Midi Breath Controller for CC

#1

So I'm experimenting with a pressure sensor going into an arduino to output midi cc values and I have confirmed that it is working. But for whatever reason (probably because I'm a noob), I cannot get Axoloti to register anything. I have a 6-pot controller that works great for sending cc to the axoloti, and it is mapped correctly... and in that box, I replaced a pot with the breath controller signal. Like I say, I have tested both using serial out to the monitor and also MidiOx to confirm it is sending data, but I'm at a loss what I am doing wrong. I use midi/cc object and have tried using a math object with it but no luck. Any ideas? thx

#2

Oh my... I just had an "aha" moment, and all is working. Anyway to delete my silly ramblings :wink:

#3

thank you @mtyas for your wonderful flute patch... I can't wait to experiment with this further. Already ideas for the features I want from it...

#5

Seems to work really well. Great Job !
What pressure sensor are you using ? is it simply plugged into the axoloti's Gpio, or do you need some small circuit as well ?
I'm pleased to see that I'm not the only one that feels like playing wind controllers is close to running up a hill :slight_smile:
I suddenly thought about an axoloti installation with 15 tubes coming out, where people need to blow into them for the sound to be produced, modulated... That could be fun (and full of saliva, that's the whole point :slight_smile: )

#6

thank you @mtyas... yes a lot of fun. I'm using the MXP5010GP sensor. I have 4 more courtesy of their maker... (so let the saliva drool...he he). I just love the sampling programs some silicon makers have (free)... it's inspiring for inspiring minds. The signal is processed by an arduino uno (328P chip) sending the cc over midi to the Axoloti. Hardwired to the gpio would indeed be better but for now I like the flexibility that midi gives me... I can plug it into the daw and control a vst for example. The tube (edit: thank you fish/tank) is currently too long in my opinion and I had to do some math in your subpatch so that I wasn't 'blowing my brains out'. I really look forward to this feature and am thinking of 1 or maybe 2 tubes nearby a microphone for easy access...15? That does sound like fun but what if more guests arrive? :wink: Maybe a main constant pressure 'header tube' that we plug into? :slight_smile: cheers

#7

I once made a very crude breath sensor with a LED - Photo resistor combo (sort of like a torn apart vactrol) where the LED and the photo resistor were in a chamber separated by a flap valve. As you blew into the tube, the valve would open and more light would shine on the photo resistor. VERY crude but easy to make and would be hardly any circuitry to get working with the Axoloti.

#8

the latest update on my prototype...

#9

Wild! Like the LED lightup in the Easter-egg looking baffle too.

#10

thank you. I was lucky to find a bright blue led in my kit... it gives it a little character I think. :wink:

#11

This is really great @timeorspace,
i wonder how this patch would sound with your breath controller.

menu: File -> Library -> Community -> tiar -> reed -> Misty

You will have to edit it so that it will be controlled by your breath controller instead of the aftertouch...

BTW and slightly off topic, the portable QuNexus keyboard has two CV inputs (it provides ground and a ref +5v), i wonder how difficult it can be to connect a custom breath controller such as yours :star_struck:

#12

buy a mpxv4006, and you are all set...

#13

Love this answer, the supply current is 10mA max, so even if it is a "ref 5v", it will be ok :smiley:

#14

they have a whole range of breath/pressure sensors, check the different models, different pressure sensitivities, second pipe for output etc...

#15

Misty sounds great, and thank you @SmashedTransistors... I am a fan of your work. I am actually messing about today with music... very loosely based, and planned to use some sort of wind instrument. I am not familiar with the keyboard you mention but can say the sensor only requires 3 leads... and wired just like a pot. So it sounds like you are half way there since you have GR and +5V... just hook up to signal in. I am supplying 5V to this sensor... and the led, (a bit of a gimmick, but fun), just 'rides the rails' so to speak.

#16

I'm using the MPX5010GP and will look into the differences... thanks.
off topic, but NXP has a good sample program... I have 4 more ...he he

#17

I've been having a bit of a hardware issue...
You see, I have a home built controller using the arduino to send midi into the DIN, and usually everything works fine. But in this big patch instance (Misty), it seems the midi box is taking too much 'overhead'. I have a controller object patch for this in preferences which maps everything, and if it is enabled, I run out of memory with "region `CCMSRAM' overflowed by 213 bytes.". But if I disable the DIY midi box the patch loads just fine at about 41% load.

I've tried setting all the preset values and targets low as suggested in other posts to no avail.
Any ideas? Posted here because of the relevance to the project.

#18

You'll need to post the non-working version of the patch.
... it's a memory issue not cpu, you need to move stuff from sram to sdram, probably tables/delays.
could also be a code issue, depending on patch complexity.

As such, midi input will not take up extra memory , and midi objects are pretty light.

Of course could be a combo of lots of small things.

#19

Library>community>tiar>reed>Misty

#20

Oops, i think it is because i abused of allpass filters in the reverb. They tend to eat a large amount of internal SRAM (that's why SDRAM should be used for long allpass and delays).

To free some RAM, you can edit the reverb object:

and suppress these two big allpass filters:

and reconnect the others:

once done, don't forget to click on the "update" button of the reverb patcher object (i often forget to do it...).

Hope it will free enough RAM.

#21

Thanks for having a look @SmashedTransistors I'll give that a try and report back.