Sunday, November 12, 2006
24 comments:
To reduce spam, comments for posts older than one week are not displayed until approved, usually same day. Do not insult people. For items for sale, do not ask if it is still available. Check the auction link and search for the item. Auctions are from various sellers and expire over time. Posts remain for the pics and historical purposes. This site is meant to be a daily snapshot of some of what was out there in the world of synths.
PREVIOUS PAGE
NEXT PAGE
HOME
© Matrixsynth - All posts are presented here for informative, historical and educative purposes as applicable within fair use.
MATRIXSYNTH is supported by affiliate links that use cookies to track clickthroughs and sales. See the privacy policy for details.
MATRIXSYNTH - EVERYTHING SYNTH
© Matrixsynth - All posts are presented here for informative, historical and educative purposes as applicable within fair use.
MATRIXSYNTH is supported by affiliate links that use cookies to track clickthroughs and sales. See the privacy policy for details.
MATRIXSYNTH - EVERYTHING SYNTH
Subscribe to:
Post Comments (Atom)
oh wow, i want one...
ReplyDeleteThis comment has been removed by a blog administrator.
ReplyDeleteThis comment has been removed by a blog administrator.
ReplyDeleteOk, I give up.
ReplyDeleteI hear if you bounce an electron off the module you can't tell where it is anymore.
ReplyDeletemy favorite part is the fancy production...oh wait no my favorite part is the model 24...next on my list of things to buy! Peter (and the crew) build awsome modules..
ReplyDeleteThat Rules! Go Eurorack everybody! You will miss out on this and all the other cool stuff Peter will develop.
ReplyDeletei don't know. it didn't sound too random to me. maybe I'm wrong, but I didn't think DSPs could produce true randomness and I believe there's a DSP at the heart of the Model 24
ReplyDeleteHow ever good it is, it's better than any other Eurorack randomizer. I have the Doepfer and ASystems and neither offers both stepped and slewed outputs and also have CV control of all parameters, especially the range of randomness. I also have seen this demo'd live and do think the YouTibe video lacks a bit.
ReplyDeletedbch,
ReplyDeleteThere's no DSP in this module - it's a micro controller, the Atmel 2051. It's not producing noise, it's producing two random pulse strings, both the sum of two independent 16 bit shift registers. they are paired off (summed) - one set for stepped and one set for smooth. Once summed they are then converted to an analog random - not via an D to A converter (there are none of these either in this module), but using an ordinary .047 uF cap. So while it's digitally driven, the noise is analog. Actually all of the processing after thatt is analog. There's a functional diagram which I think may clear up any confusion:
http://www.ear-group.net/M24block.jpg
But I am curious what exactly doesn't sound random to you. Have you listened to the sound samples on the M24 page from the EAR site? Please do and let me know, I would welcome these comments:
http://www.ear-group.net/model_24.html
Two other considerations:
1) If you have a close look, notice the Deviation, the control which governs the amount of variation form the center point, i.e. - the depth of randomess (second knob from the bottom on the left side), never goes above 2 o'clock. There's a lot more range to be had from this instrument. It's got a nine volt swing.
2) The internal pulses, while random as far as selection, are quantized into musical divisions of time. I did this purposely to give the instrument some charactor. My feeling was if you want truly random events (i'm speaking of time, not voltage) you could easily patch that from the outside world. It's been done to death in random voltage generators, I elected to be different in this regard. But because it accepts outside triggers, the periodicity can be metric (non-random as it gets) if you tell it to do that, as I have done in this demonstration.
Anyway, do let me know. Constructive criticism leads to improvement and I'm all for that. The good thing about using a micro is I can always re-evaluate the algorithm and make adjustments (introduce rules) to get things 'more better'. That's the beauty of a processor.
best,
- P
mine is in the mail as we peak ha ha
ReplyDelete> The internal pulses, while random as far as selection, are quantized into musical divisions of time.
ReplyDeleteWell, then that is very interesting, and something I missed when reading your page on it. I thought quantized in this case referred to 1/12th of an octave steps instead of timing.
wow are the controls as tiny as they look in the video? it seems all crammed together and necessary to use dainty fingertip motions to dial in anything. anyone actually used this with adult sized hands?
ReplyDeleteThis comment has been removed by a blog administrator.
ReplyDeleteReminder: link.
ReplyDeleteWe are one step closer to shutting down anonymous comments.
Peter, why am I not the only one that thinks that your posting under assumed names is bullshit?
ReplyDeletehttp://launch.groups.yahoo.com/group/livewire-synthesizers/message/174
Peter: By "not random", I assume that dbch was talking about your typical 5837-generated noise, which if you used it to drive an S&H to generate random notes, would have obvious repeats.
ReplyDeletedbch, there are much, much better PRN algorithms out there. I had occasion to study some of these a few years ago. My favorite was one that got named the "Mersenne twister"; if you had started running one at the moment of the Big Bang, producing one value per millisecond, it would not have repeated yet. It may not be mathematically random, but for ordinary uses you can't tell the difference.
Regarding amounts of random: This is always an issue. But in a purely theoretical sense, I believe it's overrated. The worst case example being the Buchla 266, which can only derive it's random from 256 possible voltage levels. Sounds limiting, on paper...but if you've ever used one you'll realize quickly it's not a problem. What is does provide is a certain personality to the instrument, a temperament which to my ears anyway, works. Ours use the sum of two 16 bit shift registers (the 266 uses a single 8-bit), so there are a lot more possible outcomes, but again - to my tastes anyway, the amount of random is the last issue to gage a random generator on. There is of course a limit which would be considered silly (for instance if the thing only selected from 16 levels or something)
ReplyDeleteenxsvuyRegarding tiny controls:
ReplyDeletethe image in this demo film is a bit squwashed. if you look at the overall height of the module you can see there's something going on there. He did this to provide a blacked out region on the bottom of the screen in which to lay in the text.
In actuality, Plan B uses a .85 inch center for it's pots - which is equal to A. Sys standard and greater than the Deopfer standard of .75 inches.
Also, I've got huge hands and i think they add to it things looking a bit cramped
- P
Quoting peter:
ReplyDelete"Once summed they are then converted to an analog random - not via an D to A converter (there are none of these either in this module), but using an ordinary .047 uF cap. So while it's digitally driven, the noise is analog."
I am pretty sure that a Sigma-Delta DAC is exactly what you are describing. Yes just a cap to decode, but this is a real DAC. If yours is different I would love to hear how. If not, I think your statements are misleading at best.
Dear Anonymous,
ReplyDeleteI am not trying to be unclear. let me quantify my original statement.
While there is a passive lowpass filter at work here, it's not Sigma-Delta by any means, also not very DSP - no decimated, interpolated values, or digital filtering and while there's plenty of conversion going, it's not by means of dedicated D to A chip which I believe was the original point drawn by 'dbch'.
Instead, the “random” noise generator was modeled upon the old National MM5837 with the major difference of implementing a 24 bit shift register with 4 taps to produce a 16 bit value which serially shifts out to a simple R/C low-pass filter. In fact, two of these bitstream generators are passively summed as part of the R/C. This passive summing and the summed serial arrival perform similar to a PWM generated waveform. The R/C is, of course the method by which the conversion occurs.
Hope this helps. I'll be happy to field any other questions on the topic.
- P
1. I wish I knew how to quantify a statement
ReplyDelete2. A 'random' module that didn't produce a repetition after running at 1KHz since the Big Bang with really bother me. LavaRnd and LavaRand tickle my fancy.
3. The soundtrack to the intro on that vid is fantastic!
On dipshit anonymous (annoying) posters (aka sinners):
ReplyDeleteCensorship = bad, Anonymous naysaying = bad &stupid, anonymity = not bad but frivolous, tagging = good!
I think Matrix should identify all posters by IP (not necessarily displaying actual IP, but perhaps an encrypted code generated by IP) and link to collections of posts by each poster.
-This would expose anons who have previously posted with a name (or names).
-We would know once and for all whether a manufacturer (ahem) posts under assumed names (which I very highly doubt)
-true anons (those who never post under a name) would not lose their personal anonymity.
-It would discourage abuse of the commenting system.
-It would get everyone more connected and make the world a happier place.
What this means:
By commenting with a name even once you forfeit your ability to dissociate that name with any posts you make as anon or any other names.
If this is too harsh:
An IP could be assigned separate IDs (one for each name and one for anon). That way anons who also use names would have their anon posts grouped but no identifying feature other than the ID. At least we'd know who's who in the anon club.
My first Anon post!
ReplyDelete- Doktor Future