Problems using MBOX (original) with apps other than ProTools

Although digidesign provides ASIO drivers for the mBox series to be able to be used with other recording software, it reports its latency incorrectly, and your recorded audio is misaligned by DAW's other than Protools.

So, you must do loopback tests for all of the buffer sizes you use and manually change the settings in your software (if your DAW has this capability) everytime you change the buffer size on the MBOX...

Its a royal pain in that ass, courtesy of DigiDesign, and is yet another reason to not bother with protools or their interfaces. In my experience, the damn mBox never worked for very long with the version of protools that it came with, nevermind this problem!

First of all, a little rant..... USB is a lousy way to interface with your computer for audio apps, it is very liable to constant interruptions from the system, and it takes alot of tweaking of the Windows OS to make it run glitch free with ANY interface, never mind one that misreports its latency, like the original MBOX...

BUT, if thats what youve got, its what youve got, so lets make it work!!!!

So first, we have to do a loopback test to measure the ACTUAL latency:

1. attach a 1/4 inch standard audio guitar cable from output one to input one on the MBOX

2. Create 2 tracks and import a sample of some kind of short, sharp transient... like a stick click or something and place it on track one

3. Arm the second track to record and select input one

4. Take note of the buffer settings for the MBOX and Record the sample onto the new track.

5. Zoom in on the waveforms and see if they are lined up completely... They are likely not, so measure the distance between the sample on the original track against that of the recording you just made. The way to do this will depend on your software... might be a "ruler tool" or a way to set your timeline ruler to samples, then you can zoom in tight and make a precise timeline selection that represents the distance between your reference point (likely a high peak in the audio sample) in the two tracks. After the selection is made, there is likely somewhere in your DAW that will show you what the selection size is...

The number you get from this test is your systems latency at whatever buffer setting you have the MBOX at. It is the number of samples that you must move new recorded audio tracks on the timeline to be perfectly in line with the other tracks in your session. If you need to change the buffer settings on the MBOX, you need to do this all over again, as the latency changes right long with the buffer settings, and possibly will change dynamically as your track count gets bigger and bigger in your session... (possibly not) so unfortunately you might end up doing this test quite often. This can be made easier by leaving the tracks from the loopback test in your session... Then you can quickly re-record and measure the offset as often as necessary.

Some software (reaper) has a provision to enter a manual offset, a number of samples that you tell the software to move the audio forward or back in time after being recorded. This has its advantages... that you dont have to do manually move things after recording them, but the downside it that if you change the buffer settings on the MBOX and forget to change the settings in the manual offset, the recorded audio is also misaligned.

So, this is what you have to do to make sure that your DAW and your soundcard /interface are dealing with recording latency compensation properly.