Controlling Analog synths from Max?
I thought I was going to be able to do this with an Arduino, but since they use pulse width modulation for the "analog" outputs, I don't think I'll be able to use the analog outs of the arduino to directly interface and control the paia 9700s i just started building.
Are there any users out there who have gotten Max to control an analog modular through CV/gate controls somehow? What kind of hardware or circuits/mods (I'm not afraid to solder) did/are you using? I'd LOVE to see Max controlling, and maybe even acting like a "analog" module.
I've read some postings about using certain audio cards to control modular synths, but haven't seen any definative answers on successes or failures.
Basically, I wanna use midi and digital audio inside the computer to interface in a analoglike way with a modular synth without using any MIDI to CV/gate hardware converters.
-Don
Any other boards out there work better for this?
Make controller? Wiring controller? Some other controller that works with OSX?
Or does anyone know have a schematic to convert 0-5 volt PWM signal into a 0-5 volt constant signal?
I had a hunch that you could do the conversion with a low-pass
filter. If the filter's cutoff is low enough, it will be doing a
running average. There will likely be some non-linearity involved, so
you may need to account for that in your control software.
Check this out:
I would imagine (because the analog filter will be essentially
averaging) that there will be some lag, and the filter will have to be
tuned to the pwm frequency and your acceptable latency. Luckily you
can model this in MaxMSP with a biquad, given that you know what the
frequency of the PWM coming out of the arduino is :)
_Mark
On Jan 2, 2008, at 3:53 PM, Don K wrote:
>
> Any other boards out there work better for this?
> Make controller? Wiring controller? Some other controller that
> works with OSX?
>
> Or does anyone know have a schematic to convert 0-5 volt PWM signal
> into a 0-5 volt constant signal?
>
> --
> -DonK
I saw that page this weekend. I haven't tried it yet, I have to finish building the synth first. :)
Thanks for the tip on biquad. hopefully I can start some testing in a couple weeks. I was hoping that there were people who were successful at this max controlling analog equipment concept.
I'd love to use Max as a super fancy analog sequencer and maybe even use Max as an effects processor/sound generator for within the analog chain.
I've been using Max to feed 2 galvos in a laser scanning system using a soundcard.
The basic issue for me was that most soundcards contain a stage that blocks DC offset. If you look on the alt.lasers newsgroup, you'll find a bunch of people who have modded cheap soundcards to take that stage out. It essentially just means bridging over a cap on the output stage.
But I've found that the MOTU 828 doesn't seem to block DC offset, so you can use this as is. You'll get a range of voltage between +/- 1v, then you need to amplify it. Voila. 8 channels of CV.
I think.....
I could be wrong about this, but that's what my oscilloscope is telling me. Probably not the most stable source though. Plenty of overshoots on the scope at transition points.
Good luck...
Mark Pauley schrieb:
> Luckily you can model this in MaxMSP with a biquad, given that you
> know what the frequency of the PWM coming out of the arduino is :)
Though its correct, that a low-pass should do it, you can't patch it, it
has to be hardware, basically a condenser and a resistor...
Stefan
--
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com
At 3:49 PM -0800 1/2/08, Don K wrote:
>Basically, I wanna use midi and digital audio inside the computer to interface in a analoglike way with a modular synth without using any MIDI to CV/gate hardware converters.
Just out of curiosity, why are you so set against MIDI. With a proper interface it works pretty well for this.
I'm using a custom Max/MSP program with a Buchla 200e and having a great time with it. A little info about my program can be found here: http://www.xfade.com/Gyre
-C
--
Chris Muir | "There are many futures and only one status quo.
cbm@well.com | This is why conservatives mostly agree,
http://www.xfade.com | and radicals always argue." - Brian Eno
Haha...
Nothing against MIDI, just trying to stay away from 0's and 1's for now. I'm just stuck in a phase where I'm trying to learn some electronics (I started building some paia kits last year), learn Max (I'm a new user), and learn how to interface between the physical world and Max using various sensors from pots to accelerometers.
I will definitely be using midi to control the synth, but I am also obsessively curious about using Max's number crunching flexibility to connect to analog equipment and control it with "digital precision" (+ some lag).
I'm jealous of your Buchla... :)
Is your Buchla controlled by midi from max?
I wish I could see you play with Max and hear what the Buchla sounds like. That DJ interface you use with it reminds me of a way slickened up version of the control box I'm working on. The arduino would be great for making a custom HW interface if anything. One lightbulb that just went off in my head would be that CV data out of the synth and into Max might be cool to use somehow. Kinda like letting the synth be "MSP".
I'm obsessed with this topic, as well, but I have only done the same research as you and have come up with the same findings. I can't really contribute anything other than interest, though. I, too, would like to control my analog modular with Max sans MIDI. And use CV from my modular to control Max...
Hi Don,
If you allready started soldering, this might be the best solution for you:
kits available here
It is proper 8 channel MIDI to CV/gate DIY interface. Cheap kits are
available. It works great.
klif
maybe you should check out the ME-72 or multi exchanger 72 from Showtec or Botex... it has 72 outputs from midi, and costs something like 200 bucks. i'm controlling a lot of analog stuff with it (mostly lights). Although it's really a good thing for making LFO's and slow stuff, the timing accuracy probably isn't good enough for fast sequences etc.
Hi Don,at least one of my gluion customers is using PWM to control analog
synths. To avoid the PWM frequency being audible he simply sets it to above
20kHz. With the gluion's main clock at 20MHz, this allows for 10bit
resolution, roughly. You'd have up to 66 channels and timing resolution down
to 1ms (depending somewhat on the network lag).
So this is a start, but a bit more resolution would be nice especially if
you want to control pitch.
A short term solution is to clock the gluion higher (or rather the PWM
modules). 100MHz should be possible, which would increase resolution to
about 12bit. This is possible in software and on my to-do list.
Eventually though I want to provide a "real" DAC module with full 16bit
resolution. Might happen as soon as mid-2008, so watch out for announcements
- newsletter and blog, as well as more information at www.glui.de
Sukandar
On 1/3/08, Don K wrote:
>
>
> I thought I was going to be able to do this with an Arduino, but since
> they use pulse width modulation for the "analog" outputs, I don't think I'll
> be able to use the analog outs of the arduino to directly interface and
> control the paia 9700s i just started building.
>
> Are there any users out there who have gotten Max to control an analog
> modular through CV/gate controls somehow? What kind of hardware or
> circuits/mods (I'm not afraid to solder) did/are you using? I'd LOVE to see
> Max controlling, and maybe even acting like a "analog" module.
>
> I've read some postings about using certain audio cards to control modular
> synths, but haven't seen any definative answers on successes or failures.
>
> Basically, I wanna use midi and digital audio inside the computer to
> interface in a analoglike way with a modular synth without using any MIDI to
> CV/gate hardware converters.
>
> -Don
> --
> -DonK
>
A device that does OSC to modular synth level signals (AC and DC) and vice versa would be a real boon, man!
Thanks for the tips Klif and Daanbr, I'll keep those converters in mind, but kartadinata seems to have a couple possible solutions. The 16-bits are definatly better than the 10-bits in and even crappier 8-bit out that the arduino spits out.
I saw the Gluion board's site a while back when i was first researching HW i/o, but the costs scared me away, especially with the US dollar so weak now. The "real" DAC planned for 2008 sounds exactly what I'm looking for and trying to hack together. Please announce it to the cycling forums! :)
Now I wish I had studied electrical engineering in college....
On Jan 3, 2008, at 12:00 AM, marcus lyall wrote:
>
> But I've found that the MOTU 828 doesn't seem to block DC offset, so
> you can use this as is. You'll get a range of voltage between +/-
> 1v, then you need to amplify it. Voila. 8 channels of CV.
>
> I think.....
>
> I could be wrong about this, but that's what my oscilloscope is
> telling me. Probably not the most stable source though. Plenty of
> overshoots on the scope at transition points.
I've had great luck using a MOTU 828 mkII to generate CVs for a MOTM
system I've built. If memory serves the scope said it was generating
+/- 5V which meant I didn't need any amplification. More often than
not I'll use a traditional MIDI-CV converter (MOTM-650) to handle
pitch/trigger CV generation and the 828 to generate "non-traditional"
LFOs and envelopes that I can't easily generate in hardware...
Cheers,
-greg
A good solution to turn PWM into an analogue voltage is to use a 4th
order (24 dB/oct) low pass filter (for example butterworth or
bessel). You will a get very clean output signal when the filter
cutoff frequency is 4-5 octaves below the PWM frequency.
It does need soldering but using these filters could turn for example
the GLUI into a great 66 channel OSC controllable analogue interface.
One of the great features of the GLUI is that the PWM frequency is
very high and can be very accurately controlled.
here is a link for somebody who would want to design a filter like this:
http://www.maxim-ic.com/appnotes.cfm/appnote_number/1795/
best,
Edwin
On 5-jan-2008, at 2:56, Sukandar Kartadinata wrote:
> Hi Don,
> at least one of my gluion customers is using PWM to control analog
> synths. To avoid the PWM frequency being audible he simply sets it
> to above 20kHz. With the gluion's main clock at 20MHz, this allows
> for 10bit resolution, roughly. You'd have up to 66 channels and
> timing resolution down to 1ms (depending somewhat on the network lag).
> So this is a start, but a bit more resolution would be nice
> especially if you want to control pitch.
> A short term solution is to clock the gluion higher (or rather the
> PWM modules). 100MHz should be possible, which would increase
> resolution to about 12bit. This is possible in software and on my
> to-do list.
> Eventually though I want to provide a "real" DAC module with full
> 16bit resolution. Might happen as soon as mid-2008, so watch out
> for announcements - newsletter and blog, as well as more
> information at www.glui.de
>
>
> Sukandar
>
for osc to voltage, have a look at this;
www.koncon.nl/ipsonlab
I made one, it's wicked
Quote: daanbr wrote on Sat, 05 January 2008 04:56
----------------------------------------------------
> for osc to voltage, have a look at this;
>
> www.koncon.nl/ipsonlab
>
> I made one, it's wicked
----------------------------------------------------
This looks like just the thing! Thanks.
Care to mention anything you've used it with?
Whatever it takes, I must make one of these with banana jacks to interface with my modular synthesizer.
I just checked the specs, it's using PWM with 8 bit resolution as well.
8 x analog outputs (Pulse With Modulation),
resolution 8 bits [0-256 / 0-100%] **
Cool device, but it seems like a osc version of something like an arduino. But some very cool work seems to be done there.
>Care to mention anything you've used it with?
i've used it to read out light dependent resistors for controlling patches.
another thing; together with lex (the developer) we expanded the pwm outs for controlling led's... using 3 dedicated 19khz outputs...
it's not much like an arduino, the x-port is actually a complete webserver, and it's working at 12 bits.
One thing that hasn't been mentioned yet, that you may consider, is
using an R2R network to do DDS-style output using however many digital
I/O pins you have access to. Here is one project that does something
similar using an AVR chip:
http://www.myplace.nu/avr/minidds/index.htm
Should be easy to apply the same concept to Arduino or whatever
programmable digital i/o device you have. The resolution limit will be
imposed by the number of output pins/resistors available (1-bit per
pin), so Arduino would be capable of 13-bit precision if you use all the
digital I/O pins for signal generation.
Best of luck,
andrew B.
Wow! Lots of great info here. Unfortunately most is over my head right now, but I see a path to follow in learning the circuitry side of things. Thanks for all your great information, help, and advice. I'm off to experimentation and edumacation land.
Quote: ngwese wrote on Fri, 04 January 2008 20:31
----------------------------------------------------
>
> On Jan 3, 2008, at 12:00 AM, marcus lyall wrote:
> >
> > But I've found that the MOTU 828 doesn't seem to block DC offset, so
> > you can use this as is. You'll get a range of voltage between +/-
> > 1v, then you need to amplify it. Voila. 8 channels of CV.
> >
> > I think.....
> >
> > I could be wrong about this, but that's what my oscilloscope is
> > telling me. Probably not the most stable source though. Plenty of
> > overshoots on the scope at transition points.
>
> I've had great luck using a MOTU 828 mkII to generate CVs for a MOTM
> system I've built. If memory serves the scope said it was generating
> +/- 5V which meant I didn't need any amplification. More often than
> not I'll use a traditional MIDI-CV converter (MOTM-650) to handle
> pitch/trigger CV generation and the 828 to generate "non-traditional"
> LFOs and envelopes that I can't easily generate in hardware...
>
> Cheers,
>
> -greg
>
>
----------------------------------------------------
This is an old thread, but I just got the 828MK2 firewire (not specifically for this, but this would be a bonus). How did you generate +/-5 volts from max? just drive the signal up to +/- 5 on the scope?
On May 11, 2008, at 8:15 PM, Don K wrote:
>
> Quote: ngwese wrote on Fri, 04 January 2008 20:31
> ----------------------------------------------------
>>
>> I've had great luck using a MOTU 828 mkII to generate CVs for a MOTM
>> system I've built. If memory serves the scope said it was generating
>> +/- 5V which meant I didn't need any amplification. More often than
>> not I'll use a traditional MIDI-CV converter (MOTM-650) to handle
>> pitch/trigger CV generation and the 828 to generate "non-traditional"
>> LFOs and envelopes that I can't easily generate in hardware...
>>
> ----------------------------------------------------
>
> This is an old thread, but I just got the 828MK2 firewire (not
> specifically for this, but this would be a bonus). How did you
> generate +/-5 volts from max? just drive the signal up to +/- 5 on
> the scope?
I simply connected the 828mkII up to a scope and then in Max connected
up [flonum] => [sig~] => [dac~ 3 4]. On my 828, a constant 1.0 signal
in Max read as +5V on the scope, -1.0 read as -5V. I've found using a
wiimote as a controller with this setup fun...
-greg
If you do not need a big MIDI-to-CV interface and are happy with 8 bit (= 25 cent pitch resolution), the MCV4 from Doepfer might be an option:
http://www.doepfer.de/mcv4.htm
MIDI in and a limited Thru output, Gate out, 4 CVs out via stereo jackets. You can make two Y-cables to convert the stereo plugs to 4 bananas. Maybe Doepfer can make the cables for you, I am not sure which plugs they use themselves for the A-100 system.
the plugs for the A-100 are mono mini jacks
Here's some good news and a successsful test and video.
I tried sending out LFO signals out of the MOTU828MKII, into my planB modular, which processes the Max signals as well.
Hey, great!
What are you scaling the LFO output to in Max? Also, do you have any idea what the voltage coming out of the MOTU at peak is?
How are you determining that the waveforms are steppy? Are you running the LFO from the output of the MOTU back into the MOTU input or something?
Thanks!
Well, I can confirm that it's working with my Fireface 800. Nice!
Quote: cebec wrote on Fri, 25 July 2008 10:29
----------------------------------------------------
> Well, I can confirm that it's working with my Fireface 800. Nice!
----------------------------------------------------
Really? I recall a conversation I had with Matthias at RME, something like there's a resistor that would need to be removed (or maybe it was added) on each FF800 analog out to pass DC. Maybe they've changed that. What are you driving with it? I have a Korg MS20 that might be an excellent candidate to use via FF800 DC output ... do let us know if your FF800 blows up after a while!
Quote: cebec wrote on Thu, 24 July 2008 06:57
----------------------------------------------------
> Hey, great!
>
> What are you scaling the LFO output to in Max? Also, do you have any idea what the voltage coming out of the MOTU at peak is?
>
> How are you determining that the waveforms are steppy? Are you running the LFO from the output of the MOTU back into the MOTU input or something?
>
> Thanks!
----------------------------------------------------
Well, I did a couple scope tests today. I didn't measure, but there's definitely less than 5 volts, looks about 1 volt on the scope. There is some aliasing or something. It looks worse on the scope at that inbetween time of the wave feeling like an LFO and right before it become a tone we can hear. I can't say it's aliasing since the signal looks fine when I have the phasor at lower tempos. It's when I start getting into like 32n divisions it looks aliased. It still sounds pretty good, so I can live with it. I'll think of it as the digital form of chaos. I'd still use a MICI2CV converter for more accurate pitch control.
In the video, the two channels of waveform go out the audio and into the Model 14 Dual Voltage processor, one on each side, one of the clicks is also mixed in the processor, as well as the envelope that the polyphonic envelope generator makes (which is why when I start twisting the envelope time base, the timings change too. The control signals get added, attenuated, and mixed together (not in that order) and then that new waveform/envelope is sent to the tripe event timer, which has the mixed input coming in, as well as the model 10's 2nd envelope on the 2nd input. So you end up with nice timing variations, almost like a Max object with 8 outputs made from sampling up to 3 waveforms. Hmmm, sounds like the triple timer actually could be a cool Max object or abstraction.
Not good if you need super accuracy, but hell, what analog modular person wants 100% accurate? :)
Max through the interface is definitely good enough to be used as a digital programmable analog almost anything ubermodule. :)
Quote: DonK wrote on Thu, 24 July 2008 17:31
Hmmm, sounds like the triple timer actually could be a cool Max object or abstraction.
----------------------------------------------------
funny! according to Peter, the triple timer is based on a reaktor ensemble!
Quote: johnpitcairn wrote on Thu, 24 July 2008 17:10
----------------------------------------------------
> Quote: cebec wrote on Fri, 25 July 2008 10:29
> ----------------------------------------------------
> > Well, I can confirm that it's working with my Fireface 800. Nice!
> ----------------------------------------------------
>
> Really? I recall a conversation I had with Matthias at RME, something like there's a resistor that would need to be removed (or maybe it was added) on each FF800 analog out to pass DC. Maybe they've changed that. What are you driving with it? I have a Korg MS20 that might be an excellent candidate to use via FF800 DC output ... do let us know if your FF800 blows up after a while!
>
----------------------------------------------------
I hope they've changed it or that it doesn't cause any damage. I think I read that the outs were DC coupled. Regardless, I'm able to drive 1v/Oct, vc waveshape, etc. inputs on my modular and it sounds fine at really low and pretty high modulation rates.
There's definitely some slewing or something, though, when trying to send discrete pitch values. Not great for precision like DonK said...
Quote: cebec wrote on Thu, 24 July 2008 18:22
----------------------------------------------------
> Quote: DonK wrote on Thu, 24 July 2008 17:31
> Hmmm, sounds like the triple timer actually could be a cool Max object or abstraction.
> ----------------------------------------------------
>
> funny! according to Peter, the triple timer is based on a reaktor ensemble!
>
>
>
----------------------------------------------------
Ahhh, the cycle of life. :)
I read that too. I see why it's so useful after playing around with it all week. This may have to be something I try to implement in Max sometime soon.
Quote: cebec wrote on Thu, 24 July 2008 18:24
----------------------------------------------------
> I hope they've changed it or that it doesn't cause any damage. I think I read that the outs were DC coupled. Regardless, I'm able to drive 1v/Oct, vc waveshape, etc. inputs on my modular and it sounds fine at really low and pretty high modulation rates.
----------------------------------------------------
I haven't tried for pitch accuracy yet, how close are you getting to 1 vt/oct? I'm just ecstatic I don't have to buy any LFOs and CV processors for now. It works good enough to tear sound apart and send wacky VC signals, and get into the middle of digital/analog chaos. It works as a CV source at least as good as my kit built Paia (which needs major tuning now that it's hot). So that's all I need for now. If I want pitch accuracy, I think I'd still use a MIDI 2 CV converter, expecially since if I'm going for accurate notes, I'd probably not want notes faster than MIDI could send which at that point, I bet the analog VC Max out FM would sound close enough.
I was a little surprised that the clicks were good enough for the MBase to catch and sounds like, keep up with.
Myth proven. :)
Just been talking to RME support about sending CV from the Fireface 400. unlike Motu they seem to think there is no danger of damaging their card with any voltage it is capable of producing itself, and suggest "giving it a go". I can get pretty accurate envelopes , although any sustained DC voltage tapers off slowly. its definately a highly usable method for sending envelopes and LFOs. just not for accurate sustained pitches.
Hello
I would like to experiment with sending LFOs, and ADSR type envelopes
(generated with Max MSP) from the ff400 to directly control the
filters and oscillators of my Korg MS20. Have you any idea if this is
a safe thing to do with my equipment? Could I damage my soundcard, or
maybe damage the synth? The Korg takes control voltages between -5v
and +5v.
thanks
JL
Hi James
I forwarded your enquiry to the head of RME R&D, Matthias Carstens, and the thread between us is below. The gist of it seems to be - give it a go!
Best regards
Martin
Synthax Audio (UK) Ltd
Unit 61, The Wenta Business Centre
Colne Way
Watford
Herts
WD24 7ND
UK
T: +44 (0) 1923 693 756
F: +44 (0) 1923 693 757
W: www.synthax.co.uk
Martin Warr schrieb:
> Hi Matthias
>
> I hope you are well.
>
> Does this gent's question make sense? I seem to recall that
> RME gear has filters to stop signals below about 1.5Hz being passed, so
> I suppose our equipment will not be damaged if he tries what he wants to
> do, but also it might not work properly since some lower freq LFO's etc
> will simply be filtered.
>
> Best regards
> Martin Warr
Hi,
of course there is no signal at the output possible that could destroy
the unit, as the unit generates the signal itself. And there is no way
to play back real DC as the DACs and the analog stage prevent this. Low
frequencies do work. He can try...
Regards
Matthias Carstens
RME
i'd put some optocouplers in between... just to isolate the whole thing a 100%.
Just came across this post about motu's & DC;
wtf is an optocoupler? is that an easy thing to do?
the head of R+D at RME seem to think its safe.....
i mean, why is it any more dangerous to send out an LFO with a frequency of 0.5Hz to the filter CV input of my synth, than it is to send a sine wave with frequency 100Hz? should i use optocouplers to send normal AC audio to my synth's inputs?
Quote: bin wrote on Tue, 18 November 2008 19:07
----------------------------------------------------
> i mean, why is it any more dangerous to send out an LFO with a frequency of 0.5Hz to the filter CV input of my synth, than it is to send a sine wave with frequency 100Hz? should i use optocouplers to send normal AC audio to my synth's inputs?
----------------------------------------------------
well, dc and synths are good friends of course.. but i believe dc and speaker coils aren't... but there are probably a dozen other reasons for cutting stuff below 20hz.
i don't think doing more or less conventional stuff will harm your equipment/soundcard or whatsoever. the thing is that i've seen smoke coming out of some of my crappy homebrewn stuff once...that's why i'd like to galvanically isolate stuff from my motu... plannning to work on that soon, i'll keep you posted.
best,
daan.
this maybe relevant to your interests
my Arduino midiDAC has 8 outputs and 16 bit resolution between 0 and 5 volts, here is video proof.
midiDAC + max + x0xb0x example
showing off the max patch.
sorry for the shameless plugging, building this thing took ages. Now to make acid.
your solution looks interesting... could you explain a bit more about how your midi control works? It is not clear to me if when you say 16bit resolution you're just referring to the output stage. For controllers can you use NRPNs to send high resolution data or is it limited to standard midi cc messages 0-127?
I think the reason most people here want to send CV in/out of their soundcard is to use the higher resolution and timing accuracy and avoid midi.
do you intend to publish your code / schematic?
cheers,
oli
@Oli
It works with standard midi input at the moment so yes you're right that limits the resolution to 128 discrete steps over a 5v range for the cc's but that can be altered by fiddling with firware arduino code which is already in the public domain on my blog. The AD5668 is a 16 bit DAC which over 5 volts gives incremental steps of 76.3 microvolts. The midi note to cv part of the code scales each incoming midi note to a cv by deducing which octave it lies in and then linearly interpolating between known correct tuning values for C in each octave. These known correct tuning values can be reprogrammed using sysex.
If you download the Arduino code you can just strip out the midi stuff and use the full 16 bit resolution via whatever proprietary solution over USB you like from MAX.
The setup I posted is part of a work in progress for a hardware pitch and amplitude tracker optimized for double bass that will allow my x0x to be played/controlled by my bass, this will make better use of the high resolution (as the double bass is fretless sending note information using midi and pitch bend seems a bit stupid, the pitch to cv conversion will be done on the arduino hardware and bypass MAX, USB and the computer completely).
In the video I posted I was more interested in showing the more standard features that could be easily reproduced using an equivalent commercial midi-cv.
As for timing accuracy and midi I think the soundcard solution is pretty cool but I'm still a big fan of MIDI and as none of the soundcards I have access to support DC output I was never going to go down that path. My MIDI setup is as tight as you like, MAX5 seems much better timing wise than 4.63 ever was without even running it in overdrive. I don't hear it lag and to be honest as long as it doesn't wander out more than 5ms out I don't care.
I have already included sysex support for retuning the midi-cv part (allowing for corrections to the x0xb0x's tuning) and will probably just write some NRPN code for high res support via midi in the future.
As for a full schematic, bill of parts, assembly instructions, etc, these will be forthcoming on my blog in a time frame that fits in with my other studies.
Hope that helps.
Robin
thanks for the info. This looks like a good solution:
i thought motu advised people not to try and send CVs out of their interfaces :-)
oli
At 3:27 AM -0700 1/17/09, Oliver Larkin wrote:
>thanks for the info. This looks like a good solution:
>
>http://createdigitalmusic.com/2009/01/16/analog-meet-digital-motu-volta-connects-the-mac-to-cv-synths-effects-graphically/#more-4785
>
>i thought motu advised people not to try and send DC out of their
>interfaces :-)
i caught this announcement too Oli, and had quite the chuckle re:
MOTU's previous stance...
Dan
--
Dan Nigrin
Defective Records
202 Hack / PC-1600 User / VSTi Host / OMS Convert / Jack OS X / Major
Malfunction
http://www.defectiverecords.com
http://www.jackosx.com
Quote: oli larkin wrote on Sat, 17 January 2009 10:27
----------------------------------------------------
> thanks for the info. This looks like a good solution:
>
> http://createdigitalmusic.com/2009/01/16/analog-meet-digital-motu-volta-connects-the-mac-to-cv-synths-effects-graphically/#more-4785
thats looks pretty cool. I dont quite understand what it means by the "calibration feature". Ive been experimenting with my RME for a while, and it doesnt seem very good for sending out pitch data, because its hard to send a constant DC. it slowly drifts back to zero volts. How is this Volta plugin overcoming that if it works with RME devices? can the Motu soundcards send out a true steady DC signal?
Surely it must be fairly simple to make a max patch to do the same as this plugin, and save 200 dollars. in fact ive already made several such patches, but wthout a "calibration feature" whatever that is.
i've just been measuring my motu mk2 firewire, but it only gives me 4.57 volt... wonder if there is a workaround for that? has any of you guys been measuring?
we have to build this calibration feature into a patch... MOTU can't get away with charging 200 dollars for something they actively tried to prevent people experimenting with!
Is it something that calibrates the DC output to be output at a steady level, or is it just a tuning thing, with no time based component? I cant see how it would work without some kind of time based compensation, at least on my RME.
hmmm, would they perhaps be doing stuff in the firmware? that'd be a bummer.
Quote: bin wrote on Sun, 18 January 2009 04:11
----------------------------------------------------
> we have to build this calibration feature into a patch... MOTU can't get away with charging 200 dollars for something they actively tried to prevent people experimenting with!
just found out that the headphone output on the RME FF400 is DC coupled. you can output 2 channels of true DC from that socket on the RME Fireface 400, as well as the Digiface, Multiface (not II) and HDSP9632. the FF800 and the MFII have a capacitor before the pot.
I have an RME Fireface 800.
I'm unclear as to what you're saying but am interested in maximising
channels, also in your opinions about 48kHz x 2 for 96kHz versus using
just 44.1kHz.
I've always opted for 96kHz, but it sucks when things like SKYPE plays
back Voice Messages in double speed.
I got the RME Fireface 800 for routing and submixing, and have tired
Aggregation with Motif XS8 MLAN (gives 16 more channels, X-Station
gives 2 more, Guitar Rig 3 gives 2 more, Presonus hooked up to TD20 V-
Drums gives 8 more via ADAT, but takes up both RME ADAT's at 96kHz
because of Presonus Digimax output at 96kHz.
I've mastered many things, but it's taken until just recently for me
to delv. into Max/MSP/Jitter (for goodness sakes I have the whole C'74
product line finally and have been following the progress for years).
But now with Max for Live (MFL), I'm ready to "Go". I'm not as
interested in "generative" music as I am in using any Indeterminate
process (my interest in that began with granular synthesis which lead
to Schoastic processed, and then went way up from there using my AI/
Cognitive Science background from College, Bs in Cs; all that money
for one thing - it was a Liberal Arts Cs degree program when most are
in Electrical Engineering, and so it primarily broadened my horizons,
I've learned most things on my own).
Great bunch of people here on the list, no flames, polite candor.
On Jan 19, 2009, at 10:25 AM, James Little wrote:
>
> just found out that the headphone output on the RME is DC coupled.
> you can output 2 channels of true DC from that socket on the RME
> Fireface 400, as well as the Digiface, Multiface (not II) and
> HDSP9632. the FF800 and the MFII have a capacitor before the pot.
> --
> www.myspace.com/binray
justed tested all outputs on my edirol FA-66 with a multimeter and unfortunately they are not DC coupled.
oli
daan brinkmann skrev:
> i've just been measuring my motu mk2 firewire, but it only gives me 4.57 volt... wonder if there is a workaround for that? has any of you guys been measuring?
>
>
Amplification'll do it ;) But hey, +-4.57 is PLENTY to control your rig,
I'd say - unless it's one of those +-10v things?
Andreas.
Quote: robert wrote on Mon, 19 January 2009 15:52
----------------------------------------------------
> I have an RME Fireface 800.
>
> I'm unclear as to what you're saying
unfortunately you won't be able to generate a stable DC voltage, like you need to play a steady pitch on an analogue synth, with a FF800. you can probably get still make interesting envelopes and LFOs though.
If i'm understanding correctly, the Voltas "calibration feature" is just a pitch detector, that tunes the output of the plugin to produce a predictable frequency from your analog gear. i made this feature with about 4 max objects earlier. i can't believe they are sellling this for 250 dollars. its a good reason to buy a motu though....
I have a motu and I'd like Max/Msp to control my analogue modular but i'm worried about doing damage. Do you think there is any special safety features in the Volta like do they limit the volume or scale the voltages or does that matter. Can I just send any frequency I want out from MAX and plug it straight into my CV inputs.
RME told me there was no way their interface could be harmed by any voltage produced it produces itself.
I remember, a long time back, Matthias at RME saying something to the effect that it should be possible to remove or bypass that DC-blocking cap on the Fireface 800 outputs. At least I think I remember...
It would be a cheaper gamble looking for a used 828. I've seen them as low as $150. But maybe they'll go up a little now.
Looks like Volta's kicked some life into this thread. :)
Quote: stefantiedje wrote on Thu, 22 January 2009 21:32
----------------------------------------------------
> James Little schrieb:
> > I cant see how it would work without some kind of time based
> > compensation, at least on my RME.
>
> It won't work with a RME, most interfaces have a condenser at the
> output, it will block dc, there is a lower frequency limit.
> For normal audio applications that is what you want (to protect your
> speakers), the motus are an exception. Even if they don't recommend it,
> as they never know what you would connect it to, it is capable of
> pulling out DC for a synth...
> (I would be careful with too low impedances, as they drive a higher
> current...)
>
> On their web site its clearly stated "Volta will work with any audio
> interface with DC-coupled outputs"
> Your RME is not, and I only know of the motu's, but there could be
> others around as well...
>
> Stefan
>
As i said earlier, the RME Fireface 400, as well as the Digiface, Multiface (not II) and HDSP9632 can output true DC from their headphone socket, I learned this from RME themselves. i have tried this and it works perfectly as far as i can tell. I can control my SH101 with stable accurately tuneable reproducible pitches from output 7/8 only. you can just about use a non coupled output for trigger voltages, leaving you one more channel for your VCF or whatever. so the FF400 is good for controlling a basic analog synth like an MS20 or a modded SH101.
The FF800 and the MFII have a capacitor before the pot, so you'd have to remove that to get true DC from those.
Hi All,
Did anyone build a trigger patch to trigger an envelope or a sequencer of their modular via a motu soundcard of any other for that matter?
When I produce a signal to trigger the envelope in my modular, I have to boost the signal in my modular. I read that Motu has build a trigger sequencer but they don't mention any amplification of the signal.
If I remember correctly, a trigger or clock signal goes from 0v to 10v. However, the motu devices only produces 0v to 5v.
Does anyone has ideas about to build this correctly?
Thanx for the help!!!!
On 25-janv.-09, at 02:14, danny de graan wrote:
> If I remember correctly, a trigger or clock signal goes from 0v to
> 10v. However, the motu devices only produces 0v to 5v.
It depends on the synth, some need negative transitions, some positive
transitions, some with 5V, some with 10V, etc. My Syrinx trigger input
behaves differently if I send +5V or +10V: one value restarts the LFOs
and enveloppes, the other only the enveloppes.
Yes, I did some researce and the arp2600 needs at least 10v or more. So I guess this is impossible with Volta to control. The signal needs to be amplified for that to work.
Quote: bin wrote on Tue, 20 January 2009 21:59
----------------------------------------------------
> If i'm understanding correctly, the Voltas "calibration feature" is just a pitch detector, that tunes the output of the plugin to produce a predictable frequency from your analog gear. i made this feature with about 4 max objects earlier. i can't believe they are sellling this for 250 dollars. its a good reason to buy a motu though....
----------------------------------------------------
Hi Bin,
Are you willing to share your calibration patch you discribed previously? I'm building a plugin similar to the Volta version. I'm almost finished. Only the calibration feature is missing. If you have something that works it saves me time to build one myself ;-)
Thanx!
Quote: - danny - wrote on Wed, 04 February 2009 15:09
----------------------------------------------------
> Quote: bin wrote on Tue, 20 January 2009 21:59
> ----------------------------------------------------
> > If i'm understanding correctly, the Voltas "calibration feature" is just a pitch detector, that tunes the output of the plugin to produce a predictable frequency from your analog gear. i made this feature with about 4 max objects earlier. i can't believe they are sellling this for 250 dollars. its a good reason to buy a motu though....
> ----------------------------------------------------
>
> Hi Bin,
>
> Are you willing to share your calibration patch you discribed previously? I'm building a plugin similar to the Volta version. I'm almost finished. Only the calibration feature is missing. If you have something that works it saves me time to build one myself ;-)
>
> Thanx!
----------------------------------------------------
I don't know how the Volta calibration feature works, but I assume it does a voltage sweep, and then creates a profile of the response of the oscillator, which is then used to generate predictable frequencies.
The patch i built was a simple tune and scale feature. it did not build a profile of the entire oscillator response at all voltages, so it only works if the oscillator responds with perfect Hz/volt or volt/octave scaling, which my synths certainly don't, especially outside of the range of CV generated by their keyboards.
to implement the calibration feature i think you'd need something like a voltage sweep output, while recording an analysis of the frequency of the oscillator, and recording that profile into a table or a buffer. then you could put in a y value (osc frequency) into that table to read the x value (voltage needed to generate that osc frequency).....
i quite like the way analog oscillators sound out of tune though...
Hi Bin,
Thanx for your reply.
I came across some problems after I posted my question.
First of all, When doing a pitch analysis, fiddle~ has BIG problems analysing the low frequencies....It just can't do it. But I used Yin~ and that works fine.
Second, When doing a sweep, like you suggested, the Motu can't generate the full freq range of my osc's. It can get about 5 octaves and thats it. But thats understandable since it can't produce 5 volts exactly.
What I'm working on right now is to store each analysed octave (octave SWEEP) into a buffer and than scale this to have the right freq's. After all the octaves are scaled, I put them back together again into the whole range and them map the midi-date accordingly. Hopefully this will work. If it works I will post the success (or failure)
another way would be to use the function object. you could analalyse a few responses to discrete voltages and use functions interpolate feature to work out voltages for particular frequencies. ideally youd use some kind of spline tool for this, but the more points you put in a function object, the more finely tuned it would sound.
Other uses for this calibration would be to tune your filter, so that you could get perfect filter tracking, which can be alot of fun with analog, like another oscillator basically. and then you could also do FM with that filter, and if your analysis was good you might get some quite accurate harmonic ratios which tend to be hard with analog stuff.
I was thinking about the multislider as a buffer, since this can hold floats and can work as a table. I would use this as a point to point analysis but I'm wondering how long this would take to finish. You don't want to wait a few minutes for your osc's to be calibrated.
Quote: - danny - wrote on Thu, 05 February 2009 15:11
----------------------------------------------------
> I was thinking about the multislider as a buffer, since this can hold floats and can work as a table. I would use this as a point to point analysis but I'm wondering how long this would take to finish. You don't want to wait a few minutes for your osc's to be calibrated.
----------------------------------------------------
why not use lookup~ or peek~?
I have the feeling that if I stay with floats I can change the data more easely then if they are converted into a signal. Maybe I'm wrong about this...please let me know!!!!
Since the data coming from yin~ are floats and the midi data are integers I can always convert later to a signal. But again, thats what I'm thinking now...so if someone has any (better) suggestions I really like to hear (read) them.
Quote: - danny - wrote on Thu, 05 February 2009 15:47
----------------------------------------------------
> I have the feeling that if I stay with floats I can change the data more easely then if they are converted into a signal. Maybe I'm wrong about this...please let me know!!!!
>
> Since the data coming from yin~ are floats and the midi data are integers I can always convert later to a signal. But again, thats what I'm thinking now...so if someone has any (better) suggestions I really like to hear (read) them.
----------------------------------------------------
well, dunno, but if your planning do lookup stuff at audio rate i'd go for signal.
Wel thats the thing, I don't think thats possible since yin~ only outputs floats. And what I saw in the Volta demo video is that they don't use audio rate calibration. I'm building a patch that calibrates the audio signal coming back from your modular system and maps it with midi notenumbers. Later I wil make this available for the filters e.a.
Quote: - danny - wrote on Thu, 05 February 2009 15:56
----------------------------------------------------
> Wel thats the thing, I don't think thats possible since yin~ only outputs floats. And what I saw in the Volta demo video is that they don't use audio rate calibration. I'm building a patch that calibrates the audio signal coming back from your modular system and maps it with midi notenumbers. Later I wil make this available for the filters e.a.
----------------------------------------------------
so if my understanding is correct: you have two stages, 1 for the actual calibration i.e filling some sorta lookup with corrected values. 2, reading this stuff out when playing. I guess for the latter it would be best to do this dsp since here speed is crucial if you'd like to do more than simple midi stuff only.
do you already some sort of version which you could share? maybe we could work on this with the people on this thread?
daan
Hi Daan,
Yes, I can post the patch I already have later today. Note that i'm still on 4.6. I hope thats okay, since a lot of people switched to 5.x.
It is my intension to build a plugin out of it, so 5.x is not an option at this moment (or did I miss something)
Hi Daan,
Just got home and I'm really tired. I post the patch tomorrow.
Goodnight!!!!
Quote: - danny - wrote on Fri, 06 February 2009 00:20
----------------------------------------------------
> Hi Daan,
>
> Just got home and I'm really tired. I post the patch tomorrow.
>
> Goodnight!!!!
----------------------------------------------------
great!
where did you get that Yin~ object ? can't find it anywhere. pitch~ only seems to analyse frequencies above about 35hz
heres a patch that uses function to calibrate the oscillator and filter of an MS20, using an RME fireface 400
audio output 3 is for the gate, output 7 is the pitch, and output 8 is the "total" input, with the total external voltage for the LP filter turned right up.
to calibrate the oscillator, open the filter wide and set up a square wave at the highest octave you can. press calibrate pitch and the patch makes a profile of that oscillators voltage response
to calibrate the filter shut off the oscillators and turn the LP to self resonante, and turn it right down. then press calibrate filter.
If you change the filter cutoff on the MS20 the tracking goes awry, so if you want to keep the tracking you have to change the cutoff from the patch.
the interval for the steps should probably be at least 150ms or it gets a bit sloppy at analysing the pitches
Hi Bin,
The Yin~ object is by Ircam. I got it from the Universaty I used worked at. I'm not sure it its public domain.
On my system it tracks as low as 15hz at 44.1Khz (I use a motu traveler with MacBook pro 2.33 running max 4.6). If its public domain I could send it to you if you like or you can try downloading it from the Ircam website.
As I mentioned in a previous post, I didn't had max 5 so I couldn't open your patch. However, I just downloaded and installed it. I try to hookup my system so I can try your patch. Get back to you later.
> The Yin~ object is by Ircam. I got it from the Universaty I used worked at.
> I'm not sure it its public domain.
It's not.
Quote: jeanfrancois.charles wrote on Fri, 06 February 2009 20:36
----------------------------------------------------
> > The Yin~ object is by Ircam. I got it from the Universaty I used worked at.
> > I'm not sure it its public domain.
>
> It's not.
>
>
>
----------------------------------------------------
Well, than I will not distribute it. 35hz from pitch~ isn't that bad.
To Bin
Hi Bin,
I tested your patch and and it works fine on my Korg MS20 and MS10, really great work. However, I first tried it with my modular system but no luck....that means about 50% luck ;-)
I noticed that the MS series responds only to a signal of about 2.5v - 5v from my motu. My modular, however, responds to the whole range 0-5v. On my modular system I only get about 3,5 octaves and absolutely no low end. The [pitch~] object seems to work fine but I try to implement the [yin~] object to see how it does.
This weekend I will try to modify your patch (if thats okay with you?) so it will work on my modular system. I'll post the result.
To Daan,
Since Bin's patch works (and mine doesn't) I'm not going to finish my old patch. Try Bin's patch and see what you can do with it. Goodluck!!!!
Quote: bin wrote on Fri, 06 February 2009 01:59
----------------------------------------------------
> heres a patch that uses function to calibrate the oscillator and filter of an MS20, using an RME fireface 400
>
> audio output 3 is for the gate, output 7 is the pitch, and output 8 is the "total" input, with the total external voltage for the LP filter turned right up.
>
> to calibrate the oscillator, open the filter wide and set up a square wave at the highest octave you can. press calibrate pitch and the patch makes a profile of that oscillators voltage response
>
> to calibrate the filter shut off the oscillators and turn the LP to self resonante, and turn it right down. then press calibrate filter.
> If you change the filter cutoff on the MS20 the tracking goes awry, so if you want to keep the tracking you have to change the cutoff from the patch.
>
> the interval for the steps should probably be at least 150ms or it gets a bit sloppy at analysing the pitches
----------------------------------------------------
Can't wait to see what you come up with, Danny. My modular also responds to 0-5v and I have yin~.
Quote: - danny - wrote on Fri, 06 February 2009 21:21
----------------------------------------------------
>
> This weekend I will try to modify your patch (if thats okay with you?) so it will work on my modular system. I'll post the result.
>
>
>
Yeah go ahead of course. I've been trying to use a MOTU 828 mk1 this weekend, and the voltages it gives out are not nearly as high as the FF400
With the FF400 set to "high gain " i can get about +-9v
With the MOTU 828 mk1 the highest seems to be about +-2.5v, so its not really very good for controlling filters especially, which needs to go up to 5v to open fully on an MS20.
Has anyone got a higher voltage out of an 828 mk1? Is the mk2 alot higher voltage? I'm using unbalanced cables out of the main analog outputs. Is there a way to get a higher voltage using balanced cables somehow?
I have the flew so I'm not in my studio but you need to send out
-1. to 1. to have the full range of modulars.
The patch you posted is only sending 0. to 1. since the korg ms series only reacts in this range.
I modified your patch a bit so it can work with modulars. Also I adjusted the calibration process to make it a 1 click process. I did this because I want to make a plugin that's easy to use for everyone.
When I feel a little better, I will finish the patch and post it. BTW, I rebuild it in max4.6 since I do not work yet in max5...also I use the yin~ object since this can be used for the lower freq's.
Quote: - danny - wrote on Tue, 10 February 2009 20:08
----------------------------------------------------
> I have the flew so I'm not in my studio but you need to send out
> -1. to 1. to have the full range of modulars.
>
> The patch you posted is only sending 0. to 1. since the korg ms series only reacts in this range.
>
> I modified your patch a bit so it can work with modulars. Also I adjusted the calibration process to make it a 1 click process. I did this because I want to make a plugin that's easy to use for everyone.
>
> When I feel a little better, I will finish the patch and post it. BTW, I rebuild it in max4.6 since I do not work yet in max5...also I use the yin~ object since this can be used for the lower freq's.
----------------------------------------------------
good stuff going on here,
didn't yet have the time to hook up the patch, but i will very soon.
ps: to whom it might concern: last week I was fooling around generating dinsync / sync24 from my motu. works like a charm!
How did you generate dinsync? I love to know since I'm building the ultimate VOLTO clone for free ;-)))
Quote: bin wrote on Tue, 10 February 2009 19:44
----------------------------------------------------
> Has anyone got a higher voltage out of an 828 mk1? Is the mk2 alot higher voltage? I'm using unbalanced cables out of the main analog outputs. Is there a way to get a higher voltage using balanced cables somehow?
----------------------------------------------------
don't think that will work, although you could maybe do some cool stuff with the inverted node which is there anyway... and the best part is, it's for free!
Quote: - danny - wrote on Tue, 10 February 2009 13:10
----------------------------------------------------
> How did you generate dinsync? I love to know since I'm building the ultimate VOLTO clone for free ;-)))
----------------------------------------------------
and it'll be cross-platform!
Quote: - danny - wrote on Tue, 10 February 2009 21:10
----------------------------------------------------
> How did you generate dinsync? I love to know since I'm building the ultimate VOLTO clone for free ;-)))
----------------------------------------------------
well, it's certainly not rocket science!
you were on max 4 as i recall:
oh yeah,
the clock pulse in the patch is maybe too wide / small..but it works for me. if you really speed up stuff it maybe helps to tweek it a bit.
Well, I'm mac based but if there is someone who works on PC and wants to compile it for PC, I'm happy to work together with him/her.
I'm not going to post the whole plugin here but if someone whats to know how something is done, I'm happy to share.
I think its outrages of MOTU to ask $250,- for something so simple. It should cost only $99,-. Thats why I'm building it and why I will post it for free.
Quote: cebec wrote on Tue, 10 February 2009 21:24
----------------------------------------------------
> Quote: - danny - wrote on Tue, 10 February 2009 13:10
> ----------------------------------------------------
> > How did you generate dinsync? I love to know since I'm building the ultimate VOLTO clone for free ;-)))
> ----------------------------------------------------
>
> and it'll be cross-platform!
----------------------------------------------------
> I think its outrages of MOTU to ask $250,- for something so simple. It should cost only $99,-. Thats why I'm building it and why I will post it for free.
It's really silly economics: if I were them I'd give it a away for free, resulting in all synthfreaks buying a motu. Charging 250$ for something like this is a really impolite way of asking people to do it themselves.
what annoys me is that they warned people away from experimenting by saying it would void the warranty to send dc form their interfaces, while at the same time developing a commercial product to do exactly that.
nice patch bin, just hooked it up to my ms-10 using a motu and it works fine.
danny, what kind of features / changes were you thinking to integrate in the-ultimate-volta-hack?
cheers!
daan.
Hi Daan,
I've been sick the last few days so I haven't bin in my studio.
Bin's patch works ONLY with the MS series from Korg and the CS series from Yamaha. I adjusted it a bit so it works with modular synths. Bin's patch sends out 0. to 1. but in order it to work with modulars you need to send out the whole range -1. to 1.. Thats it.
I really like simple user interfaces and I adjusted bin's patch so the user only needs to press 1 button and it calibrates 100% correct. Its begins at 0 Hz and ends at the highest freq of the synth you connect it to.
When I feel a bit better I will post my adjustments.
BTW do you work on PC? I work on mac and I want to release the whole plugin on both pc and mac so everybody can benefit from it.
Quote: - danny - wrote on Fri, 13 February 2009 23:59
> I've been sick the last few days so I haven't bin in my studio.
damn flu
> Bin's patch works ONLY with the MS series from Korg and the CS series from Yamaha. I adjusted it a bit so it works with modular synths. Bin's patch sends out 0. to 1. but in order it to work with modulars you need to send out the whole range -1. to 1.. Thats it.
yeah, good idea, didn't test that yet, the ms-10 is actually -5/+5 too.
I'm wondering, because of this custom calibration feature isn't it's already working in the positive range on both hz/v and volt/octave systems? that is: as long as they are 5v.
> I really like simple user interfaces and I adjusted bin's patch so the user only needs to press 1 button and it calibrates 100% correct. Its begins at 0 Hz and ends at the highest freq of the synth you connect it to.
i can remember reading somewhere in this list pitch~ was only tracking from 35hz, do you want to use yin~ or pitch~?
> When I feel a bit better I will post my adjustments.
great!
> BTW do you work on PC? I work on mac and I want to release the whole plugin on both pc and mac so everybody can benefit from it.
nope, is it about externals?
best,
daan.