Pad sensitivity not that great for medium and hard hits

[drumsdibums]

@ Ignotus thanks a lot for your very clear and detailed answers!

I think I know where I have to head next and see where it goes.

The basic setup is now clear to me.
Hopefully I find time to try it out in the not so distant time... will then report here.

[drumsdibums]

Update: I spent the whole day testing out different trigger designs..

BUT: > Nothing actually changed!! My Issues stay nearly 100% the same!!??

to recap: these Issues where:
- 1) only good playing dynamics on soft hits (bellow 20% force)
- 2) somehow squashed notes when playing something fast. (I have an audio example attached)
- 3) All (Ring) RIM Inputs NEED a Threshold of 36 whereas all (Tip) Head Inputs are fine with a Threshold of 4-6 to avoid fast ghost notes triggering
(This was an Issue I mentioned in my other Thread here:http://www.megadrum.info/forums/viewtopic.php?f=3&t=3121
I made sure that the ring Piezo (outer diameter) was free to vibrate... this did not change anything in this regard

- 4) Any hit between 40 and 100% force mostly create a MIDI Note with the max possible value.
This might be important... I forgot to explain this earlier:
I thick "high level auto" and enter a high level value of 100, then update... I do some strong hits...and some very strong hits are able to push the high level value a bit more... my result is 548. I untick "high level auto" , update and start playing.
Now even relatively low force hits (maybe 30% max force) are EASILY triggering the Module to Output the max MIDI Value of 127.
If I enter a higher "high level" lets say 600...my low force hits from before trigger a MIDI Value of 111. But this is now my new Max Value! No matter how hard I hit I will never exceed this (new limit) of 111.

What I observe is this: the dynamic range/response of my Piezos kind of stop after 30-40% force hits.
also when drumming with my finger directly on the Piezo... it is really sensitive...but reaches its maximum quite fast/easily...
This issues where tested with different side mounted trigger designs.. so hotspots are not a issue I would think.

I think my triggers would need to be way way less sensitive to work correctly, >so I get a differentiation between 40-100% forced hits
but still manage to register the soft hits..

Different trigger designs:
--------------------------------

I tested different designs, the results where almost identical and exchangeable (except for the rubber pad design):
(> please see the images for the different Trigger designs)

1) Meshhead>directly mounted Piezo>Foam on the ceramic part
2) Similar to 1) // Meshead>"chewing gum like sticker">Piezo>Foam on the ceramic Part
3) Meshhead>3mm dense foam>Piezo> Foam on the back of the piezo leaving a ring of 3 mm
4) Similar to 3) // Meshhead>3mm foam cone>3mm dense foam>Piezo>Foam on the back of the piezo leaving a ring of 3 mm
5) The design suggested by Ignotus: Meshhead> foam>ceramic part of Piezo>two foam poles on the outer side of the piezo>piece of wood

I tested Nr. 4 and 5 with a very soft piece of foam from a kitchen sponge instead of the 3M sanding foam i usually used.

my impression was: softer foam = less "signal" the pads where quieter but also less sensitive... (high level was lower)
very soft hits did not get registred... it was harder to get full value with relatively soft hits... but just a bit... the issue that medium soft and really hard hits get about the same MIDI Value persisted.

-> I also tested a 10k resistor in line with the piezo hot signal. this did not do much.
-> I also tested a 250k logarithmic potentiometer I happend to had here... this also did not do much...except that the soft hits did not get registred that easily

Rubber Pad
6) 3mm Rubber Pad>2mm wood> foamy scotchtape>Piezo>3mm Rubber Stripes>two sided tape to desk
7) Similar to 6 // 3mm Rubber Pad>2mm wood>4 dots"chewing gum like sticker" on the outer perimeter of the Piezo>Piezo>3mm Rubber Stripes>two sided tape to desk

These had a high level value of around 700. These did not solve or improve much on the issues I had...but otherwise worked well.
I could set all parameters close to their lowest value (ex. dyntime to 3) without having any issues.
I had some hotspot issues. Nr 7 improved that a bit. If I hit the table close to where the Pad was with medium light force the pad registred a hit (value 4)... but I guess this could easily be fixed with a bit better isolation, crosstalk settings and raising some of the ridiculously low settings.
-----------------------------


I forgot to record the actual output of the piezo with my audio interface...> this might help to shed some light into the issue?
Will add that later...

I attached a mp3 file to illustrate the squashed note (and dynamics) issue: Right is MIDI Drums / Left is the Real Audio from the Rubber Pads)
You first hear some stick bounce... then some hard hits then some fast drumming.
At the end you can clearly hear that the time alignment of the MIDI Drums is different to the real Inputs... It detects 6 hits correctly... but the timing is not right.

********
So far I have no Idea how to solve these Issues...??? Really need some help to get this going!

? > could it be that this Piezo >4,4 kHZ / 200Ohm / 21. pF / 27mm Found here: https://www.reichelt.de/Signalakustik/E ... PZ-27MS44W
simply does not work??
Which Piezo exactly can you recommend?

Thanks to anyone for their inputs!

[drumsdibums]

Some more Images...

[drumsdibums]

the resistor and the pot I used...

-> + the mp3 Audio file

[drumsdibums]

max file size is 256KiB???

...mp3 sounds now really bad and distorted...

[airflamesred]

This might be important... I forgot to explain this earlier:
I thick "high level auto" and enter a high level value of 100, then update... I do some strong hits...and some very strong hits are able to push the high level value a bit more... my result is 548. I untick "high level auto" , update and start playing.
Now even relatively low force hits (maybe 30% max force) are EASILY triggering the Module to Output the max MIDI Value of 127.
If I enter a higher "high level" lets say 600...my low force hits from before trigger a MIDI Value of 111. But this is now my new Max Value! No matter how hard I hit I will never exceed this (new limit) of 111.

1. Tick high level auto
2. Strike the drum a number of times with fairly hard hits.
3. You may need to press 'get' to retreve the number from MD
4. Untick high level auto

This is to adapt to your style and what the piezo is picking up. Make sure you are using a linear curve and that MD/MDM are updating each other.
You should get, with these settings, something to work with. If not then it's a physical issue.
With regards to the MP3, The last section (ruffs, rolls), Is that a real time recording?

[drumsdibums]

Hi Airflamesred,

Thanks for joining in!

Your points 1-4 > That was exactly what I did to get the high level...
(I did not press "get" but I could see the numbers going up... and then I hit sent.. I am sure this worked correctly)
I used linear curves. Had the impression I could get a better "feel"/response with log3 curve.
I mean I could get something to work with, it's not that it did not work at all, it is just that I could not solves this issues I mentioned.

The mp3 file: the last bit is the interesting part where it alternates between the real hits on the rubberpads and the resulting recording from the Megadrum output via MIDI in Logic Pro X. So both should sound identical (apart from the "voice" obviously). It is a real time recording but the last figure is the MIDI output but slowed down from 120bpm to 40bpm to illustrate the issue. (I could not slow down the audio file without getting a lot of artefacts, so there is only the slowed down version of the MIDI Drums). The Modul registers all of the hits (6 in total) but the timing alignment is not the same as the input in my opinion.

If you listen to only left and then only right channel separately... You will realize that the original (Input) recording on the rubber pads has A) some dynamics build ups/crescendo and B) is correctly timed and played at least half decently.
...the resulting MIDI Output has A) some suddenly and unexpected loud hits (out of nowwhere) and B) Sound like a beginner is playing something he can't play properly and so it sounds rushed and floppy..

[drumsdibums]

What I might try out... but I am not to confident now that that might actually eliminates my issues are:


-1)Changing the Piezo
(Are there Piezos that work really good and some that do not work properly at all??? > I think I would need one with a better dynamics response (or at least a dynamic response thats translates better.

Sure there are many people around that also build their triggers and could give me some input

-2) Use some softer Foam for the Triggers
-3) Tensioning the Mesh-heads really hard

- I tested my triggers with a 10k Resistor as well as a 250k potentiometer. surprisingly this did not change much (I left the settings the same) because I thought they could benefit from cooling down. The only change I had was that the triggers lost their sensibility for low value hits

So my guess after long testing is: - the triggers are to sensitive in a way that I will always hit the celling already with 40% force hits.
- there is something wrong with the Modul.

I don't know anyone with Edrums...
If they let me test things out at a Drumshop I might test my Modul with some Roland/Yamaha Pads and my Mesh Head triggers with some Roland Modul.

*like I said... I will record the Piezo Output with my Audiointerface... this might give some insight as to what is going on...

[drumsdibums]

@ Ignatous I just found your statement bellow in an other post.

> To me, this looks like you had exactly the same issues I had! Probably minus the squashed notes Issues and the thing that I need to turn threshold high to stop ghost notes triggering on the rim input.

by Ignatous in the Thread: DIY Piezo wiring: You might have gain at 0 and a highlevel of 500 - apparently all's fine. But then you see that the velocity goes from, say, 30 to 90-127 too easily, skipping all the mid-range velocities, no matter what curve you use, and if you set highlevel to 600 it maxes out at a velocity of 100 or 110. With a trimpot you can get back all that mid range and adjust your gain so you can have a highlevel of 800 if you want. You also have to find the sweet spot but it's easier to turn a trimpot than to mess around trying to physically dampen a piezo by "exactly" the right amount.

So I probably need a pot / voltage divider to get good results?
I have tested that, but I might have not wired the pot correctly... Should that work also with a 250kohm logarithmic Pot? I don't have anything else here for testing...

the raw ouput of the piezo might help to shed some light into that issue... will post this audio file later

[drumsdibums]

I recorded the output of my Piezos with my Audio Interface.

I used the rubber pad design shown earlier, connected the Piezo to my DI Box and then to the Audio Interface.
The recording showed that all hits get registred... so the issue I had with "squashy notes not timed correctly" which you can hear on the mp3 posted earlier might come from somewhere else...maybe some megadrum settings? What settings could cause such a issue?

It might be that my Piezos are WAY! to loud/hot and that this is the cause of all my problems??

my DI Box (Hi Z Input for passive Guitar and Bass Pickups) easily clips even with medium hits on the PAD.
If I engage the PAD -30db. It works without clipping... But I am still getting a high reading and close to use the full headroom of the DI Box.

my DI Box (Palmer PAN02) has a Headroom of +15dBu or +45dBu when the Pad is engaged.
( I know, (I have measured it), that the DI Box will lower the balanced Mic Output by -4.5db compared to the same signal to the Hi-Z Input without a DI Box.
If I get a reading of -4.5db or higher into the DAW with MicPreamp Gain set to 0 I know that I am clipping inside the DI Box)
The strongest hit on the Pad resulted in a reading of -7.5db inside the DAW. (DI Box -30db Pad engaged)

I used this callculator http://www.sengpielaudio.com/calculator-db-volt.htm to convert dBu values in Volt:
- For 45dBu I get: Level Lv 42.78 dBV // Voltage V 137.75 volts (RMS) // 389.6 volts peak to peak
- For 15dBu I get: Level Lv 12.78 dBV // Voltage V 4.35 volts (RMS) // 12.32 volts peak to peak

I think the 45dBu results are important... these seem to me absurdly high?!

So.. anyone with some insights as to how the megadrum box works, what level (dBu / Volts) it expects to work proberly can help me??

- Are these Piezo to Hot?
- Do I need to buy some other Piezos? smaller piezos that are quieter but still sensitive enough?
- If someone knows a (specific) Piezo that works well and can be ordered in Germany (from conrad or reichelt preferably) that would be great!

---------------------
---------------------

I also tested the Pad again with the pot wired correctly.
I could turn it down and it was harder (or impossible depending on the position of the potentiometer) to get the highest values.
But most notably my Pads where not sensitive enough anymore to be useful. (megadrum settings dyntime/level retrigger mask etc. where already very low)
> Without a Pot...like stated before the sensitivity and response to light (it did not register extremely light hits) hits where great.