Dave Smith Instruments SYNTHESIZERS

hi,
this is kind of off topic since its not an evolver specific issue but
i've got a clock sync problem thats been bothering me for years.
im bringing it up here today since i was just trying to sync the
evolver to cubase SL and thought perhaps some of you gurus
may know the answer.

here's the problem...
when i've got my evolver synced to my sequencer's clock, the
bpm display on the evolver jumps around some. for example, 
if i've got cubase SL transmitting at 120, the evolver will jump
around between 119, 120, and 121.

also, when i start cubase's sequencer, the evolver will first go
to 101 and then jump up into the 119-120 range.

this problem evidences its self with all my other clock syncable
gear too. the degree of instability in the clock sync varries
depedning on the equiptment in question...with my korg kaoss 2
being the most stable to my fizmo's sync being the most unstable.

for an external sequencer sync, the problem is usually not too
noticeable since it usually jumps around in the right range closely
enough and the jumps are very fast. but for synced delays, its a
nightmare because (depending on the sync implimentation) i either
get clicks or massive pitch glitches.

i've had this problem with other sequencers too so i know
its not just a cubase problem. i've also experianced this with 
3 different midi interfaces and on two OS's (XP and 98).

here are my current specs:
win XP
cubase SL
motu midi express XT (parallel)
echo layla 20 (im using the midi on this too)

anyone have any idea what the hell is causing this and how to
fix it?

thanks,
-ugo

on 3/27/04 1:47 PM, Ugo wrote:

> when i've got my evolver synced to my sequencer's clock, the
> bpm display on the evolver jumps around some. for example,
> if i've got cubase SL transmitting at 120, the evolver will jump
> around between 119, 120, and 121.

this can be caused by (at least) two factors:

1. the clock being sent is not stable, and has some
   degree of jitter...and based on other people¹s
   comments re: Cubase, this is likely in your case.

2. the receiving device does not have a good algo for
   interpolating the incoming clock, or a method of
   hysteresis...I cannot comment on the Evolver¹s code.

> also, when i start cubase's sequencer, the evolver will first go
> to 101 and then jump up into the 119-120 range.

this is common because clock is derived as the timing
between pulses...there is no clock info in a single
pulse, so the Evolver is simply deriving the clock as
best it can from your source...the solution is to make
use of a blank measure or the like so clock can be
resolved by your slave devices before they need to
produce any audio.

> this problem evidences its self with all my other clock syncable
> gear too. the degree of instability in the clock sync varries
> depedning on the equiptment in question...with my korg kaoss 2
> being the most stable to my fizmo's sync being the most unstable.

yes, it is clear to me that different pieces of gear
produce different levels of quality as it regards timing
for MIDI clock...my most stable being my Elektron
MachineDrum and Clavia Nord Modular (never used the
Evolver as a clock source).

> for an external sequencer sync, the problem is usually not too
> noticeable since it usually jumps around in the right range closely
> enough and the jumps are very fast. but for synced delays, its a
> nightmare because (depending on the sync implimentation) i either
> get clicks or massive pitch glitches.

you will get one or the other...if you get clicks, it
means the delay algo is not interpolated...if you get
pitch warbles, it means it is interpolated.

> i've had this problem with other sequencers too so i know
> its not just a cubase problem. i've also experianced this with
> 3 different midi interfaces and on two OS's (XP and 98).

you may wish to use some other device than the computer
as your clock master.

you may also wish to consider not syncing delays to MIDI
clock, but setting the delay value by hand...in general,
this will work most all of the time (if not as
convenient).


cheers,
aeon

> 1. the clock being sent is not stable, and has some
>   degree of jitter...and based on other people\ufffds
>   comments re: Cubase, this is likely in your case.

can you suggest a pc program (with a demo) that can be
expected to have a more stable clock? not that i plan to
switch hosts for this but i'd be curious to test this issue
with something thats known to be stable.

> this is common because clock is derived as the timing
> between pulses...

oh, ok. that certainly makes sense.

> you may also wish to consider not syncing delays to MIDI
> clock, but setting the delay value by hand...

thats what i've been doing so far. syncing them from the
computer has proven to be completely useless. too bad
really, it would be handy if this worked.

-ugo

on 3/27/04 9:17 PM, Ugo wrote:

>> 1. the clock being sent is not stable, and has some degree of jitter...and
>> based on other people¹s comments re: Cubase, this is likely in your case.
>> 
> can you suggest a pc program (with a demo) that can be expected to have a more
> stable clock? not that i plan to switch hosts for this but i'd be curious to
> test this issue with something thats known to be stable.

sorry, no...I am a Mac user.

>> you may also wish to consider not syncing delays to MIDI clock, but setting
>> the delay value by hand...
>> 
> thats what i've been doing so far. syncing them from the computer has proven
> to be completely useless. too bad really, it would be handy if this worked.

in general, I find that the combo of source jitter and lack of hysteresis in
many delays means they glitch like mad...my Line6 Echo Pro sure does!


cheers,
aeon

> in general, I find that the combo of source jitter and lack of
> hysteresis in many delays means they glitch like mad...my
> Line6 Echo Pro sure does!

im glad to know that its not just me that runs into this problem!
with all the midi syncable effects being produced these days
i was really wondering if it was just me that had troubles
getting/keeping a good sync.

-ugo

>> in general, I find that the combo of source jitter and lack of
>> hysteresis in many delays means they glitch like mad...my
>> Line6 Echo Pro sure does!
>
>im glad to know that its not just me that runs into this problem!
>with all the midi syncable effects being produced these days
>i was really wondering if it was just me that had troubles
>getting/keeping a good sync.
>
>-ugo

In some cases high latency used with audio drivers seems to mess with the
clock-output, and the midi timing.  Setting the latency lower may help.  It certainly
seems to help in logic OSX.

z.