MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Time For a New Specification?
MUFF WIGGLER Forum Index -> Modular Synth General Discussion Goto page 1, 2, 3 ... 14, 15, 16  Next [all]
Author Time For a New Specification?
innerclock
I’m throwing this out there for general discussion to see if there is some common ground for a potential future product specification rather than simply to have a rant. Yes this issue is obviously close to our hearts but I think this may be of benefit for the longer term in general module development.

When purchasing any synth module we all expect certain basic standards in manufacturing - module width and mounting-hole position for specific modular formats are a good case in point. Correct power supply connectors and operating voltage are another. We all expect these standards to be followed by the manufacturer before a purchase is made so that when we unpack and bolt it in our rack everything fits and works as expected.

There are many modules available that use time as a core parameter of their overall function - sequencers being the most common. In stark contrast to our well respected standards of module width and power supply voltages however, from my experience there seems almost no way currently of determining if a time-based module you are buying adheres to the International Standard of Time before you make your purchase.

The second has been determined atomically since 1967 by the International Committee for Weights and Measures:-

http://en.wikipedia.org/wiki/Second

Time and Tempo are inseparable and both form the very core of what makes music – sounds heard as a function of time.

At 120 BPM there are precisely 6000 samples (@ 48 kHz) between all rhythmic 16th steps according to the long-accepted International Standard. This is not open to interpretation. It is a fixed value like a Metre or a Kilogram. These standards are fundamental to the way we experience the world we live in.

Upon testing many modules over the years I have found that accurate tempo is not often followed in favour of other secondary features or simply the result of design limitations. Depending on the customer and the application, secondary features that compromise on the core function of a module may be desirable but not in all cases and not by every customer.

I find it strange that a recent sequencer module purchase fits in my rack slot precisely, the mounting holes line up to the micron and it powers up correctly with the standard voltage and current supplied but it fails to keep accurate time which is its prime function and the only reason for my purchase. It looks and fits great but it doesn’t work as I expected.

My purchase expectation is not a subjective one either. Features like VCO Range, Filter Response and Envelope Curves are subjective – they are variable and form the character of a module. There is no International Standard for VCO Range or Filter Frequency Response but despite this we regularly see these type of specifications clearly stated on the manufacturer’s product documentation.

There is a recognised standard for time. This has not changed in forty five years. One second is a very clearly defined value and yet we never see timing performance and tempo stability specifications in sequencer module documentation.

This issue may not be of equal importance to everyone but I do feel it deserves at least the same respect and manufacturer attention as VCO Range and Filter Frequency Response.

If a sequencer module does deviate (either by deliberate design or for other reasons) from the International Standard for Time then this deviation could be stated clearly in the module specification so that customers know exactly what to expect prior to purchase.

My proposal is for a defined time-specification to be included on sequencer and clock processing modules so that users and potential future customers can make their purchase decision based on timing performance.

Your input would be most appreciated. Regards to all - David
dude
this has never once occurred to me. i'm sure you are right but it just doesn't come up in my musical meanderings, composition performance etc. modules are like the wild west in euro. nobody really has any EXACT specific standards even if they appear to afaict.
Matos
This is an interesting point, and one I'd like to see taken into account when designing new modules. I for one enjoy quirks,errors, drifts, glitches, and various uncertainties in my system. But a general group of standards regarding timing, voltages, and power would be welcome. I know this is venturing off topic, but a voluntary set of community based standards to adhere to, along with a seal of approval might help in the long run. A little modular standards approved seal to let buyers know that it meets these basic standards would help those who seek such things.
John Noble
It's a laudable goal, but I think it's more suited to test equipment than musical instruments.

I can't imagine doing the same thing for electric guitars or tubas, so I don't see the purpose for modular synths. OTOH, there's nothing wrong with like minded people coming to an agreement about such things either.
J3RK
I both agree with you and disagree. If a module is advertised to keep precise time, then it should be held to this standard. Otherwise, it's likely performing as advertised. An analog sequencer without a quantizer doesn't follow any more of a standard than a clock controlled be an analog voltage. You turn the knob, it speeds up. No increments to be measured here, other than a voltage.

If your clock is based on an MCU, CPU, FPGA, etc. where precise time can be kept, I think it would be silly NOT to adhere to a timing standard, if for nothing else, than because it's so easy to implement in software, using a crystal for timing.

For a piece of "analog" or digital (in the CMOS sense) circuitry, I think designs would be overly complicated. If I have a clock, that provides consistent timing, I really don't care if that timing adheres to a specification as long as it stays consistent at the setting I have it at.

To me, in this case, that means using a precise enough timing method. If based on a capacitor charge cycle (like in an astable configuration) this means using polystyrene or other suitably stable cap.

If a module is set up to be a timing source with note-length divisions, then I would agree that it should definitely adhere to a recognized timing standard.

seriously, i just don't get it

Just a thought or two.
chrisso
I think you are mistaking desirable inconsistency for undesirable inaccuracy with timing.
If I want human drift, in timing or tuning, I want to control that myself, not be presented with it thru an expensive, supposedly accurate machine.
chrisso
J3RK wrote:
I both agree with you and disagree. If a module is advertised to keep precise time, then it should be held to this standard. Otherwise, it's likely performing as advertised. An analog sequencer without a quantizer doesn't follow any more of a standard than a clock controlled be an analog voltage. You turn the knob, it speeds up. No increments to be measured here, other than a voltage.

I think what Innerclock is getting at, is if you run a sequencer from the 2LS or Expert Sleepers, then you run a second sequencer by a different company, unless there is an agreed standard, those two sequences will run out of time with each other, causing smearing, and in worst case scenarios flamming.
For me this is an issue because I can be running two or three clockable devices (a sequencer, a clock divider, a drum machine) on the same song.
I was assuming they were all 100% tight with each other, but Innerclock is now saying many drift ahead and behind the clock - randomly.
innerclock
chrisso wrote:
J3RK wrote:
I both agree with you and disagree. If a module is advertised to keep precise time, then it should be held to this standard. Otherwise, it's likely performing as advertised. An analog sequencer without a quantizer doesn't follow any more of a standard than a clock controlled be an analog voltage. You turn the knob, it speeds up. No increments to be measured here, other than a voltage.

I think what Innerclock is getting at, is if you run a sequencer from the 2LS or Expert Sleepers, then you run a second sequencer by a different company, unless there is an agreed standard, those two sequences will run out of time with each other, causing smearing, and in worst case scenarios flamming.
For me this is an issue because I can be running two or three clockable devices (a sequencer, a clock divider, a drum machine) on the same song.
I was assuming they were all 100% tight with each other, but Innerclock is now saying many drift ahead and behind the clock - randomly.


100% correct Chrisso - Clock Input to Clock/Trigger/Gate Output timing integrity is the core of this discussion. When I purchase a Sequencing Module or Clock Processor Module my expectation is that the module will take my programmed source Clock or Trigger Pulses and faithfully represent the source timing/tempo precision at its generated output. What I find in many cases is that the module adds jitter and random push-pull timing at it's output. At least if there was a standard I/O timing specification that could be clearly indicated by manufacturers then we could make purchase decisions in full knowledge of what to expect. best - David
WaveRider
innerclock wrote:

Upon testing many modules over the years I have found that accurate tempo is not often followed


I am with you 100% on the need for timing accuracy, in clocks and sequencers.... but my take on that is that we need someone to test and post results, and yes shame manufacturers that do not take into account that important function of sequencers! we'll then be able to choose knowingly... but I suspect there are not many in eurorack??? so we need to put an emphasis on that to influence future designs...

now let's name names Guinness ftw!
ndkent
I don't actually own a single euro module with a BPM numerical beat or time interval available and measured in seconds. I thought voltage control generally eschews that. I guess if you have a module with a microprocessor inside that is determining tempo then it's a reasonable request.
Mitchk1989
His point is that some sequencers actually lag a bit between receiving a clock pulse and actually advancing to a degree some would find unacceptable
chrisso
I think he means drift, not a set and consistent lag which can be corrected by dialling in an offset, or nudging audio in a DAW.
I had this problem with Logic Audio, which brought me to invest in Expert Sleepers.
Recording my synth into Logic, playing tow separate passes but exactly the same part, I had terrible and audible flamming.
I tried nudging audio back and forth and never fixed the problem.
Then when I went through and inspected transients, I found to my horror Logic's timing was variable, one note a little ahead, one note in time, the next note behind.
It is so random as to be unfixable.
the problem disappeared with ES and my ES4.
You don't hear timing inconsistencies when you record one part, but when you layer, or overdub the same musical passage with two sound sources, the inconsistency is immediately apparent.
Pailo
I love hearing of this because Latency with DAW's has become a nightmare
and using sequencer modules by every different company will always
have different results. Even interfaces have no time standard.

Even for example the TR-909 and its
MIDI and Sequencer within ITSELF! Completely drift from itself because
it was a primitive form of midi and an early sequencer with no precision
at all. it was an art.

Lately I also notice from one manufacture to another different terms, and
thougts on what time is LSB MSB......leading to things never syncing up
and you having to be a master of timing and arranging and recording
because every piece terms their systems differently.

Honestly I have moved to an entire CV based system and left MIDI.
No more faith in CPU's and Latency. Always have it no matter how
much we spend ($$). CV out of Trig outs with clock dividers, and cv
keyboards. Its 1971 in here again and i stopped using MIDI because
its too slow. The CV works Perfect. No question about it.


Back to the question here....
Is that the art of being a studio obsessed person?

a. I would love a "Standard" because it would educate me in building a
sequencer along with informing me of a time standard that would assure
future modules to perform together in sync without bumps.

b. I love learning an instrument so well that i can make screwed up
sequencers work together by spending hours & hours in the studio.

Either way.........
I like what you get at David for future designs.
chrisso
I've given up on midi too. Even though I write and record traditional sounding music.
I use a midi (USB) keyboard only to program parts that go to hardware via Expert Sleepers, or go to virtual instruments.
Midi has been very good, but it's time is done (IMO).
innerclock
How do these sound as a starting point for a possible future sequencer manufacturer module timing specification:-

1: Source Clock Input to Generated Pulse Output Lag: [Minimum/Maximum/Average]

2: Introduced Tempo Jitter with a Source Input Clock Reference: 120 BPM/16ths/6000 samples between pulses.

Module Generated Output Trigger/Gate Tempo Stability:

6000/5999/5985/6000/6005/6010/5999/6000/6000/5985

Introduced Tempo Change between consecutive Steps: [BPM: Min/Max]

Regards - David
innerclock
WaveRider wrote:
innerclock wrote:

Upon testing many modules over the years I have found that accurate tempo is not often followed


I am with you 100% on the need for timing accuracy, in clocks and sequencers.... but my take on that is that we need someone to test and post results, and yes shame manufacturers that do not take into account that important function of sequencers! we'll then be able to choose knowingly... but I suspect there are not many in eurorack??? so we need to put an emphasis on that to influence future designs...

now let's name names Guinness ftw!


Well I can certainly vouch positive for the TipTop Z-8000 on the precision factor:-

This Test: External Sample Accurate 120 BPM Pulse-Sync

Number of samples [48 kHz] between consecutive Sixteenth Notes:
6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000

Maximum variation between any two consecutive Sixteenth Note intervals:
Zero Samples [0.00ms]

applause
bouzoukijoe1
personally I'm not usually that critical when it comes to pitch in modular synths, but it does bother me if things don't trigger when they're supposed to. but come to think of it, generalized tests for error/deviation would be great. maybe some kind of measurement or formula like Circle of Confusion in photography

1. http://www.dofmaster.com/digital_coc.html
2. http://en.wikipedia.org/wiki/Circle_of_confusion

- timing like you mentioned is pretty important, maybe it could be based on a standard Hz/BPM table of common measures and tempos? (extremes too)

- envelope "speed" would also be great. maybe standardized ways to measure "fast envelopes"

- standard voltage drop might be another good one

- a pitch accuracy table might also be good, maybe similar to the Dotcom tracking accuracy examples?

interesting topic!

(disclaimer: I have to say though that this is probably the least interesting topic for some designers whose modules are not really about accuracy. some modules' strengths are not in their reliability or predictability, but rather just their ability to coax musicians into creating stimulating work)

komyta
I also left MIDI several years ago because of far too much irritating and frustrating timing problems. angry
I came to the conclusion that MIDI = innacuracy. very frustrating

Now I use Reaktor (I made sort of a Silent Way clone, more suited to my needs and preferences) and my DC-coupled MOTU interface as a timing / voltage generator system and I'm pretty satisfied with it. grin

I've never tried to hook up an external sequencer to that system (all the sequencing tasks are made in Cubase with Reaktor) but one day when I do it I hope I won't be disappointed...

So I'd really appreciate a "defined time-specification to be included on sequencer and clock processing modules". thumbs up

Good to hear that Tiptop Audio sequencer is a timing beast !
Matos
komyta
Sounds like an interesting route. I love silent way, but im curious what direction you went for this. any chance you want to share your patch?
Monobass
Pailo wrote:
Even for example the TR-909 and its MIDI and Sequencer within ITSELF! Completely drift from itself because it was a primitive form of midi and an early sequencer with no precision at all. it was an art.


maybe that contributed to its funky sound?

tests are ok.. I use ears though.
komyta
Matos wrote:
komyta
Sounds like an interesting route. I love silent way, but im curious what direction you went for this. any chance you want to share your patch?

No problem Matos. PM sent.
Christopher Winkels
I have little to contribute here (it's a good and interesting topic though) other than to say that any sequencer I've bought has had better timing than I could ever manage using my own two hands. I have terrible, terrible precision. In that context any sequencer/clock is a good (or should I say "better") clock than I am.
WaveRider
Monobass wrote:
Pailo wrote:
Even for example the TR-909 and its MIDI and Sequencer within ITSELF! Completely drift from itself because it was a primitive form of midi and an early sequencer with no precision at all. it was an art.


maybe that contributed to its funky sound?

tests are ok.. I use ears though.


the 909 also synced on DINsync
WaveRider
innerclock wrote:

Well I can certainly vouch positive for the TipTop Z-8000 on the precision factor


It's been mentioned a couple of times as very good....

now who's bad???? René???? this one may have jitter..? I hope that PP has good timing! SEQ02 runs pretty stable, stays on a given BPM for long, as I warp it into live I can see that...


witch are bad timing wise???

and when does Tiptop is going to make trigger sequencers????
WaveRider
Pailo wrote:

Honestly I have moved to an entire CV based system and left MIDI.


did that in 1990 and 2010... apparently I didn't learn the first time... i gave newer devices a shot but MIDI is just bad timing. And to me, when working on a groove, that tight analog feel is everything.
MUFF WIGGLER Forum Index -> Modular Synth General Discussion Goto page 1, 2, 3 ... 14, 15, 16  Next [all]
Page 1 of 16
Powered by phpBB © phpBB Group