HD24 pitch, sample rate

  • Thread starter Thread starter foreverain4
  • Start date Start date
F

foreverain4

New member
ok, i have been having this problem. i am creating a click via computer (vegas) and sending it and recording it to a channel on the HD24. when i transfer the tracks via fireport, the tracks are actually slower. an example. click created at 92 bpm. import tracks in to sony vegas, have to slow the project tempo down to 91.962 . this is pretty consistant. i have also used my tama rhythm watch to output to HD24 and computer at the same time. HD24 tracks are again slightly slower. anyone else run in to this issue? i am positive it is the HD24.
 
Never. I have everything clocked externally with the GenX6 so it's all on the same clock in and out.
 
why would the hd24 be slow? does the clock suck? i just exported some tracks from reason on my friends mac. pulled them up into vegas everything matched tempo perfectly. played them out of the mac, recorded to hd24, transfered via fireport to vegas, they now are miss aligned... hhmm....
 
Blue Bear Sound said:
When you use the Fireport s/w, are you mis-setting the sample rate?
Click to expand...


no, if i would, it would be off by more that 92 bpm to 91.962, would it not? could you guys try it with yours? take off your master clock, and just use the hd24 internal.....
 
Actually - it sounds more like the software is picking up an oddball rate... there's really no way for the HD24 to not have the right rate unless you had pitch control on. That's the only way the HD24 could not be at either 44.1 or 48... and the internal clock is no Lucid, but it does its job perfectly well.

S/W, OTOH, could be getting its timing from a number of different places depending on your config....

If you like, you can give me a call a little later on and I can try and help you through some options... this kind of thing is difficult to diagnose on threads...

Also - the Fireport doesn't sync with anything - it's strictly a data transfer device -- it will dump the tracks at whatever SR they were recorded at (unless you manually reset the rate, which isn't advised!) So again, the only "variable" aspect is the clocking on your DAW -- everything else is pretty much set at a fixed SR!
 
Last edited:
this is very strange indeed. i understand what you are suggesting about my SW clock. i originally thought it was my software, or computer, not the hd24. but, i had some tracks exported from Reason that was running on a totaly different computer, a mac at that (i am running pc), and those tracks imported perfectly at the project tempo of 92 bpm. all i know, is that somehow, tracks recorded on my hd24 are ending up .038 bmp slower than the rest of the world. :(
 
from the HD24 yahoo group. posted by another user


"i've experienced this as well, but i just thought is was my crappy
Alesis Drum machine drifting.....ive noticed that when i generate a
clik in the computer- its more accurate- but never dead on."
 
Waituminut - are you compared a midi-generated click (ie, midi track) to an audio track? If so, then there definitely could be drift at the DAW end. Midi timing and audio WAV timing come from 2 different sources (which is why handling midi and audio together in one s/w package is so complicated!)

Audio derives its timing from the clock source (or internally, or from the digital signal itself), but Midi's timing comes strictly from the Midi Driver - which is completely separate and unrelated to the audio driver, and it's not unusual to have some drift between the two......
 
ok, i am still having problems. i have since switched to protools. i have an mbox 2. i have some prerecorded audio (pre production) recorded directly to the mbox 2. i am syncing the mbox to the hd24. here is the setup.


HD24 midi IO connected to mbox 2 midi IO.
HD24 is master, controlling transport of PT
HD24 lightpipe out to tascam DM24 lightpipe in
HD24 clock internal
DM24 clock external lightpipe.
DM24 spdif out connected to mbox 2 spdif in
mbox 2 clock set to spdif

i record 12 drum tracks to the hd24 while PT is syncd. works perfectly. tracks playback on the hd24 in sync. BUT, when i transfer the files and import them in to PT, they drift... eeeerrrr!!! should i be clocking the HD24 to the mbox instead? it seems like the HD24 clock is just short of 44.1
 
foreverain4 said:
it seems like the HD24 clock is just short of 44.1
Click to expand...
Actually, the HD24's clock (when operating at 44.1) runs very slightly faster than 44.1. AT 48K, it's more or less perfect, but at 44.1 it is very slightly fast. It isn't off enough to cause any pitch issues, and it is still clean and jitter-free, but it is about 0.04% fast. The only impact of this slight inaccuracy is that long-running tracks will drift slightly out of sync if some (but not all) of them are recorded with the HD24's clock and then played back with a different one (or vice versa).

For example, I regularly record events with my HD24, then transfer the tracks to my PC for mixdown. As a result, this issue does not impact me - even though there is a slight difference in the clocks, the same difference is applied equally to all tracks. However, if I recorded some of the tracks on the HD24 and others directly into my PC, I would find sync problems over time.

We had a long discussion of this issue on the Yahoo HD24 discussion group a few months ago.

Here's what we learned from Alesis:

[From the HD24 hardware designer]

"The HD24 sample clock is synchronously divided down from 54MHz
crystal oscillator that is ±50ppm accurate. There is no known
clock accuracy problem with the hardware or firmware. The effect
being reported on the group is simply due to limitations of integer
math. At 48kHz, the 54MHz crystal is divided by 1125, and the 48kHz
clock rate is as accurate and precise as the 54MHz crystal clock.
At 44.1kHz, the 54MHz crystal is divided by 1224, which is not
exactly 44.1kHz (44117.6Hz), but maintains the precision and jitter
performance of the 54MHz crystal clock. This is equivalent to a
frequency error of 400ppm, which is sufficiently accurate for any
application *except* trying to synchronize free-running devices over
a long duration. The Alesis recommendation if this amount of
frequency error (and resulting time drift) is not acceptable is to
use an external clock (particularly, a common clock for all devices
that need to be synchronized), or to sample at 48kHz, which will
reduce the frequency error by approximately one order of magnitude."

The bottom line is this: If you are require multiple digital devices to be in perfect sync, you should make sure that all are using a common clock.
 
i ended up switching the clocks around. i went mbox 2 spdif in to DM24. DM24 lightpipe out to HD24. this has been solid for me. everything lines up sample accurate...this is good info to know though..
 
That's great info Gilliland. That explains why no problems here. I clock with the genX6.
 
You must log in or register to reply here.
Back
Top