I'm on a mission to get my latency down and don't want to spend more $$ than I have to.  I purchased a TASCAM US-1800 about a year ago and can't seem to get the RT latency below 10ms. If I set the latency in ASIO control panel (i.e. TASCAM) to "very low latency", the driver crashes and takes down X2.  I've been able to set it to the next step up to "low latency" but the best I get is 15ms. Still unacceptable.  I've seen a few posts regarding the Focusrite devices and their low latency drivers.  I have to be able to deal with the ADAT inputs (for 8 of the channels anyway), so I'm looking at the Behringer ADA8000 which should foot the bill.  I would purchase the Focusrite OctoPre MKII but it's twice the price of the Behringer.  My goal is to use X2 in "Live Gig" situations as well as the studio.  To have a hybrid solution between software and hardware is probably going to work best.  What I'm trying to say here is that I don't need to use ProChannel  or other plug-ins for EQ, Compression, FX since I already have hardware to accomplish this instead of adding software plugins (can always add later) that increase latency.  I also am trying not to drag my 32-channel board around with me and want to record live venues and mix down later. BTW, I've turned off CPU "Parking" which makes a big difference in performance.  Any thoughts? 

The Scarlett series aren't ultra low latency but I've got mine down to around 7-8ms roundtrip for tracking at the lowest buffer settings without any problems. It doesn't like being that low for mixing though and I'm not sure I'd trust it that low live (maybe the next setting up which usually results in 13ms roundtrip). My specs are similar to yours except I have the i7 2600k. Cheers.

If you're looking for the lowest possible round-trip latency, that means using an audio interface that has a small hidden safety-buffer.  

External USB/Firewire units often use a large hidden safety-buffer to ensure glitch-free playback.  While this help to ensure glitch-free playback (under less than ideal circumstances), it comes at the expense of greater round-trip latency.  Other than increasing the sample-rate, there's nothing you can do to mitigate an audio interface's round-trip latency.  Units that use DICE-II tend to use a larger hidden safety-buffer (Alesis, TC, Focusrite, etc)

In an external unit:
RME, MOTU, and Presonus Audiobox VSL series would all fit the bill.
The RME USB units offer round-trip latency of 4.9ms at a 48-sample ASIO buffer size/44.1k

MOTU and Presonus units would be slightly above this.
The Octacapture and siblings would be slightly above the MOTU and Presonus units (7.4ms)

With PCI/e units:
Most offer round-trip latency of ~5ms at a 64-sample ASIO buffer size/44.1k

Gio,

I'm trying to understand what you are wanting to accomplish.  I tried using a stopwatch to see if I could start and stop it in less than 10 milli-seconds.  The best I could accomplish is 19 m-seconds.  I'm teasing of course, I understand when you strum a guitar through your audio card and you have that split second delay it'll drive ya crazy.  For a live setup, I'm assuming that you are going to be playing some instrument along with pre-recorded sound files during the shows and record what you are performing.  

If that is the case, I suggest turning off input monitoring on your own instrument and vocals so that you are not getting that odd continuation of instrument/vocal sound when the music stops.  It'll drive you, the professional that knows it's wrong, up the wall... the crowd may not notice at as much.  Therein, you have your tracks lined up and armed to record, you hit record - the backing tracks start playing and you play and sing along as you normally would using your floor monitor to hear the pre-recorded sounds from the computer.  If you are not using any FX or PROCESSING, don't worry about latency.  When you take your laptop back to the studio, you can nudge your audio a little to the left to line it up.

Of course you are going to want to hear yourself in your monitor....USE the Software mixer to ROUTE your vocals to the same output that sound from your computer is using.  Where you may have a headache is making certain that you are not recording feedback since your monitor will be giving your own vocals back to yourself/your vocal mike. 

Thanks all for you input.  I did purchase the Focusrite Scarlett 18i6 today with the Presonus Digimax 8D ADAT unit.  I've attached links to screenshots in X2 on how I configured my preferences:

https://www.dropbox.com/s/p4f5v9nhq0i950m/1.jpg
https://www.dropbox.com/s/6ys7758wxwoycyf/2.jpg
https://www.dropbox.com/s/6r0g9rpesw03u74/3.jpg
https://www.dropbox.com/s/t05kup2so54mkhd/4.jpg

Travis, In response to your post, I use two different configuration.  A duo format with two female vocalists where we use professional tracks. Both vocalists play percussion and one plays guitar. I'm just the lonely Harmonica player that sits in from time-to-time . The other format is a band consisting of 3-part vocal harmonies, guitar, bass and drums (no tracks).  All channels in X2 are recorded audio tracks from the instruments or vocal mics. I have hardware to handle compression, FX and EQ for stage monitoring or FOH.  


Jim, In response to your post, I was going to go with the Octacapture initially, but an expensive bugga when compared to the Focusrite.  I decided against the Behringer ADA8000 (although all my other hardware is Behringer) because I was told the pre-amps tend to be noisy.  I've heard good things about the Presonus Digimax 8D.  Ended up getting the Scarlett 18i6 for $240 US and DigiMax 8D for $330 US.


Bottom line is I'm going to have to play around a bit to find what works.  What else is new  

Gio
 
I was against the Behringer because of the age of the product and the limitations... 44.1 or 48 khz only, modern digital equipment that is reasonably inexpensive can go 96khz and some can even go 192khz.  That may not be noticeable to the ear, but I would try to maintain a certain level of compatability between digital equipment.
 
I have the SX4882 for my studio and the Invisible Mic Pre-amps are quiet.  But then again I don't toss that board around like I did my 1986 mark iv built-like-a-tank-and-weighed-just-as-much Peavey board.
 
I know Behringer gets a bad wrap, but nobody is trashing their chinese-made iPhones or X-boxes.

Travis,


I know Behringer has gotten a bad rap in the past, but I think they've come around quote a bit.  Their 32-channel board is one of the few with 6 aux sends, 4 of which are pre-fade.  Works great for separate monitor mixes.  I use their FX, compressors and 2 EQ's (one for monitor and one for FOH).  Back to the latency issue, I may be taking the Focusrite back.  To get less than 10ms RT, I had to drop the buffer width to 1ms and increase the buffer I/O to 2048.  Still hangs the SONICPDR.exe process.  I know I can increase the buffer width to 2ms, but that will put the RT to about 14ms.  What's the difference between that and the 15ms I'm already getting with my TASCAM?  Only that I've spent another $700 US (with tax and warranty) to gain nothing.  If I keep spending this kind of $$, I could have purchased the X32 digital board for $2500 US.  I think I'm coming to the conclusion that I shouldn't use a soft mixer for live gigs and stick with hardware. I don't know how Mackie is doing it with their DL1608 running on an iPad or the PreSonus Studio Live units.  Even Behringer released their iX line of mixers that, like the Mackie, run with an iPad interface.  This part of the business is so frustrating.  Am I sticking with a software product (X2) that only interfaces with legacy hardware (even the Behringer X32 with Mackie Controller won't interface), or should I move to Pro Tools and learn another DAW or will I be faced with the same issues?  I've been with Cakewalk since before they branded it SONAR, but when CW Support states that you have to have a RT of 10ms or less and won't recommend hardware that will work, it puts me off.  Am I ranting?  Can you feel my frustration ?

travismc1


Gio
 
I was against the Behringer because of the age of the product and the limitations... 44.1 or 48 khz only, modern digital equipment that is reasonably inexpensive can go 96khz and some can even go 192khz.  That may not be noticeable to the ear, but I would try to maintain a certain level of compatability between digital equipment.
 
I have the SX4882 for my studio and the Invisible Mic Pre-amps are quiet.  But then again I don't toss that board around like I did my 1986 mark iv built-like-a-tank-and-weighed-just-as-much Peavey board.
 
I know Behringer gets a bad wrap, but nobody is trashing their chinese-made iPhones or X-boxes.

Question, are you trying to use SONAR as your "digital mixer?" If so, you are always going to have the latency in a live setup: the signal is going from vocal (example) to interface to computer BACK to interface to to outputs to power amp... The way these interface manufacturers get their "low latency monitoring" that you are always hearing about is with their own software mixer. That software mixer functions to route the sound you are providing right to your outputs without going through the computer (vocal to interface input - split - signal going to sonar - and signal with no latency going to output... you can still MIX the stage sound from the software mixer provided by the interface, have that on your desktop and treat it like a board. The sound will still get to SONAR, just turn off input monitoring on each track you are wanting to record.

I wouldn't swap out to protools, personally I just keep PT for the few people that come in and say "I want recorded in ProTools. It's the industry standard." I show them I have it, let them see it, they think I'm more professional than Bob down the street that is running reaper... I probably wouldn't swap out to anything else, software wise. The end results on the recording side are still awesome. Every piece of software is going to work similarly. Presonus's hardware is capturing the sound before it gets to the computer. They are just designing the dynamics into the BOX. You control the settings from your computer but the info is in the box before it gets to the DAW. And your TASCAM software mixer should do the exact same thing. Play with that just a little.... MIX with the Tascam Software Mixer and see what you get. Have sonar running in the background and arm the tracks one at a time. You'll see what I mean.

OMG... I just went and browsed the tascam site.... you don't have a software mixer as part of the driver kit? do you???  I'm sorry, your setup is soooooooooooo totally different from mine... The maudio stuff I have has a software mixer that controls the box from the computer.  Your "zero latency" monitoring is hardware controlled in the box.  Basically - everything is hard wired to the MONITOR OUT on the back. 

Meaning - you can plug all of your mics into the front and a couple of instruments and your ten volume knobs are all you've got, and you can run it like a mixer.  The sound is split with what you record going into the DSP then the USB cable and what you hear being summed and going to the MONITOR OUT. 

As long as you run MONITOR OUT to your POWER AMP then to SPEAKERS... everything should be heard in real time, this is just a guess based on the schematic in the back of the TASCAM instructions.  Try that before you blow some major bucks.

MONITOR OUT gets everything in real time.  LINE OUT 1 and 2 Gets MONITOR OUT and from the computer and it's controlled by the MIX knob on the front.  LINE OUT 3 and 4 gets every thing processed regardless of the monitor knob or the mix knob... whew... that's a lot of info in a short amount of time.