As of September 23rd elektron-users has been replaced by elektronauts.com. Find out what this means here.
Elektron-Users Elektron Forum Elektron Gear MD SPS-1 Timing Performance Issues (1 viewing)
Go to bottom Post Reply
TOPIC: MD SPS-1 Timing Performance Issues
#18176
Cappy
Posts: 63
Re: MD SPS-1 Timing Performance Issues 17 Years ago
Well even Elektron themselves have acknowledged both your method of measuring the variation (which was your first battle in this thread!), and the fact that it most certainly matters - as Daniel himself says they put a lot of effort into finding their magic timing in the machinedrum. He quite clearly says the timing variations are important.

Then there the quite different argument as to whether super-tight timing OR slightly-sloppy timing is better...but they are quite clearly different...I just wish we could choose.
  The topic has been locked.
#18189
King Koopa
Posts: 242
0
Re: MD SPS-1 Timing Performance Issues 17 Years ago

husker wrote:
Well even Elektron themselves have acknowledged both your method of measuring the variation (which was your first battle in this thread!), and the fact that it most certainly matters - as Daniel himself says they put a lot of effort into finding their magic timing in the machinedrum. He quite clearly says the timing variations are important.

Then there the quite different argument as to whether super-tight timing OR slightly-sloppy timing is better...but they are quite clearly different...I just wish we could choose.


I'm still highly skeptical as to the 'deliberate' magic-random timing concept put up as a reason for the timing not being so spot-on in the SPS-1. However - the importance of choice regarding tempo stability/step jitter in a sequencer or drum machine was not lost on German synchroniser designers in 1984.

I have changed my 'Avatar' to a shot of the devices 'Jitter' module as proof. The manual description (forgive the poor German-English translation) I have uploaded also.

http://www.elektron-users.com/modules/wfdownloads/singlefile.php?cid=17&lid=698

It describes how to take the internal precision clock/sync signal, delay it by a frame so that the mean/average sync position is correct and then use the 'Jitter' value knob to add increased amounts of random (stochastic) clock/tempo/step errors.

The very existence of this feature alone shows two important points:

1. That strict tempo precision as a core time-base in sequencing anything was considered significant in the first place and;

2. That sometimes you may indeed wish to soften the edges a bit but you make this blur optional not the baseline precision reference of the device.

Regards - David
www.innerclocksystems.com
  The topic has been locked.
#18193
Kong
Posts: 1312
mr. applehead
Re: MD SPS-1 Timing Performance Issues 17 Years ago
hey , i found it curious that Daniel Elektron included this in the very informative post on yahoo groups..


the internal "magic" is only applied when
you run on internal clock on the internal sequencer and is/can not be
applied to incoming MIDI triggering data.


soo.. has anyone been able to test this with access to a stable midiclock device? i tried briefly with my jomox as master and got pretty sloppy results.

perhaps the answer for some is to slave the MD?
  The topic has been locked.
#18194
King Koopa
Posts: 242
0
Re: MD SPS-1 Timing Performance Issues 17 Years ago

neonleg wrote:
hey , i found it curious that Daniel Elektron included this in the very informative post on yahoo groups..


the internal "magic" is only applied when
you run on internal clock on the internal sequencer and is/can not be
applied to incoming MIDI triggering data.


soo.. has anyone been able to test this with access to a stable midiclock device? i tried briefly with my jomox as master and got pretty sloppy results.

perhaps the answer for some is to slave the MD?


Any talk of magic aside - it is true that a machine with a rough internal clocking engine can sometimes be 'tightened up' by providing external sync that has less jitter than the device itself. However - if the internal step/clock tempo jitter is at the very limits of the hardware/software then it is impossible to correct from even the most precise external sync source.

This holds true for the SPS-1. My initial tests were done with both the SPS-1 running on its own and slaved to very precise external Midi Clock. In both cases the same step/event/tempo random variation was evident which indicates baseline CPU/Code/DSP limitation.

What backs up my initial belief that the errors are not deliberate is the fact that the outgoing Midi clock is so very reliable on the SPS-1. The same internal SPS-1 clock/tempo/event generator drives both the outgoing Midi Clock and the Internal Sequencer/Voice DSP triggers. The outgoing Midi Clock has a 'clear pipeline' to the outside world. The same tempo/clock stream that also drives the internal sequencer and voice DSP generators has to deal with a lot more 'traffic' and is therefore subject to greater drift.

regards - David
  The topic has been locked.
#18195
Cappy
Posts: 87
Re: MD SPS-1 Timing Performance Issues 17 Years ago
David read this please,

These are the tools I've/we've got and I/we use em, pretty simple? I won't use the old saying about blaming but I'll get close

This thread should be closed. Its only going in circles.

Mods please?
  The topic has been locked.
#18196
Chain Chomp
Posts: 521
Dubby music & free samples
http://leocavallo.bandcamp.com
Re: MD SPS-1 Timing Performance Issues 17 Years ago
I don't think this thread should be closed.
And it's not going in circles either.
I think it's a highly informative resource for current/new/potential MD users and it's taking us closer and closer to the source of the issue.

I'll try triggering the MD machines with my MPC and report the results. That might reveal if it's the synthesis engine getting in the way of timing tightness or not...

Later
leo
  The topic has been locked.
#18197
King Koopa
Posts: 242
0
Re: MD SPS-1 Timing Performance Issues 17 Years ago

dreg wrote:
David read this please,

These are the tools I've/we've got and I/we use em, pretty simple? I won't use the old saying about blaming but I'll get close

This thread should be closed. Its only going in circles.

Mods please?


I read every post and respond where possible Dreg.

The very reason these are the tools we have right now is precisely why the thread is important and possibly why the topic has had nearly 5000 views in just over a single month.

It may also be why the moderators are still leaving it open. I hope so.

The importance of timing stability is core to what we all do. If the thread gives you nightmares or causes you distress it is much simpler if you do not read it rather than seeking to gag healthy debate. The issue will still generate interest and creative discussion regardless and it may benefit you someday if you allow it space.

I feel fairly confident now nothing can be done to improve the SPS-1/UW in its current form, however, I am still a big Elektron supporter and I am certain the interest and topics raised will be of interest to them in the future.

In the end I am still a potential customer for future products and that will always be of importance to any manufacturer in any business.

It is not going in circles either - many of the hysterical outbursts by people who took personally my views as an attack on Elektron, themselves, their music and creativity in general have either calmed down enough to see rationally, grasped the concept finally or dropped off altogether because they have nothing constructive to contribute.

On the other hand there have been many posts and I have had many private messages from individuals who feel just as strongly as I do and agree that if we remain silent on these issues then we keep getting more of the same and our tools don't improve.

I'm not blaming my tools for creative frustration - I am not frustrated at all. I understand well that accepting limitation can be a good thing sometimes.

My motivation for this thread is not blame or axe-grinding for something impossible or unnecessary.

I have tools already from years gone by that honor the concept that a stable tempo/event clock was/is the foundation of sequencer design.

The importance of that concept has never changed but our equipment designers have let it slip down the priority order.

Many people feel the same as I do about this.

I'm not saying I can't make music with what we have.

What I am saying is that precisely the attitude of 'just live with what you have and stop moaning' over the last 20 years in the area of electronic music production has directly contributed to the overall slide in focus - most certainly in the area of clock/event/tempo stability.

Without debate no-one learns anything.

Without pressure, nothing changes.

In 1984 - dedicated controls for adding jitter to precision click tracks, dedicated I/O on all hardware for synchronisation and sub-millisecond drum/trigger alignment were considered an essential and expected part of making good records.

Why is it in 2007 we are happy to accept that 2.2 millisecond random clock/event/step slop in 'The World's Best Beat Box' is 'feel' and should in some strange way be seen as a feature?

Because over time we forget.

We forget what good timing in electronic music actually sounds like. We are so used to hearing software and hardware doing impressions of Linns and TR-808s that we start to believe the new stuff is just like the old stuff and any subjective difference is just nostalgia for the past.

Bullshit.

Yes of course sound quality has a role to play in this too and the electronic instrument industry has developed and nurtured ever increasing sample frequency and bit rates in its quest to make digital a convincing clone of analogue.

However, almost no effort has been made over that same period to ensure that our machines that play these sounds back to us do so with the same precision and consistency as we used to take for granted in the past.

Due to very clever marketing mostly rather than an understanding of the processes involved most consumers these days expect their DAW/Sound Card to at least record and play back at 96 kHz.

In my line of work many individuals with perfectly good ears will quietly admit to not hearing much difference between 48 kHz and 96 kHz in most listening environments and yet they all would have no problem throwing that 48 kHz sound card in the bin and insist they must always record at 96 kHz.

Think of the serious time, engineering, reasearch, money and investment globally in developing precision Word Clock Generators with ever lower jitter figures to give digital audio mixes increased clarity and depth.

Would anyone with a brain walk into the Apogee or Prism head offices and tell them all to 'just be happy with what you have and make some music guys'.

Why is that same level of mostly blind, often peer-pressure driven commitment to higher sonic fidelity not seen when discussing timing stability in sequencing hardware and software when it is at least equal in significance when making music?

Baseline sequencer tight timing is worth striving for because it makes a huge difference to what we do and unless we, as the equipment consumers, make a noise about it then it just slips off the radar even further.

If we were to follow your advice and just be happy 'with the tools we have now' - in 5 years time, who knows, maybe 8 milliseconds of random clock/tempo/step slop will be considered the perfect human groove.

I'd rather reverse that trend if we possibly can.

Regards - David
www.innerclocksystems.com
  The topic has been locked.
#18198
Chain Chomp
Posts: 521
Dubby music & free samples
http://leocavallo.bandcamp.com
Re: MD SPS-1 Timing Performance Issues 17 Years ago
Another thing to consider, mostly directed to who keeps repeating "stop complaining and make some music".

The MD - as every other Elektron product - is definitely a top-end, professional product (even just for the price of it!).

So why shouldn't it deserve a professional and constructive attitutude from its users?

Most of us are not here to merely criticize the MD and spread bad reputation across the web.

We spend our time here because we'd like to understand where the issue (that many of us feel as an important point) is and hopefully help Elektron improve the situation.

On any hi-end, pro gear forum/list nobody will sidetrack the technical discussion about how a A/D-D/A converter, a boutique preamp or a summing box perform saying "shut the f**k up and go make some music".

Why shouldn't be the same thing here, regarding the MD?

The SPS-1 is marketed as a high end percussion synthesizer and rhythm sequencer.

In my own experience the synthesis part is simply brilliant: digitally cold and sharp as a razor blade (I definitely didn't buy my MD to emulate those analog TR boxes) but so incredibly versatile and inspiring.

Like many others I found that the sequencing part of the MD could be improved, mostly regarding its timing aspect.

So please tell me, where's the "heresy"?

Peace and Perfect Timing for everybody
leo
  The topic has been locked.
#18200
Game & Watch
Posts: 2163
Once you go Elektron...
Theres no going backetron!

-----------------
www.geirhelgi.com
Re: MD SPS-1 Timing Performance Issues 17 Years ago
^I do appreciate this thread

but one thing I'm wondering about, is this problem fixeable with a os update?
or is it strictly hardware?

because the md doesn't apply this 'magic' to incoming midi signals so it must be in the os, right?

oh well, I'll just wait for my md and I'll try it out :-D
  The topic has been locked.
#18202
Boo
Posts: 173
Re: MD SPS-1 Timing Performance Issues 17 Years ago
I guess that they could get rid of the push/pull timing that was measured here in the thread but cant eliminate the random factor...
so without the pushpull everything would probably sound worse than now because the random derivation becomes than more audiable...
  The topic has been locked.
Go to top Post Reply
Powered by FireBoardget the latest posts directly to your desktop

Login Form

start Player