Flexor Audio to CV modules

[fra77x]

i think in 2 hours it will be ready

Ok, i'll need 2 days

[Eanna]

i think in 2 hours it will be ready

Ok, i'll need 2 days

[Eanna]

I would like to use my hex pickup straight into an ADA8000, and use that in modular for direct-to-synth sounds.

Yep, this is what my original post was about...
http://forums.planetz.com/viewtopic.php?f=9&t=31871
The little breakout box from ubertar looks really simple and effective as a means to capturing the per-string output from a hex pickup.

But I'm hoping to get a pile of info from the strings...envelope detect, pitch, energy, spectral content (upper harmonics), etc, and use that to add more dynamics to the synth sounds.

So you're taking a different approach - taking the raw waveform from each string, and transforming that to a synth-oscillator-type output by hard sync and other techniques of adding and other types of rectifier-type circuits that produce a harmonically-rich sound that can be subtractively filtered later? Like that the GR300 did? Or am I reading you wrong?

If you are going to capture that amount of data measures from the raw audio, how were you going to extract that? I can see how envelope detection, energy (amplitude?), spectral content (high-pass-filter rms amplitude), but the pitch following algorithm is the 'hard bit'. How did you see your pitch tracking working, if you were going to control instances of ProWave?

I play fretless too - so I would definitely be interested in getting 'raw pitch CV' out of a module as well as an option for Midi out (which Wolf's Midi Toolbox can handle), since the inaccuracies of pitch from fretless play is part of the organic magic of a fretless.

Hexaphonic pickups. Great idea. But now I want to install one in one of these!
http://www.fender.com/en-IE/squier/seri ... -sunburst/
Yep folks, that's a bass guitar. With six string and a whammy bar. Mmmmmm...

[Eanna]

Hi fra77x, looking again at the Flexor modules, there is quite a bit of work behind them...
http://www.cwmodular.org/cv.html

And jhulk, I could find the Env Follower and Audio Trigger modules in bcmodular, but not any Pitch module..
https://bcmodular.wikispaces.com/Downloads

If anyone cares to try using Flexor's modules for this lark, that'd be great...

[Eanna]

I was thinking about people who might be bored...
Boredom is such a shocking, destructive force!

The antidote to boredom is the creation and tuning of audio-to-cv/gate modules...

What you think?

[jhulk]

there are a few pitch modules as i use them

theirs octave modules take an incoming frequency and set it to 16'8'4'2'

there are pitch modulation modules take in frequency and the take in control signals that are added to the pitch module

theirs frequency to note

and if you missing a type custom ones can be created

one such module is a frequency splitter

for a string osc/string filter im making of the logan mellodyII string machine where on the logan you have an upper and lower section each with there own vca and envelope so splits can be created

so we are going to make a note frequency splitter so that any note on midi keyboard is below c3 will go to one set of draw bars

and any key from c3 to c5 to the upper draw bars

as the bbd were noisy and bad aliasing happened they did the keyboard as 4 octaves and the antialiasing brickwall filter was set to 12khz

ill have a look at the flexor and build a clm using what you need do you have a diagram of what your trying to achieve

[Eanna]

Hi jhulk,

I was thinking of Audio In, with Pitch CV / Freq out, and Gate out for envelopes and triggered lfo's.

And the device should measure RMS or Peak Amplitude, and send that out - i.e., an Envelope Follower, to follow the dynamics of your play.

You'll need some 'sensitivity' knobs probably, bit like a Gate's controls maybe. Probably high and low pass filters too before the detector circuit.


When I think of it, it really only makes sense that this device is created if it can send Midi Out. I really ought to bite the bullet and buy Flexor, rather than appear to be coaxing individuals to build devices - this is/was not my intention - this thread was a hope to hear back reports on individuals who have indeed tried these Flexor modules.

The one thing that Flexor doesn't do is a Pitch/Gate/Amplitude to Midi Note transformation.... and Wolf's MidiToolbox has the val2Note module for exactly that (http://wolf-audio.com/dev/mtb_manual.pdf), altho it's a significant outlay for just that one module....

What I do know is that the ability to do "Midi Guitar" exists in Scope, but no device exists that is designed to perform just that function...

[jhulk]

theirs a val2note in bc modular

there are a lot of good flexor modules but a lot only allow mono because they are on one dsp for phase reasons

the ramp sequencer modules are nice

[fra77x]

Eanna you 'll have your mono audio to midi convertor soon. It's not so easy, as it sounds. I wait the time i'll got interest again which means soon... Only the pitch to midi note part is left. I have the trigger to midi part ready. I will make it also for myself due to personal interest. (i don't really miss it as i have the axon, but i want to see the procedure)

[jksuperstar]

The Axon does something very different. It doesn't even use an ADC..it just has comparators that are sample at 2MHz or so...so it monitors the transient speed and +/- value to determine frequency. That's how they estimate that initial frequency so fast.

[fra77x]

It doesn't even use an ADC..it just has comparators that are sample at 2MHz or so..

that doesn't make sense

The phase locked loop that is used for the frequency tracking (in my device) is just made of comparators and a reference osc. In 96 kHz it happens < 1 ms. I believe they are identical. Thanks for the info by the way. Can you point me to the source of it?

[fra77x]

Trule where did you found that info?

For the AXON system, the composite signal of all plucked strings is detected by its six-channel electronic
pickups, and is fed out of an output jack via a connector cable to an interface. The composite signal is then sent
to an A/D (Analog-to-Digital) converter to convert the waveform signal into a digital signal. All attack transient
information is analyzed by zero-crossings to determine the pitch. Then, the information is sent to a neural
network whereby the determined pitch data is converted into MIDI information and can trigger a synthesizer or a
computer (Szalay 1998a, 1998b, 1998c, 1998d). The manufacturer of the AXON system describes this as the
“Transient Early Recognition System”

[jksuperstar]

I cracked mine open to look at it some time ago (An AX50-USB). I was looking into the possibility of tapping into the ADC, and converting it to ADAT to send into the XITE.

What I found, was no ADC. All analog comparators, connected to an FPGA. I probed the signals I could, and remember things running at about 2MHz. I assumed they did this because at audio-rate sampling, the response is not only slower, but quantized to the sample rate.

But, from the structure of the comparators, it looked like they were configured similar to a delta-sigma ADC. At that point I didn't dig too much deeper since I couldn't find what I was looking for. But I didn't see the sample&hold needed for a full delta-sigma converter, only the comparators.

I can pull it apart and re-analyze if it helps!

(BTW - this is when I picked up an ADA8000 to just convert the hex pickup to ADAT, but haven't gotten around to it yet. I did receive a 13-pin DIN connector, so I have some soldering to do soon!)

[jksuperstar]

Oh, and the original AXON was superceded by the AX100, and that was later followed by the AX50 (after Terratec aquired the design).

So all three probably have some level of differences.

[fra77x]

My info is from an empirical study so you must speak right. Cool. I have opened my nova after i burned it accidentantly one day. I use it's pots and its opamps now for some amateur experiment with analog circuits. I think axon is regarded the fastest. I'll have to see in what accuracy and how fast it can get in only digital.
So you opened your axon? To get the signals by ADAT? cool. So to use 6 audio channels? the midi?

[jksuperstar]

My initial plan was to have both MIDI and each string individually available for processing. Playing bass, I think it would be very cool to have two low strings for locking to the kick (more analog style processing), and three others for synth sounds and such (midi processing). The hex pickup allows for that, and the Axon allows for each string to have it's own midi channel...so I can filter midi out some strings, and use the rest for synth triggering.

But other times, I want to also have only a deep synth sound, and midi triggering wasn't the solution: too much latency on low notes, and if I played higher notes and octave shifted down, I lost range of notes available. So then I started thinking about having a bass synth, no midi processing. But having the pitch in a CV or MiDI is useful, since synth sounds often track the note played on the filter cutoff. So back to needing both analog and also pitch detection!

[fra77x]

I think with a guitar you could work better. I remember i played on the upper strings with the sound shifted down, because yes the low strings had worst latency. I haven't checked a bass. I don't have any. The low string should be very problematic. Very few herz. I think it would be easier use a guitar and for effect just connect the normal line out to the scope and use that signal to modulate the midi sound. It also works with a violin and with a microphone for mono recognition. Axon is cool, i 'll use it one of these days. I have forgoten it at the library.

[fra77x]

i remember one of the coolest sounds i have made with the axon was to use the novation nova vocoder and play chords with the midi guitar each string assigned to a different channel and part on the nova, listen to the dry signal of the guitar, and sing in the vocoder audio in. A pink floyd theme. The vocals get the harmony of the chords and it sounds great...

[jksuperstar]

I don't really play guitar or own one though

But, if everything happened in XITE/ModIV, I wouldn't need the axon always connected, and I'd loose the additional MIDI latency of 1ms for each note (on top of pitch to midi conversion)

[fra77x]

for me 1ms is ok. 3-4 ms are ok. More getting bad.