Different audio interfaces for recording

[Doum]

Hello,

I have an M-Audio Delta 1010 (8 analogs I/O + 1 word clock I/O + 1 spdif
I/O on RCA jacks) on my desktop computer and an Edirol FA-101 firewire
interface (8 analogs I/O + 1 spdif I/O optical) for my laptop.

I need to record 14 audio tracks and play a stereo mix of already recorded
tracks (monitoring) simultaneously.

Is it a workable solution to install the firewire interface on the desktop
and use its inputs along with the PCI interface inputs?

The monitoring would go through the Delta 1010, no output used on the
Edirol.

Can I expect some sync problems from the track recorded with the PCI
against those recorded with the FW.

The software is VegasPro 9.0.

Any thoughts?

TIA
Dominique Geoffroy

[Doum]

"Soundhaspriority" > écrivait
:

>
> "Doum" > wrote in message
> .. .
>> Hello,
>>
>> I have an M-Audio Delta 1010 (8 analogs I/O + 1 word clock I/O + 1
>> spdif I/O on RCA jacks) on my desktop computer and an Edirol FA-101
>> firewire interface (8 analogs I/O + 1 spdif I/O optical) for my
>> laptop.
>>
>> I need to record 14 audio tracks and play a stereo mix of already
>> recorded tracks (monitoring) simultaneously.
>>
>> Is it a workable solution to install the firewire interface on the
>> desktop and use its inputs along with the PCI interface inputs?
>>
>> The monitoring would go through the Delta 1010, no output used on the
>> Edirol.
>>
>> Can I expect some sync problems from the track recorded with the PCI
>> against those recorded with the FW.
>>
> Yes.
>
>> The software is VegasPro 9.0.
>>
>> Any thoughts?
>>
>> TIA
>> Dominique Geoffroy
>
> Not with Windows, if that's what you're thinking. There are sync
> problems, and there are OS problems. It's just not feasible with
> Windows.
>
> Bob Morein
> (310) 237-6511
>
>
>

Yes it's Windows (I'm using Vegas), OK for the sync problems, I'm
suspecting it, but what do you mean by OS problems?

TIA
DG

[Sean Conolly]

"Soundhaspriority" > wrote in message
...
> The OS simply isn't set up to do it. Each card represents a complex
> realtime process that demands prompt service in a competing way. With some
> conniving, a Windows programmer can arrange for his realtime app to work.
> It's impossible to coordinate two unrelated apps. The internals of Windows
> aren't up to it.

There's only one app at work here, taking streams from two different
drivers. It shouldn't be hard to set up sufficient buffers for both to
record smoothly. I've done it with a PCI and a USB interface with no
problems from the OS or application side, on an older single core system.

> From the computer science perspective, it's a difficult problem to design
> an OS with guaranteed realtime execution of multiple applications. That
> capability actually isn't compatible with a graphical user interface with
> the responsiveness users have come to expect. In the future, when all
> computers have many cores (a lot more than two), there will be an option
> to allocate some of the cores to realtime.

Again, you're conufsing what the app does with what the drivers do. And the
place where true realtime is needed is the interface itself. Every stage
downstream is buffered.

Sean

[Sean Conolly]

"Doum" > wrote in message
.. .
> Can I expect some sync problems from the track recorded with the PCI
> against those recorded with the FW.

The problem is that the interfaces won't share a common clock, and the
Edirol has no support to send or receive word clock. Unless you are very
lucky to have two internal clocks that are almost exactly the same, the two
sets of tracks will end up not just out of sync but out of pitch.

Do a quick test - record two tracks, one through each interface, using a
guitar or keyboard or just singing, and then see if you can make the tracks
fit together. That will let you know if it's going to to be workable or if
you need to find another option.


Sean

[Sean Conolly]

"Soundhaspriority" > wrote in message
...
>
> "Sean Conolly" > wrote in message
> ...
>> "Soundhaspriority" > wrote in message
>> ...
>>> The OS simply isn't set up to do it. Each card represents a complex
>>> realtime process that demands prompt service in a competing way. With
>>> some conniving, a Windows programmer can arrange for his realtime app to
>>> work. It's impossible to coordinate two unrelated apps. The internals of
>>> Windows aren't up to it.
>>
>> There's only one app at work here, taking streams from two different
>> drivers. It shouldn't be hard to set up sufficient buffers for both to
>> record smoothly. I've done it with a PCI and a USB interface with no
>> problems from the OS or application side, on an older single core system.
>>
> Sean, may we have some details on what you did and how you did it?

Um, see above? I did some tests with recording tracks via a USB interface
(E-Mu) and a PCI interface (MOTU core) at the same time. There's no issue
with the OS unable to support the interfaces - servicing one additional
driver adds little on top of all the other drivers on the system. If you can
record X number of tracks you should be able to do it with two interfaces
about as easily as one.

The real problem was exactly what I described in my other post - the clocks
weren't sync'd so the tracks couldn't be mixed together.

Sean

[Peter Larsen[_3_]]

Doum wrote:

> "Soundhaspriority" > écrivait
> :

>>> Can I expect some sync problems from the track recorded with the PCI
>>> against those recorded with the FW.

>> Yes.

Affirmative, you need to use a common clock source, one way to do it is to
link the sp-dif out from the firewire-card to the sp-dif in on the 1010 and
select sync to sp-dif in the 1010 monitor mixer application.

>> Not with Windows, if that's what you're thinking. There are sync
>> problems, and there are OS problems. It's just not feasible with
>> Windows.

Nonsense, scaremongering as a general statement. The actual setup may
however require tuning and some of the time things just can not be made to
work reliably. Check by recording a 200 Hz sine wave, any discontinuities
should be easy to spot, be they under- or overflow, buffer settings may need
to be adjusted. Stop and restart recording multiple times and recheck ....
this here laptop is a great recorder only if left running, some issue with
the xp audio drivers and memory management and -paging that I haven't quite
worked out.

>> Bob Morein
>> (310) 237-6511

> Yes it's Windows (I'm using Vegas), OK for the sync problems, I'm
> suspecting it, but what do you mean by OS problems?

There's a real morein and there's a stalker morein, the real morein usually
writes sense, the stalker impersonating him usually doesn't.

> TIA
> DG

Kind regards

Peter Larsen

[Sean Conolly]

"Soundhaspriority" > wrote in message
...
>
> "Sean Conolly" > wrote in message
> ...
>> "Soundhaspriority" > wrote in message
>> ...
>>>
>>> "Sean Conolly" > wrote in message
>>> ...
>>>> "Soundhaspriority" > wrote in message
>>>> ...
>>>>> The OS simply isn't set up to do it. Each card represents a complex
>>>>> realtime process that demands prompt service in a competing way. With
>>>>> some conniving, a Windows programmer can arrange for his realtime app
>>>>> to work. It's impossible to coordinate two unrelated apps. The
>>>>> internals of Windows aren't up to it.
>>>>
>>>> There's only one app at work here, taking streams from two different
>>>> drivers. It shouldn't be hard to set up sufficient buffers for both to
>>>> record smoothly. I've done it with a PCI and a USB interface with no
>>>> problems from the OS or application side, on an older single core
>>>> system.
>>>>
>>> Sean, may we have some details on what you did and how you did it?
>>
>> Um, see above? I did some tests with recording tracks via a USB interface
>> (E-Mu) and a PCI interface (MOTU core) at the same time. There's no issue
>> with the OS unable to support the interfaces - servicing one additional
>> driver adds little on top of all the other drivers on the system. If you
>> can record X number of tracks you should be able to do it with two
>> interfaces about as easily as one.
>>
>> The real problem was exactly what I described in my other post - the
>> clocks weren't sync'd so the tracks couldn't be mixed together.
>>
>> Sean
>>
> Sean, what app(s) did you use in the test setup?

Reaper, using the ASIO interface.

Sean

[Doum]

Doum > écrivait
:

> Hello,
>
> I have an M-Audio Delta 1010 (8 analogs I/O + 1 word clock I/O + 1
> spdif I/O on RCA jacks) on my desktop computer and an Edirol FA-101
> firewire interface (8 analogs I/O + 1 spdif I/O optical) for my
> laptop.
>
> I need to record 14 audio tracks and play a stereo mix of already
> recorded tracks (monitoring) simultaneously.
>
> Is it a workable solution to install the firewire interface on the
> desktop and use its inputs along with the PCI interface inputs?
>
> The monitoring would go through the Delta 1010, no output used on the
> Edirol.
>
> Can I expect some sync problems from the track recorded with the PCI
> against those recorded with the FW.
>
> The software is VegasPro 9.0.
>
> Any thoughts?
>
> TIA
> Dominique Geoffroy
>

Thanks all for your comments, I'll install the FW on my desktop and do some
tests.

My system is a Core2Quad with 8 GB RAM running Windows7 x64 and VegasPro
9.0 x64, if that matters.

[Doum]

"Sean Conolly" > écrivait
:

> "Doum" > wrote in message
> .. .
>> Can I expect some sync problems from the track recorded with the PCI
>> against those recorded with the FW.
>
> The problem is that the interfaces won't share a common clock, and the
> Edirol has no support to send or receive word clock. Unless you are
> very lucky to have two internal clocks that are almost exactly the
> same, the two sets of tracks will end up not just out of sync but out
> of pitch.
>
> Do a quick test - record two tracks, one through each interface, using
> a guitar or keyboard or just singing, and then see if you can make the
> tracks fit together. That will let you know if it's going to to be
> workable or if you need to find another option.
>
>
> Sean
>
>

How about syncing through spdif, as suggested by Peter Larsen, is it a shot
in the dark?

[Doum]

"Peter Larsen" > écrivait
k:

> Doum wrote:
>
>> "Soundhaspriority" > écrivait
>> :
>
<snip>
>
> Affirmative, you need to use a common clock source, one way to do it
> is to link the sp-dif out from the firewire-card to the sp-dif in on
> the 1010 and select sync to sp-dif in the 1010 monitor mixer
> application.
>
<snip>
>
> Kind regards
>
> Peter Larsen
>
>
>

So far, one (2) of the inputs I was using was the spdif on the 1010, I
have 2 cheap condenser mics that I use as overheads on the drums, they go
to a tube mic preamp, the preamp goes to the XLR inputs of a Tascam
stereo DAT recorder and the spdif OUT of the DAT goes to the spdif IN of
the 1010. The clock is set to spdif in the Delta control panel. I have
no problem recording 8 simultaneous tracks with that setup.

If I use a device like http://www.m-audio.com/products/en_us/CO2.html

Is such device any good or sometimes it works and other time it doesn't?
It's not too expensive and it seems that it might do the trick (syncing
spdif, I mean).


Could this work?

DAT spdif OUT->Delta spdif IN, Delta spdif OUT->CO2 coax IN, CO2 optical
OUT->Edirol optical IN, set the clock to spdif in the Delta control panel
and press the Digital In sync switch on the Edirol.

I ask all those questions because I know I'm going to have to spend some
money either on the spdif converter or another Delta 1010. The 1010 is
more expensive than the converter and since I already have the Edirol...

TIA
Dominique

[Peter Larsen[_3_]]

Doum wrote:

> "Peter Larsen" > écrivait

>> Affirmative, you need to use a common clock source ...

[Doum using 1010 sp-dif in for signal from a mic pair via a dat]

> If I use a device like http://www.m-audio.com/products/en_us/CO2.html

> Is such device any good or sometimes it works and other time it
> doesn't? It's not too expensive and it seems that it might do the
> trick (syncing spdif, I mean).

Looks like one of the gadgets it is nice to have in the tool-kit.

> Could this work?

> DAT spdif OUT->Delta spdif IN, Delta spdif OUT->CO2 coax IN, CO2
> optical OUT->Edirol optical IN, set the clock to spdif in the Delta
> control panel and press the Digital In sync switch on the Edirol.

Here's what I would do:

mic pre's -> Edirol -> firewire in on computer

mic pre's -> 1010 -> pci-bus on computer

edirol sp-dif out -> 1010 sp-dif in, set 1010 to sync to sp-dif

use the dat machine as paperweight on desk.

> I ask all those questions because I know I'm going to have to spend
> some money either on the spdif converter or another Delta 1010. The
> 1010 is more expensive than the converter and since I already have
> the Edirol...

As I see it with the what remains of human intelligence, if any, on a hot
day your concept should work, I just suggest a simpler hook-up that gives
you 16 channels in. You may have a mic pre issue, ie. not enough. A
Soundcraft EPM or a Mackie or similar may be the cheapest way to get a
bundle of acceptable micpres with phantom.

With all such things you mileage may vary wildly, I don't think there is a
"correct solution", but perhaps you should consider adding a 1010lt tho ....
might be a neater solution since the delta series things are meant to
co-exist.

All of the above idle attempts at speculation during a heatwave, you'll be
the first to know whether it works with your specific setup.

> TIA
> Dominique

Kind regards

Peter Larsen

[Arny Krueger]

"Doum" > wrote in message

> Hello,
>
> I have an M-Audio Delta 1010 (8 analogs I/O + 1 word
> clock I/O + 1 spdif I/O on RCA jacks) on my desktop
> computer and an Edirol FA-101 firewire interface (8
> analogs I/O + 1 spdif I/O optical) for my laptop.
>
> I need to record 14 audio tracks and play a stereo mix of
> already recorded tracks (monitoring) simultaneously.

> Is it a workable solution to install the firewire
> interface on the desktop and use its inputs along with
> the PCI interface inputs?

> The monitoring would go through the Delta 1010, no output
> used on the Edirol.

> Can I expect some sync problems from the track recorded
> with the PCI against those recorded with the FW.
>
> The software is VegasPro 9.0.
>
> Any thoughts?

The general trick for synching multiple audio interfaces is to mentally
designate one of them as the master, and daisy-chain the others off of it,
via digital inputs and outputs. You then lock the clocks of the subordinate
intefaces to their digital inputs. This usually works even if you aren't
recording or playing via the inputs and outputs that you choose to use for
this.

Virtually every digital audio interface that has digital I/O has some
provision for locking its internal clock to one of its digital input(s).
Most every digital output recieves clock information that it passes on via
its digital output, whether you're playing music via it or not.

If your digital interfaces actually have dedicated inputs and outputs for
external clocking, then so much the better - use those.

That takes care of the clock synching problem. The remaining problem is
small variations in how long a digital interface takes to get going once you
start recording or playing. Usually those are small. If they are on the
order of several milliseconds or more, then you will need to hand synch the
tracks from the various interfaces. Ordinarly being off by a few clock
cycles is more of a theoretical problem than a practical problem. Do try to
get both halves of your coincident pairs on the same audio interface!

[Arny Krueger]

"Peter Larsen" > wrote in message
k

>>> Not with Windows, if that's what you're thinking. There
>>> are sync problems, and there are OS problems. It's just
>>> not feasible with Windows.
>
> Nonsense, scaremongering as a general statement. The
> actual setup may however require tuning and some of the
> time things just can not be made to work reliably. Check
> by recording a 200 Hz sine wave, any discontinuities
> should be easy to spot, be they under- or overflow,
> buffer settings may need to be adjusted. Stop and restart
> recording multiple times and recheck .... this here
> laptop is a great recorder only if left running, some
> issue with the xp audio drivers and memory management and
> -paging that I haven't quite worked out.

Agreed Peter, you can use multiple audio interfaces and they will start
fairly close together and stay in synch as long as you solve the clock
synching problem. That requires digital I/O or external clock connections,
which not every audio interface has.

It's always a lot of fun to see computer science geeks spouting
high-sounding theoretical balderdash saying that you can't do something that
you have done successfully and routinely for years.

I won't guarantee that recording with multiple interfaces will work
perfectly in *every* case but I've seen it work often enough to know that
when it doesn't work, its not some global rule asserting itself, it is due
to very practical situations that aren't the general rule.

[Mike Rivers]

Doum wrote:
> I have an M-Audio Delta 1010 (8 analogs I/O + 1 word clock I/O + 1 spdif
> I/O on RCA jacks) on my desktop computer and an Edirol FA-101 firewire
> interface (8 analogs I/O + 1 spdif I/O optical) for my laptop.
>
> I need to record 14 audio tracks and play a stereo mix of already recorded
> tracks (monitoring) simultaneously.
>
> Is it a workable solution to install the firewire interface on the desktop
> and use its inputs along with the PCI interface inputs?

Since you're using Vegas, I assume your computer runs
Windows. You will have trouble with this approach. The short
answer is that Windows doesn't allow more than a single ASIO
driver to be active at a time, and most multi-channel
interfaces that have a WDM driver don't offer full
multi-channel recording capabilities in that mode. It's just
there for convenience of playback with games and such.

> Can I expect some sync problems from the track recorded with the PCI
> against those recorded with the FW.

That's a secondary situation, yes.

> Any thoughts?

If you want to stick with your software and computer, you'll
need an interface with more channels, or you'll need to move
into a product family that has a driver which support
expansion by adding multiple units running under the same
driver. PreSonus has a line of FireThings that alllow you to
use up to three devices with one driver. Focusrite has
something similar. I'm not aware of such a capability with
M-Audio or Edifol drivers.

If your need is to record live shows, I would recommend that
you go the simplest route possible, involving the fewest
pieces. Look into a mixer that can send all of its mic
preamps out through a single Firewire interface (the Mackie
Onyx i series, PreSonus StudioLive, Allen & Heath ZED-R16,
and I think Alesis has something like that now, Of if it's a
studio recording kind of project, take your group to a
studio that has the setup to record all the simultaneous
tracks that you need, and then take the files home, import
them into Vegas, and mix away.

--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

[Mike Rivers]

Sean Conolly wrote:

> Um, see above? I did some tests with recording tracks via a USB interface
> (E-Mu) and a PCI interface (MOTU core) at the same time. There's no issue
> with the OS unable to support the interfaces

Given the age of those devices, you were likely using the
WDM driver. This may not be satisfactory (or even available)
for the original poster's hardware. Nothing wrong with
believing that it will work, and trying it to see if it
does, however.

> The real problem was exactly what I described in my other post - the clocks
> weren't sync'd so the tracks couldn't be mixed together.

As far as word clock sync goes, if both interfaces have
S/PDIF I/O, that can usually be used as a clock source to
synchronize the two interfaces.


--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

[Mike Rivers]

Sean Conolly wrote:

> Reaper, using the ASIO interface.

That's interesting. I've never tried connecting two devices
to a computer and running Reaper. My experience with other
DAW programs is that while you may be offered a choice of
any ASIO device for which there's a driver installed, you
can select only one ASIO device for recording and playback.
Are you perhaps using ASIO4ALL? I've heard that this can
work since in essence it uses a WDM driver.

--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

[Peter Larsen[_3_]]

Mike Rivers wrote:

> If you want to stick with your software and computer, you'll
> need an interface with more channels, or you'll need to move
> into a product family that has a driver which support
> expansion by adding multiple units running under the same
> driver. PreSonus has a line of FireThings that alllow you to
> use up to three devices with one driver. Focusrite has
> something similar. I'm not aware of such a capability with
> M-Audio or Edifol drivers.

M-Audio delta series cards are designed for using up to 4 cards in a
computer, it was for this reason that I suggested getting a 1010lt to
supplement the 1010 if possible as one of the means o making the setup
neater.

Kind regards

Peter Larsen

[Neil Gould]

Mike Rivers wrote:
> Sean Conolly wrote:
>
>> The real problem was exactly what I described in my other post - the
>> clocks weren't sync'd so the tracks couldn't be mixed together.
>
> As far as word clock sync goes, if both interfaces have
> S/PDIF I/O, that can usually be used as a clock source to
> synchronize the two interfaces.
>
I've recorded from multiple interfaces that were locked together via
wordclock, but each had an input for wordclock (and/or timecode) sync as
part of their hardware interface. I don't see how it would work reliably
without the hardware-level locking, though, since everything would be too
far downstream.

--
best regards,

Neil

[Arny Krueger]

"Peter Larsen" > wrote in message
k
> Mike Rivers wrote:
>
>> If you want to stick with your software and computer,
>> you'll need an interface with more channels, or you'll
>> need to move into a product family that has a driver
>> which support expansion by adding multiple units running
>> under the same driver. PreSonus has a line of FireThings
>> that alllow you to use up to three devices with one
>> driver. Focusrite has something similar. I'm not aware
>> of such a capability with M-Audio or Edifol drivers.
>
> M-Audio delta series cards are designed for using up to 4
> cards in a computer, it was for this reason that I
> suggested getting a 1010lt to supplement the 1010 if
> possible as one of the means o making the setup neater.

I've put 100s of hours and dozens of gigs on a setup that was composed of a
1010, 2 each 1010LT, and a Delta 66 for a total of 28 channels.

Results were the same as 8 channels, all on the same interface. The general
methodology used was as previously described - digital port chaining.

[Mike Rivers]

Peter Larsen wrote:

> M-Audio delta series cards are designed for using up to 4 cards in a
> computer, it was for this reason that I suggested getting a 1010lt to
> supplement the 1010 if possible as one of the means o making the setup
> neater.

Yes, I remember the ability to use multiple Delta 1010 cards
back when they were new, which was before ASIO was invented.
Perhaps M-Audio's current ASIO driver (like the Focusrite
and PreSonus drivers I mentioned) is capable of supporting
multiple cards. If so, that's good to know, assuming he has
an available PCI slot in his computer.


--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

[Mike Rivers]

Neil Gould wrote:

> I've recorded from multiple interfaces that were locked together via
> wordclock, but each had an input for wordclock (and/or timecode) sync as
> part of their hardware interface. I don't see how it would work reliably
> without the hardware-level locking

Most S/PDIF input devices offer the option of synchronizing
its word clock to the incoming data stream. That's your
hardware-level locking. It works. Net wisdom is that it's
not a good idea to daisy chain S/PDIF ports for clock sync.
Using the output from #1 to sync #2 is OK, but using the
output of #2 to sync #3 will impose #2's jitter on the sync
source. Better to use #1 to sync #2 and #3, but there may be
some issues with signal level with the double load.

On the other hand, with today's re-clocking high stability
converters, incoming jitter is essentially erased, so maybe
it doesn't matter. There are no guarantees when you try to
mix too many things digital, just good principles to try to
follow and see what happens.



--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

[Doum]

"Peter Larsen" > écrivait
k:

> Doum wrote:
>
>> "Peter Larsen" > écrivait
>
>>> Affirmative, you need to use a common clock source ...
>
> [Doum using 1010 sp-dif in for signal from a mic pair via a dat]
>
>> If I use a device like http://www.m-audio.com/products/en_us/CO2.html
>
>> Is such device any good or sometimes it works and other time it
>> doesn't? It's not too expensive and it seems that it might do the
>> trick (syncing spdif, I mean).
>
> Looks like one of the gadgets it is nice to have in the tool-kit.
>
>> Could this work?
>
>> DAT spdif OUT->Delta spdif IN, Delta spdif OUT->CO2 coax IN, CO2
>> optical OUT->Edirol optical IN, set the clock to spdif in the Delta
>> control panel and press the Digital In sync switch on the Edirol.
>
> Here's what I would do:
>
> mic pre's -> Edirol -> firewire in on computer
>
> mic pre's -> 1010 -> pci-bus on computer
>
> edirol sp-dif out -> 1010 sp-dif in, set 1010 to sync to sp-dif
>
> use the dat machine as paperweight on desk.
>
>> I ask all those questions because I know I'm going to have to spend
>> some money either on the spdif converter or another Delta 1010. The
>> 1010 is more expensive than the converter and since I already have
>> the Edirol...
>
> As I see it with the what remains of human intelligence, if any, on a
> hot day your concept should work, I just suggest a simpler hook-up
> that gives you 16 channels in. You may have a mic pre issue, ie. not
> enough. A Soundcraft EPM or a Mackie or similar may be the cheapest
> way to get a bundle of acceptable micpres with phantom.
>
> With all such things you mileage may vary wildly, I don't think there
> is a "correct solution", but perhaps you should consider adding a
> 1010lt tho .... might be a neater solution since the delta series
> things are meant to co-exist.
>
> All of the above idle attempts at speculation during a heatwave,
> you'll be the first to know whether it works with your specific setup.
>
>> TIA
>> Dominique
>
> Kind regards
>
> Peter Larsen
>
I like your connection scheme, I didn't think about it, as far as pre's
are concerned, no trouble there, I have a 32 channels (24 mics) Yamaha
console. I was using the preamp and the DAT to get 2 more inputs on my
Delta.

But the more I think about it, I'll probably go for another Delta 1010 so
I'll have my 14 recording channels, my 2 monitoring channels, the spdif
will be available for syncing the 2 interface and there won't be headache
involving drivers.

Thanks

[Doum]

Mike Rivers > écrivait -
september.org:

> Peter Larsen wrote:
>
>> M-Audio delta series cards are designed for using up to 4 cards in a
>> computer, it was for this reason that I suggested getting a 1010lt to
>> supplement the 1010 if possible as one of the means o making the setup
>> neater.
>
> Yes, I remember the ability to use multiple Delta 1010 cards
> back when they were new, which was before ASIO was invented.
> Perhaps M-Audio's current ASIO driver (like the Focusrite
> and PreSonus drivers I mentioned) is capable of supporting
> multiple cards. If so, that's good to know, assuming he has
> an available PCI slot in his computer.
>
>

Yes, I have two PCI slot left and I will probably get another Delta 1010,
the current drivers does support multiple Delta cards and ASIO.

Thanks to all.

[Sean Conolly]

"Mike Rivers" > wrote in message
...
> Sean Conolly wrote:
>
>> Reaper, using the ASIO interface.
>
> That's interesting. I've never tried connecting two devices to a computer
> and running Reaper. My experience with other DAW programs is that while
> you may be offered a choice of any ASIO device for which there's a driver
> installed, you can select only one ASIO device for recording and playback.
> Are you perhaps using ASIO4ALL? I've heard that this can work since in
> essence it uses a WDM driver.

You know, it's been a couple of years since I did this, and nowadays I'm
lucky to remember last week .

I had it completely wrong, I wasn't using Reaper or ASIO. I was using Cool
Edit, which in turn uses the individual wave inputs registered with Windows.
You select inputs/outputs that could be used from a list of all inputs, and
then you can select any of those for each track.

It makes it pretty trivial to record mixing interfaces - as long as you can
keep the clocks in sync somehow. Never tried the sp/dif sync trick, since I
don't have anything uselful equipped with sp/dif right now.

My apologies to all for the misinformation.

Sean

[Doum]

"Soundhaspriority" > écrivait
:

>
> "Doum" > wrote in message
> .. .
>> "Peter Larsen" > écrivait
>> k:
>>
> [snip]
>>
>> But the more I think about it, I'll probably go for another Delta
>> 1010 so I'll have my 14 recording channels, my 2 monitoring channels,
>> the spdif will be available for syncing the 2 interface and there
>> won't be headache involving drivers.
>>
>> Thanks
>
> You may well save a headache. Some felt my statement was too dogmatic,
> but here's some technical detail.
>
> When a Windows driver needs to work with a hardware device on a
> sub-microsecond basis, it invokes what is known as a "spin lock." The
> lock stops anything else from using the CPU during a time interval
> which Microsoft specifies must not exceed 35 microseconds. This is not
> friendly to other processes on the machine, but the duration is
> limited.
>
> However, many device drivers break this rule, causing an issue known
> as "latency". The more channels you put on a device driver, the more
> likely it is to break the rule. As long as only one device driver on
> the system is a rule breaker, things tend to work. This is why DAW
> workstations and laptop field recorders are configured to run just the
> audio app, frequently with drivers such as wireless and battery
> monitor disabled. It eliminates troublesome spin locks.
>
> If you have two drivers from different manufacturers, and each wants
> to stretch the spin lock, there is the possibility of drops. This is
> typically not acceptable to the user. If a single device driver runs a
> bunch of cards, this problem does not exist, because there is only a
> single instance of the driver for all those cards.
>
> Please pay no attention to my impersonator, Brian L. McCarty. We're
> all here to help, with our different styles and modest disagreements.
>
> Regards,
> Bob Morein
> (310) 237-6511
>
>
>

Useful info, Thank you.

Dominique Geoffroy

[Peter Larsen[_3_]]

Doum wrote:

[quoting morein]

>> You may well save a headache. Some felt my statement was too
>> dogmatic, but here's some technical detail.

>> When a Windows driver needs to work with a hardware device on a
>> sub-microsecond basis, it invokes what is known as a "spin lock." The
>> lock stops anything else from using the CPU during a time interval
>> which Microsoft specifies must not exceed 35 microseconds. This is
>> not friendly to other processes on the machine, but the duration is
>> limited.

>> However, many device drivers break this rule, causing an issue known
>> as "latency". The more channels you put on a device driver, the more
>> likely it is to break the rule. As long as only one device driver on
>> the system is a rule breaker, things tend to work. This is why DAW
>> workstations and laptop field recorders are configured to run just
>> the audio app, frequently with drivers such as wireless and battery
>> monitor disabled. It eliminates troublesome spin locks.

>> If you have two drivers from different manufacturers, and each wants
>> to stretch the spin lock, there is the possibility of drops. This is
>> typically not acceptable to the user. If a single device driver runs
>> a bunch of cards, this problem does not exist, because there is only
>> a single instance of the driver for all those cards.

And once gain the ocntext is that the OP suggests getting another 1010 or
1010lt, cards that are designed so that four cards run from the SAME driver
instance. Your comments are interesting but not to the point, if anything
they are in favour of the suggestion of using the delta cards in this
context, assuming pci bus is available.

Kind regards

Peter Larsen



>> Please pay no attention to my impersonator, Brian L. McCarty. We're
>> all here to help, with our different styles and modest disagreements.
>>
>> Regards,
>> Bob Morein
>> (310) 237-6511
>>
>>
>>
>
> Useful info, Thank you.
>
> Dominique Geoffroy

[Preben Friis]

"Soundhaspriority" > wrote in message
...
> When a Windows driver needs to work with a hardware device on a
> sub-microsecond basis, it invokes what is known as a "spin lock." The lock
> stops anything else from using the CPU during a time interval which
> Microsoft specifies must not exceed 35 microseconds. This is not friendly
> to other processes on the machine, but the duration is limited.
>
> However, many device drivers break this rule, causing an issue known as
> "latency". The more channels you put on a device driver, the more likely
> it is to break the rule. As long as only one device driver on the system
> is a rule breaker, things tend to work. This is why DAW workstations and
> laptop field recorders are configured to run just the audio app,
> frequently with drivers such as wireless and battery monitor disabled. It
> eliminates troublesome spin locks.

You are confusing DPC latency (where a driver can hold the interrupt for too
long) with the use of spin locks. Audio cards most often deliver audio in
blocks that are longer than 1 millisecond, so 25 us. does not really matter
in the big picture. A wireless driver that hawks the interrupt for several
milliseconds is a completely different story.

If a system has high DPC latency you can be forced to record audio in bigger
chunks (larger buffer sizes), but it is still possible and has nothing to do
with using multiple audio devices in one system.

You really shouldn't give advices about operating systems based on the
little that you know...

/Preben Friis

[Mike Rivers]

Soundhaspriority wrote:

> [The correct word is hog, not hawks]

I think he mistyped "hawgs" (the bus) ;)




--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

[Preben Friis]

> You do not know English very well. You really shouldn't try to express
> yourself in a language you do not know. Stick to Danish. I presume you are
> fluent in that language?

Godtaget. Hvis du afholder dig fra at sprede vildledning omkring
operativsystemer så lover jeg højt og helligt at undlade at give dig råd
omkring engelsk. Sjovt nok kunne din klon faktisk ikke skrive noget værre
end dig denne gang.

/Preben Friis