Artesia

(shrug) ..its a box that (eventually) will control multiple modules via CV & route audio - an expansion on the already adiquate original mbcv project for filter gluttons like me :) ...what it becomes, is down to the users choice of filters, oscillators and all that jazz.. So, naming the project/topic is tricky ...CV Modular sort of accounts for the vagueness ..the rest points to possible applications; all it does is provides a framework to build ontop of.. Finished the user interface, psu and box ..been too lazy & distracted to finish the last bits on it ..will fix that, provide some snaps & maybe think about furtheringthis project shortly :) ..yeah cv module revision 2 ? ..wtf ? ..news to me.

lol.. ahh :o :) well.. i'm all spent buying synths atm ..so maybe when we're making some money using them ...i'll dash out for a nord wave too ;) :)

how much delay tho ? ..i cold check thru the codec manual ...tho really only a fly would be able to tell :) nothing wrong with the ssm2044s :) ..i might just pick some up myself ..always good for a little project ..and buildign a dedicated filterbank rn too :)

just updated my previous posting :) ..ahh ic. ...also talking of playing in live situations, where instrument latency is a possible issue; id love to see musician's who can play togeather, with less than say 20ms diviation from the beat ;) ...and besides ...digital music creation technology has clearly illistrated that having rock solid timing & ahearence sounds dull & boring when it is acheived ;) ..thus why humanizer features appeared ;) Also on that note, another reason why music preformed / recorded from real people playing sounds far more interesting ..is that firstly, clearly perceveable timing domain ..variations in overall beat are fun ..and secondly the fact that no two instruments sound at exactly the same time ..or exact same 'way'; results in a ever changing mix of sound each time ....a variation on a 'theme' if you will : ))

aaah ..so the new nord wavetable machine filters in the analog domain ? ..now i'm interested :) There maybe competition with the dsi, tho i wouldnt put it past him to recode some of the stuff to compete if theres competition in some respects ;) Also it largely depends on what the nord filters are like.. Then again, the whole point of the evolver, is that it was supposed to roughly cover the sounds the old prophet & prophet vs could make (not quite got the fulll potential for vectors the vs had tho) ..and of course be capable of other things too. Ultimately its a fat machine.. tho assuming nord made a machine with a similar topology (not another VA) ..then trying to quote one as being 'better' than the other isnt really productive. Their differences are more about 'character and personality' of the instrument ..and thus preference is more about what sounds you desire ..rather than how well it can do them. Oh on the note of codec and adjustable delays, thats something thats done entirely in the dsp chip & thus is only a matter of code. all a codec chip does, is converts between the analog and digital domain. There is nothing segnificant happening here to introduce any segnificant delay. the only delays which would be segnificant, would be those introduced by the length of the 'code chain' used in processing the signals (which is doing 160,000,000 cycles a second / 2,666 cycles a millisecond !). And hey, even if there was a 1ms delay, as you speculate ..thats imperceptable ;) The smallest delay a human can hear between two signals arriving at the ear sepperately is 20ms btw ...this is a phenomenon determined by the nature of whats going on in your brain (psychoaccoustics). Anything less than that, will be perceived as a spatial distortion of the sounds placement ..along with some other accoustic artifacts. if this happens in mono, its perceived as a specific colouration of sound ...both of these things make the classic building blocks of some well known and used effects. Ultimately tho, you maybe able to detect delays longer than say 5ms between a keypress and a sound arriving at you ear ..but it all gets marginal / speculative in this region.

well, theres plenty of ways of doing that ..should you want it ..diy dacs are easy enough ..and so are analog filters ;) Mind, provided signal antialiasing & oversampling are turned off on a synth which can play waveform samples ...and i fully expect they are on the evolver & the xt ...you can create waveforms with any apparent 'bit depth' you want. Infact ...look on the carbon111 website for the 'bit reduction' wavetable ...which does just that: http://www.carbon111.com/mwxt.html oh, and the link reminds me, the xt has the facility to adjust apparent sample rate as a modulateable value via a Sample and Hold filter... so you can patch that too, should you wish.

http://www.soundonsound.com/sos/1996_articles/feb96/waldorfpulse.html Also worth reading this :) quote: A WORD FROM WALDORF I contacted Waldorf designer Wolfram Franke with a couple of technical queries during the course of this review, and he offered some interesting insights into how certain aspects of the Pulse were designed, which are reproduced here. "Our first consideration was how many oscillators a synth should have. One oscillator is boring; you can't create, detune or transpose effects. Two oscillators are better, because you have lots of possible ways to detune or transpose them, or to use different waveforms. But you have one problem. If the phase of one oscillator is shifted by 180 degrees to the other, you will lose sound. This happens every now and then, when you simply detune oscillators by a slight amount. But if you add a third oscillator, it fills these small silences, and the sound gets much richer and fatter. Also, a third oscillator is useful if you plan to offer oscillator synchronisation, cross modulation and so on; you'll still have one oscillator left for doing interesting things. So, the Pulse had to have three oscillators! However, the three waveforms are made differently, and only have one thing in common -- they aren't processed by a D/A converter. The oscillators are not digital, but analogue. The pulse waveform is controlled digitally; the clock stipulates when the pulse waveform has to be at its maximum or minimum point. If you use pulse width modulation or cross modulation, this is also generated by the clock itself. The clock has only a 0 and 1 position, so the Pulse's cross modulation is mathematically identical with ring modulation. The sawtooth waveform is voltage-controlled; the voltage is generated from the D/A-converted clock signal, with additional parts creating the ramp. The triangular waveform is the most complex one; it's a combination between the sawtooth waveform and additional parts that process the down-ramp. :) Ahh... you have the same quote ..tho basically my point being that the oscillation process is happening in the analog domain ...the waveform itself is not generated in the digital domain ...then sent to the analog one with 12 bit conversion artifacts.

well, i don't read german (tho google does ;) ) ...however things on that post are a little ambiguious; based on 'measurement based' speculation. i wouldnt call it a definate conclusion.. and from what i remember of an interview with the pulse designers, it's all analog ...and it would be a little misleading to say so otherwise.. will try to find that post again. Tho at the end of the day, its unclear - maybe query waldorf themselfs ?

its easy to mis-read 'digitally controlled analog oscillators' as dcos. True dcos are those like seen in the matrix's cem3396 chip, whereby the frequency of the oscilator inside the chip is set by a digital pulse train (width of the signal sets the width of one freq cycle of the oscilator). I dont see any signs of a custom oscillator chip ..or read anysuch things elsewhere. Digitally Controlled oscialltors can also be interprited as a DAC (digital to analog converter) turning a stack of numbers from the cpu into a specific & changing analog voltage level, this can then be used to drive the analog oscillator in the same way that the voltage read from the keyboard of say an old moog ..would determine its frequency :) i mean, how else are you going to control an analog oscillator on a synth with a digital brain; other than through digital control ! ;) :) also ..the reason it's oscilators sound a little static sometimes, is because with modern digital control.. its often an overacheiver in preventing things drifting.. (old analogs tend not to follow their controlling circuitary as well, and drift out of tune ...this was on paper considered a bad thing, tho it proved pleasing until it got out of hand) Note: even analogs accused of sounding 'not quite as fat' in certain ways, as say a moog, due to DCO's and other overly precise modern engineering techniques.. will still outdo Virtual Analogs for sonics ..take a comparison on exactly the same patch programmed on a matrix 1000 & the exact same one programmed on a nord lead 3 - which is labelled specifically as a matrix sound (..this i did myself when i had a lend of one) ...and theres a huuuuge difference. anyways, i can see this turning into a two 'man' flame war ..and i can't be bothered with that.. This was just supposed to be a post about the contents of a number of notable synths ...other people im quite sure are as curious as myself as to what goes into these commercial products :)

er, pardon.. The 12bit dac is used for control signals (CV) ..to control the oscillators & filters ..etc ..everything that makes and shapes sound in the pulse is analog :) the pulse, to the best of my knowlege, is waldorfs -only- ..all analog synth. Designed to drive a wedge in a market, which at the time was flooded by 'almost analog' Novation basstations & where there was rising interests in things like moogs again.

btw playbo$$, i think you'll find the dac conversion for the digital waveforms on the evolver is more than adiquate... just checked in on the specs: * 4x Analog Devices AD1836AAS 96khz 'codecs' - 16>24bit 3x stereo dacs 2x stereo adcs (http://www.analog.com/UploadedFiles/Data_Sheets/AD1836A.pdf) The chip above has more than ample high fidelity ins and outs ...no doubt those 24bit 96khz ones are also used for the digital voicing !

if its your bag.. go for it :) Tho not entirely sure myself.. no.. offence intended ..tho it would seem you are just looking for reasons to justify follow the urge to buy a nord wave (over anything else ?).. mind, i can hardly speak.. it doesnt take much to convince me that another analog might be interesting. then again they dont usually set me back 1.5k a pop ! :o :)

i may resurrect it & do just that at somepoint.. maybe later this year i will rebuild it around a formidable spec dual core board (hmm new, not so neat case panel redesign for that :/) ...and sling in a slightly bigger new screen for good measure. and use it as the 'studio in a box'. ..that machine is only a p3 733-133 with a gig of ram ...but my, it ticked over quite well indeed :)

i'm soso about the nords too, tho i prefer them to the virus in some respects ..neither of which, having had the time to pay with em' particularly excite me ...at the end of the day i'm a really picky sod & have a real love for analogery; so it's very hard to sell hardware va's to me. The waldorfs really hold my interest tho.. but they just seem to be a totally different kettle of fish frankly.. Agressive, naughty machines :)

nah.. its fine ..no bother. i'm just about as organised as a bowl of spagetti half the time.. okis, yeah it runs real quiet ..due to thoughtful use of ventilation and the right fans. btw the rollerball was a custom paintjob ..i detested the standard greyish silver microsofft provided it in ..so out with the metalic, pearlescent paints ;)

multitimberal, as in it allows you to layer patches into one sound ..as for being able to map different sounds to different keyboard sections & process them sepperately externally ...this could be really annoying. Im personally not too sold on most virtual analogs ..for the money you get alot less actual hardware inside; and when you can get the real grit and warmth of the genuine thing for the same money.. and far less second hand... well.. Also many va's start looking like an obtusely overpriced 'plugin in a box', when you compare them to such items as the Terratec Komplexer plugin synth ...which,to my ears is very, very good & seems to cover the tonalitys and options coveredd by almost all the VA's available. Oh and it doesn wavetable synthsis. Don't of course forget its 1/10th the price of the Virus ! ...hmmm

hmm .. i want to know what the solaris does ...looks sweet & very late oberheimish ...off to find out rn ...hmm and waldorfs new linup looks alot less plasticky than i had first imagined ...all that shiny metal ...whooo ! ...ok i'm puzzled, i cant find anything out about the solaris ...looks like another monster wavetable machine tho...

hi, as promised ..er ages ago ..i found my pics of the flightcase pc i built ..would take some new ones ..however its out of serive atm & in need of a new screen and the scanners dead. Nicked the screen for a video projector project ;) These where taken quite some time ago, in a friends old house.. doesnt live there anymore ...let alone the same county !

Hmm.. i guess i should really start some midibox.org wiki posts for all these 'under the lid' expose's ...would be a little tidyer to wade thru :)

And Yeuup The Alesis Andromeda * tons and tons of circuitboards... * The much hyped (why ?) Motorola Coldfire 32 bit 90mhz processor - mcf5307ft90b (http://www.freescale.com/webapp/search/MainSERP.jsp?SelectedAsset=Orderable%20Parts) (Replaced by MCF5307AI90B ..so they must of had a stockpile of old chips..) (my first pc in 1996 had as much poke ;) ) * burr brown pcm54hp 16bit 333khz capable dacs - used on the voice board. (http://focus.ti.com/lit/ds/symlink/pcm54.pdf) (probably used to provide the drive signals for the osicllator & filter chip, which multiplex the signal inside to save an unrealistic ammound of external circuitary to control -ALL- those internal CV drives :) ) * custom made and designed ASICS labeled - alesis - aso 0625a & asf 0625a - no info onine, however these internal schemetics: [ See end of section !] * proprietary dsp made by ami for Alesis, probably aslo used in their effects racks - has some dram tied to it/ * tl082c opamps throughout the audio stages * Wavefront al1101 ADC - 24bit 48khz, 64x oversampling (http://www.wavefrontsemi.com/DataSheetsFolder/WavefrontAL1101.pdf) (must also use this forthe calibration routine ;) ) * Wavefront al1201 DAC - 24bit 48khz, 128x oversampling (http://www.wavefrontsemi.com/DataSheetsFolder/WavefrontAL1201.pdf) * 14x (at least) achct16374 d-type, tri state 16bit flip flop chips.. (http://focus.ti.com/lit/ds/symlink/74ac16374.pdf) * AMD Flash Ram for OS ..etc. * 2x is61lv25616-15t - 'fast ram' ..can't be asked to find that, however its probably 256kb 16bit 15ns sram. * Samsung k6t4016v3c-tb70 cmos sram - 'slow ram' 16 bit 256k 70ns, tft technology ram (related to tft tech for displays ? standard stuff anyways tho ...hmm) (http://www.samsung.com/Products/Semiconductor/LowPowerSRAM/30_2533V/4Mbit/K6T4016V3C/ds_k6t4016v(u)3c.pdf) * Sepperate microcontroller with old style eprom for reading & controling the control surface. * philips p87c52x28bd 8 bit microcontroler (control surface i/o) - 128-1024 ram, upto 33mhz clock. (phillps no longer make semi's !) (http://www.datasheetcatalog.com/datasheets_pdf/8/7/C/5/87C52.shtml) * connected to old style eprom :) *other chips i cant be bothered to look up at this time. Here for some random info: (http://www.sequencer.de/syns/alesis/AndromedaA6.html) Note the two 22" Widescreens below are cheap as chips and great quality if anyones after such things.. Made by Belinea, called the artistline 22W ..has aa contrast ratio of 1000:1, res of 1680x1050 ..and sells for £180 !

A slightly grubby Waldorf XT * Motorola mc68331cfc20 processor - 32bit 20mhz (http://www.freescale.com/files/microcontrollers/doc/data_sheet/MC68331TS.pdf) * Motorola xc56303pv80 - 80mhz (480mips) 24 bit dsp * 2x am29f010-70pc Flash Roms - 1 megabit 70ns (http://noel.feld.cvut.cz/hw/amd/16736f1.pdf) * 5x nec d431008le-15 sram chips - 256kb 15ns ? * ne5532 opamps * cirrus 5330aks - 18bit stereo ADC (http://pdf1.alldatasheet.com/datasheet-pdf/view/58142/CIRRUS/CS5330A-KS.html) * 2x cirrus cs4329ks - 20bit stereo DAC, 115db signal to noise :) (http://pdf1.alldatasheet.com/datasheet-pdf/view/58106/CIRRUS/CS4329.html)

Dave Smith Instruments - Poly Evolver * Specs pretty much the same as the PEK ..except for a slight change in board design, colour & and some additional notes son the legend masking ..hmm. * DSI-120 / PA-397 analog synth chips * a number of Analog devices op275 dual j-fet opamps (...hmm curious) (http://www.analog.com/UploadedFiles/Data_Sheets/OP275.pdf) * a number of Texas Instruments 2272c dual lin-cmos opams (..odd) (http://focus.ti.com/lit/ds/symlink/tlc2272.pdf) * 4x Analog Devices AD1836AAS 96khz 'codecs' - 16>24bit 3x stereo dacs 2x stereo adcs (http://www.analog.com/UploadedFiles/Data_Sheets/AD1836A.pdf) * 4x Analog Devices adsp-2191m DSP Chips - 160mhz ! 24bit memory structure, 40bit accumaltors.. (fairly nuts all in all..) (http://www.analog.com/UploadedFiles/Data_Sheets/ADSP-2191M.pdf) (surely dave, dave that dsp is underutilised & can do some insane things.. more features pleeease ;) ) * sockted chips & cpu which i dont want to peel the labels off whilst i still have a warranty !

Oberhem Matrix 1000 ..yes we've seen these before ..tho why the 'ell not ;) * 6x CEM 3396 'synth on a chip' chips (...goes without saying really) ..contacts prone to corroding, may need reseating & cleaning occaisionally. * 3x 7805 & 1x 7912 & 7812 regulators - it would seem to eat alot of 5v juice ...hmm. * Fairly sizeable cap bank & psu 6800uf (5v) & 2x 4700uf for the +/- 12v analog rails. * tl081, tl082, lm311 opamps. * hitachi hd68b09p 2mhz 8bit processor ! (http://www.cpu-world.com/CPUs/6809/Hitachi-HD68B09P.html) * 5x UMC um82c54-2 8mhz programmable interval timer with 3x 16 bit counters. (http://www.datasheetcatalog.com/datasheets_pdf/U/M/8/2/UM82C54.shtml) * 1x ST EF68B50P 2mhz 8 bit asincronous communications interface adapter (http://www.datasheets.org.uk/datasheet.php?article=1408420) * a ton of 74 series logic chips & 4051 analog multiplexers. My, the control & processing functions certainly seemed to sprawl out from the conservatively specced cpu ;) ...mind if they made this these days, my guess is that it'd be controlled by a single microcontroller, with one or two ancilliary chips & the cem's ..all packed into a board the size of a video cassette ;) :) Matrix schematics unearthed, from somewhere else on the web: http://homepage.ntlworld.com/nausicaa.valley/ucapps/oberheim_matrix1000/matrix_1000_scem.zip More hardware notes here: http://wolzow.mindworks.ee/analog/m1k-hardware.htm

Waldorf Pulse A very fat 3 oscillator, monophonic 'all' analog synth.. designed with moog/roland type sound characteristics in mind. Uses transistor cascade filters & transconductance vca's in a fasion which is very similar to those found in the moog. Ultimately has a sound of its own tho.. if you get a chance, check one of these out ..alot of fun :) Contains * 2x CA3046 Transistor array chips stabalising each end (?) of a stacked transistor moog style filter. (http://www.intersil.com/data/fn/fn341.pdf) * 3x CA3080e Transconductance amps (also used in some moog vca's ?) (http://www.intersil.com/data/fn/fn475.pdf) * 1x tl074cn opamp * 5x tl064cn opamp * High quality polystyrene/polyester/metal film capacitors throughout * AD7545 12 bit DAC - for cv control ? (http://www.analog.com/UploadedFiles/Data_Sheets/AD7545.pdf) * 2x 8 bit 3mhz ? (xtal says 12mhz) Motorola mc68hc11e1cfn3 Processors (http://www.freescale.com/files/microcontrollers/doc/data_sheet/M68HC11E.pdf) * 1x very strange programmable cmos logic array (hmm what is this being used for ?) - palce16v6h-26 pc/4 (http://www.latticesemi.com/lit/docs/datasheets/pal_gal/p16v8.pdf) * 3x 93c66 EEPROM chips. * a curious L165 Circuit used to generate an artificial -12v rail from a single 12v supply. (this was dead in the one i adquired (another case fo psu abuse !) ..and was a little bit of a pita to adquire) Note: the basic topology of the analog sections is crying out to be cloned, tweaked midi cv'ed & diy'ed :) ..anyone want to have a go ?

Yes, the waldorf puse is very, very snappy & if you want something increadibly fat, with 'minimal specs' ..for not too much money ..its one to keep an eye out for; great little synth !