Author |
Message |
zthee
Joined: Feb 20, 2008 Posts: 414 Location: Stockholm
|
Posted: Wed May 13, 2009 8:58 am Post subject:
|
|
|
fonik wrote: |
i am far from a PCB, but i envision to implement following features (this year at least ):
- 10 octaves
- selectable bi-polar -5V to +5V OR unipolar 0V to 10V
- selectable step width (semi, 2x/3x/4x semi?)
- at least two channels
- one channel could work as a "overall" quantizer to shift the output of the 2nd channel |
This sounds great! _________________ http://www.thehumancomparator.net/ |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Wed May 13, 2009 10:52 am Post subject:
|
|
|
loss1234 wrote: | i wonder if my DAC0808 would work with some mods to the circuit? |
the datasheet of the DAC0808 references the DAC0800 for high speed applications. the 0800 is 50% faster than the 0808. i did not look closer at the pinout, but i guess they are identical in all other terms.
however the speed (or settle time) of the DAC influences the performance... _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
loss1234
Joined: Jul 24, 2007 Posts: 1536 Location: nyc
Audio files: 41
|
Posted: Wed May 20, 2009 1:02 pm Post subject:
|
|
|
since this thread is about quantizers, i figured id ask here
i have been having a really hard time getting my home-made breadboard quantizer calibrated and one of the main reasons (i think) for this is that i am using an ADC0808 which only takes up to a 5volt input
now when you consider that a normal modular might have signals of 0-5, +/-5, and +/- 10 volts, it is really really hard to get it set to work with all three and keep your calibration in check.
i am basically just hooking the outs of an ADC0808 into the ins of a DAC0808 and then putting that though an opamp with a 100k trimmer to get the right calibration.
so i was wondering if anybody could give me some advice on how to
A: get the input conditioning setup to work with all three voltages and offsets
and
B: how to calibrate the darn thing for 12 notes over at least a few octaves
right now i AM getting nice stairstep outs but they are not quite in tune.
and as i said, when i put in a +/- 10 volt signal, it chops off the top. but if i then set it to NOT chop it off, my 0-5 volt input then gets a strange offset
i am so lost.
sorry for asking this in a few places (sdiy as well) but i think the more answers the better.
thanks guys _________________ -------------------------------------------- check out various dan music at: http://www.myspace.com/lossnyc
http://www.myspace.com/snazelle
http://www.soundclick.com/lossnyc.htm http://www.indie911.com/dan-snazelle |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Wed Jul 01, 2009 7:32 pm Post subject:
|
|
|
dead thread ? |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Thu Jul 02, 2009 1:05 am Post subject:
|
|
|
on hold at least?
since there has not been that much interest in the ARPstyle DAC based circuit i am working on, and since other projects became more "urgent", i put it aside. but it is still on breadboard... _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
kkissinger
Joined: Mar 28, 2006 Posts: 1354 Location: Kansas City, Mo USA
Audio files: 42
|
Posted: Mon Nov 16, 2009 9:51 am Post subject:
Anyone want to build a voltage quantizer? (thread rescue) |
|
|
This post is a thread rescue.
I would like to build a voltage quantizer and a PCB would definately be a big help.
Is anyone else interested in this or am I alone here? Had I seen this thread last spring I would have jumped in.
I've never made my own PCB. I have built some simple circuits (i.e., op-amp mixers, inverters, etc) on pegboard and just hand-wired everything.
I could do similar for a quantizer however a PCB would be much more elegant. _________________ -- Kevin
http://kevinkissinger.com |
|
Back to top
|
|
|
bf
Joined: Feb 28, 2009 Posts: 22 Location: mid-west
|
Posted: Mon Nov 16, 2009 12:18 pm Post subject:
|
|
|
Interested as well. Awaiting patiently. |
|
Back to top
|
|
|
hadal-drop
Joined: Nov 25, 2008 Posts: 10 Location: Brisbane
|
Posted: Mon Nov 16, 2009 6:41 pm Post subject:
|
|
|
yes quantizer needed here. i'm sick of using the miniwave as a quantizer when it could be doing more fun things. |
|
Back to top
|
|
|
Funky40
Joined: Sep 24, 2005 Posts: 875 Location: Swiss
Audio files: 1
G2 patch files: 5
|
Posted: Fri Nov 20, 2009 4:06 am Post subject:
|
|
|
yup, i need DIY quantisers, but only if the unit has 12 switches to set scales up on the fly.
since Docvague brought this Idea up i'm in love with that idea ( or was it anybody else? ) ) |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Fri Nov 20, 2009 4:19 am Post subject:
|
|
|
Funky40 wrote: | yup, i need DIY quantisers, but only if the unit has 12 switches to set scales up on the fly.
since Docvague brought this Idea up i'm in love with that idea ( or was it anybody else? ) ) |
for designs we currently speak of 12 switches won't be any use, i think. if you switch the first 4 bits you get a kind of scale.
actually you won't get 'scales' by switching the bits, but 'steps', i.e. whole note steps instead of half ntoe steps. since harmonic scales have both: half and whole note steps you won't get very far.
same for 5th's. if you switch the step width to 5ths you would get (in regards to 'C' as base) C - G - D - A - E and so forth.
however, i don't say it won't make any sense to switch some bits on or of, you just won't get harmonic scales. _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
clee
Joined: Dec 08, 2009 Posts: 42 Location: Chicago
|
Posted: Tue Dec 08, 2009 7:50 pm Post subject:
|
|
|
Hello all. I've been lurking about here for a while now... particularly on this thread. Finally loging on with a few questions.
I am currently working on a design for a PIC based quantizer and would like a few suggestions.
The current idea is to have 8 channels of quantiziation. A rotary pot and a voltage control input that would select 1 of 16 scales to quantize to. An LED would be lit when the output of channel 1 is quantized to the root note of the selected scale.
So I have 2 questions...
1. If anyone is familiar with PIC programing/ digital design. Is there any advantage to either using a dedicated R/2R D/A converter or using the PWM/1bit DAC routine built into the PIC chip? The D/A output will have a 1v/Oct trim but is one type of D/A more accurate? Also how accurate is the average Quantizer that is currently available as far as +/- cents of error for a note.
2. There will be 16 scales to select from. Most of my modular stuff is nontonal so I'm looking for feedback on the scales. Here is what I have so far: Quartertone/All notes, Halftone/All notes, Wholetone/All notes, Major, Minor, Harmonic Minor, Major Pentatonic, Minor Pentatonic, Phrygian, Aolian, Lydian Dominant, Lydian, Lydian Quartertone and octaves only (for adjusting Osc & filter tracking.
That's 14 scales, so I need a few more suggestions. Also, if people think there are some totally useless scales in there let me know.
That was way more than 2 questions. Thanks, Craig |
|
Back to top
|
|
|
forbin
Joined: Jan 29, 2009 Posts: 120 Location: Fremantle, Australia
|
Posted: Wed Dec 09, 2009 1:49 am Post subject:
|
|
|
Hi Craig
I have also been thinking about this for a while and looking at doing a dsPIC version of this using the internal 12 bit ADC and an SPI 12 bit DAC out. I was looking at some of the scales that you mention but also some of the more "interesting" ones that things like Ableton build into their MIDI effects on Live. I was also looking at a sort of Glissando effect between note changes that was fed from either an on board Oscillator or an LFO. I don't think that a PWM will give you a particularly clean signal for frequency control -- but i am willing to be proved wrong? the MCP4921 12bit DAC is only $2.25 from Futurlec. My investigations into this have been a bit slow and prompted from an Elektor article that has beeb mentioned earlier i think... I was actually looking at a simple LCD display from a Nokia phone to display the 12 x 12 mapping as in the Ableton Live plugin for a bit of eye candy... |
|
Back to top
|
|
|
clee
Joined: Dec 08, 2009 Posts: 42 Location: Chicago
|
Posted: Wed Dec 09, 2009 9:22 am Post subject:
|
|
|
forbin, thanks for pointing me at the DAC chip. I've been thinking about using a 12bit DAC to increase accuracy. The one I was thinking of is more like $10 so that one would be good.
Obviously I want the voltages to be as spot on as they can be but I'm not sure what is really required. The Serge quantizer I started this whole project modeling claims like 2 cents error, that's .0016 volts . Aside from wanting to be as accurate as possible I don't know what is NEEDED to get a quantizer that plays in tune. _________________ life in many forms...
http://clsound.com |
|
Back to top
|
|
|
forbin
Joined: Jan 29, 2009 Posts: 120 Location: Fremantle, Australia
|
Posted: Wed Dec 09, 2009 7:20 pm Post subject:
|
|
|
The whole behavior of Quantizers seems to vary a bit! There are some which seem to depend upon a trigger or gate and they just quantize on the voltage at that moment. There are others which monitor the voltage continuously and just provide steps out as it changes in a continuous manner. This is what I was looking at doing. If you take the likely range of the signal to be 0->10V or -5->+5V then with a 12 bit ADC/DAC pair you are looking at getting a step of about 2.4mV. In reality it is easier to loose a little bit of accuracy for an even step so we go with a voltage step of 2.5mV and get an integer number of values per note. This allows the micro-controller to implement a little bit of hysteresis. The speed shouldn't be an issue as I have easily run the dsPIC to 200kHz so you could theoretically use the quantizer as an audio effect/bit reducer as well. I was going to get it to produce a gate/trigger on a step change and maybe time quantization as well? (sort of sample and hold).
With respect to the Serge -- i suspect that the specification is that the accuracy of the 0.1V steps is to 2cents -- this is dictated to by things like the linearity of the DAC and the noise etc. and is often described in fractions of a LSB or minimum step.
Accepting that the linearity of the ADC is going to be pretty good and the offset is from the VCO the only real adjustment is going to be the slope which will just be the gain of the output buffer. |
|
Back to top
|
|
|
clee
Joined: Dec 08, 2009 Posts: 42 Location: Chicago
|
Posted: Thu Dec 10, 2009 8:53 am Post subject:
|
|
|
Like yours, my quantizer also will just constantly step an analog input voltage (8 actually). It has no trigger/gate ins or outs. I'll run the PIC at at least 4MHZ, it can go up to 20mhz so speed is not a problem. Even multiplexing the 8 analog channels they should all change at practically the same time.
I think 12 bit on the DAC side is the way to go. My thinking is to use the PICs 10 bit A/D converter on the way in. With a 0-10v input you need 240 steps to represent 10 octaves of quartertones, and 8 bits will cover that.
I forgot about the hysteresis problem. With my analog design I just added feedback to the comparators.
Funny you should mention the bit reduction. If I get this working my next idea is a bit/sample rate reducer "lo fi" module. Nice thing is, I should be able to lay out the circuit board so I can use the same one for a lot of modules (quantizer, lo fi, waveshaper) and just change the panel connections and the code. _________________ life in many forms...
http://clsound.com |
|
Back to top
|
|
|
roman_f
Joined: Jul 03, 2007 Posts: 63 Location: Moscow
Audio files: 1
|
Posted: Mon Nov 19, 2012 6:06 am Post subject:
|
|
|
fonik wrote: | Funky40 wrote: | yup, i need DIY quantisers, but only if the unit has 12 switches to set scales up on the fly.
since Docvague brought this Idea up i'm in love with that idea ( or was it anybody else? ) ) |
for designs we currently speak of 12 switches won't be any use, i think. if you switch the first 4 bits you get a kind of scale.
actually you won't get 'scales' by switching the bits, but 'steps', i.e. whole note steps instead of half ntoe steps. since harmonic scales have both: half and whole note steps you won't get very far.
same for 5th's. if you switch the step width to 5ths you would get (in regards to 'C' as base) C - G - D - A - E and so forth.
however, i don't say it won't make any sense to switch some bits on or of, you just won't get harmonic scales. |
dear matthias
do you think if i can trim it to 1/10V per halftone? not 1/12V |
|
Back to top
|
|
|
|