Sensing Velocity for MIDI using Arduino

Bruce, do you already have a key-on/key-off system in place? Or in simpler terms, a key-press detection system without any velocity detection?

Hi,

Yes, there's a system in place in case I can't develop the VS ability. It's a VS-capable midi controller board made by midiboutique.com, model MKCV128SN, which offers 128-note capability. The nice thing with this company is that at time of purchase I can describe the various functions I need (in my case this'll include octave up and down switches, a level control, and Pitch amd Mod' wheels) and they'll burn the processor with these options. So I'm assuming they can also burn in whatever's needed to accept VS data from an Arduino/force resistor circuit. The only thing that's bugging me is what I'm calling the release factor. Perhaps hard to explain if you're not a keyboard player, but I'm assuming that if I press a key, and the Arduino circuit reads and sends the key's velocity data, will that key need to be released before VS data can be read again? I'm guessing so and this may cause problems if you're plating arpeggios that require the notes to be held

down as succesive note are added. Like I said, it's a little hard to explain to non players, but if you do play keyboards then you can likely imagine what I'm getting at. The VS data for the first key played will likely establish the VS for any subsequent keys played if the first key is not released.

This fellow has done exactly what I'd like to do, using 4 force resistor strips under a toy keyboard, probably to establish some independeny along the keyboard in dealing with the situation I descibed above. Actually, since these strips can be cut to specific lengths, couldn't I do one for each key?

Here's a YouTube video by a guy named Jen Shen who's installed the force resistor strips in a toy keyboard and achieved not only velocity sensitivity but also after touch:

http://www.youtube.com/watch?v=yIBBxDHOa0g

Once a key is held down, after touch allows the player to alter the sound by changing the pressure on the key.

Here's Jen's Jen's tutorial on cutting the strips:

http://www.youtube.com/watch?v=x-pQ651_iZY

By the way, this is the same fellow I saw originally that lead me to the idea of using these strips, unfortunately he won't answer any of my inquiries--probably inundated with emails and busy. Go figure. There's a ton of info on this. On Google I searched "midi velocity diy arduino" and got all kinds of returns, but not being familiar with Arduino and tghe associated verbage I could very well be looking at exactly what I need and not know it. Arrrg...

Well, I appreciate your interest in this and hope you can assist me in making it work.

Best regards,

Bruce Nelson

Toledo, OH

Hello there Bruce, I understand a little bit of keyboard playing, so I can perfectly understand what you're getting at! . I'm not entire sure if I can help you, but here is my two cents about the subject;

I'm familiar with the project that you've linked, did you've checked the blog post about it? in it you can have a schematic and even a part-list. The main problem is that for every Velocity/Aftertouch detector, you'll have to expend an average of 4$. He uses those "slide" pads that i'm sure you're familiar with. So, I'm not sure this is the way to go, if you're looking for a keyboard with a decent amont of keys.

I'm not entirely sure if I will help you or confuse you even more, but here it goes.

The reason why I've asked if you already have a "Note-On/Off" system in place, is because for velocity calculation you would simply have to duplicate it. You would put two pressure pads on each key, one would be activated if the key is just touched and the other only when the key is fully pressed. Using an Arduino you would then calculate the delay between the two Inputs and have the "Velocity".

Not to barge in on your conversation.....

Would an accelerometer be of any use.?

It will detect the force of the key hitting home and transfer that into a 'z' axis.(up/down)

You know what key and from the accelerometer, how much force was applied.

At worst you might need three to cover the whole organ more evenly.

I wouldn't have thought this was possible, until I ran across the "secret knock' example on mbed.org using a KL25z board.

With a setting of about 1/2 g the knock on the table is not at all loud, so the sensitivity is there.

Mark

Hi again,

Well, with the Midi Boutique controller in place, I'll have Note On function. It was my hope to be able to augment it with the resistive force strip(s) in order to provide the VS data. In a perfect world I'd have my CNC up and running and would simply machine 61 small SPDT lever switches to reside under each key and the whole resistive strip idea could be chucked out the window. But I can't afford the CNC. haha Maybe next year. I guess my hope here is to learn how to collect and process analog data into digital data using Arduino. With that ability any number of methods could be tried. I still like the optical method, too. A guy named Sebastian Steppler, a German engineering student, has perfected an array of optical sensors, called Epick, that reside under the keys and provide VS and after touch data. It's really cool but he won't bring it to the US for fear of copyright infringements with companies doing the same.

Well, I appreciate your advise and interest in this. I'll talke another look at Jen Shen's blog and see what open source material is available.

Thanks again,

Bruce

Hi again,

Well, with the Midi Boutique controller in place, I'll have Note On function. It was my hope to be able to augment it with the resistive force strip(s) in order to provide the VS data. In a perfect world I'd have my CNC up and running and would simply machine 61 small SPDT lever switches to reside under each key and the whole resistive strip idea could be chucked out the window. But I can't afford the CNC. haha Maybe next year. I guess my hope here is to learn how to collect and process analog data into digital data using Arduino. With that ability any number of methods could be tried. I still like the optical method, too. A guy named Sebastian Steppler, a German engineering student, has perfected an array of optical sensors, called Epick, that reside under the keys and provide VS and after touch data. It's really cool but he won't bring it to the US for fear of copyright infringements with companies doing the same.

Well, I appreciate your advise and interest in this. I'll talke another look at Jen Shen's blog and see what open source material is available.

Thanks again,

Bruce