalk, etc.

Having said that, I don't think it will sound very good at all if you mix, for example, bell/bow or bell/edge sounds. With the acoustic rides I've owned, there is in fact quite an abrupt change in sound between the bell and bow; I wouldn't say there is a gradual crossfade of both depending on where you hit the cymbal. With bow and edge maybe there is a bit more of that effect, but again, I don't think combining them will sound very good.

After all, the samples recorded in Addictive Drums (and other VSTs) are of real cymbals. That means that the subtle interactions between zones are already there. If a bow hit is a combination of bow + hints of bell and edge, it's already there, in the recording. I can understand you might find the onboard sounds of a module artificial, but, to my ears at least, VSTs do a pretty good job (with obvious limitations) of imitating the real thing. In AD, there are some ride cymbals where, as bow hits get harder, the sound starts to open up and become more 'washy', precisely like hitting it with the shank of the stick, producing a sound that is a combination of bow and edge - like the real thing because it is a recording of the real thing.

However, I would be delighted to be proven wrong, so if you can come up with a setup that works, all the better!

[Brian Jackson]

Thank you, Dmitri and ignotus.

@Dmitri: No, I did not know that. This is all new to me, though I'm pretty astute and now grasp the concept. Thank you for that bit of info.

@ignotus: Yes, that is precisely what I was describing. The module wouldn't see it as "positional sensing" at all, rather treating it as 2 separate and completely independent triggers that just happen to be fastened to the same cymbal. That's the idea anyway, to eliminate the abrupt "zones" when transitioning from bow to bell.

It may very well sound terrible but should be somewhat easy to test. The thought came to me because after talking with XLN (makers of AD2) they explained to me that they accomplish their ride positional sensing in exactly the same way; by crossfading the 2 samples together at various levels/ratios. Regrettably (for me anyway) they chose to use Shank>to>Tip crossfades rather than Bow>to>Bell.

I wholeheartedly agree that it would be preferable to have individual samples recorded at each strike position from edge to bow and mapped out somehow in software, along with a true positional sensing ride pad to trigger them. I imagine that's possible given enough time to record, build and program such a system. But I have to be realistic about my expectations given my current knowledge and resources.

Lastly, what makes me suspect that this "poor man's PS" described earlier just might work is that bell and bow sounds are both heard together regardless of where you strike an acoustic cymbal, just in different proportions. Try cupping your hand over the bell and strike the bow to hear the effect. So the bell is already part of the bow samples and vice versa. The idea then is that by adjusting the sensitivities of each independent peizo, we can artificially exaggerate the balance between bow and bell similar to how we can adjust separate left/right stereo channels with a balance knob, or two faders on a mixer.

Or it could sound like crap. Worth trying though.

Respectfully,
Brian Jackson

[airflamesred]

I do like your theory here Brian, though I do agree with Ignotus that the current state of affairs (ie the piezo) rather limits you to velocity based articulations.

I suspect with FSR you could get a more realistic (in the acoustic sense) feel, but I would say the VST I use does, as Ignotus describes, introduces the shank play with velocity . Alas, they are still ridiculously expensive.

[Brian Jackson]

Agreed, airflamesred.

What I've described is not an elegant, correct solution, but more of a fake-out to eliminate the sharp zone changes. I'm willing to entertain all solutions but unfortunately I'm ignorant of how other VST/Pad systems would accomplish this rather complex task. If you or anyone can suggest a VST pairing with a ride pad design that offers these natural, smooth gradations I would be extremely interested.

My only experience is with AD2 and their PS is limited to shank/tip. Thanks for reading.

Regards,
Brian Jackson

[angr77]

Hi!

We made some tests a year ago to really nail the algorithm for CY-15R (which supports PS) but we failed to really see the characteristic of first half wave. (May be too poor equipment). The three included settings in MD firmware don't really work on the CY-15R from a PS standpoint.

Then we have the AD2 implementation of PS...which after some simulations in AD2 do prove what you are mentioning above...the PS function on the ride cymbals is very equal in sound when moving over the surface.

Best Regards

ANGR77

[Brian Jackson]

Hi!

We made some tests a year ago to really nail the algorithm for CY-15R (which supports PS) but we failed to really see the characteristic of first half wave. (May be too poor equipment). The three included settings in MD firmware don't really work on the CY-15R from a PS standpoint.

Then we have the AD2 implementation of PS...which after some simulations in AD2 do prove what you are mentioning above...the PS function on the ride cymbals is very equal in sound when moving over the surface.

Best Regards

ANGR77

Hello Angr77. Was hoping I might get to chat with you at some point, having read many of your insightful posts on this forum and elsewhere.

I gather that Roland has a different type of PS detection for their ride pad than is commonly done DIY, and modeled accordingly in their module. I know of no VST equivalent, or even how well Roland pulls this off. But my gut tells me that, unlike the high-amplitude wave pattern of a drum head, Roland may not be detecting a wave at all on their ride. If that's true, it would seem to be confirmed by your inconclusive test measurements.

On a hunch, what if all Roland is doing is comparing the difference in spike strengths between 2 piezos on opposite ends of the PS range? If the coincident spikes were, say, 30/70%, or 50/50%, or 90/10%, would that not be enough information for the module to deduce approximately where on the pad it was struck?

Anyway, it's nice meeting you and I hope to maintain a productive dialog.

Respectfully,
Brian Jackson

[ignotus]

On a hunch, what if all Roland is doing is comparing the difference in spike strengths between 2 piezos on opposite ends of the PS range? If the coincident spikes were, say, 30/70%, or 50/50%, or 90/10%, would that not be enough information for the module to deduce approximately where on the pad it was struck?

I actually proposed this method a few years ago (see here) to determine position and even get 3 zones using 2 piezos, and did some tentative testing to see if it might work - but it was never followed up.

[angr77]

Great Stuff - Brian!

Just a note regarding Roland CY-15R. The cymbal do only have one piezo and two switches. (Bell/Edge for the switches and the Piezo for the Bow) So I am guessing they really is using PS on the CY-15R.

Best Regards

Angr77