Does TD-12 or TD-20 send midi cc for positional sensing on toms?

[GW1000]

Hi. Can anyone with a TD-12 or TD-20 tell me whether it sends midi cc information for positional sensing on all toms, and if so, is it just on rimshots or for head shots too? I know the module has built in positional rimshot sounds for the toms (as well as the more commonly used snare and ride positional stuff), and I assume the only reason Roland didn't supply positional head shot sounds for toms was that they didn't think people would find them very useful so were not worth the sample memory, rather than due to any technical differences between how the toms detect position compared to how the snare does it. If I'm right then an external VST could potentially be configured to use the cc info to provide positional samples. (I know it would only work well with large mesh pads being used for the toms). Anybody know for sure if the cc information is coming out of the TD-12 or TD-20 for tom head shots?
Thanks

[Tazed]

I don't think it does. It's not positional sensing like the snare. I've read about it before and I think the answer was no. It wouldn't be CC data, anyway. PS is derived from the first half wave form off the head sensor, so it's a set signal, nor variable (like a hi-hat controller).

[GW1000]

Thanks Tazed, but I don't think you're quite right. The module uses the first half wave from the trigger to calculate the position, yes. But when it sends the midi data to another device like a computer VST, it does not send separate notes for different positions (like it would for head vs edge), instead it sends a value corresponding to the position on a particular midi CC number, just before it sends the note of the hit, but the same note number is used regardless of the position. The VST stores the positional number from the CC, and uses it a moment later when the next midi note arrives (the note also contains the velocity information), to decide which positional sample to use for that note. For example if you hit the mesh half way between the centre and the edge, it would send 63 as the value on the positional CC number, then a tiny moment later it would send the normal note information, with the velocity. The CC data is not pouring out all the time, just one value for each hit, sent a tiny fraction of a second before the note is sent. I think Roland use the same controller number for any positional pad, so the timing is crucial, the first note received after the CC value is assumed to be the note that the CC value is for. My question is does this CC value only get sent before snare and ride hits, or is it also sent before tom hits.

[jpsquared482]

GW100 is correct regarding positional sensing. It is a separate MIDI value sent prior to the note value/#. VSTs such as Superior Drummer interpret PS and reproduce it just like the TD12 20 and 30, so it's definite that it is sent via MIDI.

To address the question, no, Roland modules do not send PS for toms; the technology is only applied to the snare and ride where it matters most to most of us.

Hope this helps.

JP

[GW1000]

Thanks very much jp. I don't think that kind of information appears in the official documents, so I finally gave up looking and asked in a forum. It seems odd though. I would have thought that when Roland went to all the trouble of making the module interpret the position for the tom rimshots, they would want to make it all available from midi too, but there may be some technical/cost limitations that are beyond my knowledge.

[BarT]

Toms CC is covered on Page 74 of the TD-12 manual: Control change used for transmitting/receiving the strike position of the snare, ride, and

The TD-12 MIDI Implementation manual says it also applies to AUX inputs. By default it's CC16 for Snare, CC17 for Ride and CC18 for Toms.

A TD-10 Expanded with TDW-1 had positional sensing on tom heads, with appropriate module sounds included. With the introduction of the TD-20 and TD-12, Roland switched to tom rims.

(It was discussed here at the time, but I only remember one person claiming that tom head position would be missed much for internal module sounds; The change enabled tom rim clicks and shallow tom rim shots which were not available on a TD-10X.)

It's not explicitly stated in TD-12 or TD-20 documentation that CC18 is transmitted for head and rim of tom and aux inputs, but it is in TD-30 MIDI implementation and I think the same applies to a TD-12/20.

So I believe the answer to the original question is: Yes, both.

Bruce

[jpsquared482]

It turns out the complete answer is sort of a hybrid of BarT's and my original response. I left out the part that the tom rims do have "positional sensing", but the tom heads do not support positional sensing. Positional sensing on the toms allows you to articulate soft rim clicks and louder rim shots. However, there is no tonal variation offered for the tom heads, as you have with the snare.


For example, from the Roland Support site, for the TD-20 (from what I have found, the same is true for the TD-30 and 12):

- SNARE HEAD (head shot) - SNARE RIM (rim shot) - TOM1-4 RIM (rim shot) - RIDE HEAD (bow shot) - AUX1-4 RIM (rim shot)

[GW1000]

Thanks guys, it's a pretty confusing subject. I think BarT is probably correct that the tom limitations relate to the module's internal sounds rather than the midi capability.
JP, Roland have not provided tom head shot sounds in this module for different positional areas, but I think the detection technology it is using internally is the same for the snare and the toms, so it could still output the midi positional data for tom head shots. I would think that the positional sensing on the toms is actually always relating to the mesh head in some way, so regarding rimshots it distinguishes between rimshots where the stick hits near the centre of the mesh as well as the rim, and "shallow" rimshots where the stick hits near the edge of the mesh as well as the rim. The positional value is still coming from the tip of the stick hitting the mesh somewhere.
I can't imagine that hitting only the rim would allow any kind of actual position to be detected, because the positional sensing is all done from analysing the shape of the first wave from the piezo at the centre of the mesh head. The module is probably able to use the velocity to choose between rim click, cross stick or whatever those terms are, when you only hit the rim, but I don't think that is actual "positional sensing" in the official terminology.
One thing that does still confuse me is how positional sensing works on a CY-15R or CY-12R/C ride cymbal. Perhaps the rubber in the bow area vibrates differently depending on where you hit it, like a mesh head does, so the module can analyse the wave and work out a rough position where the tip struck the bow. These cymbals only have switches at the edge and bell, so it can't be comparing the strength or timing of the signal from two piezos to work out the position.

[jpsquared482]

Hi again,

GW1000, I should have read your post/question more closely. (I thought you wanted to know if positional sensing is applied to the toms generally, assuming the head was most important) I now see what you are specifically addressing, and your question is a good one. Based on the fact that the toms do detect position but don't use it in the module for the head sound is a different point.

As you say, the module must be detecting position of the stick tip on the head and then combines that with the rim strike to provide rim shot nuances. So, it stands to reason that VST programmers should be able to harness that position data and use it for tom head positional detection.

JP

[GW1000]

If I get my hands on one of these modules I will take a close look at the midi data in a DAW and post the findings here. I've heard that positional cc data can be a nuisance when editing midi, because if you drag a note earlier in time, without also moving the cc data for that note the same amount, the positional value may get applied to the wrong note (or become irrelevant depending on what comes next). It's a pity the midi standard did not leave some spare properties in the definition of a note, so it could be easily expanded with things like a positional value while keeping the data neatly together with the velocity and key. Using cc data for an attribute of each note is a nasty kludge, but I guess midi has lasted longer than its inventors would have imagined, so we shouldn't be too harsh!