Dongle replacement.

Alright - I looked around for a bit, and I'm finding all the stuff I thought I remembered again - the difference is the architecture of firewire was originally built for high-performance blah blah, USB was originally built for simple stuff, and although it got a big upgrade in performance, it's architecture still relies on it's original design of a master/slave system where the host computer's CPU bears the brunt of the load of the processing required, whereas firewire devices share the load amongst themselves. In addition, it has been shown (well... alleged, at least. I can't actually find a conclusive, believable experiment, which kinda sucks) numerous times that firewire outperforms USB in every way there is, including data transfer, despite the fact that USB has a technically wider bus (or something like that).

The part that really matters, though, is the isochronous guarantee of firewire that a device can guarantee it will be able to transmit XX amount of data without interruption, whereas the isochronous guarantee of USB is much, much lower. I can't really find any authoritative sources that just jump out and scream "I prove that firewire is better!" - but here's some good places (well, the best I could find, anyway) to start:

http://en.wikipedia.org/wiki/High_Definition_Audio-Video_Network_Alliance

http://1394ta.org/consumers/FAQ.html

I find it odd that I can't find a clear, and authoritative source that states that firewire is better. On the other hand, I didn't see one single (non-authoritative) claim that USB was even a 'good choice' or a 'good idea' for audio work. There's definitely a general consensus against it...weird. There's several places that seem almost-athoritative that show a chart from an Oxford-Semiconductor experiment, but not on single link to the original research article, or published findings or anything of the sort - lame.
 
Wow, comprehensive posts indeed. Thanks. Originally when I started to think about setting up a DAW my uneducated gut told me that Firewire would be better than USB. A tech guy in a shop recommended the Edirol UA25EX and when I asked "What about Firewire? I thought it would provide faster and more accurate data transfer", his reply was confused. All this guy does is set up DAWs and sell all the gear so I kind of just trusted him. I guess I should have questioned him further but we didn't have time for the conversation. Maybe it is a cost thing, perhaps USB interfaces are just cheaper? I don't need more than two channels and want to spend around what the Edirol UA25EX costs. What interface would you recommend?

I just compared the Edirol FA-66 Firewire interface with the UA25EX. I think the UA25EX is newer and may have more features. It has a limiter/compression control on the front that the FA-66 does not. It also has aground lift switch on the back which I think helps reduce noise. The FA-66 costs a bit more.

One thing the tech guy said was that he has sold a lot of the Edirol UA25EXs and in his experience they just work. I have read so many posts about compatibility issues and "ghosts in the machine" when setting up DAWs using laptops and interfaces that his experiences with the Edirol were attractive to me.
 
Last edited:
enuenu said:
Maybe it is a cost thing, perhaps USB interfaces are just cheaper?
Click to expand...
They certainly are - they also have a reputation as not being...well, serious. Kind of like taking a nerfball gun into war is the only analogy I can think of. Sure - it's a gun - it fires a projectile and everything - hell you might even get lucky and kill somebody with your nerfgun, but even if the odd fluke happens here and there, it's still not really the right tool for the job in the opinion of the overwhelming majority of people in the field - and that says something, right?

enuenu said:
I don't need more than two channels and want to spend around what the Edirol UA25EX costs. What interface would you recommend?
Click to expand...
I actually used to have a little Edirol thing, and it worked ok for recording, I guess - but the latency was HORRIBLE HORRIBLE HORRIBLE HORRIBLE HORRIBLE HORRIBLE HORRIBLE HORRIBLE HORRIBLE!! I mean like quarter-note delay at around 150 BPM. This has also been the case with every other USB device I have ever seen (which is really just a bunch of different flavors of mboxes, to be fair) - you could record a track or two that didn't have any obvious, flaming "I WAS RECORDED ON A PIECE OF SHIT USB DEVICE" characteristics to them (such characteristics don't exist - there is no "USB sound" or "Firewire warmth" :D), but the owners always had a hell of a hard-time recording themselves if they needed to monitor in real-time. That's just the nature of the USB devices, apparently - an unusably horrible latency.

enuenu said:
PS - One thing the sales guy said was that he has sold a lot of the Edirol UA25EXs and in his experience they just work.
Click to expand...

I think you had a typo - fixed it for ya!

Seriously, though - almost every device out there uses the same chipsets and building blocks. For example, it's not like there's actually hundreds of unique designs for firewire audio devices... there's like two. Ok, maybe there's like 5, but I can only think of 1 to be honest, and that's the bebob format (or whatever it's called - the thing that a TON of *apparently different* firewire devices actually have in common, and thus have the *same driver* for use in Linux, hahahahaha....)

People who have problems are just people who have problems - they don't understand their shit or they break it. Simple as that, really - a wizkid computer geek is never going to encounter a device that he can't make work (by RTFM if nothing else), but his grandfather, on the other hand, is going to think that everything newer than the facsimile machine is fuckin broken because he can't figure it out - then grandpa posts his horrible reviews on sites, right next to the great review from the kid, and the confusion mounts. Of course, there are exceptions to this, but it really holds true...really.

Any tech-savvy person will tell you how many times he's been told "The internet's broken!" because some dipshit deleted his browser shortcut off his desktop, when, in fact - the internet does not ever "break". Imagine if people read reviews of "the internet being broken" before they picked a browser - it would be chaos because there would be tons of vile, horrible reviews about *whatever browser happened to be installed on the computer that the dipshit sat down at* as being the cause of all kinds of diabolical problems. This same, sad-but-true phenomenon occurs with reviews of poorly understood technology all the time.

I'll go look and see how much that edirol thing is, and see if there's anything that's not USB based with the same features - be back later.
 
http://pro-audio.musiciansfriend.co...erfaces-computers-pro-audio?N=100001+344253+9

There's quite a few firewire devices on there for around the same price as I could find the Edirol thing you mentioned - all the usual suspects are about the same price -

M-Audio: 410, and apparently the 1814 got a pricecut - it's now the same price as a 410, and has way more ins and outs, Firebox, Firepod

Presonus: Firebox

Focusrite: Saffire

These are all around the same price, and they will all probably perform/sound/sing/dance/tell jokes on about the same level with each other. I think the apparently-recently-reduced-in-price 1814 may be the best deal of the lot at the moment...
 
Thanks so much. I need the brutal truth. I sense it would be worth my while spending a bit more on an interface. I think I might up the budget to around USD$350-$600. I am sensing the investment may be worth it. I don't need the channels but I want the quality and lowest possible latency. I don't want to go to all the trouble of learning a sequencer package, buying a laptop and mic and record some real gold to find I am being let down by a toy interface and maybe have to re-record at a later date. I am a digital music newb but am very serious about quality. The stuff I record now will be in the vault for a long time.

EDIT - Done a bit of research since this post. The latency issue looms large. I want to be able to lay down a beat and then while listening to it record my electric guitar. I want what I am playing to be recorded exactly in alignment with the beat I am hearing with zero latency. Does zero latency monitoring achieve this? If I can't have a rhythm section playing the song at the same time I am while I record my guitar, I can't see how any of this stuff can work. I have read about people having to play their notes "early" to compensate for latency issues. Feel is everything in music. This business of having to artificially play out of time is crapola, not interested in that at all.
 
Last edited:
enuenu said:
This business of having to artificially having to play out of time is crapola, not interested in that at all.
Click to expand...

Yep, total crapola, indeed. As long as your PC isn't a POS, any decent firewire or PCI device is going to have ~8 ms latency or less, so don't worry about it too much, really. USB devices, on the other hand, I have personally worked with on the same computers as with 8ms firewire devices, and they have had over 100 ms latency or more - total shit, unusable. 8 ms is an undetectably short amount of time - I think humans can only really notice a difference down to about 20 ms (if I recall correctly).

As for "Zero latency", I'm not sure what context you read that in, but I can tell you that truly zero latency monitoring is only acheivable with direct monitoring, ie: not passing through the software - in which case it wouldn't matter what interface you were recording or how delayed it was, just don't listen to the input stream in the software, and mix the playback with the direct monitoring feed - To further complicate things, though, you can put ASIO devices into ASIO Direct monitoring, which still means they don't really pass through the software (you can't use EQ and effects on them) but you can usually control their volume, at least, as it plays back in the mix using your software - this won't really be zero latency, I don't think...but it will be so close (like less than 1 ms, I think) as to be completely negligible.

Really, though - the only way that is nice is just having extremely low latency that does pass through the software - it's awesome, with this setup you can throw a reverb plug on a singer's voice in (apparently) realtime (8 ms or less) and get better performances from him, etc. while still printing a dry take to disk. Go for that, don't be ghetto or you'll wish you hadn't been later. :p
 
Thanks again. I would have to do some experiments to see what sort of latency the ear can pick up. Sorry, but I have a penchant for perfection which can become a little manic. More experienced operators would already know what can be perceived and what can't. I just can't help but think you really want your notes where you initially put them. Maybe this is a reason some musicians still use tape.

Can Reaper do the ASIO thing you mentioned to reduce latency? Do only certain interfaces do the ASIO thing? Forgive the wording of these questions. I am trying to paint a room while blindfolded that I will live in when my blindfold is removed.

Maybe I can play my guitar along with an already recorded rhythm track and then have the software automatically move the guitar recording back in time an amount equal to the latency that eventuated. Can this be done?

Maybe the latency value fluctuates while recording. If this is the case the software would have to not only move the guitar track back but also stretch or shrink it in places. Complicated. In this way all latency would be removed and the initial performance could be played naturally.

Can the software dynamically measure the latency it and the hardware are creating?
 
enuenu said:
EDIT - Done a bit of research since this post. The latency issue looms large. I want to be able to lay down a beat and then while listening to it record my electric guitar. I want what I am playing to be recorded exactly in alignment with the beat I am hearing with zero latency. Does zero latency monitoring achieve this? If I can't have a rhythm section playing the song at the same time I am while I record my guitar, I can't see how any of this stuff can work. I have read about people having to play their notes "early" to compensate for latency issues. Feel is everything in music. This business of having to artificially play out of time is crapola, not interested in that at all.
Click to expand...

Take a look at this:

http://www.americanmusical.com/Item--i-PHO-H12FWMKII-LIST

This is what I use. The thing I like most about it is I can monitor what I'm tracking in realtime with effects (like reverb) but the signal going to the computer is dry. The a/d signal taps off after the trim knob and inserts but before it goes to the eq, aux's and fader sections. When I'm tracking, I'm listening to the previously recorded tracks while recording the current one. Latency and buffer sizes are not an issue. I like a little reverb to help me sing, but of course, I don't want it on the track until I'm ready to mix down. During mixdown, I add effects, compression, eq, etc to suit the song. You don't know what you'll need when tracking.

The Alesis firewire mixer is similar with one major difference. The signal is sent to the a/d converter AFTER the channel strip (but not the aux's) so you get eq and fader adjustments in the signal. Any changes there is part of the recording. I don't like that feature.

Hope this helps.
 
enuenu said:
Can Reaper do the ASIO thing you mentioned to reduce latency? ....

Maybe I can play my guitar along with an already recorded rhythm track and then have the software automatically move the guitar recording back in time an amount equal to the latency that eventuated. Can this be done?

Maybe the latency value fluctuates while recording.

Can the software dynamically measure the latency it and the hardware are creating?
Click to expand...

I have no experience with Reaper, but I would imagine that it uses ASIO interfaces.

I'm pretty sure all DAW have latency adjustments. It's not a dynamic function. You set one value, whatever that might be for your particular application/ setup, and you leave it alone. The latency adjustment will automatically move your track to align with previously recorded tracks.

I think some DAW's might be able to measure latency, but my version of cubase does not (Cubase SE3). Or maybe there is a download somewhere that measures latency. I remember reading where people had actual measurements.


Have fun,
 
enuenu said:
I just can't help but think you really want your notes where you initially put them. Maybe this is a reason some musicians still use tape.
Click to expand...
Naaa... The only reason anybody still uses (EXPENSIVE) tape is their "taste" (because they believe in some magical voodoo crap about it sounding "warm".)
enuenu said:
Can Reaper do the ASIO thing you mentioned to reduce latency?
Click to expand...
Yes
enuenu said:
Do only certain interfaces do the ASIO thing? Forgive the wording of these questions. I am trying to paint a room while blindfolded that I will live in when my blindfold is removed.
Click to expand...
Yes, only ASIO-compliant devices can do this. Look in the spec-sheets if you want to make sure, but I think it would be extremely difficult to find a device made in the last few years that is not ASIO-compliant.
enuenu said:
Maybe I can play my guitar along with an already recorded rhythm track and then have the software automatically move the guitar recording back in time an amount equal to the latency that eventuated. Can this be done?
Click to expand...
Yes, this is exactly how it works! It takes no input, or even any thought, from you to make this happen, it just happens.
The only meaningful issue of latency is about nothing other than being able to hear the notes you are playing/singing, as you play them, within the mix you are hearing in your headphones - every other aspect of latency is automatically compensated for in every DAW application there is (except for ProTools LE/M-Powered, if I'm not mistaken, which are purposely crippled by DigiDesign in this way).
enuenu said:
Maybe the latency value fluctuates while recording.
Click to expand...
Like Chill said: It doesn't. If you have exactly 100 ms latency in your system, you have exactly 100 ms latency in your system, end of story. With Steinberg products, when you start up the software, a little thing will pop up if it determines there's a new device connected, and ask you to run a test on it to determine the system latency - you click "ok run the test", it runs the test, and sets your latency. That's all there is to it.

I wrote up an in-depth, but possibly-not-exactly-scientifically-accurate description of how I believe this works that you may find useful as a non-authoritative reference. But I'm omitting it in the interest of trimming down this post. PM me if you want it.

enuenu said:
If this is the case the software would have to not only move the guitar track back but also stretch or shrink it in places. Complicated. In this way all latency would be removed and the initial performance could be played naturally.
Can the software dynamically measure the latency it and the hardware are creating?
Click to expand...
There's no time-distortion/stretching trickery involved in a properly working system. The only time the actual latency playback time will change is if you throw a plugin on the monitor-out bus that is so CPU-intensive that it introduces it's own, tangible delay, in which case - don't use that plugin on the monitor-out bus while recording - simple as that.

I can understand, and even appreciate, your concern with the exact workings of the stuff, but here's what you really, practically, need to know: 8 ms is a fairly consistent latency amount (in my experience, at least) with firewire devices in the price-range we are discussing here, provided that the computer in use is not a 40 year old vaccuum tube machine (really anything newer than a P3, lol). Nobody - NOBODY - can tell an 8 ms difference, so it's a non-issue. Anything more in-depth than that, and I think you might have to go to engineering school to *truly* understand 100%.

Consider this:
Traveling 1 foot per millisecond = 304.8 meters per second.
The speed of sound = 340.29 meters per second
Thus, it is a *conservative* rule of thumb that sound travels 1 foot per millisecond

If you are so sensitive to time-differences, that jamming out with a drummer who is 8 feet away from you is impossible because the time it takes for that sound to get to your ears throws you off-beat, then you're an amazing freak of nature, really, and you should go make some money for being a labrat or something, seriously... :p
 
TyphoidHippo said:
Nobody - NOBODY - can tell an 8 ms difference, so it's a non-issue.
Click to expand...

This should be clarified that I mean that nobody will be thrown off-beat by this - it has been known to happen that some singers hear a weird sort of "out of phase" kind of sound because of the difference in polarity between the soundwave coming from their throat, traveling through their bones/body to their ears, and the signal that's a few ms behind that they are monitoring in their phones... it's rare, but it happens.

It's also a non-issue for the most part (especially since no amount of worrying or pre-planning will change it), but it occurred to me that saying NOBODY can tell any difference wasn't entirely true because of this (rare) phenomenon - as for being thrown off beat, it still stands that NOBODY will have that issue with 8 ms of latency.
 
Thank you very much, you have taught me a lot. I will start shopping for a laptop and firewire interface. I will check out the interfaces you recommended. Looking to spend USD$400-$550 on an interface now.

The M-Audio 1812 looks interesting. I guess with this device you also have the ability to use Pro Tools if you ever so desired. It requires a 6 pin Firewire connection so I will have to see whether many laptops have this connectivity.
 
Last edited:
The 6-pin connection is only needed if you want to run it on bus-power. You can use the wall-wart power adapter in combination with a 4-pin to 6-pin adapter cable, if need be - ;)
 
enuenu said:
Hard to choose an interface. The Edirol FA-66 is looking good now. Lot's of people seem to complain about M-Audio gear. Suffering from paralysis by analysis.
Click to expand...

Well, I hate to simply repeat myself, as that won't be of much help for you... so here's a little insight into M-Audio - I have, personally, used the M-Audio 1814 and also the ProjectMix with zero problems on many, many different computers...running all manner of operating systems - with all kinds of different software. Hell, I even had to use a Mac on one extremely rare occasion of my laptop being out of commission for a live-recording - I don't see how people mess up such straight-forward devices........... Then again, I had to *literally* tell people that they must first turn their computer on before their "internet will work" during my brief employment as a tech-support-guy for a few different ISPs.

My main point here is that I strongly advise to read reviews from people that you can talk to, read about, or at least do *some* kind of reference check on -

Ted Perlman is a fantastic example - this guy has recorded #1 Billboard Hits on M-Audio hardware - http://tedperlman.com see for yourself - Look at his (friggin AMAZING) list of credits while you're there. Everyone from Bob Dylan to 50 Cent. Why would a guy like that use M-Audio hardware? Pretty simple answer, really... because it works. If you find it hard to identify with a grammy-winning producer you can listen to a few tracks I did with M-Audio hardware here: http://myspace.com/casperpro - You can also see my humble little list of miscellaneous credits on there (it has a few names on it that you just might recognize... hopefully at least one of them...) :D

Am I saying that M-Audio is some amazing secret-weapon? No... I only suggested it because of the pricecut - which made the 1814 a really good deal at the moment with 8 usable converters, and two usable preamps for $299... Like I said before - in general, M-Audio and any of the other devices around the same pricerange (probably even the one you mentioned there) are going to perform/sound/sing/dance/do the macarena on about the same level -
 
Thanks again. I guess you can find a negative review on anything if you search. I have seen the 1814 advertised online here in Australia for just over AUD$700, we usualy get ripped off blind when buying anything to do with music here. I might just get my Canadian relatives to send me one unless I can find an online US store that ships internationally.
 
Damn.. That's just too much - If you can get something similar for less, go for it, heh.. I didn't realize you were in Australia.
 
Just to pitch in my 2 cents re M-Audio. I used to run off a Delta 1010 and thought it was great but the capacitors fried and the unit developed terrible buzzing. From trying to find out what the problem was I found quite a few similar stories online. Now I use an Echo Layla 3G and it is a lot better. The sound is noticeably more 'alive' than with the Delta - which I thought sounded fine until I A/B parts recorded direct through the same signal chain. Anyway, just thought I'd pitch that in there - not sure if it helps at all though!
 
Superhuman said:
Just to pitch in my 2 cents re M-Audio. I used to run off a Delta 1010 and thought it was great but the capacitors fried and the unit developed terrible buzzing. From trying to find out what the problem was I found quite a few similar stories online. Now I use an Echo Layla 3G and it is a lot better. The sound is noticeably more 'alive' than with the Delta - which I thought sounded fine until I A/B parts recorded direct through the same signal chain. Anyway, just thought I'd pitch that in there - not sure if it helps at all though!
Click to expand...

Yea, man - that's useful insight if I've ever seen it :) How old was yours when the caps went out? I've never seen that happen to anything other than poweramps...
 
You must log in or register to reply here.
Back
Top