Apogee X-Firewire 400 update

[Max]

Apogee X-FireWire card availability update.

The chip provider Apogee has partnered with for the X-FireWire card has experienced a severe delay in production. This has unfortunately forced us to push back the release of the card. We will have a limited supply of pre-production cards for certain configurations such as a single Rosetta 800 or Rosetta 200 (or multiple units on a PC at lower sample rates) by September 30th. At this time, full production cards have been pushed back to late November.

For our part, we are doing the following to address the issue:

1.) We have lowered the retail price of the initial card from $595 to $395.

2.) We will be providing a free upgrade to the full production X-FireWire 400 card (either a card or chip swap) for those customers that receive a pre-production version and require a full-production card.

3.) When we do get S800 compliant chips that are stable and reliable, we will release the X-FireWire 800 card and offer an upgrade for $200.00 to owners of the X-FireWire 400 card.


A Tech Note regarding compatibility and performance of the pre-production X-FireWire 400 card will be available September 9th on the Apogee website here:

http://www.apogeedigital.com/pdf/xfirewire_technote.pdf

Use this to determine whether or not your configuration will work with the pre-production X-FireWire 400 card. If your configuration is compatible with the pre-production cards and you would like to receive one or more, please email us at info@apogeedigital.com with your contact and dealer information and we will put your order in the queue.

We sincerely apologize for the long delay and any inconvenience this may have caused (it is frustrating for us too). Feel free to contact us with any other questions.

Best regards,

[subspace]

http://www.apogeedigital.com/pdf/xfirewire_technote.pdf

returns this:

Sorry...

The page that you have requested is missing or does not exist.
For more info or to report it please contact us and specify "Apogee Web Site" as the topic.


Thank you.

[Max]

That is because it is not September 9th yet. Check back on Thursday.

[xqtion]

Any rough release date for X-FireWire 800 card ?
Thanks.

[signit]

Thanks for the update Max.

Now to try and get my hands on one of those X400 Cards....

[signit]

Apparently these pre-prod cards etc are for the US only ?

AAARGGHH!!

[Max]

Originally posted by signit
Apparently these pre-prod cards etc are for the US only ?

AAARGGHH!!

No, the pre-production cards are available worldwide.

[signit]

Not according to arbiter the sole importer in the uk :-(

Is there anyway I can buy from apogee directly ?

[pendejo]

Is 15.6 msec latency the best we can hope for on Mac OS X? . That seems a lot. Thanks.

[Kestral]

As pendejo mentioned, the latency seems high. I wonder what the latency would be if one were to just use the digital i/o on the G5 on a Rosetta 200.

[Max]

Originally posted by pendejo
Is 15.6 msec latency the best we can hope for on Mac OS X? . That seems a lot. Thanks.

It is actually better than most and we expect it to continue to improve. That being said, do not forget that the X-Firewire 400 card will feature a mixer application allowing for zero latency monitoring, among other things.

[jimmymio]

Max: How does that rate compare w/ the latency I get w/ the mlan firewire card I bought for my Trak2? JP

[pendejo]

Originally posted by Max
It is actually better than most and we expect it to continue to improve. That being said, do not forget that the X-Firewire 400 card will feature a mixer application allowing for zero latency monitoring, among other things.

I've had my eyes on a Rosetta 200 and I guess I was hoping for a somewhat lower latency than that. My aging 900 Mhz PC/Aardvark 24/96 combo (which I'm in the process of replacing with a G5) somehow manages around 10-11ms latency (if what the driver reports is accurate). I often record guitar with amp modeling in-the-box and I've found that if the latency is greater than 12ms it starts becoming noticable and distracting.

[Max]

Ok, let's get a few things straight here, so that we are all on the same page.

First of all, latency is caused by the driver in the computer not the card itself. Latency is needed with native systems to insure that your audio processes correctly when you are recording. Multi-track audio, as we all know, is quite taxing on a computer. If your computer is busy and doesn't process the audio in time, the consequences are severe glitches (i.e.drop outs, pops and clicks, etc.) in your recordings. To eliminate these glitches audio drivers use a buffer to read ahead and store the samples in RAM; as one sample is played, another is added to the buffer. When the computer is busy there’s hopefully enough samples stored in the buffer to keep playing uninterrupted. This creates a waiting queue and causes the playback to be delayed. As you increase the latency you give the computer more time to process the audio, but increase the delay between what you are recording and playback.

So the latency of the X-Firewire 400 card itself is negligible. The latency of Apple’s Core Audio driver is 15.6 msec. The latency of the X-Firewire ASIO driver for PC is 6.5 msec.

Now, no computer based system is capable of eliminating latency altogether, but there are ways around it. The X-Firewire cards feature a mixer application that provides zero-latency monitoring by allowing you to route your incoming tracks directly out of the hardware, eliminating any driver latency (other than the hardware itself, which is negligible as far as your ears are concerned). So no matter what the actual driver latency is, it will not be a problem with the low latency mixer utility.

The big problem with Apogee’s mLAN card was that it did not have this zero-latency mixer application and so you were stuck with 16 msec of latency and no way to compensate for it. With the X-Firewire card, you're covered.

[OzNimbus]

Originally posted by Max


So the latency of the X-Firewire 400 card itself is negligible. The latency of Apple’s Core Audio driver is 15.6 msec. The latency of the X-Firewire ASIO driver for PC is 6.5 msec.

As someone used to working at 1.2 ms of latency, that's a pretty bitter pill to swallow. Sure, you've got your workarounds, but if you want to run a live rig thru your DAW software of choice, you're pretty screwed if you can only get down to 6.5 ms.

Sounds like you need to work harder on those drivers! Perhaps you should give a serious look at the DWAVE protocol.

-0z-

[Max]

Originally posted by OzNimbus
As someone used to working at 1.2 ms of latency, that's a pretty bitter pill to swallow. Sure, you've got your workarounds, but if you want to run a live rig thru your DAW software of choice, you're pretty screwed if you can only get down to 6.5 ms.

Sounds like you need to work harder on those drivers! Perhaps you should give a serious look at the DWAVE protocol.

-0z-

As I said, with the mixer utility, you are getting less than 1.2 msec of latency. I would be very interested to know what you are using that is giving you 1.2msec of latency without a compensator....

[Max]

I should also make it clear that the latency on the Apogee is actually measured analog in to analog out. This is not the same as the 'buffer setting' latency you would read in your application. So many people with other products out there actually have more latency than they think if they were to measure it in this fashion. Our numbers also includes the converter latency which is about 2msec total in and out worse case. So even if the driver was infinitely fast and had no latency, you'd still get up to 2msec latency from the converters alone. Note that even with hardware based systems such as ProTools there still is latency of between 4 and 6 msec measured analog in to analog out.

But again, latency is only a factor when listening back to your source from the computer. With the mixer app, you bypass the computer playback, eliminating the problem.

[OzNimbus]

Originally posted by Max
As I said, with the mixer utility, you are getting less than 1.2 msec of latency. I would be very interested to know what you are using that is giving you 1.2msec of latency without a compensator....

Max, with all due respect, I really don't have any use for a utility ZLM mixer scheme. I'm using SawStudio, which is written in Machine Code, so belive me, it has more than enough power to monitor and run EQ's, Comps, verbs, and anything else you can throw at it, at 1.2 ms..... and that's with a pair of Delta 1010s using ASIO!

Now only if more manufacturers would take a look at DWAVE, we might be on to something.

I'm sure your converters are brilliant, but I'm also sure I could get better latency results running the Rosetta 800 into an RME card.
-0z-

[jimmymio]

max wrote:

"The big problem with Apogee’s mLAN card was that it did not have this zero-latency mixer application and so you were stuck with 16 msec of latency and no way to compensate for it. With the X-Firewire card, you're covered."


But I use the direct monitoring in the Trak2 so I'm assuming this feature would be of no advantage to me. What advantages would this card offer to someone who uses the Trak2 and the Apogee mlan card? JP

[Max]

Originally posted by OzNimbus
Max, with all due respect, I really don't have any use for a utility ZLM mixer scheme. I'm using SawStudio, which is written in Machine Code, so belive me, it has more than enough power to monitor and run EQ's, Comps, verbs, and anything else you can throw at it, at 1.2 ms..... and that's with a pair of Delta 1010s using ASIO!

Now only if more manufacturers would take a look at DWAVE, we might be on to something.

I'm sure your converters are brilliant, but I'm also sure I could get better latency results running the Rosetta 800 into an RME card.
-0z-

With all due respect, I do not think you are getting the numbers you claim. The latency on your I/O is higher than 1.2 msec (the Apogee hardware at around 2msec analog to analog has pretty much the lowest latency on the market). Add the latency of the ASIO driver and you are probably somewhere between 4 and 6 msec. in real time.

The way to accurately measure latency is to send an analog pulse into the A/D on one channel of a scope and send the D/A output in to the second channel of the scope and measure the difference. For reference, the RME 96/52 card measures around 7 msec. analog to analog with the Rosetta 800 using this method. So yes, a Rosetta 800/RME combo is faster on the Mac, but not as fast as the Rosetta 800 X-Firewire combo on the PC. It should also be noted that RME uses a ZLM app as well, for similar reasons.

As for driver development, Core Audio is Apple's driver. We have no control over Core Audio development, although we do encourage Apple to continue to strive to achieve better performance and they are continually improving things where they can. What this means is that any device using Apple's Firewire Core Audio driver has the same basic latency as the X-Firewire card or higher depending on the hardware used.

As there is more control over ASIO development, more can be done outside of Windows to improve performance on our end, which is reflected in the spec obviously.

[OzNimbus]

Originally posted by Max
With all due respect, I do not think you are getting the numbers you claim. The latency on your I/O is higher than 1.2 msec (the Apogee hardware at around 2msec analog to analog has pretty much the lowest latency on the market). Add the latency of the ASIO driver and you are probably somewhere between 4 and 6 msec. in real time.

Ok, fair enough. Chalk that one up to a misunderstanding. But I am curious, what is the lowest stable buffer setting you guys have got the card to run at on a PC, in # of samples?

-0z-

[OzNimbus]

Originally posted by Max


As for driver development, Core Audio is Apple's driver. We have no control over Core Audio development, although we do encourage Apple to continue to strive to achieve better performance and they are continually improving things where they can. What this means is that any device using Apple's Firewire Core Audio driver has the same basic latency as the X-Firewire card or higher depending on the hardware used.

As there is more control over ASIO development, more can be done outside of Windows to improve performance on our end, which is reflected in the spec obviously.

Sounds like all is not well in Apple land. Meanwhile, on the PC side, WDM is a freaking nightmare, and ASIO ain't much better. I'd highly recommend checking out THIS thread Max!

http://www.sawstudiouser.net/forums/...read.php?t=526


Thanks,
-0z-

[blaugruen7]

max,
how about playing vst-instruments and monitoring in mac os x?
same latency?
greets,
gunnar

[sparker]

Hi Chaps,
We will bring the F/Wire 400 card into the UK if our customers require it. What we need to make sure happens is that all customers for the card realise that this is not the 800 card that will follow later - and so expect the correct level of performance and functionality for their particular system.

As the 400 card is lower priced, we'll introduce it as a different product code to distinguish the models from each other.

Hope this helps - keep your eye on our site for further news - www.arbitermt.co.uk

Cheers

Steve Parker
Music Technology Manager
Arbiter Group

[signit]

In that case I will be buying one as soon as you have them available. Any idea on pricing and timeframes ?

[sparker]

We just have to check our purchase price with Apogee and then we'll set a UK price. Timeframe would be about 10 days from shipping at worst so we'll try to organise this on Friday and then ship them as soon as they're ready.

Cheers

Steve Parker
Arbiter Group

[sparker]

Hi -

UK SRP for the Firewire 400 card will be £220 ex VAT. By the looks of things, we should see these early in October.

Hope this helps

Steve

[signit]

Thanks for the info Steve, look forward to getting one.

[Max]

Originally posted by blaugruen7
max,
how about playing vst-instruments and monitoring in mac os x?
same latency?
greets,
gunnar

No, latency will be different here, as you are now measuring the latency of the MIDI controller to the computer, through the VST instrument, back out of the computer and through the D/A.

[rnbeatz]

Originally posted by Max
Apogee X-FireWire card availability update.

3.) When we do get S800 compliant chips that are stable and reliable, we will release the X-FireWire 800 card and offer an upgrade for $200.00 to owners of the X-FireWire 400 card.

Will there be a performance gain by using the 800 card if you have a firewire 800 port over a 400 card on a 400 port?