Unwanted double actuations

Erae Touch
1

I’m using drum grids or keys in four different presets I’ve created to play BFD’s drums hosted in Bitwig. I disable all Pressure, Vibrato, and X- and Y- related stuff. Playing straight eighth notes at about 110 BPM, I get frequent double actuations at all the grid positions and keys that I’ve checked. I’d say about 1 in 5 times, a non-rhythmic flam comes out instead of 1 eighth note. I’ve verified this using MIDI Monitor. I can sometimes make it happen when playing one note or two notes, but when I play groups of three notes, MIDI Monitor verifies that–consistantly–about 1 in every 3-5 times, there are four MIDI note on’s. A stream of eighth notes results in random doublings.

Playing the same BFD/Bitwig setup with my other connected controller, a Zendrum set up to transmit the same MIDI note numbers on the same channel, I don’t get a single error of this type. This error is a dealbreaker for me since I’m all about the drum controller, with everything else being gravy. I can’t think of anything else to troubleshoot. It would be nice if the Erae had a way to suppress channel pressure since I’ve suppressed everything except that, and channel pressure seems to be the only difference from the Zendrum, which is triggers only.

The Zendrum always transmit snote off immediately after its note on, with no way to control the note off’s timing. The Erae, of course, doesn’t transmit the note off until you release, and then it transmits note off velocity. I wonder if they could make a dedicated drum element for situations where you only need the note on, and that only transmits note on followed by a note off after a fixed interval, or an option for no note off at all because percussion plugins like BFD make no use of the note off.

2

id say most daws wouldn’t like not getting a note-off (when recording)

but given midi is serial, they could send the note-off immediately after the note-on.
that would be as close as midi gets to a ‘trig’ in modular.

I wonder about the double actuations…Im surprised this has not been reported more frequently - perhaps its some kind of rebound on the surface… perhaps a dedicated drum element - could have a higher ‘debounce’.

I’ve not seen double notes, but I guess if you playing melodically then you hold the surface down for longer, so less likely to bounce?!

1 Like
3

I’ve been examining the issue closely to see if there’s some playing technique or unexpected or unintended action or timing of the release (finger lift) that causes the doubled note. I find nothing in particular. If I do single taps in isolation (say, quarter notes at 60 BPM), it never happens. When I go to a regular pulse at a faster speed, whether playing naturally or tapping in a very controlled manner with the same finger going up and down, I get the frequent doubled note. I didn’t notice it until now because the doubled note is even shorter than a flam or regular grace note, so it can come off as a sonic quality rather than two notes. The problem with percussion is that some sounds (e.g. closed hat) have very short releases so that the doubled note is prominent. Up till now, I think I was “hearing around” the error, accepting it as a sonic variation rather than a rhythmic error, but it was just too obtrusive with some drum kit pieces in dry patches so that I finally investigated.

Can you try this experiment? Using MIDI Monitor (on Mac), play groups of three notes as a constant rhythm, which is easy if you do a piano-style trill of fingers 2 and 3. Trying different tempos and velocities, see if MIDI Monitor lists four notes played. I boiled it down to these groups of three, hitting MIDI Monitor’s “clear” button on the fourth note with the other hand so that I’m always looking at a clean log and won’t make a visual mistake. For a while, I thought my ears were fooling me, but MIDI Monitor corroborates what I’m hearing.

4

Here’s a video demonstrating the glitch:


Here’s the Erae Touch preset .emk file I’m using in the video. I had to rename it to .txt in order to upload it:
5x5-BFD-R.txt (140.1 KB)

2 Likes
5

i had a similar problem with double actuations on the erae as well. my first unit got replaced, because there were problems with the calibration and on that one i had the problem quite frequently. i noticed it more when playing melodic patches, when using drumsticks it was mostly fine. on the replacement unit i got, the problem still exists, but is much less frequent. since that new unit is more responsive to lighter touches than the first one, i assumed, as @thetechnobear suggested, that it is some kind of rebound on the surface. the pressure becomes a bit lighter after the inital hit, so that it is no longer detected and then is detected again and produces another note on. not sure if that really is what is happening, but since the problem only happens very infrequently on my new unit, i stopped looking into it.

what happens if you hit the erae harder? in the video you seem to hit it quite softly. in my case hitting it harder led to less double actuations. of course that is not a solution for the problem, but could give some insight on why it is happening.

1 Like
6

Actually, I believe it doesn’t happen when I strike harder. I’ll double-check that. I agree adjusting velocity isn’t a solution because full dynamic range live percussion is my main purpose, which is why I have several Zendrums. I can’t accept a MIDI controller that doubles up randomly no matter how infrequent. It’s such a basic defect that if Embodme can’t affirm that they’re making it a priority and that it’s addressable in firmware, then I’m out. My fear would be that, in order to ensure one note per strike, they make the device less sensitive, and it’s already borderline for my purposes.

7

Hmmm. I have the same problem again with melodic, I get quite a few ghost notes and have to keep removing them. It’s quite annoying over time when I was doing a few sessions. I am absolutely certain that my unit will have to be replaced because there is clearly something wrong, not to mention the inconsistent velocity in certain areas, right in the middle of the board, using consistent force I get values of 100 on one note, and 62 on another note, I get a value of 107 on one note and 73 on another. It’s bizarre and is in no way allowing me to get a good performance of a piano patch. I’ve already contacted support, so I’ll hopefully receive a reply soon. I am so, so horribly disappointed. I would have hoped that my unit would be perfect. I don’t know. I mean, what happened here? There must be quite a few users who are not happy, surely you and I aren’t the only ones who are experiencing this problem.

I am pretty sure the replacement will fix the problem. There is just no way I can use my Erae Touch, certainly not this unit. Sorry.

8

I can pretty much accept the inconsistent velocity in different areas. Somehow I resigned myself to that going in, that it couldn’t be perfect, that I could compensate for it by adapting playing technique (like a “dead” spot on an acoustic instrument), and that recalibration would take care of the worst of it. Still, I’ve got a prominent under-sensitive spot right where I put one of my snare pads, and that’s the most important expressive kit piece in conventional drum parts. Another workaround has been to enlarge my pad size from 4x5 to 5x5 so that there’s a larger sweet spot to hit per kit piece. That’s fine for a one-handed drum grid, but my two-hand 48 pad grid requires the smaller 4x5.

These ghost notes, however, are a dealbreaker. In the case of drum parts and live drumming, they make your results unmusical with no way to compensate. As pointed out above, depending on the sound module patch, and the style of of music and playing (longer melodic stuff), the ghost notes may not be noticeable or even happen. I expected we’d be working with the developers and a device that was not musically perfect, but the Erae Touch’s surprise defect stops me from doing 95% of what I got it to do.

9

I work on assembly lines in a factory and the stuff we QC is as cheap as dirt. We’re talking $5 communication semiconductors. We fail stuff routinely if it doesn’t fulfil minimum spec parameters. I have failed dozens of items that didn’t meet minimum spec. If it isn’t good enough we don’t pass it. Simple as that.

1 Like
10

small production runs, small team…
whilst its disappointing for those that have faults, let’s remember, quite a few of us here have the Erae and are not having issues with it. we don’t really know the failure rate… or tolerances we are talking here.

personally, Id say… It’s ok for a company to have a small number of faulty units (this is regrettably inevitable) - what is much more important, is then how the company deals with the issue.

so far, from reading above, it seems @embodme have been happy to replace ‘faulty’ units, so I don’t really see an issue… frustrating as I’m sure it is.

anyway, I suggest anyone that has an issue contacts contact@embodme.com , and give them the opportunity to help you.

I think an important point here, that we need to get to the bottom of is…
is this only from ‘faulty’ units, or is there some tolerance that calibration may deal with?

my experience in this area with other similar controllers using similar surfaces, is they all have noise and bleed - so, ‘calibration’ is an important part of making these usable, and those calibration routines tend to evolve over time and get better… so some of this may be down to ‘less’ mature firmware.
(there is also still a possibility of more mundane bugs causing issues e.g. in touch detection)

but of course, others may have faulty units or units outside of ‘acceptable’ sensor range … in which case they need will need replacement units.

and I don’t think any of us are in a position to say categorically how many fall into each of these groups… since we only have our own experiences, and what we hear here…which is very limited.

4 Likes
11

I completely agree with that, and just to point out: While the replacement in my case took some time, embodme customer support has been very friendly and helpful along the way. Actually the uncomplicated way they handled my issue and the mail contact i had with them so far, gave me the impression, that they really care about people being happy with the erae in the long run.
So no complaints from my side at all.

2 Likes
12

Is there anyone who can verify that their Erae’s Drumpad element does not have the doubled actuations problem? I mean verify with something like MIDI Monitor because a pitfall for players is hearing what you want to hear rather than what you play.

13

I’ve since contacted them about the velocity inconsistency problem. I’m not sure what is going on. How long did it take for you guys to get a reply?

14

I’ve finally gotten a reply, and it took two days. I guess that’s fine as long as we can rely on them to respond, but I think it would give users more confidence if they had a threaded message board for the serious users so that we and they could keep track of things more systematically.

They responded that my case appears to be that my strikes are light enough to be at the limit of detection. I won’t be able to carry out the tests they suggested for a few days. Those strikes you see in my video are soft, but standard for me. They’d be about velocity 50 on a Zendrum, and are giving 17-20 on the Erae, which would be a borderline acceptable limit even if it didn’t glitch. I go down to 7 on a Zendrum, but I at least need down to 15, or else Erae isn’t worth it compared to Zendrum. The problem is that the whole world likes to whack their drum controllers much harder than they need to (including most Zendrummers), and Embodme has designed for that (for example, sticks work). My fear would be that they can’t adjust in firmware. If they can, then maybe this one-size-fits-all velocity curve is a no go, and they need to enabled user-side adjustment of velocity response. I mean, the current one-size is a surprisingly good fit, but not in the end good enough.

My favorite Zendrums are set up on their most sensitive velocity curve, with the lowest possible floor, which is value 1. I have a couple that can’t handle such a low floor, giving false actuations, so they must go to 2, which filters false strikes out, and is still an acceptable feel. With the Erae, it’s unnacceptable that a light touch will double the triggers in a way that can’t be filtered so that you have to self-fillter by always striking with a minimum vigor. It’s a musically unnatural solution and leads to bad musical results when doing a decrescendo. It’s a disaster to suddenly move to a quieter passage only get flurries of ghost notes. I’d say filter more aggressively, but that might lead to decreasing the already borderline sensitivity.

15

I believe Embodme are working on something like velocity curves - they recognised the ‘issue’ when I spoke to them about it.

there is a minimum pressure possible on these things, since below the you get to possible noise issues and false triggering etc… I don’t think they can go very much lower than where they are, but perhaps still room for small improvement.

I did talk with them about the possibility of users ‘tweaking’ this threshold, its a bit of a tough one, since in most circumstances, I think ‘as low as possible’ without false triggers is what we all want… so not sure adjustment is very useful … unless if varies on different units… but then its arguable that its part of calibration… not sure if they have come to a conclusion on this.

I agree with what you say about drumpads been ‘whacked’ , I find this on things like the Push 2, I find it borderline playable for melodic play since it has got such a heavy ‘touch’. (and I heard MPCs are even worst)

anyway, I hope they can improve this all over time…

I do think a velocity/pressure curve will really help, since I personally felt, currently, there is quite a range of pressure available… from point of activation to full force, so I think changing how that reacts, and outputs midi will yield quite a marked difference.

you could perhaps play with this using a ‘midi tool’ that can remap velocity curves… you’ll get limited results due to the 7 bit nature… but might be illuminating.

Ive actually been meaning to test this out, but utilising the API … since that gets the full sensor resolution, so that could let me get a better feel for what’s possible. but so far not had time.
though I will say, when I looked at the raw api data, it did look very promising (in terms of resolution)
(note: it may be the api is also a little limited, since I suspect it might throttle/filter the data a little to keep it at a reasonable data bandwidth)

they are working on having their own forum, so I think that will help. … not sure, when they are planning on releasing this.

1 Like
16

4 days and still counting from my email to “support”.

2 days ago I forwarded that email to “contact”. Still no reply as of yet. Maybe this weekend? I’m starting to get worried to be honest. It normally doesn’t take this long!! Strangely enough they’ve often replied within 24 hours on other queries. I sure hope they get back to me, as I am pretty sure my Erae has to be replaced.

Perhaps they’ve got a few units with these issues and it will take a bit more time to get back. I know I’m not the only person with this problem.

I certainly don’t think mine can be fixed with software/firmware. Mine is certainly a sensor issue.

17

Another angle, using the Zendrum as an example: it has both a floor setting and a limited set of velocity curves, and perhaps there’s an interaction between the two that make the settings usefully customizable on the user side rather than one size fits all. In my case, I do set everything up as hot as possible. There are only 3 sensitizing curves with the other 4 being desensitizing for “whackers,” so it’s a no brainer to choose the most sensitive of the 3. The floor setting does vary a bit between instruments, which is what I mean by interaction. Being triggers only, though, it’s a much less complicated device, and each trigger can be physically accessed separately for re-seating and adjusting. Plus, they’re all handmade, and the inventor takes my requests for hottest possible construction.

Perhaps create a video, and that will have additional impact on them? As a Kickstarter device, I’ve been surprised at Embodme’s lack of user interaction from even before the device was shipped. It’s not like there are that many users, and even less with bothersome issues that are more user failure than device failure. A lot of user support can be tedious and draining and unproductive for the creator, and simply amount to attitude management, but if they want to handle communications like a big company that absorbs a certain fail rate, then Embodme needs to be willing to issue cheerful refunds to those who simply want out. Otherwise, a greater presence would be welcome.

18

Why should I create a video to convince them? I mean, other people had their units replaced so why can’t my unit be replaced?

Well it’s now 5 days and no reply. At this point I am not sure what’s going on. I almost feel like just giving on up on this thing.

It doesn’t take 5 days to respond to a customer problem. I asked them about XY pads and they replied the same day.

I then mentioned this problem and then…crickets…

When will I get a reply because it feels as though I might not be getting one at this point. Maybe I should just sell it and move on. I don’t like the feeling of being ignored and that’s exactly how it feels.

I contacted Alex 5 days ago and then sent the same email to support@embodme.com. Then I sent it to contact@embodme.com

If I get a reply I’ll be pleasantly surprised. If I don’t ever get a reply I’ll sell it and move on. I cannot deal with something like this obviously and at this point I’m willing to go “f… this” and sell it and take a loss. I paid good money for this and I’m starting to get quite angry.

I am sure you understand.

19

I’ve got quite a few double notes on Erae touch when I tried to use it with Melodics for finger drumming lessons. It was not as severe, as in your videos though. I’ve got a similar issue with Sensel Morph. Both surfaces are unusable with Melodics. This is probably related, as @thetechnobear said, to finger bouncing while playing fast. Some kind of filtering must be implemented in the firmware by @embodme. Double activations are not so noticeable during the real playing, without the MIDI monitor. Also I’m mostly using Erae for playing expressive melodic instruments, which is usually much slower, so I don’t usually notice any trouble from double activations.

20

Id give them a few more days… we really don’t know how many of them there are, and also how many are working full-time on this project… my experience with music tech companies , are often there are very few full time resources on the project, and they are spread over development, production, shipping, support…

this can make interactions a bit inconsistent…sometimes you get immediate replies, other times things can fall through the gaps, and you have to wait or try again.

Ive seen this a little with embodme , often getting next day replies to my emails, others took quite a while to get a response … but then Ive even had a video call with them, so were very supportive and really listening to feedback … so, they are trying, but its difficult.
(and of course, they are not like the big companies who have specialist in support etc)

of course, I do get its frustrating… as you just one an answer, and a course of action.

I’ll have a go with mine and see what I get…
previously when I was looking at the midi (for other reasons), I didn’t see any major issue.
that said, I do seem to remember raising an issue with Embodme that there was some ‘excess’ data being sent e.g. CC / pressure values being duplicated… which I suggested could be optimised to help reduce the data bandwidth.

as for sensitivity…
I remember at some point seeing some areas at the extreme edge of the board being less sensitive, but I saw that as natural (the membrane will be tighter at the edges where its fixed?)

also occasionally, I get ghost notes in some areas of the board… but its never in the same area of the board, so not sure what thats about.

finally, it seems that I needed to have the Erae placed on a firm firmware, sometimes I put in on its (soft) case… just to get the case out the way, and that seemed to cause issues, I put this down to perhaps the surfaces flexing?

but overall, the issues were minor, and I think they probably can be worked through.
but it sounds like others issues may be different…

anyway, will give ‘drumming’ a better look, I cannot say it’s something I really use that much , and just for a bit of quick fun, (and frankly im no drummer, so probably won’t notice issues anyway :wink: ) … so… its quite possible there are ‘hidden’ firmware issues lurking that Ive not spotted.
I assume if I just record a midi track extra notes will be very obvious !