COCKOS
CONFEDERATED FORUMS
Cockos : REAPER : NINJAM : Forums
Forum Home : Register : FAQ : Members List : Search :
Old 07-21-2008, 03:27 PM   #1
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default Edirol FA-101/ FA-66 Asio Problem: Need help

Hello there!

I just bought an Edirol FA-101 Audi interface (identical to FA-66 with more I/O). I bought it mainly because some experienced user told me its on of the most lower stable round-trip latency. (specifically the well known Jim Roseberry). Putting all sliders to the minimum in the control panel, including the 'FireWire buffer" slider, the ASIO shows 2.9 ms. Not bad here.
Then I run the Centrance's Latency Test Utility:
http://www.centrance.com/downloads/ltu/

basically it measures the REAL WORLD ROUND TRIP LATENCY, which usually is:
ASIO is (3ms)
ASIO out (3ms)
A/D (1ms. in general)
D/A (1ms. in general)
Hidden or safety buffers (i was told Edirol doesnt have those)

At 44.1khz I get almost 13ms. of latency! Taht is WAY TOO MUCH for live EFX processing.
I would like to ask all FA-101 and FA-66 users out there a simple favor:

Run the Latency Test Utility and post your results. Many, including Jim Roseberry, have said Edirols work fine.
You would realize how well or bad your interface performs in your current system. Try it with minimum buffer settings. AAMOF, all users who monitor thru software should run that program to see 'the truth' about the interface. Just plug the output to the input, and follow the instructions. It sends a pulse and measures the time t receive again.

I appreciate your help!

Danny Bullo
bullo is offline   Reply With Quote
Old 07-21-2008, 03:36 PM   #2
stratman
Human being with feelings
 
stratman's Avatar
 
Join Date: Oct 2007
Location: Woking, England
Posts: 2,685
Default

I did try Centrance's Latency Test Utility and couldn't get it to work. Do you have any tips?

I'm not sure if this is any use but, with my FA-66, Reaper reports an Input/Output latency of 4.6/6.4ms (i.e. a total of 11ms) at the minimum settings.

Pete
stratman is offline   Reply With Quote
Old 07-21-2008, 08:25 PM   #3
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

strat man:
Thank dude!

Just connect the output 1 to input 1 (not the pres, jsut line level out to line level in).
And try different settings at the latency test utility. (-12dB,-6dB, etc)
Anyway, my test shows similar results as yours with the minimum latency (around 12 ms.) which I consider to be too high for real time monitoring thru software.

regards,
danny
bullo is offline   Reply With Quote
Old 07-22-2008, 06:30 AM   #4
bullshark
Human being with feelings
 
bullshark's Avatar
 
Join Date: Jan 2006
Location: traîne mes guêtres en Québec...
Posts: 5,390
Default

I get 10 ms round trip latency @ 44100 on my FA-101 (444 samples) and it's honestly the best I could find for Firewire solution, short of the RME stuff of course (which is 5 time more expensive here, and out of reach for me).

If that's too much for you, you might consider a PCI solution.
bullshark is offline   Reply With Quote
Old 07-27-2008, 05:05 PM   #5
frog_jr
Human being with feelings
 
frog_jr's Avatar
 
Join Date: Sep 2007
Location: Here at the Western World
Posts: 1,253
Default

I don't see any indication in your post of which 1394 chipset you have on your computer. I bought a FA-66 for my laptop (Thinkpad T61) and had to purchase a PCMCIA adapter since the onboard 1394 chip's performance was utterly crappy. There's a lot of talk that you MUST have the TI chipset on your computer or you don't get good performance with the Edirol boxes. Adaptec seems to be a goto for that.

Oh, incidentally, Reaper has worked great with the FA-66 otherwise. Sorry if I'm too lazy to download, install, and run your benchmarking software.

Last edited by frog_jr; 07-27-2008 at 05:09 PM.
frog_jr is offline   Reply With Quote
Old 07-28-2008, 07:30 AM   #6
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

Thank you for your response.

I DO PERFECTLY know about the chipset problems, that the TI is the best for audio bla bla bla...

That is not an issue at all. In the wort case scenario, an audio interface wont work with a crappy chipset or you get a tons of dropouts.

What I'm talking about is about Round-trip latency.
Please, check the software I posted. It requires NO INSTALLATION.
It explains exactly what it does and the truth about latency.

It will take you 1 min. Please post your result with the minimum asio buffer latency.

Regards,
Danny Bullo
bullo is offline   Reply With Quote
Old 07-28-2008, 07:47 AM   #7
bullshark
Human being with feelings
 
bullshark's Avatar
 
Join Date: Jan 2006
Location: traîne mes guêtres en Québec...
Posts: 5,390
Default

I think you should bring your laptop to a dealer and make the same test using a few different Firewire interface from different makers. My guess is you'll be surprised how well the FA-101 fare when compared to the competition.
bullshark is offline   Reply With Quote
Old 07-28-2008, 10:59 AM   #8
frog_jr
Human being with feelings
 
frog_jr's Avatar
 
Join Date: Sep 2007
Location: Here at the Western World
Posts: 1,253
Default

Hey guys, thx for the replies. I am in fact satisfied w/ the FA-66. I knew I'd need the PCMCIA adapter before I went in so I purchased it before I got the laptop. Bullo, I'll have your results in a few minutes.
frog_jr is offline   Reply With Quote
Old 07-28-2008, 11:17 AM   #9
frog_jr
Human being with feelings
 
frog_jr's Avatar
 
Join Date: Sep 2007
Location: Here at the Western World
Posts: 1,253
Default

OK, results:

(Everything set at the lowest of course)
At -18db Measurement results: 476 samples / 10.79 ms
At -24db Measurement results: 474 samples / 10.75 ms

Howzzat?

ps, I tried different volume settings on the front vol and the back input level. The 10.79 only happened once; it was otherwise at 10.75. And BTW, I'm getting pops now that I've got the buffer so low, which I've never done before. Gotta turn it back up.

Last edited by frog_jr; 07-28-2008 at 11:40 AM.
frog_jr is offline   Reply With Quote
Old 07-28-2008, 11:45 AM   #10
stratman
Human being with feelings
 
stratman's Avatar
 
Join Date: Oct 2007
Location: Woking, England
Posts: 2,685
Default

Ah, seeing frog_jr's settings, I finally got it to work when I set my sensitivity lower.

I just got this result from my FA-66.



12.59ms is a bit higher than Reaper reports (see my earler post). I wonder why that is?

However, these days I use my Line 6 Toneport UX1 and ToneDirect monitor using the Line 6 sounds but record dry. Then I re-amp later. I'm sure the Line 6 ToneDirect method gives me lower monitoring latency and seems to be independent of the buffer settings. This means I can use high buffer settings, have loads of other stuff running and not get any pops or clicks.

Pete

Last edited by stratman; 07-28-2008 at 11:54 AM.
stratman is offline   Reply With Quote
Old 07-28-2008, 11:49 AM   #11
frog_jr
Human being with feelings
 
frog_jr's Avatar
 
Join Date: Sep 2007
Location: Here at the Western World
Posts: 1,253
Default

Hey, just for kicks I put the buffer settings back up to the 'middle' settings. Results:

-24db Measurement results: 1329 samples / 30.14 ms
-18db Measurement results: 1330 samples / 30.16 ms

Me, I originally purchased this for 2-6 track recording mostly, so this is not a big problem.
frog_jr is offline   Reply With Quote
Old 07-28-2008, 12:02 PM   #12
stratman
Human being with feelings
 
stratman's Avatar
 
Join Date: Oct 2007
Location: Woking, England
Posts: 2,685
Default

How come this is in the spam trap? It seems like an interesting discussion. Well it's very interesting to me, but I'm a bit of a nerd about these things.

Pete
stratman is offline   Reply With Quote
Old 07-28-2008, 12:09 PM   #13
bullshark
Human being with feelings
 
bullshark's Avatar
 
Join Date: Jan 2006
Location: traîne mes guêtres en Québec...
Posts: 5,390
Default

I was asking myself the same question.
bullshark is offline   Reply With Quote
Old 07-29-2008, 10:28 AM   #14
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

Thank you all!!!

Don't misinterpret myself: The FA-101/66 are GOOD interfaces. Converters and Preamps are more that decent. The thing is that I'm after the lowest round trip latency. For that, I was told it would be RME, MOTU, and EDIROL. Who said that was wrong. To answer to the question "why the reported latency is higher that in Reaper?"

The reason is this: (as explained before)
Total Round trip latency, in case of firewire and USB interfdaces is this:

ASIO is (3ms)-->Whatever the software buffer shows
ASIO out (3ms)-->Whatever the software buffer shows
A/D (1ms. in general)
D/A (1ms. in general)
Hidden or safety buffers (Fire wire and USB have in general those buffers. PCI dont have those since they DO have DMA. The other dont).

So the best PCI can give a TOTAL of 4.xx/5ms. asio 2+2. AD, DA 2.

As explained 13ms is a little bit up. U wont notice that but playing guitar thru it will be sloppy.

Regards,
Danny

PS: stratman: DO you have a Texas Instuments Fire Wire chipset?
bullo is offline   Reply With Quote
Old 07-29-2008, 11:23 AM   #15
stratman
Human being with feelings
 
stratman's Avatar
 
Join Date: Oct 2007
Location: Woking, England
Posts: 2,685
Default

bullo,

Actually I can definitely hear the latency from the FA-66, which is why I mostly use Tone Direct monitoring with my Line 6 Toneport UX1, when I’m recording properly.

I just use the Edirol FA-66 to get an idea of the sound I’m eventually going to get when I re-amp with Revalver MK3.

I’m embarrassed to admit that I don’t know the chipset. In device manager is says “AGERE OHCI Compliant IEEE 1394 Host Controller”. Does that give any clues?

I’m just using a fairly bog standard Compaq computer running Vista, so I expect it’s far from optimized for music. I’m starting to think that the reason I’ve been getting away with things is because my Tone Direct monitoring from my Toneport carries most of the load when I’m monitoring.

btw. still wondering why this is in the spam trap. This thread is answering questions I've had for a long time (e.g. see link below to a post I made a while back).

http://forum.cockos.com/showthread.php?t=20983

Pete
stratman is offline   Reply With Quote
Old 07-29-2008, 11:35 AM   #16
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

Stratman:

In this case your chipset is "AGERE". This chipset is the piece of hardware that controls, in this case, the FireWire bus.
Regarding latency..Well 12/13 ms. You can hear it of fell it. Does not matter. In any case is not optimum for plying thru Revalver/ guitar rig, Amplitube, etc. Just fells strange.

Note that is you are using a soft synth/ sampler, the lantncy is half of the round trip.

Example: triggering a note on the Native Instruments "Pro53" Synth:

Once the software receives the MIDI NOTE ON message the folloing happens:

ASIO out (3ms)-->Whatever the software buffer shows
D/A (1ms. in general)
Hidden or safety buffers (Fire wire and USB have in general those buffers. PCI dont have those since they DO have DMA. The other dont).

A/d conversion and Asio in buffers dont take place. So latency is bettere in that case. So 6.5 ms, for jsut soft synths is VERY VERY GOOD. Almos as good as hardware synths.

Regards

Danny Bullo
bullo is offline   Reply With Quote
Old 07-29-2008, 11:48 AM   #17
bullshark
Human being with feelings
 
bullshark's Avatar
 
Join Date: Jan 2006
Location: traîne mes guêtres en Québec...
Posts: 5,390
Default

Still no answer why this is in spam trap...

Why is bullo round trip latency 30% worst than mine for the exact same Firewire interface?

And, just to put thing in perspective, 10ms round trip latency is the same as sitting 11 feet from your guitar amp at sea level when playing.
bullshark is offline   Reply With Quote
Old 07-29-2008, 12:09 PM   #18
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

Stratman:

I dont really get why u mean by SPAM TRAP. I know..It is very frustrating... Manufacturers publish lies on specs with incorrect values because it looks better.
What I can tell you is what I learned thru forums, reading and talking to people. If your question is why there is a difference of +/-1ms with my latencies, this is why i think:

I made the test with an FA-101 @ 44.1khz. Changing sample rate changes the latency too. Higher sampling rate-->lower latency.

this is because the latency is in samples actually. So x samples at 44,1 pass 'slower' that the same x sample at 96khz.

Also, maybe the driver for the FA-101 is slightly different that the one for FA-66. I have no idea. In any case. I returned the unit. I have my emu1820 and Im very happy with it. This interface would be for latptop live playing.

Regards,
danny
bullo is offline   Reply With Quote
Old 07-29-2008, 12:34 PM   #19
stratman
Human being with feelings
 
stratman's Avatar
 
Join Date: Oct 2007
Location: Woking, England
Posts: 2,685
Default

bullo, I think you accidentally put your post in the spam trap part of the forum. As it seems to be a very worthy subject, I was hoping that someone might move it if I keep going on about it (but nobody ever listens to me )

I just did the same test on my Tomeport UX1 and got the result below.



Now I'm very confused.

edit. with this setting Reaper reports 5.1/10ms for my UX1. I did a loopback test against the click source in Reaper and the recorded track is spot on in time with original click source in Reaper. So Reaper is definitely getting things right.

Pete

Last edited by stratman; 07-29-2008 at 12:44 PM.
stratman is offline   Reply With Quote
Old 07-29-2008, 01:35 PM   #20
bullo
Human being with feelings
 
Join Date: Jul 2008
Posts: 15
Default

Pete:

I do not know the TonePort products. but that test is incorrect. That latency value is IMPOSSIBLE. As I explaied, the best PCI cards give around 5ms. round trip @ 64 samples buffer size (that is 1.5ms ASIO IN + 1.5ms ASIO out + 1ms AD + 1ms DA.


To make the loop back test MAKE SURE that the audio interface is outputting ONLY WHAT COMES FROM SOFTWARE. That is, is, the signal must go to the interface, beign digitalized, passed thrud the ASIO driver and then return to the outside world thru the D/A converter. For this test you need not to have the imput signal present without passign thru the computer. Some interfaces call it direct monitoring, and I guess teh toneport is similar, since the ultra low latency is achievable only thru its software, not thru any ASIO client. Does it make any sense to you?


Hope it helps

Dany Bullo
bullo is offline   Reply With Quote
Old 07-30-2008, 09:15 AM   #21
bullshark
Human being with feelings
 
bullshark's Avatar
 
Join Date: Jan 2006
Location: traîne mes guêtres en Québec...
Posts: 5,390
Default

Quote:
Originally Posted by stratman View Post
bullo, I think you accidentally put your post in the spam trap part of the forum. As it seems to be a very worthy subject, I was hoping that someone might move it if I keep going on about it (but nobody ever listens to me )
Why not start another topic in one of the regular forum? One more generic in it's title asking people to test the round trip latency of their interface? This thread, being in spamtrap, will disappear shortly anyway.

I think it would very useful for two reason:

The first is that that data is not available from manufacturers, so it would help people in the market for an interface. My needs are presently covered, but I would have liked to read about that when I was in the market not too long ago.

The second is it would help identify potential problem with one's interface or sub-system. Just take, for example, the performance test thread I started, well, a few people found, by doing this test, that they had multi-CPU turned off in Reaper on their multi-core machine.

Oh, and don't worry, nobody ever listens to me neither, heck the op in this thread didn't even read my post, yet I go along just fine

Last edited by bullshark; 07-30-2008 at 09:18 AM.
bullshark is offline   Reply With Quote
Old 07-30-2008, 09:28 AM   #22
frog_jr
Human being with feelings
 
frog_jr's Avatar
 
Join Date: Sep 2007
Location: Here at the Western World
Posts: 1,253
Default

Quote:
Originally Posted by bullshark View Post
Oh, and don't worry, nobody ever listens to me neither, heck the op in this thread didn't even read my post, yet I go along just fine
He's absolutely right, I agree %100. I ignore him all the time myself.
frog_jr is offline   Reply With Quote
Old 07-30-2008, 12:37 PM   #23
stratman
Human being with feelings
 
stratman's Avatar
 
Join Date: Oct 2007
Location: Woking, England
Posts: 2,685
Default

Quote:
Originally Posted by bullshark View Post
Oh, and don't worry, nobody ever listens to me ...
Actually I listen to you, which isn't always a good idea.

I still haven't forgotten how you got me with that April Fool.

Pete
stratman is offline   Reply With Quote
Old 10-22-2008, 08:47 AM   #24
SiKo
Human being with feelings
 
SiKo's Avatar
 
Join Date: Aug 2008
Location: dusty hot place
Posts: 1,492
Default

Quote:
Originally Posted by bullshark View Post
Still no answer why this is in spam trap...

Why is bullo round trip latency 30% worst than mine for the exact same Firewire interface?

And, just to put thing in perspective, 10ms round trip latency is the same as sitting 11 feet from your guitar amp at sea level when playing.
Isn't anything that has latency below 25-30ms un-hear-able ? (HAAS effect)
__________________
... yOu aNd mE are ...
SiKo is offline   Reply With Quote
Old 10-22-2008, 11:09 AM   #25
Ollie
Super Moderator (no feelings)
 
Ollie's Avatar
 
Join Date: Dec 2007
Location: On or near a dike
Posts: 9,742
Default

I think it's not that easy. Oh, and AFAIK it has nothing to do with Haas (which is more about the precedence effect/law of the first wavefront).

1. These figures vary from source to source, some say 10ms.
2. These figures are about the time two identical signal impulses have to be apart to be distinguishable as separate signals. But these vary individually and have not much to do with the effects of latency.

Latency of e.g. modeling amps is much lower, around 5ms maybe. But many guitar players can actually "feel" this latency - the brain seems to be able to detect such a small delay between action (plucking the string) and the reaction. This may as well be influenced by the brain's "time resolution", which is not static and can be trained. "Microtiming" is happening in similar low time variations and good drummers have actually control over them, you can try that yourself: Insert a 30ms hesitation into a drum loop, it will probably throw you out of the groove. Delay a snare by only 5ms and it starts getting a "laid back" feeling, if you delay it by 20ms (@120bpm), the groove will sound pretty crappy. You see, these figures are widely misunderstood and are not relevant to round trip latency. Play a bit with a delay at such low values and the dry signal turned off and see yourself.
Ollie is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 02:06 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, vBulletin Solutions Inc.