Fireface, or other 'digital interface' - why?

[cfmsp]

 

Chris,

 

I've always been curious when I hear that people use a Fireface 400/800 in front of their DACs. Anyone have any idea why this is done?

 

I'm guessing that perhaps it is being used to clean up jitter before passing the signal off a NOS DAC?

 

thanks,

Clay

 

 

 

 

 

 

 

[PeterSt]

Hi Clay,

 

I guess Chris missed your post, and since I (AFAIK) was the first one to do so anyway ...

 

Today it all seems very very easy and logical, but back in the days that you would be afraid to show interest in audio playback via computers because you would have been shot by the real "audiophiles" ... things just had to be "invented".

 

The main reason would be to get the soundcard electronica out of the (HF) noise of the computer.

And FYI this was in the days the first network players started to emerge (think of the SqueezeBox from today), which actually was the proof of "hey, this thing plays outside of the computer, with the data from that computer !".

 

Today, where USB DACs are common and not even a soundcard is there, we tend not to think about this anymore. Back then, when sure more of these Fireface-like devices existed, I think it needed explicit approach, and after me thinking about it an trying things, the Fireface came out as the actually only one device being able to do the job. Ok, do the job WELL. Why ?

 

Generally SPDIF pass through devices colour the sound just the same. For example, take it out from your motherboard and it will sound like sh*t. All digital, but real crap.

Although you could dedicate all sound differences caused by this part of the chain to jitter, there are some underlaying elements that each attribute to that jitter;

 

The coax or glass fiber cable for SPDIF is one of them. And generally, the longer they become, the worst it will get.

 

El cheapo sound devices (mobo chips included) indeed can pass on the data unmangled, but because they are el cheapo, nothing is done about jitter. Many (or most) of the soundcards are not made for hi end audio, so why bother about anything. It was only two/three years ago that the runner up of the good old soundblasters still upsampled to 48KHz by standards. Why bother.

 

Non pro devices (pro = what is used at the live stages) have no explicit means to control latency. Although today still latency is not recognized by everybody ase sound quality influencing, you bet it does (says me hehe). The shorter the latency, the better SQ.

Because the non-pro (a few exceptions are there these days) soundcards can't control latency, they THUS have to be pre-set to a huge latency, in order to be able to overcome the diverse hurdles that your PC might have and grabs CPU time (the physical means to have longer latency is a buffer, and the longer buffer will allow for better glicth free playback once your PC has capacity problems).

Btw, note that the latency in the pro world too has to be as low as possible, but that is because of otherwise a midi keyboard player hears is played tone 1 second later.

 

If you imagine a DTS receiver in front of your main amps, and you have an HTPC playing movies, the soundcard must be able to pass DTS. Few soundcards can do that, though more and more can these days.

 

The external soundcard, passsing through SPDIF only, must sound inherently good.

 

The drivers must work flawlessly.

 

All of the above topics do not often emerge in one soundcard at the same time. Certainly back then the Fireface was the only one, although a few of above topics now given as a kind of fact, back then had to be explored as topics themselves. For example, some Behringer was said to have DTS passthrough, but it did not work, and nobody could help with it. Or, some Maya seemed to do the job as well, but the drivers s*ck so much it couldn't be handled. This was all buying, trying, and bringing back.

 

The part of the cable towards the Fireface is just data. Btw, this cable can be 10 meters long (which few know). It carries SPDIF, but also ADAT (which just is a pro spec) or IOW 8 channels at one time (think surround and the date decoded BEFORE the soundcard, that passing through 8 SPDIF mono channels to the (pre)amps).

 

The Fireface reclocks the data, so it eliminates jitter. Careful here, because the "good specs" from back then are 800ps (by heart) and todat we laugh about it. But still.

 

Well, I can go on by mentioning each topic explicitly, but in the end all is there in the Fireface. The most important "feature" though it, that it is proven good for SQ, or the oter way around, it does not deteriorate it (like a cheap soundcard just would).

Since it also does 24/192 I think it is still the cheapest passthrough solution today in FF400 appearance. The FF400 additionally has a digital volume knob (which some on this forum call an analogue volume, which it really is not), and all together it is a great device to experiment with connections (the FF800 has 56 or so input/output channels).

 

I hope this is something for an answer. In the end it is my fault.

:-)

 

[wavac]

I suppose you mean firewire....

 

This is what Weiss believes:

 

Firewire is a peer-to-peer protocol, meaning that every device on a Firewire network is equally

capable of talking to every other device. Two video cameras on a Firewire network can share data

with each other. A Firewire audio interface could save sound data directly to a Firewire hard drive.

Your computer is just another peer on this network, and has no inherent special status.

Firewire is always implemented in hardware, with a special controller chip on every device. So the

load it puts on your CPU is much lighter than USB communications load, and you're much less

likely to lose any sound data just because you're running fifteen things at once. Specialized

hardware usually makes things faster and more reliable, and this is one of those times. But the real

reason Firewire is more reliable than USB is more fundamental than that. It's because Firewire

allows two operating modes. One is asynchronous, similar to what USB uses. The other is

isochronous mode, and it lets a device carve out a certain dedicated amount of bandwidth that

other devices can’t touch. It gets a certain number of time slices each second all its own. The

advantages for audio should be obvious: that stream of data can just keep on flowing, and as

long as there isn't more bandwidth demand than the wire can handle (not very likely) nothing will

interfere with it. No collisions, no glitches. From a practical perspective, this also makes it safer to

send a lot more audio via Firewire. That's why most of the multichannel interfaces (18 channels, 24

channels, etc.) are Firewire devices, and USB devices usually just send a two channel stereo signal.

For hooking up your mouse, keyboard or thumb drive, USB is plenty fast and plenty cheap. For

hard drives, either one will do (although Firewire is somewhat more reliable). For audio devices,

USB will do fine if no other devices are competing with it and if you have processor room to spare.

But Firewire will always be able to handle more load with lower latency and no glitches, because it

has resources it can set aside to make sure your audio gets where it needs to go.

 

 

[PeterSt]

Just typed this into a thread that got closed during my typing, while IMO it is very valid to put it in here now ...

 

From that other thread, a quote from sastusbulbas :

 

One of the reasons I frequent this site is due to my interest in building a computer source on a budget, but I still want to see why £700 sound cards are necessary and if there is better things to spend money on? I am here to learn, but am not going to throw away my old legacy kit and embrace PC audio until it has grown along side my current sources and became my matured choice in the home.

[...]

My old transports sound better than my old computers through my old Dacs, but I get constantly told a new Dac will make my old computer sound as good, therefore my Dac is rubbish, even though I enjoy it's performance with a suitable CD transport. If I use a modern computer such as an iMac or Powerbook I am told the same thing, my preference for a dedicated CD transport, and opinion that it sounds better than the Powerbook or such, is due to my Dac being rubbish!

 

Strangely enough you must have gotten this perception from the gang telling you to get the new "modern" pro-audio stuff. I mean, that side of the debate can have been the only side so much confusing.

I must say, I don't read that in between the lines, nor do I see the side I'm on intends such a thing.

 

This crazy "How sounds jitter" thread is a one on its own, and it should inform you about how jitter sounds, or whether it sounds at all for that matter. Yeah, well ...

 

You understood quite well that my side of the debate, or me only when it comes to it, suggests to have a £700 soundcard to pass through SPDIF in a good fashion. So yes, you got it, and why is in this other thread (the one right here by now), IMO sufficiently explained. OT for the "How does jitter sound" thread, but completely in context, is that one should use a good player to begin with. This by itself is totally unrelated to whatever DAC is used or is needed. On that matter I don't see *at all* why computer playback would need another DAC to begin with.

 

It does not.

 

Peter

 

PS: ... And the SPDIF passing through soundcard is just to keep your old DAC going. Not much of use when your old DAC would be USB connected ... which it won't wit your CDP ...

 

 

[audioengr]

Similar to USB devices, firewire devices move the digital conversion out of the computer to an external box with separate power supply. The electrical noise can then be controlled and there is a potential for jitter reduction, although it is not always achieved. I believe some of the firewire devices also support 24/192, although there is not much content available yet at this sample-rate.

 

There is also an opportunity for using a longer cable to the device than would be possible with S/PDIF or AES interfaces. In the firewire case however there is no support for asynchronous protocols, so the cable will still impact sound quality, just like a S/PDIF cable would.

 

Steve N.

Empirical Audio

 

[DanRubin]

A friend is using a Fireface 400 out of his Mac Mini. I'm using Steve's Offramp Turbo3. I plan to borrow the Fireface and do a comparison of the sound of the two devices, although it's hard to imagine that the Fireface could be any better than the Offramp. We'll see.

 

[BEEMB]

 

I'm hoping Ashley can give me some feedback here.

 

See, I output to some ADMI 9.1's, and I don't bother with an expensive interface to eliminate jitter before the DAC as Ashley always says that jitter is not a problem with the ADM9.1's...

 

So ...

 

Peter - you'll say the fireface would improve my sound.

 

Ashley - you'll say it wont.

 

Could you both give technical reasons as to why this is, or is not, the case.

 

Matt.

 

[Purite Audio]

I believe the ADM's allow you to use another dac as well don't they? That could make another interesting comparison. Keith.

 

[shenzi]

One advantage is being able to daisy-chain devices instead of running them all back to a hub or computer. Probably more applicable to the hard drives than the DAC, tho.

 

[ted_b]

?? I notice a lot of folks here are responding to "firewire" and one even thought the OP's "fireface" reference was a typo. The guy is NOT asking about the advantages of firewire per se; he is asking the advantages of external boxes like the Fireface 400/800. If a normal pc has a firewire interface, and folks like Weiss et al believe firewire to be a perfect digital vehicle, then why use an external box and more monies (the 400 is $1k, not chump change).

 

Peter and Steve, thanks for your responses. They are relevant, as always.

 

Ted

 

[DanRubin]

In the case of the Fireface, you need this because there are very few DACs (are there any? Minerva?) that will accept Firewire directly. What alternative do you have? In the case of USB, the same was initially true, so converters like the HagUsb allowed you to take USB out from your computer and go into any DAC. A number of DACs now have direct USB inputs, but we've learned that there's more to it -- it's not enough to simply put a USB port on the back of a DAC, which you will quickly discover if you listen to some of these DACs. There's quite a bit more required to get that interface right. Hence the continuing need for products like as the Empirical Audio Offramp and the new Bel Canto box. The same will probably be true of Firewire inputs if we start seeing them on DACs. But apart from Weiss, there's not a lot of indication that we will. Ethernet may be the dark horse.

 

[audioengr]

"The Fireface reclocks the data, so it eliminates jitter. Careful here, because the "good specs" from back then are 800ps (by heart) and todat we laugh about it. But still."

 

The word reclock is a stretch. It is an advanced PLL. I believe the chip used is the DiceII, which has an advanced PLL to reduce jitter. I've read the spec and tried some devices that use this chip.

 

The problem with Firewire is that it does not support Async mode or a "pull" protocol, only a "push" protocol. This puts the master clock in the computer rather than the device. Async puts the master clock in the device, which allows better jitter reduction.

 

I feel that Firewire is a flash in the pan for high-end audio. the Dice II chip was developed to address the latency and multiple device connectivity problems in recording studios. Soon there will be more Async USB interfaces and at 24/192, so it wil be moot IMO.

 

Steve N.

Empirical Audio

 

[weiss2496]

The problem with Firewire is that it does not support Async mode or a "pull" protocol, only a "push" protocol. This puts the master clock in the computer rather than the device.

 

Not true. Our Minerva DAC is the master clock when playing from Firewire, the computer is slaved. I can't think of anything better in terms of jitter issues.

 

Daniel

 

 

[cfmsp]

 

"The guy is NOT asking about the advantages of firewire per se; he is asking the advantages of external boxes like the Fireface 400/800."

 

As the author of the original post, I can authoritatively state that you are correct.

 

I already DO use firewire to connect to my DAC (Metric Halo ULN-2), although I can also route my Apple TV Toslink output to the DAC as well (for wireless streaming).

 

thanks to all who have responded, slightly off-key or not.

 

clay

 

 

[audioengr]

"Not true. Our Minerva DAC is the master clock when playing from Firewire, the computer is slaved. I can't think of anything better in terms of jitter issues."

 

Gordon Rankin told me that this was not supported on PC or Mac. I guess he needs correction. Thanks.

 

Steve N.

Empirical Audio

 

[PeterSt]

"The same will probably be true of Firewire inputs if we start seeing them on DACs."

 

I am fairly sure this won't happen. The market for any individual audiophile DAC manufacturer is too small to ever get started with developing a Firewire interface. No standard products exist, and the only solution offered is that board manufacturers could ask their customers to buy some back which in turn could be sold to e.g. me (as audiophile DAC manufacturer). On that matter I could have called Mr Weiss myself, but of course the answer is known in advance (and the question is stupid anyway).

 

But since we are here ... Mr Weiss, would you ? (and I *am* serious).

 

Peter

 

[weiss2496]

Well, there are standard products (chips and SDKs) available for audio over firewire. Design work is greatly simplified with those solutions.

Daniel

 

 

[Ashley James]

See, I output to some ADMI 9.1's, and I don't bother with an expensive interface to eliminate jitter before the DAC as Ashley always says that jitter is not a problem with the ADM9.1's...

 

So ...

 

Peter - you'll say the fireface would improve my sound.

 

Ashley - you'll say it wont.

 

Could you both give technical reasons as to why this is, or is not, the case.

 

Matt.

 

For the purposes of playing music from a computer or streaming device an appropriate digital output is required and the Consumer Electronics Industry has settled optical digital as the best compromise. This means that for most, unless they want the computer next to their hi fi, a DAC with optical digital input will prove the most versatile. There is no sound quality difference between any of the digital interfaces. They all work of they don't as science predicts. Gross errors or not errors is the rule of the day.

 

The arguments in favour are simply that optical digital cables avoid electrical connection between source and DAC or in our case Active Speakers with built in remote controlled preamp and DAC. ADM9.1s re-clock the data so jitter is not an issue. We are happier if our speakers are not electrically couple to a product we might not have heard of and that may have all sorts of digital crap on its chassis and I'd guess that much of the industry felt that way too.

 

FWIW wi fi is what everyone is moving to because it's simplest. Even Carphone warehouse and the up market Department stores are selling streaming devices and Apple, PS3 or Sonos and Logitech. Understandably Apple seems to be taking the hi end market, but the PS3 is as yet not fully appreciated for what it can do.

 

Hi end shops that I'm talking around the world report little interest in any of this but increasing sales in Valve?Tubes and vinyl and other exotica. It looks as though they are becoming more marginalised rather than less so and it suits them because the margins are massive by comparison with Apple for example.

 

Ash

 

PS. Matt to put things into perspective it's worth remembering that the damping factor of your speakers is at least 100 times higher than passives and the electronic crossover has one hundredth of the distortion. Big Audible differences.

 

 

 

[Purite Audio]

'the Consumer Electronics Industry has settled optical digital as the best compromise'

I always miss the important memo's.

 

[PeterSt]

Do I ? well, maybe not. :-)

 

From the side of having to use a lousy small jack digital interface from the mobo, sure yes, but from the side of the DAC used, YMMV. If Ashley tells you that his DAC(s) are 100% immune to incoming jitter, so be it, and the only thing you can do is check it if you like.

 

Putting the story a little upside down, I too have a DAC which is immune to incoming jitter, and has a spec of less than 1ps of jitter behind the SRC.;

As I told elsewhere, this DAC can be SPDIF connected (which I would do via the Fireface) and it can be I2S connected (homebrew solution from off the PC). And you know what ? a difference of day and night between the two, in favour of I2S. No additional conversion behind either, and both go into the SRC.

 

You could think like this :

a. No DAC in the base being synchronous can be made 100% immune to incoming jitter;

b. That will siple through right through all PLLs and SRCs;

c. What about that resulting in e.g. 40ps of jitter at the output, that being 40 times more as the intrinsic 1ps.

 

I don't say it is, nor do I claim that 40ps of jitter is audible let alone the difference with 1ps, but it might be a theoretical explanation for these kind of very audible differences. 40 times more = 40 times more, where of course the super low rate of 1 is to blame. I mean, might that be 35, mathematically the difference would be way more low.

 

In either case practical listening tests (or decent measuring if you like) should point out the real truth on the net result. For example, the Fireface might reclock to 800ps worth of jitter (or whatever number it exactly is), when you use an el cheapo firewire board the sound again becomes plain bad. Here too, onboard (mobo) firewire can't be used, because it just worsens the sound.

I could make it even more complicated if you like, because it doesn't necessarily take jitter alone to create a different sound quality, because the electrical influence just counts as well. So, I am not talking about noise incurring for additional jitter (which also will occur), but about influences on the power supply in the receiving apparatus. Remember, a DAC works with microvolts and less, and the stability of that has a great impact on SQ. This indeed can be fought with a glass connection, but the transmitter - or ar least the receiver introduces jitter again. When the DAC is really immune to that, no problem. But is it ? And if not, is it audible ? And if so, would it be audible in your system, for you ?

 

Peter

 

 

 

[Ashley James]

It's worth pointing out that several times over the last few years we've obtained Evaluation PCBs from the leading DAC manufacturers. These are the best that the company that designed the chipset can do with their own product and their goal is to persuade us or any manufacturer to use theirs instead of any other. We rigged up four or five with output levels carefully matched and asked a number of visitors who we knew were likely to spot differences, to shuffle between them using an IR handset. We did nothing to predispose them to anything and yet surprise surprise, all concluded, as we had, that any differences such as existed, were not enough to distinguish any one from any other. All agreed that they all sounded very good too.

 

Posts on this Forum support that view, to the effect that the differences between DACs are now quite or very small. This doesn't mean all DACs, because companies are still screwing up, but as a general principal it is established. You need to buy a good DAC and you'd be wrong if you were expecting big differences between competently implemented ones.

 

We've also see posts from others like us who stream from a variety of sources and do not hear any differences between them and yet people are still concerned that different digital interfaces may sound different. Digital interfaces are designed to have gross errors or no errors, so I'd assume that it doesn't matter which you use, but that it's a good idea to pick the most common and also to consider likely future trends.

 

Ash

 

PS. To avoid using a hat in my avatar, but to at least have some hair, I've selected a four times Great Grandfather from the early eighteen hundreds.

 

PPS. Just got this from a lurker who doesn't post and does know what he's talking about.

 

Ash,

 

Reading through that CA thread on jitter, it occurred to me that no one had bothered to check out the science.

So here it is. The effect of jitter on a sampled signal may be expressed as a signal to noise ratio :-

 

SNR(jitter) = -20*log(2*pi*tor*freq) dB

 

Where

tor = the jitter in seconds.

freq = frequency of the sinewave test signal in Hz.

 

So the degradation is proportional to frequency, and the worst case would therefore be freq=20kHz. However, as the ear's sensitivity has rolled off by at least 20dB at 20kHz, a more realistic test would be ~4kHz.

 

If we set the SNR target due to jitter at 100dB at 4kHz, so as to be masked by the 16 bit quantisation noise, plugging the numbers into the equation above gives us the absolute minimum audible jitter = 400ps. This is way above the level deemed to be necessary by the armchair experts. Maybe you could run it past Martin.

 

Chris.

 

 

 

[Wavelength]

Steve,

 

The Minerva and other firewire products don't use the Audio Stream method for getting data to the DAC. They have a driver that allows them to send blocks of data async to any clock then stream it out to the dac chip using the local oscillator.

 

You can do this on any platform even USB for that matter BUT it does require that you have a device driver and therefore is not a native device to the operating system.

 

Therefore I was correct in saying that Firewire does have an Async Audio Stream mode it just is not supported under any operating system.

 

Heck Steve you could get someone to write a bulk mode for TAS1020 and a device driver and get 24/192 today.

 

Thanks

Gordon

 

[The Computer Audiophile]

Hi Ashley -

 

"... armchair experts ..."

 

When it comes to jitter you can consider me The Grand PooBah of armchair experts. I know very little and prove it every day.

 

That said, I closed down the never ending jitter thread for many reasons. I think we've all had enough and need to let it rest for a while. Please respect this decision. We have far more enjoyable discussions ahead of us.

 

[Texas25]

Hi,everyone,

 

I want to buy a Imac, to use it as a music server. I am thinking of using a firewire into a word clock, AES out into my tactaudio preamp.

The word clock is a Brainstorm DCD 8. The DCD-8 uses the TC Electronics JET PLL Technology which is very different than the Direct Digital Synthesis (DDS) method (I mentioned it, because I absolute don know what that means :^D)

 

What do you think of this set-up?? Will this work on a highend level?? or am I missing something?

 

I own digital speakers so there is where the d/a convertion happens. (Focal SM8) modified

 

Any feedback will be great

 

Thx for the feedback

 

Alex