13 Jul
2003
13 Jul
'03
5:04 p.m.
| Message: 1
| Date: Sat, 12 Jul 2003 10:20:15 -0700
| From: toby(a)tobiah.org
| To: linux-audio-user(a)music.columbia.edu
| Subject: Re: [linux-audio-user] Multiple (delta xx) cards
| Reply-To: linux-audio-user(a)music.columbia.edu
|
|
| > Ideally, I'd like to be able to buy a delta 44 now (for financial
| > reasons), and then buy another,
|
| You can't sync two 44's, they will drift as you record.
| Also, the 1010 does the analog to digital conversion in
| the breakout box, while (contrary to what some believe)
| the 44 does not.
|
I don't understand this whole 'drift' thing. If I'm able to listen to a
previously recorded drum track and record a guitar track over it (in
sync with the recording) with either card, how is it possible for them
to be out of sync with each other?
| As for the Drum kit, I know it is desireable to
| record many channels of drum separately, but would
| it do to simply use say a Mackie 1402-VLZPro mixer
| to funnel the drum mics into two or three channels?
| You would have to get the mix right the first time,
| I know, but the 1402 would give you six very nice
| channels of mic pre, which should ease the budget
| problem. You also get phantom power, which you must
| not forget that you need to power any decent condensor
| mic.
I'll consider this, but my original plan was to spend my money on
getting the best soundcard, preamps, and mics and doing as much as
possible in software. The idea was to get the highest quality crucial
hardware i could afford, and do any and all mixing, processing, etc, in
software (which is putting a lot of faith in linux ;) ). I know it's a
fine line, though.
|
| If you get in a pinch, you can pick up a pair of
| Marshall MXL2003's for a very reasonable price, and
| in my limited experience yield a natural bright sound
| that would be very tough to match for the same price.
| Anyway, go for the large diaphragm condensors.
|
| If it is to be a $20 mic, I recommend the radio shack
| 33-3017. I have a pair, and they don't compare well to
| the Marshall's, but they are also hard to match for the price.
|
| http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2544797054&categ…
|
I was exaggerating about the $20. I'll end up with at least an sm57 or
sm58 (or comparable) for starters.
13 Jul
13 Jul
6 p.m.
New subject: [linux-audio-user] Multiple (delta xx) cards
I don't understand this whole 'drift' thing. If I'm able to listen to a
previously recorded drum track and record a guitar track over it (in
sync with the recording) with either card, how is it possible for them
to be out of sync with each other?
Peter,
It's easy for people to not see this problem the first time they look at
it, but it's usually not that difficult to explain, given a bit of basic
engineering background.
All of these sound cards have oscillators on them to allow them to sample
and playback at a given rate, such as 44100 Hz. the problem is that no two
oscillators (or the crystals that drive them) produce EXACTLY 44100 cyles
per second. They are spec'ed to be within a certain tolerance, like 1%, or
+/- 100 parts per million, but the point is that they are not creating
exactly the 'right' number of cycles per second.
With this in mind, if you have two sound cards (same models, different
models, it doesn't matter) and they run from separate clock sources, then
one of them might be operating at 44099 cyles per second, while the other is
operating at 44101 cycles per second. If this was the case, then (in
differeing by 2 cyles per second) after 22050 seconds they would be out of
sync by one complete second. (The slow one would be one second behind.)
Since the human ear can pick up very minute differences in time, the effect
becomes noticible much more quickly than that though, and unfortunately,
most crystals will be much further off than 1 cycle per second from the
spec.
The reason you can play the same track back, over and over, and even on
different sound cards, is that while 44100 samples 'represents' on second,
it may not 'take' on second to play back. You ear is pretty insensitive to
the idea that a song played on your CD player in the house takes 3:49.203
seconds, while in the car it might take 3:49:542 seonds. We are just not
sensitive enough to hear that, although the pitch in the house would be
measureable as slightly higher.
All this said, it is NOT true that you cannot use two (or more) sound
cards at the same time. You can, but they have to be clocked from the same
clock. you can accomplish this on any pair of sound cards that support the
Word Clock interface, and I've been able to do it over ADAT interfaces with
multiple cards, but that's not for the uninitiated. ;-) It's harder to set
up.
I hope this helps clarify the physical reasons that this is difficult.
There is a software reason also. Linux/Alsa will onlt interrface to a
single clock, so clocks being generated by two sound cards are not something
that Alsa is designed to handle.
Cheers,
Mark
15 Jul
15 Jul
1:45 p.m.
New subject: [linux-audio-user] Multiple (delta xx) cards
On Sun, Jul 13, 2003 at 12:46:53PM -0700, Mark Knecht wrote:
...
I don't understand this whole 'drift' thing. If I'm able to listen to a
previously recorded drum track and record a guitar track over it (in
sync with the recording) with either card, how is it possible for them
to be out of sync with each other?
Peter, I hope this helps clarify the physical reasons that
this is difficult.
There is a software reason also. Linux/Alsa will onlt interrface to a
single clock, so clocks being generated by two sound cards are not something
that Alsa is designed to handle.
:-/
->
!
Would not a software/driver-solution to synchronize the
clocks/adjust the drift be handy here? At least theoretically,
somthing similar to ntp.
This would need
1: a way to measure the clock-difference exact enough
2: a way to delay the the stream running over the hardware
within the driver within very small units.
3: ... ?
But I doubt that this idea could get far, probably timing would
become a rather statistical than exact procedure in this region.
And expensive too. Also, maybe clocks are not even stable over
time (temperature, etc.) and would have to be remeasured
periodically. Also, I bet this idea has been around earlier and
if possible we would already have it realised.
The psycho-acoustic boundary for discerning discrete acoustic
events/clicks is about 1/20 s, AFAIR, so 50 ms would have to be
beaten to satisfy the ear.
Björn
2:46 p.m.
New subject: [linux-audio-user] Multiple (delta xx) cards
Would not a software/driver-solution to synchronize the
clocks/adjust the drift be handy here? At least theoretically,
somthing similar to ntp.
I don't think it's possible. My thoughts:
1) To the best of my knowledge sound cards tell you when the frame occurs.
Not only are the clocks different frequencies (by a small amount) but they
run completely unsynchronized, so the start of frame on one sound card could
be anywhere within the frame on the other, and I've never heard that this
can be controlled at all.
2) Most important, every time this conversation has come up on Alsa-Dev, my
impression is that the developers have less than zero interest in tackling
the problem since the Word Clock/ADAT clock/spdif clock solutions do exist.
Technically I think there may be solutions, but practically speaking the
problem get immense. Should you ever advertise that it works, you know you
are going to have some young guy with a weird pair of cards, like a SB-16
and a USB device, asking why it doesn't work for him. No one else will have
the same hardware, so no one will be able to debug why.
Heck, we can't even get Alsa drivers to work reliably for single cards -
why go after this issue? ;-) ;-)
Mark
7793
days inactive
7795
days old
linux-audio-user@lists.linuxaudio.org
3 comments
3 participants
participants (3)
-
Björn Pfeiffer -
Mark Knecht -
Peter Groves