Emu XL-7 & MP-7 User's Group

I took a simple 16-note pattern and recorded it into Digital Performer
as a waveform, so that I could compare and measure the timing
accuracy. There's routinely a slop of 10 msec or so in many of the
hits, sometimes as high as 15 msec.

This is the XL-7 on internal clock. There are other tracks in the
pattern, but with volume set to zero. If the other tracks are muted,
the timing improves. The slop does seem to occur when several tracks
are triggering a note at that point, but it's at least partly
random. For the measurements above the tracks were generating MIDI;
with the tracks set to "int" the timing still sounds bad, although
I've not measured it. The click track is also high-numbered (#13); I
seem to recall that lower-numbered tracks have priority, so I should
try that at some stage.

-- 

  nick rothwell -- composition, systems, performance -- http://www.cassiel.com

--- In xl7@y..., Nick Rothwell <nick@c...> wrote:
> There's routinely a slop of 10 msec or so in many of the
> hits, sometimes as high as 15 msec.
> 
> This is the XL-7 on internal clock. 

Thanks for the info Nick.

Clearly, less slop the 'tighter', does anyone know the timings on a
MPC unit?  Or even how it compares to other hardware sequencers or
grooveboxes?

Also, are people finding the 10-15ms delay noticable to their ears? 
I appreciate this is subjective...

Again, Nick thanks for taking the time,

-b

Maybe they could market it as having a secretly integrated "live 
groove" feel...um, yeah.
ByronE

--- In xl7@y..., "balcra" <balcra@n...> wrote:

> --- In xl7@y..., Nick Rothwell <nick@c...> wrote:
> > There's routinely a slop of 10 msec or so in many of the
> > hits, sometimes as high as 15 msec.
> > 
> > This is the XL-7 on internal clock. 
> 
> Thanks for the info Nick.
> 
> Clearly, less slop the 'tighter', does anyone know the timings on a
> MPC unit?  Or even how it compares to other hardware sequencers or
> grooveboxes?
> 
> Also, are people finding the 10-15ms delay noticable to their ears? 
> I appreciate this is subjective...
> 
> Again, Nick thanks for taking the time,
> 
> -b

> Clearly, less slop the 'tighter', does anyone know the timings on a
> MPC unit?  Or even how it compares to other hardware sequencers or
> grooveboxes?

I've got an R-8 II drum box, which I'll time at some stage. I've not
used it for a while, but as I recall it has some audible slop (I'd
guess around the 5-7 msec mark).

> Also, are people finding the 10-15ms delay noticable to their ears? 
> I appreciate this is subjective...

I suppose I could export and upload a chunk of the recording as an
MP3, with an indication of where I can hear it - it would be
interesting to know how subjective this is.

-- 

  nick rothwell -- composition, systems, performance -- http://www.cassiel.com

I'd like to hear that too.

rEalm




I suppose I could export and upload a chunk of the recording as an
MP3, with an indication of where I can hear it - it would be
interesting to know how subjective this is.

-- 

  nick rothwell -- composition, systems, performance -- 
http://www.cassiel.com


To unsubscribe from this group, send an email to:
xl7-unsubscribe@yahoogroups.com

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

Hey Nick, interesting results.  If you don't mind, a couple questions:

1. Do you get the same slop with Logic providing the master MIDI clock?
2. How many tracks or notes WERE sounding when you recorded the results? 
Given that MIDI is a serial transfer protocol, it's impossible for two 
notes to trigger at exactly the same time.  Just the delay introduced by 
this fact can be around 2-3ms PER NOTE if I remember correctly.  Thus, if 
you have a lot of info being sent at the same time, I can see that 
happening.

rEalm 






"Nick Rothwell" <nick@...>
06/23/02 06:41 AM
Please respond to xl7

 
        To:     xl7@yahoogroups.com
        cc: 

        Subject:        [xl7] Sequencer Timing


I took a simple 16-note pattern and recorded it into Digital Performer
as a waveform, so that I could compare and measure the timing
accuracy. There's routinely a slop of 10 msec or so in many of the
hits, sometimes as high as 15 msec.

This is the XL-7 on internal clock. There are other tracks in the
pattern, but with volume set to zero. If the other tracks are muted,
the timing improves. The slop does seem to occur when several tracks
are triggering a note at that point, but it's at least partly
random. For the measurements above the tracks were generating MIDI;
with the tracks set to "int" the timing still sounds bad, although
I've not measured it. The click track is also high-numbered (#13); I
seem to recall that lower-numbered tracks have priority, so I should
try that at some stage.

-- 

  nick rothwell -- composition, systems, performance -- 
http://www.cassiel.com


To unsubscribe from this group, send an email to:
xl7-unsubscribe@yahoogroups.com

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

> Given that MIDI is a serial transfer protocol, it's impossible for two 
> notes to trigger at exactly the same time.  Just the delay introduced by 
> this fact can be around 2-3ms PER NOTE if I remember correctly.

I know, which is why I want to do the timings again with MIDI
transmission off. MIDI is 31.25kbaud, which means 3 msec per note in
the worst case, 2 msec if there's running status (which generally
won't be the case for multiple XL-7 pattern tracks since they're on
different MIDI channels).

-- 

  nick rothwell -- composition, systems, performance -- http://www.cassiel.com