SPECTRALIS TIMING /CLOCK TEST

Totally addicted to your speckie? Not getting the results you expected? Experiencing Problems together with other gear? This is the place to share good and bad experiences.

Moderator: Jörg Schaaf

SPECTRALIS TIMING /CLOCK TEST

Postby via » Thu Apr 18, 2013 8:59 pm

Just preformed a MIDI clock performance test with the spectralis, similar test can be found at http://www.innerclocksystems.com/New%20 ... itmus.html

the spectralis score 4 STARS 8)
6007 / 5976/ 6008 / 6007 / 6007 / 5976 / 6007 / 5976 / 6007/ 6008 / 6007 / 5976 / 6007 / 6008 / 6007

Maximum variation between any two consecutive Sixteenth Note intervals:
32 Samples [0.67ms]

when synced to a MPC4000
6007 / 6008 / 5976 /6007 / 6007 / 5986 / 6008 6007/ 6007/ 5986 / 6007 / 6008 / 6007 / 5976 / 6007 / 5976

the same 32 samples but its just a tad tighter at 5986 at some places but we are talking 0.0x ms ...


MPC4000
6000 /5998 / 5998 / 5998 /5998 / 6000/ 5997 / 6000
my test ^
6000/ 5996/ 6001/ 6000/ 5999 / 5996 / 6003 / 5999
^interlock

got the same :Maximum variation between any two consecutive Sixteenth Note intervals:
1 Samples [0.02ms] so its accurate, btw this is no rocket science (just accurate measurements of samples between two points) and the test was on a pattern that had bunch of stuff running on mute 8)



only hardcore modular sequencers and the Cirkon score perfect zero sample accuracy score, but on the whole vs other sequencing workstations it ranks up there with the best of them..

for reference here is the timing of two classic
TR 909
214 Samples [4.46ms]
TR 808
99 Samples [2.06ms]



Hardware workstation top 6 + others


MPC4000
1 Samples [0.02ms]

Octatrack
2 Samples [0.04ms]

MC909
4 Samples [0.08ms]

MPC 3000
5 Samples [0.10ms]

DSI Tempest
16 Samples [0.33ms]

MPC60
18 Samples [0.37ms]

SPECTRALIS
32 Samples [0.67ms]

Analog4
37 Samples [0.77ms]

MPC2500
70 Samples [1.46ms]

Elektron Machinedrum
104 Samples [2.17ms]

Roland MW8800
102 Samples [2.13ms]

Akai MPC 5000
138 Samples [2.87ms]
Last edited by via on Thu Apr 18, 2013 11:07 pm, edited 1 time in total.
via
 
Posts: 711
Joined: Sat Aug 19, 2006 3:19 am

Re: SPECTRALIS TIMING /CLOCK TEST

Postby jizzmaster2000 » Thu Apr 18, 2013 10:39 pm

Nice work, thanx :P
Jeremy
Melbourne
jizzmaster2000
 
Posts: 258
Joined: Fri Jan 13, 2012 9:59 pm
Location: Melbourne

Re: SPECTRALIS TIMING /CLOCK TEST

Postby via » Thu Apr 18, 2013 11:26 pm

thanks, yet I had to do a little correction, my count was off, correct is spectralis: 32 samples, but its has this push to its clock,ll like its drifts off for 3 beats and then resets itself on the 1st again or something like that .. but I´would have to sample a longer loop to confirm that pattern.. but it is swinging in there a little on its own.

I knew it sounded a little to good to be so close to the MPC4k, as I can defiantly feel the difference between the MPC4000 and the spectralis, the 4k being ultra tight vs the spectralis more relaxed...

still tighter than many mpc´s just not as tight as the 4k 3k & 60..
via
 
Posts: 711
Joined: Sat Aug 19, 2006 3:19 am

Re: SPECTRALIS TIMING /CLOCK TEST

Postby via » Fri Apr 19, 2013 12:53 am

the 4k being ultra tight vs the spectralis more relaxed...

also regarding that ^ zero sample accuracy is 0-60000 samples
and as you can see the MPC4k ruses the 16th by a few samples but is other wise spot on 6000
6000 /5998 / 5998 / 5998 /5998 6000/

Spectralis is more laid back; most of the 16th are 7-8 samples behind sample accuracy yet, with a twist;
6007 / 5976/ 6008 / 6007 / 6007

and even though we are talking samples here; I can hear ´´mix with´´ feel the difference clearly,.. and I think that´s just when you put your mind to it, no-one in the studio could tell the difference on a normal day, or care less if it was coming from the spk or the mpc to put that into perspective.. but being conscious about this = you might make choices based on it, specially if you have many sequencers..


Btw you can do this test on any sequencer, just record 16bit 48k, (record a click sound or a short perc/pulse with a spike), and measure from the peak to peak, any wave editor can display samples, I used mpc4000 trim mode, the margin of error is in the LCD contras.. (I did get 3ms on the 4k but it was one 50/50 call on the LCD)..but anyone can do this so feel free to add to the list
Kudos,
via
 
Posts: 711
Joined: Sat Aug 19, 2006 3:19 am

Re: SPECTRALIS TIMING /CLOCK TEST

Postby via » Fri Apr 19, 2013 2:15 pm

OK did some more test on the clocks,

First I played both sequencer for 10 minutes both triggering a BD then OH decaying sound and listen to their phase and its the same thing the numbers predict; they stay in phase a few beats and then a new phase might drop in at random intervals but throughout 10 minuets both sequencers running Internal Clocks individually ran in perfect sync!! :shock: … that was a real surprise, never seen that before, usually two independent sequencers are drifting at that point.

a worthy mention here is that; I can cue spot on tempo when I press play on the spectralis ´´instant´´,, vs mpc never! (pressing play is always delayed…

Ok so now we get onto to the feel of the clocks the MPC4k defiantly has a steady push/ stable feel, one could argue its to mechanical but on the flip side the beauty of it is this purity.

The spectralis keeps clock very stable between some 16th: (1 sample) then some 16th´s come in at 22-24 or 32 samples.. not as pure but on the flip side its more romantic..

Funny thing I had a drummer drop by tonight and made a A/B B/A blind test, it was a very difficult test, we are talking just above human sensitivity:f 8-16 ms, so its very hard to spot this but with 30 years of experience he was quite accurate in his response; I´ll quote him

MPC4000 like an alarm clock something very correct about it
Spectralis: warmer or more spooky somehow ..


Now Midi clock is not just only about start time, this affect decay of sound aswell (in this case up to 32samples omg) but we are talking per 16th notes so even though we are in micro timing land this movement is frequent enough to affects the entire perspective of time, thus feel of the unit.. so I´d also say the sectralis clock is …kind-of spooky yes emident on say 4/4 kick …but on 8th note hihats with synced delays…oohhh man.. the beauty of this just opens up to full affect, mpc just sounded to prefect even the swing felt a little to perfect in that test. the spectralis just sounded more organic and alive..

not so spooky after all, vintage is not the right word eater; house´y seams like the most fitting to be honest; house= sloppy sample chops TR-809´s Emu Midi,. where house originated fro and an era where that bit of mechanical imperfection shined, not that the spectralis isn´t way tighter than those units; but in that fluctuating window of 32 samples per 16th, its this not so organic clunky´ness that just gives the spectralis a touch of that vibe.

if you wanna hear this blatantly trigger full 16th note HH pattern and you can really hear the feel of the internal clock..


I would like to point out that the sounds triggering directly from the spectra's vs from the mpc where stronger think this is psychoacoustics affect that happens: impact to audio transient, its quicker internally, thus the sound has tad bit more power.

Not very scientific I know,and may over the top details but you can actually hear/feel this and this is at the core so it matters to some of us.

OK thats about as deep as I´ll go on this, yep..done with midi clocks for now..
via
 
Posts: 711
Joined: Sat Aug 19, 2006 3:19 am

Re: SPECTRALIS TIMING /CLOCK TEST

Postby Jörg Schaaf » Thu Jun 20, 2013 12:35 pm

Hmmm - do you muted the other parts of do you created an empty pattern on all instruments and created just one part with that 16th notes? I am sure, the results would be different.

I am not sure, whether such a kind of test really makes so much sense. Most timing issues are becoming audible under stress conditions. Instead of running one single instrument (what mostly never occurs in a real song) please repeat that test on different instruments with complex patterns. In those patterns select one part, create these 16th notes with a sharp transient sound and feed that part to an independent output. You still measure one instrument and the 16th notes but a lot of instruments are running in the background.

Than do the measurement again.

Create at least two of such patterns and jump from one pattern to the other while running the timing test. Count the samples again.

Then you get something much more interesting....
New users please check the following link for forum activation:
http://www.forum.radikaltechnologies.com/viewtopic.php?f=8&t=921
http://www.myspace.com/jorgschaaf
http://www.youtube.com/user/SCHAAFJOERG
User avatar
Jörg Schaaf
 
Posts: 1096
Joined: Sat Aug 20, 2005 1:46 am
Location: Germany

Re: SPECTRALIS TIMING /CLOCK TEST

Postby via » Sat Jul 06, 2013 12:58 pm

Hey Jörg, thanks for the feedback :wink:

it was a busy pattern, I just soloed the sound I was recording,.. I love the feel of the timing; is not as perfect as the MPC4000 but it has this romance that is just gorgeous, very evident on delays ..

if you know something I dont know please share the wisdom, I can´t do the test again as is ... I´m sort-of in-between places & the studio is boxed..yet again .. :|
via
 
Posts: 711
Joined: Sat Aug 19, 2006 3:19 am

Re: SPECTRALIS TIMING /CLOCK TEST

Postby flowdesigner » Sat Aug 10, 2013 9:26 pm

in bwtween montreal and reykjavik? :)
flowdesigner
 
Posts: 31
Joined: Mon Dec 17, 2012 8:59 pm

Re: SPECTRALIS TIMING /CLOCK TEST

Postby via » Fri Aug 16, 2013 12:40 am

Yep I´m out here on a Kanu in the middle of the Atlantic ocean ..

surprisingly good reception out here ..
via
 
Posts: 711
Joined: Sat Aug 19, 2006 3:19 am


Return to Spectralis experiences and opinions

Who is online

Users browsing this forum: No registered users and 1 guest

cron