Turtlerock Forum

Gday all,

just beeen talking to a mate about timing in sequencing using Cubase. My mate reckons hardware sequencers have way better timing....accuracy, etc.


just wondering, is there a way to make an ITB sequencing solution , like cubase Sx3, or 4...more accurate?? I've read about the innerclocksystems Sync Shift thingo....but im not quite sure how thatd interface with cubase...

on the other hand,..maybe ITB isnt bad?

interested to know what others opinions are that are more clued in about this than i am.

I was going spare a few years back doing MIDI sequencing in Nuendo, the timing was awful. I would play a piano part or some fast sequenced percussion and it would sound messy on replay, So I decided to do a test. I recorded a fast keyboard part that I played on a Kurzweil to audio and MIDI at the same time, and then played the two back together, and the MIDI part was so bad that it was flamming all over the place. I bought their MIDEX MIDI interface, which has time stamping and it made things dramatically better, but as usual Steinberg dropped support for it so I can't use it any more. I am told that Cubase / Nuendo has some kind of software timing system now instead, but I haven't tried it. I find using computers for MIDI to be uninspiring at best. It's fine for recording pads and basic stuff, but now I just play stuff in directly to audio and do drop-ins just like in the days of tape, I just got sick of the vibe disappearing as soon as I recorded certain things to MIDI.

jkhuri44 wrote:Gday all,

just beeen talking to a mate about timing in sequencing using Cubase. My mate reckons hardware sequencers have way better timing....accuracy, etc.


just wondering, is there a way to make an ITB sequencing solution , like cubase Sx3, or 4...more accurate?? I've read about the innerclocksystems Sync Shift thingo....but im not quite sure how thatd interface with cubase...

on the other hand,..maybe ITB isnt bad?

interested to know what others opinions are that are more clued in about this than i am.

Hi Jamil,

There are people more qualified than myself to describe the overall degradation of computer sequencers since the days of the Atari 1040STe running Notator or Cubase or the MPC3000 and earlier CV/Gate sequencers such as the Roland MC4B. I haven't heard definitive answers in the form of some thesis which may be put forward to draw at least constructive discussion on this topic but if you can possibly grab their attention, 3 people who would be able to shed some light on these issues are; Graham Hinton http://www.hinton-instruments.co.uk/, Rohan Mansell http://www.mansell-labs.com/about.html and David Lackey (Innerclock). I have had many discussions over the years with David and, having a background as a drummer, I was always interested in things to do with timing and the accurate placement of rhythmic events. I have done measurements with any hardware device which happened to end up in my home studio using the system that David describes on his Clock Watch page: http://web.webhost4life.com/innerclock/index.asp?action=page&name=35 This system is accurate enough to be repeatable and show the same results consistently with a given sequencer.

As far as I know there is no repository of information regarding the amount of push or pull exhibited by software or hardware sequencers other than the Innerclock site. I've never come across a page listing such deviations anywhere else so it's hard to definitively say that software sequencers have less accurate timing than hardware ones. However, I would be on the side of hardware maybe because that's mainly what I use :) Saying that, not all hardware sequencers have equally accurate timing. At 44.1/16 I've measured variations of between 1-3 samples with the MPC3000 up to 192 samples with an Elektron Machinedrum. I did a measurement with Cubase on an AtariSTe which sat around 5-7 samples with around 35 samples at the loop point. Newer gear seems to suffer under load, possibly because there have been more and more features crammed into gear over the last 10-15 years without the consideration that timing might be an issue.

One theory I heard from David Lackey which I think had been passed down to him by Rohan Mansell was that in the days before fancy Windows-based computers the scheduling responsible for determining which events come first in a data stream used to be written in machine language. This would be the equivalent of writing each line of code manually. At some point instead of the painstaking process of going through each line of code the instructions were put into data compilers responsible for the scheduling. The problem was that the data compiler could not anticipate that certain important (rhythmic) events would get pushed aside by other events when some dazzling new feature was added. That last line is mostly my own speculation but it seems to be backed up by what Graham Hinton has to say. Here's an interesting quote of from his website:

"Scheduling Delays: most computer based sequencers operate on an internal time interval, usually around 2.5ms. If the tempo is accurate to several decimal places this technique is undoubtably used and any MIDI event that misses one interval has to wait until the next causing a quantisation dependent on how busy the sequencer is. One very popular piece of software is incapable of getting a note within a couple of its own intervals from the clock that caused it!

Tell the software producer that accurate timing is more important than flashy graphics."

http://www.hinton-instruments.co.uk/reference/midi/promidi/index.htm

I believe it was sloppy sequencers that led Brian Transeau to start making electronic music by moving chunks of audio around on a sample accurate grid from the late 90's.

As far as Innerclock's Sync-shift goes, it will not solve any inherent problems of a software sequencer's data scheduling. However, the Sync-lock which is still in development is designed to allow you to drive any dinsync, midi or trigger sequencer external to your computer from the computer's sample accurate grid. I hope that helps shed some light on your question but as you can see it's definitely a pandora's box and beyond my resources to explain sufficiently.

yeah, i sorta came to be pissed off about this after i read Daves cool explanation on his site....and though, "f@#$, how much better can good timing make my stuff?"

and then thought, "why do house beats sound so cool" and then when i try the same shit, i have to do so much fancy wierd polyrhythmics and fancy shit to make "it" work...and sound mad..

then came 808 32nd hi hats...that sound like ass. or, a wierd buzz, rather than a machine gun effect :P

thanks for your lengthy response beatmad, if only there was a magic box that made software sequencing tighter...thats what i'd really like....

i really like the idea of having my arrangements on the screen, and coz im a child of the 90s..and a bit of a point click fellow, step sequencing doesnt really tickle my fancy...ideally i'd like to perform a part with MIDI percussion pads, keyboards, etc....either quantise, or leave....and have it sounds sick...

oh, and steve, i need hardware synths to actually record in my synth parts, hehe....i cant see it being too long before i own a few..they are getting very affordable...for some strange reason, dont tell anyone!!!

I've just gone and bought myself a Genoqs Octopus, which I have done instead of a new computer and Logic 8. The way I have set it up for now, I am using Logic purely
as a MIDI router to all my hardware synths. I have only had it 2 days and I can already see the immediate difference a hardware sequencer makes to timing. The
triggering of notes via MIDI is far snappier even if it is still going via Logic, but at the moment I am not actually syncing anything - just letting the Octopus play its
own internal clock and Logic is stopped, if you know what I mean. I am going to try to sync them tomorrow and record some MIDI streams and I expect
it to be much better. I got the Octopus to reduce my dependency on the computer for MIDI timing, I intend to do most of my MIDI tracking via Octopus from now on
and use Logic as the digital audio recorder and finish tracks that way. It's a big change to my learning curve, because I've only used Logic for the last
few years. But it is sort of getting dumbed down for hardware and becoming more softsynth oriented and MIDI is not the strategic option for Logic, I can see in its future.
Logic 8 itself doesn't support MIDI Clock for it be a slave so this is not a good sign at all. So the Octopus is my great musical experiment but already it just sounds
snappier and tighter somehow.



rachel

Hey Rachel, did you take out a mortgage on your house to buy the Octopus, or just win the Lotto? :-)

Hi Rachel,

Congrats on the Octopus....beautiful piece of gear.
My midi days are now less and less, but I still have a MC 4b and an Atari 1040.
Could never get a tight clock out of one of my macs !

Cheers
N
Y
M
O

I am using Logic purely
as a MIDI router to all my hardware synths

hey rachel, does that mean that your sequences play right out of Logic...purely using the octopus clocking, or are you triggering step sequences on ur Octopus?

ohh, and after more reading, this leads to another question...

hypothetically, if im running a session comprising of only VST synths, could I simply slave cubase to a master MIDI clock, which would be coming from an as of yet unidentified piece of hardware?? :P

I assume that would solve the timing and accuracy blues??

jkhuri44 wrote:ohh, and after more reading, this leads to another question...

hypothetically, if im running a session comprising of only VST synths, could I simply slave cubase to a master MIDI clock, which would be coming from an as of yet unidentified piece of hardware?? :P

I assume that would solve the timing and accuracy blues??

I don't think so, are the timing problems that you are experiencing as I described, ie: a general sloppiness? If you are using a later or current version of Cubase, have you looked in the help file to see if there is a MIDI timestamping option that you can turn on? I am told it is in there somewhere, I will boot Nuendo and have a look. Wait, no I won't, it's Sunday morning and we all need to get a life instead of being on web forums...

Sunday?? what weekend?

im just about to start a session !!

haha

i will check it out Steve, wasnt sure that anything internal would do anything....and ps. im not experiencing anything THAT suspicious...im actually more so looking for a revelatory experience...

coz as of yet, i dont really know better than the timing ITB.

jkhuri44 wrote:Sunday?? what weekend?

im just about to start a session !!

haha

i will check it out Steve, wasnt sure that anything internal would do anything....and ps. im not experiencing anything THAT suspicious...im actually more so looking for a revelatory experience...

coz as of yet, i dont really know better than the timing ITB.

Which version of Cubase are you using?

SX 3 at the mom.

would like to upgrade to 4 asap though

are the timing problems that you are experiencing as I described, ie: a general sloppiness?

actually...on further thought....i ve been playing keys for about 15 yrs...

and if i cant play something tightly by now, i think there's something very very wrong....but thats what happens everytime i record a part..i have to slide shit around....or quantize.


and hey, i thought that was all just me thinking im god when im playing...but i think what we're talking about could be attributing to the problem...coz when im playing, i'll be mid take and say..."YES, this is the one"...then playback is either a) sloppy, b) telling me im a bad player..

im pretty sure its A.
hehe

jkhuri44 wrote:

are the timing problems that you are experiencing as I described, ie: a general sloppiness?

actually...on further thought....i ve been playing keys for about 15 yrs...

and if i cant play something tightly by now, i think there's something very very wrong....but thats what happens everytime i record a part..i have to slide shit around....or quantize.


and hey, i thought that was all just me thinking im god when im playing...but i think what we're talking about could be attributing to the problem...coz when im playing, i'll be mid take and say..."YES, this is the one"...then playback is either a) sloppy, b) telling me im a bad player..

im pretty sure its A.
hehe

Well, if you are playing virtual instruments then you have the extra problem of latency, and that feeling of disconnection that comes with it. I would suggest that if you can get hold of any MIDI keyboard that has internal sounds, or if you can plug a hardware sound module into your controller, or borrow a MIDI drum machine, then do the same test that I did: set up 1 MIDI track and 1 Audio track in Cubase, and record the audio from the synth at the same time as you record the MIDI from it, play lots of fast notes on a short percussive sound, or a cross stick or closed hat, preferably not lots of chords, and then see how they sound when you play them back.

If they phase then all is good, if they flam then welcome to computer hell.

time to crack out my roland Electric Piano...

will experiment on monday...muhahahaha.

im expecting scarily bad results, haha

Thirteen wrote:Hey Rachel, did you take out a mortgage on your house to buy the Octopus, or just win the Lotto? :-)

Hmmm, I know it is an expensive toy for someone like me! I actually have been saving for a Mac Pro but decided that it would be yet another boat
anchor in 5 years time so I spent the money on the Octopus instead. I see it as a lifetime purchase now and something I would probably
never get another chance to do. And I think I am glad I did.... ;)


rachel

jkhuri44 wrote:

I am using Logic purely
as a MIDI router to all my hardware synths

hey rachel, does that mean that your sequences play right out of Logic...purely using the octopus clocking, or are you triggering step sequences on ur Octopus?

Well before the Octopus, I was doing it all in Logic and using it to trigger all my hardware synths. Now my plan is to only use the Audio side of Logic
and record the output into it, but Octopus will be the master sequencer for all the gear. Because I am teaching myself how it works, I am just using
Logic for monitoring at the moment and routing the MIDI to the appropriate instruments. I intend to use the Octopus as the master MIDI clock
from now on. Logic 8 doesn't support an external MIDI clock master which is why I am stuck on version 6.4.3 now. So the final intent is to not
rely on the computer as much, store my sequences in the Octopus and use it to compose/perform and not be tied to one software product on the
Audio/MIDI side so much.

Nymo - I've seen your MC4 on ebay a couple of times. I would've liked one but I have an MC-8 already, which is a long term project to
get fully operational if I can ever source the bits. Steve had a look at it and said it probably works OK, but we can't diagnose further as I am
missing the 60 pin Hirose cable that attaches to the OP module. I bought the MC-8 cheap as a fixer-upper but until I can get the cable it
languishes in the wardrobe. Now I have the Octopus though, I might sell it on, but I do love that old school approach to sequencing!


rachel

In Nuendo and Cubase you have to check the system timestamp in device setup. I have no problems with midi timing in Nuendo 4 now after I checked this box.

Hope this helps.

Cheers

tunetown wrote:In Nuendo and Cubase you have to check the system timestamp in device setup. I have no problems with midi timing in Nuendo 4 now after I checked this box.

Hope this helps.

Cheers

For Windows only, unfortunate for us Mac users....

Ahhh, Bugger. Maybe try the Nuendo or Cubase forum. I can't believe Steinberg would have shafted the MAC users like that. Midi timing is critical. If you are not a member of either, let me know and I'll ask the question for you.

Cheers

thanks pete,

now that i know wtf time stamping is...(it used to be something i looked at and didnt know what it was)...

i will read about it in the manual, then let you know if im still confused :)

jkhuri44 wrote:thanks for your lengthy response beatmad, if only there was a magic box that made software sequencing tighter...thats what i'd really like....

Thanks Jamil, I know what you mean,. With all the advances in computers you think that someone would have come up with a better solution to integrating audio and midi. Seriously, I think the best solution at this point in history is to just keep them separate. Especially when you consider that the best sequencers are damn near close to sample accurate in terms of clock stability. Of course the serial nature of midi means that there will always be delays, particularly in real-time performance and even the theoretical possibilities of midi are rarely matched in the real world of synthesisers and samplers. Your'e welcome to come over to my place and see what it's like to record a Juno 106 part into an MPC3000 sequencer. You may be a child of the 90's but that might atleast give you some point of reference by which to measure the performance of your computer sequencer. Unfortunately this doesn't solve the problem of trying to make the most of sequencing your softsynths from ITB and as far as I know it's not as simple as taking a good hardware sequencer and driving your computer softsynths from it. I think some weird !@(^%#* happens somewhere between the midi out, the interface and the audio representation of your softsynth. Perhaps the delays of the screen redraw, combined with the crunching of other various binary numbers that happens in some particular order unbeknownest to myself is the cause. I wouldn't even hazard a guess but if you can find a good sequencer stick with it.

Hi Rachel,
I think with the setup you have now Innerclock's Sync-lock is going to come in real handy. I spoke to Dave today because I wanted to get his ok to talk about the Sync-lock although I'm still going to limit how much I disclose because there are still a few things to be ironed out and really he can, if not more eloquently than myself, put things more succinctly.

I went over to his place last Thursday night and saw Sync-lock #1 in action. Have you ever heard a CR8000 swing, driven by a clock pulse set up in Ableton? How about hearing a midi, dinsync and trigger sequencer all playing at the same time driven from that same clock all with the same degree of swing, with the ability to stop/start in real-time and never lose even a clock cycle and near perfectly locked to eachother? That's what it can do. What this means to me and something I have been pondering since the beginning of the year is that the precision timing of your sequencer of choice will be preserved whilst deriving it's clock source from the computer. As a real-time tool, you'll be able to stop any external midi/trigger/din sequencer at any point you like and restart, knowing that it will be in sync with whatever audio you have ITB. Buying a quality hardware sequencer makes sense now because the Sync-lock provides a means of integrating it with the computer you record the audio to. I hope to be beta-testing a Sync-lock within the next few weeks so if you're interested maybe we could have a demo night? I don't think there'll be many beta units but it might be worth hitting Dave up if you're really interested. :) I think it's really exciting, and I know it's a big moment for Dave too with all the work he's done over the last few years to find solutions to sequencer synchronisation problems.


Hi Nymo, I just thought I'd say that I've seen your MC4B too and my advice is don't sell it! Or the Atari. Wait until you hear them swing with any amount of shuffle you choose to send them via a Sync-lock from whatever audio program you use. And if you've sold the minimoog, buy another one. Hey, I can give that kind of priceless advice, in the Vintarge Thynth Phorum :)

Your'e welcome to come over to my place and see what it's like to record a Juno 106 part into an MPC3000 sequencer. You may be a child of the 90's but that might atleast give you some point of reference by which to measure the performance of your computer sequencer.

hey man, i've had the pleasure of recording hardware synths as audio many a time, and remember those times very fondly..thanks for the offer to play, rather than learning anything, i think it would just be moreso fun, hehehe

until now, i thought the only thing letting down VSTs and such was Latency....timing seems to be a bitch now as well.....

my problem is further compounded by the fact that i need to use VST samplers alot, especially for sampling world instruments, and orchestral sounds, which thus far, are much better in VST form than hardware samplers...unless off course you're Hans Zimmer....

Hi Jamil, Rachel, Steve, Beatmad, Nymo et all – we’ve been busy in the Innerclock Bunker these last months - still a few weeks off production run but the Sync-Lock does exist finally:-

http://web.webhost4life.com/innerclock/ ... ge&name=16

Our sole focus at Innerclock Systems from day one has always been getting rhythmic timing back where it should always be – in the pocket.

When we designed and released both the original Sync-Shift in 2004 and the Sync-Shift Mark II in 2006 we had real needs that required a real solution - Start Lag Time in tempo-synchronized systems. Drove me crazy – had to find a fix so I could sleep at night!

One long-term issue of continual frustration over many years for us has been the poor stability of external Midi Clock Tempo synchronization in a software/computer environment. A full analysis of this problem is way beyond the scope of this post of course but the issue is an old one and familiar to all of you of that I am certain.

In our experience, electronic music producers these days usually fall into two main groups - those that choose all hardware over software and those that choose to do it all inside a single computer using plug-ins and softsynths.

There are a number of reasons for this of course but a big one is the difficulty in achieving reliable synchronization between computer music applications and hardware in the outside world. Why is this so hard?

Two reasons:

1. Latency – The difference in time it takes for your soundcard to generate its internal audio and the start time of your external gear.

2. Midi Clock Slop - PC/Mac (and some hardware too) generated Midi Clock is mostly soggy custard and even the best timing external Hardware (think MPC-3000) will run and sound like an old wet blanket when driven from even the best software application.

We have always appreciated the myriad of advantages that modern computers offer to electronic music production (Think Live, Reaktor/Reason etc) but the poor timing integration to the outside world where so much quality vintage and current hardware and hands-on creative interactivity thrive has left us mostly no choice but to leave them well alone.

18 months ago we set about trying to find a no-compromise solution to this problem so we could close the gap between computers and external hardware in rhythmic electronic music production.

The Sync-Lock lets you still use a PC/Mac - any application of your choice - and generate sample/grid accurate simultaneous Midi Clock, Din Sync and Voltage Triggers.

All of this is done without USB or Midi Clock generation from the software and is therefore immune to jitter/slop and because the sync signals are provided via your soundcard audio outputs – there is effectively zero latency offset between your computer’s internal audio and the start times and tempo sync of your external equipment. Sync-Lock Pulse to Trig conversion time is under 20 microseconds!

True Plug and Play. You can have your cake and eat it too!

There are some neat features – Footswitch Manual Quantized Stop/Force Re-Start and of course you can program in Midi Clock/Din/Trigger tempo changes, pauses, stutter restarts, you can very easily program in swing to your sync track in your application of choice – set all your external hardware to play straight and let the Sync-Lock shuffle all your external devices at the Sync Source rather than try and work out bounce settings by feel. No latency hard swing funk between Ableton Live Audio, a TR-808, an Elektron Machine Drum and an Analogue Step Sequencer by just pressing the PC space bar is a wondrous thing to behold! And not a Midi/USB cable from the computer in sight.

Best as always to all.

David
www.innerclocksystems.com

Hi there,

David....mmmmmm.....I need me one of those thingz ;-)

Cheers

N
Y
MO

No problem Nymo - a few weeks I think at worst - I did the original concept tests driving an MC-4B off Ableton Live - never heard anything like it with absolutely no effort at all....... I'll keep in touch when I get close to ship time......:) Thanks for the positive vibes too.....

Best - David
www.innerclocksystems.com

I too got very annoyed with MIDI slop and did some investigating. USB1.1 MIDI interfaces by default use bulk data endpoints, which means MIDI is relegated to lower priority than interrupt and isochronous endpoints. The slop I measured using Cubase SX3/Ableton5/6 and MidiSport2x2 can be +/-1ms. USB2.0 bulk endpoints however, have a much lower latency. Ideally, MIDI should have used an isochronous (just in time) endpoint (like audio) for data.
David, than sync-lock sounds like it'll solve all our syncing woes!

Cheers

my CME 76 key controller im plugging into my comp VIA usb....

should I be using MIDI cable?? According to the above statement BOB SNARE?

i started being by being a midi cable snob, then i said...screw it, ill use USB...

should i reconsider??

jkhuri44 wrote:my CME 76 key controller im plugging into my comp VIA usb....

should I be using MIDI cable?? According to the above statement BOB SNARE?

i started being by being a midi cable snob, then i said...screw it, ill use USB...

should i reconsider??

It depends. Do you use System Realtime, MIDI Clock or similar?


rachel