The No.1 Website for Pro Audio
Rm1x vs. RS7K vs. EMU XL7 for general sequencing
Page 2 of 2
1 2
📝 Reply
Old 12th May 2012
  #31
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
This was straight up 1/32nd notes, no swing, 120 bpm, going straight from MIDI OUT to MIDI IN of a bunch of different synths (was trying to make sure it wasn't just the target synth's response that was messed up). No interface in the way.

szf, was your jitter measurement doing MIDI or the internal sound generator? If the latter, then the timing is just jacked on the RS7K =|

Reptil, I'm not looking for a discussion about jitter in general, I've read all those threads already =) I'm looking to see if the RS7K is as bad as I think it is (I was running my M8UXL as a benchmark, saw how bad IT was, then decided to run the RS7K and see if it was better, and it's definitely better but still not what I'd call "tight").
Share Reply Quote
Old 12th May 2012 | Show parent
  #32
szf
Lives for gear
 
szf's Avatar
 
6 Reviews written
🎧 10 years
Jitter test was on the internal engine, woudn't expect the midi jitter is any better.
Share Reply Quote
Old 12th May 2012 | Show parent
  #33
Gear Head
 
Alkaline3's Avatar
 
🎧 5 years
I have used the RS7K as my main live-sequencer the last 10 years.
It works very well on the stage!
Amazing for live-improvisation and onthefly manipulation, rearranging etc.
It is great fun and very stable, and you can do a lot of tricks with the mididelay/transpose/groove functions combined with the scene memories.

In the studio, the jitter + sync problems becomes obvious.

Midi clock (or MTC) from Logic->RS7K is a mess:
A 4 onthefloor for 3min drifts so much that it practical becomes useless if you have to edit or align additional recordings afterwards.
A sync-lock from Innerclock systems has solved the alignement issue,
but the individual hits/downbeats is still drifting.

Also, the internal looping of the phrases can be sloppy.
Every first downbeat of a phrase has a tiny glitch..a short delay.
This means there is a difference between a 1bar loop, and the same pattern copied to itself to make a 4 or 8 bar loop (up to 256).
In the studio, I have get used to make the patterns longer than needed to avoid this problem.
I never use the internal drum sound, so I can´t recall if this problem also occurs with its own sounds.

Anyway the machine offers so much, that I´m willing to do all the workarounds, -also when recording.

The old RS7K forums has closed down, but there is a pretty active facebookgroup.

Hans
Share Reply Quote
Old 12th May 2012 | Show parent
  #34
Gear Head
 
🎧 5 years
Quote:
Originally Posted by Reptil View Post
here's a very good alternative. It's not difficult to make.
I use an MPC1000 and Emu MP-7, both are good machines. The combination is a lot of fun.

do a search this sequencer subject has been beaten to death!

I used to use an mpc4k and an mp-7. Theyre both awesome my favorite seqeuencer/ drum machine is def the command stations. And they have two midi outs. For the price for all they can do control and create it cant be beat. I love the workflow and i always sequence on there when i dont want to use ableton and boot up my pc. Creatting patches on these is a pain but its a deep machine.
Share Reply Quote
Old 12th May 2012
  #35
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
Well it sounds like I'm not alone on this assessment (especially if it does it internally). It's a holdover until I get my Cirklon, in the mean time I may just do 1-4 bar loops and do everything in audio for now.
Share Reply Quote
Old 12th May 2012
  #36
Lives for gear
 
asynchro_nous's Avatar
 
🎧 10 years
FWIW I love the tight feel of my RM1X -- I wonder if it has cleaner jitter specs than the more elaborate RS7k?
Share Reply Quote
Old 14th May 2012
  #37
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
For posterity I thought I'd throw out some findings.

I did a straight 16th note, 120 bpm, no swing, MIDI OUT pattern (1 bar) into a TX81Z and recorded the audio, then measured the samples (24-bit, 48KHz, Mackie 1604i).

The two measurements I cared about were note-on intervals (how consistent is the timing) and, more important to me, how far it deviates from the grid over time.

What I found really interesting is that the note to note consistency was generally pretty good (~10 sample average) but that it was consistently a little late. So instead of 6K samples between hits it would average more like 6010. This would obviously lead to some pretty epic timing drift, BUT near the end of the loop it corrects, overshoots, then corrects again so that when the phrase starts again it's locked back on with no deviance from the grid.

Okay, picture thousand words etc. etc.



You can see clearly how it gets farther off the grid then suddenly goes "Whoops!" and compensates in the other direction, only to overshoot again and recompensate.
1
Share Reply Quote
Old 15th May 2012
  #38
Here for the gear
 
🎧 10 years
Don't know if this is causing the jitters on your RS or not, but I had to chime in and say to make sure there are no PC events at the start of the events list for all tracks of each phrase loop on style/pattern. Also, if you are getting hicups in chain mode which will convert over to song mode as PC events too, you may want to stay away from sequencing section or style changes if you are running in to glitchy, stuttery, jittery timing issues on it. Sequencing sections is usually OK though if you don't include the style, pattern/style loops don't always have defined length and is dependent on the phrase lengths which can be diffrent lengths on every track. The RS has to re-buff all voice data settings for each track when you do a style change, tough for it to do while playing even though it attempts to do it anyway, it is better at switching those out when not playing or little is going on like at the start of a song, but not during the rest of it while playing. I believe it also likely takes up polyphony to do it too which just adds more stress to an already heavily taxed processor when playing. It is kind of a known issue with how the RS works and stores internal voice data settings to the track rather than the voice banks. Settings are stored to sample banks though which brings up another set of issues I won't go in to, but it is a bit strange concept to wrap the old noggin around compared to how most seq/synths work, but they chose to do it that way, and it does have the benefit of having your songs/patterns always sound the same no matter how tweaked out you change the voices on other songs and patterns. That don't happen on a lot of other seqs, you have to store and switch to new bank or use PCs and other events to recall similar voices with minor changes to them, which probably would likely cause hiccups and jittering too if you did it every loop. So make sure that is not happening on your RS if can by removing those events, the track will still remember the active voice settings every loop and cool thing is it won't jump back to stored settings either at the start of every loop, it stays smooth while you do manual sweeps through end of one loop to the next even if you are changing phrases/sequenced note loops by switching sections, and also if those events are there and you get rid of them it will likely help a lot to stabilize it's clock and timing problems. Or perhaps maybe not.

On the other hand, over the years I have heard complaints on the RS's timing, so I can't discount that as a valid concern especially if you are looking to make long samples for aligning up to other long samples. However, if you use the RS as stand alone, great great machine, make sure you have latest OS installed, extra sections and extra filters too, can even use them on the drum kits with LFO if you know the secret RS7000 handshake.

Also want to add all my buttons, switches and knobs still work great and I've used my RS almost daily, well a lot, over nearly ten years now.
Share Reply Quote
Old 16th May 2012
  #39
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
...and because I'm a completist OCD kind of guy, I decided to compare the internal sound generator of the RS7K vs. a direct MIDI OUT to my TX81Z.

These are 16th notes, no swing, running simultaneously on two tracks, one sending to internal sound generator and the other to MIDI OUT connected directly to a TX81Z, with audio going into my Mackie 1640i recorded at 24-bit, 48Khz into Reaper on Win7. A total of 4 bars in the patter to differentiate between pattern vs. measure artifacts.

You can see how the MIDI port's grid error (deviation from the grid) converges back to normal, then is thrown off by a ridiculously early note (132 samples early, almost 2.75ms). This happens roughly every measure.

The internal sound generator is really solid, with an inexplicable laggy note right at the end of the pattern (followed by a compensation on the following note to get it back in sync). It did have incrementally worse grid error, which if it continued to trend down would be an indication that its clock is running slightly faster than 120bpm and that it would likely sync back up via MIDI sync if operating with other gear.

tl;dr - MIDI timing is predictably bad, with a major deviation every measure, and the internal sound generator has pretty solid timing (with 10 samples) with a little hiccup at the start or end of the pattern.

Last edited by BaconTastesGood; 16th May 2012 at 01:55 PM.. Reason: whoops, wrote 's' instead of 'ms'
1
Share Reply Quote
Old 16th May 2012 | Show parent
  #40
Here for the gear
 
🎧 10 years
Well, we are following each other around the web on this, sure you already seen my replies:

I just got home and ran some tests too on the internal sound of a high hat like you mentioned. I couldn't do it internally that way since it only goes up to 44.1, but went phones out through my MW10 usb mixer. I see you are correct, the first note in any phrase is consistently off by a large factor in comparison to the rest of 32nd notes with 32nds. However I have more to add which may be interesting and maybe favorable or not for the RS.

I tested with 3 seperate phrases playing at the same time. phrase 1 was 2 bars, phrase 2 was 1 bar, phrase 3 1/16th so I got two 32nd high hat hits on that one. I played them simultaneously together however I muted each track individually.

What I found is that the off note was always the first of each phrase no matter what, having the tracks playing together did not help the other. Also, majority of the time I noticed it was about a 32 sample points delayed most of the time. I did get one once when I had just the one track alone 1st track with nothing else going on that it was advanced about 32 sample points. I was unable to replicate that, it was always late after that first time. So WTH on that. On the 1/16th phrase every other note was consistently lagged about 32 sample points.

Indeed interesting. To me it seemed pretty consistently off, always first hit about 32 samples delayed, the rest always seemed pretty well on target for internal hi-hat voice.

I also copied a track all tracks, it stayed consistant while listening to just one track. With multiple tracks going though and the same hi-hats, I could hear a phase difference on the daddy hits, and I've actually noticed this effect before but never knew exactly why or what was causing it until now.

Well again, thanks for bringing this to my attention. I can see that being an issue for folks, especially if it gets worse with sending out to midi other gear. I still hope the RS will find it uses for you though, if not I can take it off your hands for cheap

Zam
Share Reply Quote
Old 16th May 2012
  #41
Gear Addict
 
🎧 10 years
Quote:
Originally Posted by BaconTastesGood View Post
You can see how the MIDI port's grid error (deviation from the grid) converges back to normal, then is thrown off by a ridiculously early note (132 samples early, almost 2.75s).
I think the "ridiculous" is on the scale of 2.75 milliseconds , not seconds, ??

Whilst Inner Clock Systems Litmus TEsts show some gear to be tighter, they also show Akai's latest MPC5000 flagship to have an equally "ridiculous" 138 sample variation.

Interesting results, though I wonder what effect other parts in the chain such as internal midi handling of TX81Z
and firewire audio /drivers and software contribute when counting final results at sample level.

That said, I would be interested if it is any tighter slaved to a reliable midiclock ? ( though it seems many pieces aren't any tighter when clocked to the Inner Clock ,on their own Litmus test !)
I would also be interested to see how the Emu Command Sequencer compared, and the highly lauded CirKlon itself ?

Edit.a further musing.
Share Reply Quote
Old 16th May 2012
  #42
Lives for gear
 
2 Reviews written
🎧 10 years
I can't speak for the midi timing but I have a decked out MP7. It also as a very respectable synth engine and a lot of knobs/buttons. I've used it as a master sequencer on some small projects and it was great. The downside is that it's a complicated machine. The manual is, like, 350 pages and you'll want to read it and then keep it close at hand. Plus it has a great free online editor (not a VST).
Share Reply Quote
Old 16th May 2012 | Show parent
  #43
Lives for gear
 
grumphh's Avatar
 
🎧 10 years
Quote:
Originally Posted by Zamise View Post
Don't know if this is causing the jitters on your RS or not, but I had to chime in and say to make sure there are no PC events at the start of the events list for all tracks of each phrase loop on style/pattern.

Also, if you are getting hicups in chain mode which will convert over to song mode as PC events too, you may want to stay away from sequencing section or style changes if you are running in to glitchy, stuttery, jittery timing issues on it.

Sequencing sections is usually OK though if you don't include the style, pattern/style loops don't always have defined length and is dependent on the phrase lengths which can be diffrent lengths on every track. The RS has to re-buff all voice data settings for each track when you do a style change, tough for it to do while playing even though it attempts to do it anyway, it is better at switching those out when not playing or little is going on like at the start of a song, but not during the rest of it while playing.

I believe it also likely takes up polyphony to do it too which just adds more stress to an already heavily taxed processor when playing. It is kind of a known issue with how the RS works and stores internal voice data settings to the track rather than the voice banks.

Settings are stored to sample banks though which brings up another set of issues I won't go in to, but it is a bit strange concept to wrap the old noggin around compared to how most seq/synths work, but they chose to do it that way, and it does have the benefit of having your songs/patterns always sound the same no matter how tweaked out you change the voices on other songs and patterns. That don't happen on a lot of other seqs, you have to store and switch to new bank or use PCs and other events to recall similar voices with minor changes to them, which probably would likely cause hiccups and jittering too if you did it every loop.

So make sure that is not happening on your RS if can by removing those events, the track will still remember the active voice settings every loop and cool thing is it won't jump back to stored settings either at the start of every loop, it stays smooth while you do manual sweeps through end of one loop to the next even if you are changing phrases/sequenced note loops by switching sections, and also if those events are there and you get rid of them it will likely help a lot to stabilize it's clock and timing problems.

Or perhaps maybe not.

On the other hand, over the years I have heard complaints on the RS's timing, so I can't discount that as a valid concern especially if you are looking to make long samples for aligning up to other long samples. However, if you use the RS as stand alone, great great machine, make sure you have latest OS installed, extra sections and extra filters too, can even use them on the drum kits with LFO if you know the secret RS7000 handshake.

Also want to add all my buttons, switches and knobs still work great and I've used my RS almost daily, well a lot, over nearly ten years now.
Paragraphs, how do they work?
Share Reply Quote
Old 16th May 2012 | Show parent
  #44
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
Quote:
Originally Posted by subhertz View Post
I think the "ridiculous" is on the scale of 2.75 milliseconds , not seconds, ??
Whoops, bad typo there, fixed, thanks for catching that!

Quote:
Whilst Inner Clock Systems Litmus TEsts show some gear to be tighter, they also show Akai's latest MPC5000 flagship to have an equally "ridiculous" 138 sample variation.
I do wish their test results were a bit better organized. Some of the gear they list has some pretty incredible timing but it's hard to find since all the tests and test types are mixed together.

Note that the "ridiculous" comment is mostly pointing out the timing relative to itself. It's very solid (<10 samples) then suddenly has a huge spike, and it's very repeatable.

And to be clear, whether that 2.75ms is relevant depends largely on the style of your music. If you're doing fast tempo mechanical stuff then it can definitely be an issue, particularly if you're trying to layer things precisely. In that situation you may have to revert to operating in the audio domain.

If you're doing ambient music, pads, slow instrumentals, etc. I doubt it'll make any difference at all.

Quote:
Interesting results, though I wonder what effect other parts in the chain such as internal midi handling of TX81Z
and firewire audio /drivers and software contribute when counting final results at sample level.
Since the tests were done with both internal sound generator and MIDI and there was a big difference between the two I think it's safe to say you can factor out the PC elements.

The TX81Z could have poor MIDI handling, but I had similar results from an MKS-50 and Wavestation SR, so I don't think it's unique to them. They're actually in a MIDI chain so I can superimpose their audio tracks and see how much latency exists through the THRU.

Quote:
I would also be interested to see how the Emu Command Sequencer compared, and the highly lauded CirKlon itself ?
I have a Cirklon on order and will post results when I get it but it seems like end of summer at earliest for that =(. I don't have a Command Sequencer available. I have a Schrittmacher showing up soon and will post results of it as well.
Share Reply Quote
Old 16th May 2012 | Show parent
  #45
Gear Nut
 
rarara's Avatar
 
🎧 10 years
really interesting - its probably something minor that needs to be corrected within the OS. what a shame these things are not made open source after a while to allow people to then make enhancements

anyway - i don't suppose you could try this after a factory reset, and without any autoload? just to see if it changes things in any way, since i have seen slightly unexplained things in the past with these.

Quote:
Originally Posted by BaconTastesGood View Post
For posterity I thought I'd throw out some findings.

I did a straight 16th note, 120 bpm, no swing, MIDI OUT pattern (1 bar) into a TX81Z and recorded the audio, then measured the samples (24-bit, 48KHz, Mackie 1604i).

The two measurements I cared about were note-on intervals (how consistent is the timing) and, more important to me, how far it deviates from the grid over time.

What I found really interesting is that the note to note consistency was generally pretty good (~10 sample average) but that it was consistently a little late. So instead of 6K samples between hits it would average more like 6010. This would obviously lead to some pretty epic timing drift, BUT near the end of the loop it corrects, overshoots, then corrects again so that when the phrase starts again it's locked back on with no deviance from the grid.

Okay, picture thousand words etc. etc.



You can see clearly how it gets farther off the grid then suddenly goes "Whoops!" and compensates in the other direction, only to overshoot again and recompensate.
Share Reply Quote
Old 16th May 2012
  #46
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
This was all done after a factory reset and a totally empty style.
Share Reply Quote
Old 17th May 2012
  #47
Gear Addict
 
🎧 10 years
@ Bacon

Forgot to ask previously, for the record , were the RS7000 tests done with last O.S v1.22 ?
Share Reply Quote
Old 17th May 2012 | Show parent
  #48
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
Quote:
Originally Posted by subhertz View Post
@ Bacon

Forgot to ask previously, for the record , were the RS7000 tests done with last O.S v1.22 ?
Yep!
Share Reply Quote
Old 12th June 2012 | Show parent
  #49
Lives for gear
 
tyler477's Avatar
 
1 Review written
🎧 10 years
how you tried running these tests with the rs7000 set as slave?
Share Reply Quote
Old 12th June 2012 | Show parent
  #50
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
Quote:
Originally Posted by tyler477 View Post
how you tried running these tests with the rs7000 set as slave?
It shouldn't make a difference -- external sync should never be more accurate than internal clock when measuring note to note consistency. If it is, then your internal clock is effectively not trustworthy for even basic timing.
Share Reply Quote
Old 13th June 2012 | Show parent
  #51
Lives for gear
 
tyler477's Avatar
 
1 Review written
🎧 10 years
Quote:
Originally Posted by BaconTastesGood View Post
It shouldn't make a difference -- external sync should never be more accurate than internal clock when measuring note to note consistency. If it is, then your internal clock is effectively not trustworthy for even basic timing.
shouldn't and doesn't are 2 separate things. I'm going to try syncing my RS7000 to MPC 2000xl, I'll report back.
Share Reply Quote
Old 13th June 2012
  #52
Gear Addict
 
BaconTastesGood's Avatar
 
🎧 5 years
Let me put it another way. If I synced up and found that the RS7K was significantly tighter in terms of timing but only when acting as a slave, I'd probably sell it on general principle. That said, its days are numbered any way while I wait for my Cirklon...
Share Reply Quote
Old 7th July 2012 | Show parent
  #53
Deleted be0c6fb
Guest
 
Quote:
Originally Posted by TempoTanken View Post
I have used the RS7K as my main live-sequencer the last 10 years.
It works very well on the stage!
Amazing for live-improvisation and onthefly manipulation, rearranging etc.
It is great fun and very stable, and you can do a lot of tricks with the mididelay/transpose/groove functions combined with the scene memories.

In the studio, the jitter + sync problems becomes obvious.

Midi clock (or MTC) from Logic->RS7K is a mess:
A 4 onthefloor for 3min drifts so much that it practical becomes useless if you have to edit or align additional recordings afterwards.
A sync-lock from Innerclock systems has solved the alignement issue,
but the individual hits/downbeats is still drifting.
Hmmm, I've had my RS7000 slaved to an Octatrack and it was SOLID. No problems with drift whatsover.

Quote:
Originally Posted by TempoTanken View Post
Also, the internal looping of the phrases can be sloppy.
Every first downbeat of a phrase has a tiny glitch..a short delay.
This means there is a difference between a 1bar loop, and the same pattern copied to itself to make a 4 or 8 bar loop (up to 256).
In the studio, I have get used to make the patterns longer than needed to avoid this problem.


Hans
I've *never* noticed a delay on the downbeats and I've owned an RS7000 for 10 years now... Is this when you have it slaved to Logic perhaps?
Share Reply Quote
Old 7th July 2012
  #54
Deleted be0c6fb
Guest
 
The RS7000 puts priority on track 1 over track 2, and so forth... so if you're using all 16 tracks at once, putting your kick drum on say track 10, using MIDI delay on tracks, using the onboard sounds with long release times, etc., you're going to definitely hear bad timing... but if you know it's limitations, you can work around these problems. I use the first four tracks for external MIDI drums, and rarely notice issues in the studio. Then again, I don't use 32nd note patterns.
Share Reply Quote
Old 8th July 2012
  #55
Deleted be0c6fb
Guest
 
*deleted*
Share Reply Quote
Old 8th July 2012
  #56
Gear Nut
 
rarara's Avatar
 
🎧 10 years
Quote:
Originally Posted by boomaudio View Post
The RS7000 puts priority on track 1 over track 2, and so forth... so if you're using all 16 tracks at once, putting your kick drum on say track 10, using MIDI delay on tracks, using the onboard sounds with long release times, etc., you're going to definitely hear bad timing... but if you know it's limitations, you can work around these problems. I use the first four tracks for external MIDI drums, and rarely notice issues in the studio. Then again, I don't use 32nd note patterns.
what do you mean by priority - never heard mention? same pattern on tracks one and two - one will be spot-on and two could be 'out'
Share Reply Quote
Old 8th July 2012 | Show parent
  #57
Deleted be0c6fb
Guest
 
Quote:
Originally Posted by rarara View Post
what do you mean by priority - never heard mention? same pattern on tracks one and two - one will be spot-on and two could be 'out'
the internal processing or CPU puts MIDI timing priority on track 1 first, then 2nd, and so on...

so your timing will be better on track 1 vs. track 16 in some cases.
Share Reply Quote
📝 Reply
Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to Reddit Reddit  Submit Thread to LinkedIn LinkedIn 
Page 2 of 2
1 2

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 71 views: 39379
Avatar for Muser
Muser 8th November 2019
replies: 25798 views: 2247409
Avatar for Rcmusic35
Rcmusic35 18 hours ago
replies: 132 views: 15140
Avatar for beenwiser
beenwiser 31st August 2020
replies: 6835 views: 581603
Avatar for olafmol
olafmol 4 hours ago
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
🖨️ Show Printable Version
✉️ Email this Page
🔍 Search thread
🎙️ View mentioned gear
Forum Jump
Forum Jump