usb / firewire

K

kaliyuga

New member
I am looking for a low budget interface to use with a new PC, up to now I have used pci but will need to change, I am trying to weigh up the pros and cons of firewire vs usb 1. and usb2 ( been looking into the likes of the presonus firebox, alesis io2, m-audio 410)

I've also been looking at contoller/interfaces such as the m-audio ozonic, novation x-station, as I use cubase and reason alot and the combination would be useful.

I would like to know how usb and firewire compare, I was told usb only goes one way at a time, does that mean you can only hear what you are playing, or can you hear what you have already recorded,

what are the merits of each type of interface in practice, things like odd noises, interferance from other parts of computer like usb mouse, etc

any input appreciated.
 
Brzillian is right, this is a regular topic, but here's the long and short of it:

Don't worry about it, the manufacturers have already done the worrying for us.

If they build a box where USB just won't provide the spec'd performance, they use Firewire. If they build a box where f Firewire is more than enough, and USB will deliver just fine for less cost, then they use USB. Either way, they'll work fine and do everything they promise just fine.

Your biggest concern in this regard is making sure you have the ports needed on the computer end, and that the software you use can compensate for latency. As most compuers (other than the occasional laptop) come with plenty of I/O these days, and virtually all recorder/editor software for this racket can handle any latency problems OK, you're worries are minimal.

G.
 
I've had a lot better luck with firewire than USB. I owned a FireBox for a while and really liked it. Make sure that whatever interface you buy, that your computer is one that is supported. Like go to the Presonus web site and make sure you are using the right OS, CPU, etc.

A lot of times when people have trouble with an interface it's because they are using a system that is not recommended by the manufacturer.
 
SouthSIDE Glen said:
If they build a box where USB just won't provide the spec'd performance, they use Firewire. If they build a box where f Firewire is more than enough, and USB will deliver just fine for less cost, then they use USB. Either way, they'll work fine and do everything they promise just fine.
Click to expand...

There are caveats to that, though. It would be disingenuous to suggest that you can trust USB to always work if it is available.

First, USB by its nature eats a lot more CPU to do the same amount of work. If you're the kind of person who runs their computer to the breaking point with plug-ins, you do not want a USB interface. Also, you do not want to use USB for storage and audio at the same time, even if you are using separate USB busses.

There are severe bugs in some USB motherboard chipsets that can cause all sorts of clicks and pops. If you buy a USB interface, be sure to buy from a company with a good return policy, and if you run into this with your computer, send the interface back and buy a FireWire interface.

Finally, not all USB interfaces are created equal. You are much more likely to get a lousy USB interface than a lousy FireWire interface because there are so many more companies making USB interfaces, many of whom frankly don't know what they're doing. Many USB audio vendors buy pre-built device guts from some OEM manufacturer, rebrand it, possibly change the enclosure, and ship it. Such devices often suck massively. Of course, I'm mostly talking about things like USB speakers and other similar low-end consumer devices. The point is that I haven't seen that sort of crap hardware in the FireWire world (mainly for cost/economy-of-scale reasons).

My opinion is that USB tends to be more problematic than FireWire in general, but to be fair, many people use them and say they work just fine. Like I said, if you go USB, just buy from somewhere with a good return policy so that if it doesn't work, you can return it.
 
dgatwood said:
There are caveats to that, though. It would be disingenuous to suggest that you can trust USB to always work
Click to expand...
It would be equally disingenuous to suggest the same for any other transport protocol. There's always caveats in the form of system usage and configuration outside of the control of the manufacturer that can cause problems. You can run in via FW to ultra fast and wide SCSI RAID and still have dropouts because of something dumb in the rest of the computer.

The fact is I have never experienced or heard of or read of a problem with audio I/O via USB that was caused by the shortcomings in the actual USB protocol. I know or have discussed this with a LOT of people that use 2 track devices like the Tascam or the mAudio or the Focusrite or any number of other brands, *every single one of which use USB (some of them 1.1, even)* day in and day out, along with USB keyboard, mouse and external drive, and have never experienced a problem related to the USB protocol itself.

The simple fact of the matter is that if you want a 2 channel interface, you don't have a choice. It's USB or the highway. And the 2-channel boxes work just dandy. And if you don't want USB, and insist on FW, then you MUST get an 8 channel interface or greater (there may be a 4-channel one out there, but if so , those are few and far between.)

It's a non issue because we have no choice for any given number of channels.

G.
 
Southside, I hate to disagree, but in my own real-world experience I've had *a lot* more problems with USB devices than firewire devices. I refer to USB as the "unreliable serial buss". There are frequent issues with USB devices losing contact with each other, something that rarely if ever happens with firewire. I used to own a Digidesign Mbox and it was literally useless. I got rid of it and bought their Digi 002R, and no problems at all.

I'd never buy a USB audio interface. The Presonus FireBox worked great for me, and is priced around what many USB audio interfaces cost anyway.

USB is great for mouse, keyboard, cameras, things like that. But for audio, no way, not in my experience. It's all firewire and PCIe cards here for that.
 
SonicAlbert said:
Southside, I hate to disagree, but in my own real-world experience I've had *a lot* more problems with USB devices than firewire devices. I refer to USB as the "unreliable serial buss". There are frequent issues with USB devices losing contact with each other, something that rarely if ever happens with firewire. I used to own a Digidesign Mbox and it was literally useless. I got rid of it and bought their Digi 002R, and no problems at all.
Click to expand...
First I have to ask if the problem with your mBox (an iffy device to begin with) was actually due to inherant problems with the USB protocol or whether it was due to problems with the mBox itself?

My own experiences with several different USB audio devices (Tascam and Lexicon) are well documented as being entirely postive with zero issues in any way. Additionally I have a musician friend who home records via his Alesis USB mixer, has been doing so for over a year now with nary a single problem. Then there's a video editor associate of mine using a Tascam for some 6 months now; again no issues whatsoever after almost daily use. He even was able to record audio at the same time he was capturing hi-res video into a seperate app via FW. Suimultaneous. No dropped video frames on the FW, no audio pops, clicks, etc. on the USB-fed recording. Beautiful.

Then there's the hundreds of small home recordists on this board with 2-channel USB interfaces that use them daily without issue. Do you really think that if USB were *that* unreliable that it wouldn't have been all over these boards two years ago that those little USB interface boxes were nothing more than a crapshoot? But no, instead we have people here daily who use them happily and constantly recommend them to each other.

Sure there are problems on occasion, and some boxes have worse repuations than others (I personally shun mBox and Presonus myself, but that's more a personal bias related to more general issues than anything else.) But the vast majority of problems are due to issues other than the inhrerant technical shortcomings in the USB design itself.

And I can also testify that after working with FW video for some 10 years now in both a production and a development environment, that the rate of issues with FW is not necessarily any lower than it is with any other communication protocol, and that 9 times out of 10 the issue is because of configuration problems outside of the communication path itself.

Now, both of my FW-based video production systems have worked flawlessly for me since 1999 when I went out on my own, not a single dropped frame or lost transport command even once during that entire time. Wonderful. But that's because I have my system configured properly and I use it properly. OTOH, I also know some folks who couldn't capture a clean high res video with FW to save their lives. But I don't blame FW for that; that's because they don't have their system and operation handled properly. It's usually the same thing with most USB devices; it's not necessarily the USB transport path's fault that things aren't running right, but rather that there's something or someone else getting in the way.

I have nothing against FW (I love it), nor do I have any interest in defending or pushing USB for any particular reason. I just find the amount of both mis- and dis- information regarding USB - usually coming from those staunch Apple fans or those so buried in the techgeek trees that they don't see the forest from the end-user perspective - generate and perpetuate the kind of myths that sales people love and that bear little resembelance to reality.

It just seems ridiculous to me to tell someone that their only choices in audio are a Firebox from a company which I would not normally recommend anyway, or that they have to go 8 channels to get something relaible, just because to do otherwise would expose them to the evils of USB, when there are millions of people around the world using those USB devices just fine every day. That just smells a whiff of FW snobbishness in the air to me. (And NO, I'm not talking about you or dg, I'm talking in general :).)

G.
 
Last edited:
I'm only speaking from personal experience with a limited amount of gear. I've have a fair amount of dropped communication over USB, but none with firewire. To the point that I simply will not trust USB with any of my audio data transfer.

That's just me and my experience. I don't doubt that there are others that have good results with USB. The key is finding what works with *your* system. That can take some trial and error, and troubleshooting.

I've also A/B'ed firewire versus PCI/PCIe setups, and for whatever reason the PCI based setup sounded better. All equipment from the same manufacturer, MOTU, and the firewire interface is the newer and theoretically should sound better. But no, I felt a clear enough difference that I stuck with the PCIe based rig as my primary one.

I think there must be jitter issues with firewire and USB transfers of audio. I don't believe it's as *clean* or what have you, as a proprietary format transfer over PCI. Not sure I said that right, but you get the idea.
 
SonicAlbert said:
I've also A/B'ed firewire versus PCI/PCIe setups, and for whatever reason the PCI based setup sounded better. All equipment from the same manufacturer, MOTU, and the firewire interface is the newer and theoretically should sound better. But no, I felt a clear enough difference that I stuck with the PCIe based rig as my primary one.

I think there must be jitter issues with firewire and USB transfers of audio. I don't believe it's as *clean* or what have you, as a proprietary format transfer over PCI. Not sure I said that right, but you get the idea.
Click to expand...
I know that I have read - or maybe I should say have tried to read ;) - several articles on FW and jitter. I must admit that I have never quite been able to fully wrap my head around the deeper technicalities. I was of the general understanding (I freely admit that this may be wrong) that FW did/does have inherent jitter problems, but that on a (higher? later?) layer of the protocol that they have been largely, though not necessarily completly, addressed.

Your experience with the A/B of FW vs. PCI is interesting. I have not heard or read about that specific subject before. I have a gut suspicion that as with almost everything else, there can always be other factors (quality of PCI or FW interface design, for example) that may creep in and fudge the results. Similar to the old samle rate debates where one mfr's 44.1k may sound better than another's 96k (and vice versa), but largely for design reasons other than the sample rate itself.

Nonetheless it makes me feel good that my good ol' old-fashoned MOTO 2408 (before they had to put a mk series at the end of the model number :) ) with the old PCI-324 still has a chance ;) :D (I DO need to replace it nonetheless; it's pure age is obsoleting it on a couple of levels :o.)

And yeah, I'm sorry about your experiences, Al. I don't doubt them, either, or blame you for being soured by them. But you're a fair and observant, even-handed guy on this board and have been here for a long time; based upon the general overall response, your experiences don't seem to be the rule with these boxes, right?

G.
 
SouthSIDE Glen said:
Nonetheless it makes me feel good that my good ol' old-fashoned MOTO 2408 (before they had to put a mk series at the end of the model number :) ) with the old PCI-324 still has a chance ;) :D (I DO need to replace it nonetheless; it's pure age is obsoleting it on a couple of levels :o.)
Click to expand...

My comparison was between a MOTU 2408mkII and Traveler. I had just bought the PCIe 424 card so I could use the 2408 with my new intel mac, and I wanted to run a comparison between it and my Traveler, which I use for live gigs. The thought being that maybe I'll just go with the Traveler for everything.

The difference was extremely subtle, but I decided to hang on to the 2084mkII/308 that I use. The PCIe interface just sounded a bit better to me, a little wider and deeper, just a little nicer. Those characteristics are common indicators of minor jitter problems, i.e. a narrowing of the stereo field and loss of depth.
 
SonicAlbert said:
The PCIe interface just sounded a bit better to me, a little wider and deeper, just a little nicer. Those characteristics are common indicators of minor jitter problems, i.e. a narrowing of the stereo field and loss of depth.
Click to expand...
Agreed...and intersting info on the comparison. Thanks! :)

G.
 
SonicAlbert said:
My comparison was between a MOTU 2408mkII and Traveler. I had just bought the PCIe 424 card so I could use the 2408 with my new intel mac, and I wanted to run a comparison between it and my Traveler, which I use for live gigs. The thought being that maybe I'll just go with the Traveler for everything.

The difference was extremely subtle, but I decided to hang on to the 2084mkII/308 that I use. The PCIe interface just sounded a bit better to me, a little wider and deeper, just a little nicer. Those characteristics are common indicators of minor jitter problems, i.e. a narrowing of the stereo field and loss of depth.
Click to expand...

Thatt's definitely not a PCIe vs. FireWire issue. Once you're past the converters, it's just a ring buffer full of data, and no change in bus protocol is going to change that.... The bus can cause dropouts and glitches, but it can't have any other effect on the sound quality. Data is data.

What you're experiencing is probably either a diffference in the quality of the converters or, more likely, in the quality of the analog electronics that feed the converters.
 
dgatwood said:
Thatt's definitely not a PCIe vs. FireWire issue. Once you're past the converters, it's just a ring buffer full of data, and no change in bus protocol is going to change that.... The bus can cause dropouts and glitches, but it can't have any other effect on the sound quality. Data is data.

What you're experiencing is probably either a diffference in the quality of the converters or, more likely, in the quality of the analog electronics that feed the converters.
Click to expand...

I love my Mbox 2 via USB with my Macbook pro and a good OWC firewire audio disk. No problems with latency while recording using the magical "mix" knob on the box.
 
dgatwood said:
What you're experiencing is probably either a diffference in the quality of the converters or, more likely, in the quality of the analog electronics that feed the converters.
Click to expand...

No, I eliminated the converters from the comparison by using the same converters for each. I ran both digitally to my Presonus Central Station, so it had nothing to do with converters. I also clocked both off my UA 2192, so the clock was the same.

It really was a bit of a mystery why PCI sounded better, but it did. After eliminating the clock and the converters, the major factor that different is PCI versus firewire.
 
SonicAlbert said:
No, I eliminated the converters from the comparison by using the same converters for each. I ran both digitally to my Presonus Central Station, so it had nothing to do with converters. I also clocked both off my UA 2192, so the clock was the same.

It really was a bit of a mystery why PCI sounded better, but it did. After eliminating the clock and the converters, the major factor that different is PCI versus firewire.
Click to expand...

Wait... are you saying you used an outboard converter connected via ADAT and got different results depending on which hardware you fed it into?

Playback path differences? Beyond that, it's either all in your head or you've found a bug in MOTU's drivers. ADAT data is just raw data and there cannot be any difference in the data being passed through ADAT no matter what device is doing it.
 
dgatwood said:
Wait... are you saying you used an outboard converter connected via ADAT and got different results depending on which hardware you fed it into?

Playback path differences? Beyond that, it's either all in your head or you've found a bug in MOTU's drivers. ADAT data is just raw data and there cannot be any difference in the data being passed through ADAT no matter what device is doing it.
Click to expand...

That's yes to your question.

No playback differences and it's definitely not in my head. I actually *wanted* to get rid of my PCI setup, but once I tested both I felt I couldn't. I own and use both systems, so it's not like I need to justify a purchase to myself.

The drivers are MOTU's in both cases. Same installer installs the same version of the driver, just one is PCI and one is firewire.

It's a bit of a puzzlement. I think the issue occurs before the hardware, in other words, the streaming of the data to the interface from the computer. I think PCI handles that differently than a firewire device. Like maybe there's a buffer in there, or a bigger buffer. Something. Like I said, it's puzzling.
 
Al,

When you described the difference as being in image focus, that is a classic description of a potential problem somewhere in the clocking/timing/data jitter arena. As you were using the same (very reliable) master clock, I'd agree that jitter being introduced in the signal timing outside of the clock itself is a real suspect to be looked at.

This issue pushes the envelope of my own personal undertstanding. I still haven't fully absorbed Katz's - or is it Lavry's, I forget - thesis tht external clocking can cause it's own set of problems, and until I do, I have my questions about that;.but there is that factor, perhaps.

But I do know that I have read in the past articles about how FW can introduce it's own jitter into the data. This is several years ago that I read about this, so I don't remember details about the mechanism that caused it, nor am I sure about resolutions since then, but I have it in my mind for some reason that somewhere along the line this may have been partially (if not fully) addressed...though I won't swear to that.

I'm kinda bogged down in a different technical thread right now, so I'm not sure if I'll have time soon to find references to these subjects to try and absorb what's up. But it might behove someone out there to do a searches on "firewire jitter" and maybe "master clock controversy" to see what The Big Boys have to say.

G.
 
SonicAlbert said:
That's yes to your question.

No playback differences and it's definitely not in my head. I actually *wanted* to get rid of my PCI setup, but once I tested both I felt I couldn't. I own and use both systems, so it's not like I need to justify a purchase to myself.

The drivers are MOTU's in both cases. Same installer installs the same version of the driver, just one is PCI and one is firewire.

It's a bit of a puzzlement. I think the issue occurs before the hardware, in other words, the streaming of the data to the interface from the computer. I think PCI handles that differently than a firewire device. Like maybe there's a buffer in there, or a bigger buffer. Something. Like I said, it's puzzling.
Click to expand...
It's basically streaming the equivalent of an AIFF file across the wire, so there's not any good explanation for any difference. Were you clocking the Presonus through the two MOTU interfaces by any chance? If so, you might be getting clock jitter from a poor PLL circuit in the interface on the way to the Presonus. Beyond that, my guess remains "software bug"....

Sent from my iPhone.
 
You must log in or register to reply here.
Back
Top