real bass drum by midi

Status
Not open for further replies.

iRt

New member
Hello, first post here.

I'm thinking about a droid drummer that would have electronics (drum machine) for everything except the kick drum would be real. I'd be using a calf head because that's the sound I'm looking for.

Ideally what I'm looking for, although I doubt it exists is: a device with a beater I'd clamp onto a bass drum that has a MIDI IN jack on it. When it receives a MIDI note on the beater hits the drum.

Ideally it would have as few parts as possible. It needs to respond to MIDI velocity.

Another possible solution is to have a woofer speaker (playing a kick drum audio sample) mounted backwards (facing inside) of a drum, like the batter side of a kick drum and use a calf head on the front (reso) side and the sound would be coming off of the calf and that might get me there and be easier...

Any ideas please!
 
The electronics part is not hugely complex since there are a number of midi libraries around for either classic serial or USB midi

The tricky part is getting the forces needed to play the notes
Stepper motors might do it
https://learn.adafruit.com/adafruit-arduino-lesson-16-stepper-motors/overview
or solenoids, much like your speaker idea (or possibly even a gutted speaker if you used ones around)
https://learn.adafruit.com/midi-solenoid-drummer/overview
For non drum instruments servo motors are easy (but possibly slow)
https://learn.adafruit.com/adafruit-arduino-lesson-14-servo-motors

Suggestion would be to leave the midi part for the moment, get some basic hardware and explore the various methods to operate the drums in isolation looking for a robust and powerful enough solution you can replicate enough with your skills and available tools. Also check for how much power the final solution might consume when you multiply the project up.

A lot also depends on what you want the final project to look like. A steppor motor driven arm that moved between multiple drums would be cool but hard. A row of solenoids on a massive beam with the drums bolted to them in a straight line easier to build but inflexible.

One important thing to think about is that any mechanical solution will be slower than electronic reproduction, so you will either need to accept a fractional delay on your bigger instruments or work to shape all delays to be equal.
 
A pulling solinoid connected to the chain of a standard pedal.

Teensy controls higher voltage/current to the solinoid thru a MOSFET.

Latency may be an issue.
 
Having tried to make pinball machine flippers with many speeds and torques of servo and failing, I feel a solenoid is likely the right path for percussion. They offer impressive speed and power, and come in a large variety of throws and sizes.

I needed a dedicated power supply for my solenoid based flipper mechs because they needed ~24V; I think I used a laptop supply. IIRC I used a IRL540N on a (similar) ball launch solenoid, though I see a TIP120 used in a lot of homebrew arduino pinball ( https://www.bc-robotics.com/tutorials/controlling-a-solenoid-valve-with-arduino/ ). Don't forget those snubber diodes.
 
Thanks everyone for your replies.

As far as what the final project would look like: I was a drummer in bands from the late 60's until '83 when the MIDI thing came out I started sequencing and that went on for plenty of gigs and recording. I ended up sequencing a lot of songs (like 1000 or so) but they never seemed like tracks like some people thought. To me they were more like artificial intelligence. I ended up using the same instrumentation on near every song, so it was like a band that had been together for a long time. The bass, drums and "muted guitar" tracks seemed to turn into actual individuals over the years. I've always, well since the 80's, pictured that I'd end up with a few droid musicians on stage. It seems like to be the simplest I'd need a drum guy, a bass guy and another "everything else" stereo guy. These would be piles of gear, but ideally very pleasing to look at. Since I'm in Hawaii it would no doubt have a tiki, bamboo, grass skirt theme like most of my stuff has. So for instance the drum guy would be a kick drum, a snare and a stereo pair of speakers for the rest (toms and cymbals). The kick would have a large calf kick and if electric would have a say 100W amp (I'd use car stereo amps) for just the kick. The snare would probably be another 100W amp with a 12" speaker and "the rest" would be some sort of left/right amp/small speaker setup. I would aim to make it so you saw nothing but wood, bamboo and grass skirt stuff. The electronics would be hidden so it would look like a "what is it?". The bass guy would be just a 15" speaker and amp dressed up in similar tiki garb. Lots of carved palm tree stuff. And so on for the "everything else" guy. I've made a lot of the gear I use on stage so the carpentry and electronics is something I think I suck at but I always seem to work it out.
I've been watching a YouTube vid about this place in Nashville that records bands live with vintage gear, often with just 1 - 3 micas. So that's my idea. Get my musician friends over and record along with my MIDI droids live, no overdubs and do all the mixing by mic placement. Use 3 mics, Neumann U87 for the middle, and 2 Neumann KM84's panned hard left and right. No overdubs, a recording of a live event. No headphones, lots of track bleed, no problem. I've grown very tired of the "laying down songs track by track with headphones" thing.

Right now I'm thinking it would be best to build the drum droid (better name please) with a speaker facing into a kick with a calf front head vs using robotics. I need to try it to see how it works. I have a beautiful 28" (thunder sounding) old Rogers marching bass drum shell that would be a possible candidate. Years ago I bought a pair of Infinity speakers at Goodwill for $10 or so. They were a Made in USA not-cheap product. I re-foamed the speakers and each cabinet had a woofer (8" or 10") and also a resonant speaker of the same size. The resonat speaker had no voice coil or electronics going to it, but just pumped with the pressure the real woofer made. It was unbelievable how much bass it put out (although the sound of the speakers, the cones being non-paper, was horrible). Since then I've wonder if someone could mount something like that in a bass drum, in the from head. Well, this would be a bit like that with a speaker facing into the drum and a calf front head. For now it seems like I should go with the electronic (sound sample, I'm using a Roland SC-8850 SoundCanvas mainly) method and see if it is ok. I'm concerned about the latency of robotics and if it would vary with the tempo of songs and fluctuating ac voltage, stuff like that. The sound I'm looking for with the kick is a super low sound with hardly any high end, like you get when you use a huge lambswool beater, not played loud but having tons of bottom. The much I do is sort of radio jazz/whatever/Zeppelin/Miles/Hendrix/Jobim if that makes sense. I use a lot of Fender Rhodes and don't need to play real loud.

So for now I'm going to concentrate on making it sound and look good, not weigh a ton and be as simple (wish me luck) as possible, but long term it has to involve robotics.
 
It may be worth looking at some of what Martin of Wintergarten is up to, especially the discussion around latency and timing here https://www.youtube.com/watch?v=dNv8AeZfuUc

While your methods of control are different some of the discussion and testing happening with the team behind the marble machine may be of interest in terms of what mechanical effects produce useful interesting notes reliably.
 
Thanks, that is very interesting stuff. I'll bet Martin had a Mousetrap game when he was a kid, he probably had six and combined them into one that worked.

Latency with digital audio was hell in the late 90's. I still don't like digital audio, it's horrible to my ears. I've always noticed that since the sounds mix inside the computer the timing with MIDI tracks needs to be spot on. The other side of this is a band like the Grateful Dead where the timing is what some might call sloppy compared to say Tower of Power but since the sound of the kick drum and the bass player is being mixed in the air it's light years more forgiving. The air acts like a compressor and makes everything blend in a nice way.

If you think about playing a tuba, the player has to take a breath and start playing a considerable amount of time before the note comes out. The musician ends up doing it by instinct but what's involved actually takes a lot of mental calculating. It's amazing how sensitive our ears are to timing, it can be 1/1000 of a second off and it's noticeable. When MIDI started out we were dividing the quarter note into 384 divisions and that was nowhere near accurate enough. Reminds me of camera megapixels, where you'd need over 500MP to start approaching what our eyes see (and at least double the amount of colors, probably quadruple).

Another huge factor that makes the speaker/electronic angle more appealing is tonal variation. A human has a hard time ever playing two notes exactly the same (you can't really) whereas the robot has a hard time doing anything but that. We as humans, are always looking for patterns, like Orion's belt. Hearing two notes the same is something that can sound very unnatural. You never hear it in nature. That needs to be addressed. That's why I'd only be looking at the kick drum because in popular music the kick can sound almost the same a lot of the time and still work. But you do want a different kick sound in bossa nova than in disco so that's another place that the electronic angle has an advantage. I'd be trying to change the electronic back to analog. You always do that in music because in the end you are playing the air in the room, whether by acoustic or electronic instruments.
 
Last edited:
Status
Not open for further replies.
Back
Top