How Accurate is the internal clock?


#1

So I have the Pyramid set to it’s internal clock at 110bpm 4/4 I then entered a 1bar 4 on the floor kick in step mode. Then I recorded the audio of that kick into Reaper also set 110bpm for 60 minute’s. The result at the end of 60 minutes was almost a complete 16th out of time.


#2

did you use MIDI sync between the two sequencers?


#3

+1 - dealing with similar stuff here and trying to find what works best for sync. I think if reaper is not clocked, you should have a proper tempo after recording and setting the DAW Tempo to 110… do you?


#4

maybe the carrot on a stick the little reaper hamster on a wheel chases got too chewed up so he gave up…sry


#5

Tried the same thing - recorded a noise synth click in 120 BPM with the pYramid as master and ableton was just recording audio without being synced. After one hour the click was off beat… Think without a stable master clock like ER-M this is common behavior.


#6

maybe get a better computer.been doing this all day on my piece of shit i3 …you are imagining things. TheFranks . the click was not off beat.
i don’t mean this an angry fuck you way.
drive your car…make a turn…use you blinker…time your blinker to the car in front…
time does not and will not bend to our rules.not even 2 stopwatches started at the exact moment in time will stay in sync.
get over it.
again there is no money you can spend to avoid this fact.
It is the nature of the universe and we can’t change it.


#7

Wow so helpful Buzz with statements like maybe hamster or carrot on stick. Look my computers have nothing to do with it. If you record 4 on the floor just using the internal clock from Pyramid the clocks drifts out by a 16th in 60 minutes time. I have tried this now on many machines and even old school digital recorders. The result is the same, Pyramids clock drifts all on it’s own. The clock in Pyramid is inaccurate and will make your beats drift. Thankfully when it is synced it adjusts for that inaccuracy. Now as for your bad english and wtf are you going on about blinkers & stop watchs … Let me be clear, the clock in Pyramid is not accurate.


#8

@Panason No I didn’t use sync this was a straight out test of Pyramids internal clock.
@TheFranks Your question “if reaper is not clocked, you should have a proper tempo after recording and setting the DAW Tempo to 110… do you?”… Reaper was not sending or receiving clock, this was simply a test for drift. So the result was a 16th of drift from Pyramids internal clock in 60 minutes. Now the good news is that whatever algorithm Squarp is using to make up for slop when it is synced/clocked is fairly good.


#9

@David all computers have time drift.
this is because they derive ‘time’ from their crystal - which can be affected by a number of factors.
depending on how firmware is setup, circuitry and the crystal used this can be better or worst.
we do not have the schematics and source code so its pure speculation on how well this is done with the pyramid, but id say its ‘in the ball park’ or what you might expect.

(one good thing, is Pyramid is probably using a RTOS , like Chibios , so at least it wont have the jitter we see in desktops)

of course, we don’t notice this on desktops/phones etc, because they all use time synchronisation services (ntp)

also often its not that important , your example is showing a tiny drift (in % terms) , if you weren’t comparing it to the grid, you wouldn’t even know (or the time your desktop computers says has elapsed)

anyway this is why we have to use syncing to keep the drift under control…

(if you think about it, this is how a real band/drummer works… they do not keep perfect time (to mS) , they just keep time within themselves, a constant feedback of micro adjustments to their timing… hell is you consider how ‘slow’ sound is, even then everything is out of time :wink: )


#10

I have 10 synths currently running off of the Squarp (7 monophonic anolog 2 FM 1 rompler) the drift is my fault in the ADSR’s in the mono synth and FM synths …I dial it in and man i like it.ya ok your gunna say FM don’t use filters (i ain’t going there.)
thetechobear was much more concise .Yet your initial inquiry was but ! reason I turned a $1500 laptop into a frizzbee.


#11

@Buzz - how can you contradict my timing check? I tried several DAWs and did the maths as well (how long is a 120BPM Bar Loop and how long are multiples of this). I think the pyramid doesn’t have 100% accurate timing when setting it to a specific BPM level. Your answer seems like you have a 100% proper timing with your pyramid - Panason and me don’t. We’re not imagining things. The clock in pyramid is not accurate. I hop getting this solved with the expert sleepers sync device. I’ll let you guys know as soon as I have changed my setup and did some checks on the timing.


#12

Im really a bit confused about this topic…

two computers without sync will drift
( bare in mind this means desktop computers need to be disconnected from networks, to stop them ‘cheating’ ;))

so having two instruments without any sync at all is, well, pointless. why would you even do this?

so i did at test, so i could given ‘practical’ rather than just the theory
I had Live connected via USB , using midi clock as sync… the timing was identical after 90+ minutes (+/- jitter)
which is exactly the behaviour i would expect (from the theory perspective)

basically with two sync devices, they should remain in time +/- jitter. thats the thing with an E-RM clock, it doesn’t do anything for drift, it just helps jitter…
i.e. if you have +/- 2ms Jitter on (usb) midi clock then after 60 mins, your beat should still be +/- 2ms, the E-RM will just make it +/- 1 sample.

as to how accurate a individual clock is, that pretty irrelevant, surely, in a musical context, time/timing is relative.
no one sits listening to music and says… oh my god thats 120.1 bpm, and its varying by 100ms over a 30 mins… what they notice is that things become out of sync with each other

ok, they might notice it switch from 120 to 130 bpm, but no where here are we talking about such inaccuracies… 1/16 note at 120 bpm is 125mS … over an hour, that is a tiny shift :wink:

seems to me too much obsession with ‘the grid’ :wink:


#13

Hmm, I can kind of see where it could be a problem if the requirement was to fit a piece of music exactly to a specified length of time, but even then such a small discrepancy isn’t likely to be a problem in most cases.

I use a Roland SBX-1 as my master clock and all my gear including the Pyramid stays locked for days (I tested it)
As @thetechnobear said you are not going to notice such a small drift in isolation and if synced then everything will still be in time anyway.

If you were talking about jitter (erratic timing) I’d agree that it should be fixed, but that is a different subject entirely, as far as I can tell the Pyramid does not suffer any significant jitter.


#14

I haven’t synced the daw - just Recorded Audio and wanted to see whether 120 BPM is 120 BPM on a different machine. When jamming for an hour or two which could also take place at an event live is critical to have stable sync over time. That is the main reason why some people try improving sync between machines to have less drift. I had setups before which remained in sync for days and want to get as far back to that situation as possible.


#15

How do you know the Pyramid is drifting rather than your computer or Reaper? What other MIDI gear have you tested with this set-up?

The reason they invented sync is that timing drifts. I have never owned any two pieces of MIDI gear that would stay in time without one piece acting as the master clock. They need to be synced, that’s the reality. Sometimes it’s frustrating.

I sync to Logic because I know I’ll want to drop the audio into Logic for vocal overdubs and I might want the tracks to fit the grid at that point.

You can always buy a stand alone MIDI clock. They make them because, well, you have to sync to something. Expecting your computer and the Pyramid, or any other piece of MIDI gear, to be in perfect time after an hour of unsynced play is unrealistic.


#16

Crap I don’t even know where to start with the the obtuse “slow to understand” reply’s here. Go try it for yourself folks, the clock is not accurate. If you dont understand why one might need an acurate clock then this post is not for you. I have now tried this test on many recording devices… including an Atari St, Zoom recorders, Fostex etc. The result is the same the internal clock is not accurate. We are talking about jitter (erratic timing).


#17

If you’re going to throw around wild accusations, you’re going to need to be more scientific in your responses,
Exactly how are you measuring the audio file ? what is the precise length in hours:minutes:seconds. Can you demonstrate with images ?
How are you measuring using an atari st ? how are you doing it with a Zoom ? which Zoom, which Fostex ?
What is your precise methodology, How are you gathering the data (you say you record an audio file - of what ? the pyramid does not generate audio… could the device you’re triggering not be introducing error ?)
Jitter is NOT what we’re talking about if you say you are getting a consistent 16th note variation every single time surely ? jitter by definition is erratic ?

There’s no ‘slow to understand’ from the response from what I am reading…just some valid points being made as to why this should even matter if your focus is making music. Dismissing the responses as “If you dont understand why one might need an acurate clock then this post is not for you” is not addressing the point mbeing made that PC’s themselves have jitter. They do not prioritise timing above other processes.


#18

whoa… real community attitude… why post to have a discussion, if you don’t want one?

perhaps you can explain to those of us “too slow” to understand:

a) what do you think jitter has to do with drift?
(bare in mind both terms have proper definitions with respect to timing)

b) why would you not sync two (or more) devices if you require them to share a common time base? what is your purpose? what are you trying to achieve? what is your musical goal?

also fyi, re-read my reply, i did not say that the Pyramid does not drift. i never doubted your ‘experiment’ … in fact, i pretty much said “of course this happens”.
the bit I’m “slow to understand” is, if your using the available tools correctly, whats the musical relevance ?


#19

I think what the thread started with is just the assumption and some tries to proof that a fix BPM value on pyramid is not accurately translating to timelines in daws or other audio programs. It’s not about sync between devices ( there are different problems and solutions addresses in other threads), just the fact that 120 BPM should be 120 beats per minute and some of us measured that pyramid doesn’t have the some on the output side as is shown on the display. I have the feeling that this is a community driven device and also forum here so let’s try being polite and helpful here.


#20

You may be right about the initial idea behind this thread, I wouldn’t know, I don’t read minds. I would like to point out that my reply was, and all the replies in this thread so far are, polite and very much trying to be helpful.

The apparently unpopular truth is MIDI needs to be synced. Any two MIDI devices left to run by themselves will lose sync. The OP did not clarify the problem- is it jitter or drift?- did not explain his/her testing procedure, and didn’t logically eliminate the computer or Reaper from the equation. There is nothing wrong with the MIDI timing on the Pyramid, that’s not a defensive remark, just honest.

So to answer the question, how accurate is the Pyramid MIDI clock? More accurate than most MIDI gear I’ve used, no jitter, syncs perfectly. Less accurate than a purpose built sample accurate MIDI clock. If you want the Pyramid to line up on your DAW grid, sync it like everybody else on the planet does.