I received my 990 this month and although it sounds fine I'm having terrible trouble with the accuracy of the AutoCal software, even with the new firmware. Here's what's puzzling. I'll AutoCal with the 990, watching on an SPL meter as it works it's way around the speakers and noting the db level it sets for each. When I manually check it using the same noise and same meter the results are different... not even close. Then I'll run the test program on the S&V disc and get yet another set of readings. The last time I ran AutoCal I got an error message stating my center was small, the surrounds were large and I needed to swap. No matter what combination of X/over points I use I can't save the AutoCal results. My room is on the "live" side but I had no problem with my previous recvr's autocal in the same room.
Any ideas?

Since you've got an SPL meter, I'd tend toward using the manual test tones to set the levels - the auto-setup seems to do well with distances typically, but in a fair number of cases it has trouble with the crossover points (sometimes exascerbated by microphone placement) that gets it confused about whether speakers are large or small.

When I have used auto-setup, I've generally gone back and tweaked my channel trim settings +/- 1dB or 2dB. Those sorts of variations could be attributed to placement of the microphone vs the SPL meter or to the judgment of the ol' Mark One eyeball versus the algorithm used by the 990. The sub's a little bit of a different case - I've been using the SMS-1 to get the sub sync'ed with the mains lately, but prior to that I got close with the SPL meter and test tone, then just adjusted "to taste" (typically a couple dB lower than what the meter would suggest).

I've actually had issues with 990's test tones in general, whether using auto calibration or manual. My left and right surrounds calibrated with an SPL meter to +7dB and +9dB respectively while the left and right mains were 0dB and +1dB. This was surprising since my surround speakers are only a bit smaller than my mains and are about the same distance from the listening spot.

When I used the THX optimizer on one of my DVD's to balance the channels, the surrounds calibrated to -3dB and +0dB while the mains remained the same.

I ran with both setups for awhile and found the latter calibration to be much more balanced and natural than the former.

So, based on the experience I had, I'd echo gonk's advice and suggest you just calibrate manually.

As for the difference between the S&V disc and the 990's test tones, it seems reasonable to expect that different source noise could yield very different results. That was certainly my experience, and I've read similiar stories from others who compared different calibration discs to one another like Avia and Video Essentials.

Obviously, the only way to tell what's right (assuming one isn't wildly off) is to just try both configurations for awhile and see what sounds best.

I spoke with Steve at Outlaw a few times about this... for one thing he says you must set your crossover points BEFORE calibration. That helped.
Tone defeat ON helped tremendously, I was able to save the configuration. Outlaw needs to take another look at the autocal program and make it more accurate. Right now we're paying for a program that really doesn't work as advertised.

You know what made a HUGE difference? The Delay/Distance calibration simply adjusts relative phase of all your speakers. Go back to actual measured distances instead of what autocal suggests and while running a music DVD step the Center distance up and down until the "image" snaps into focus. Try it... (thanks to the folks at Genelec for that suggestion)

Just for giggles I made an attempt using a calibrated mic instead of the disc provided by Outlaw. A little better, but not much. Today my LFM arrives so I'll have to recal all over again.
Another fine afternoon shot to hell!

Gonk, How does the SMS "sync" to your mains, and is it really worth the extra $$? I have an outboard parametric eq already in-line between the 990 and my sub.

Rather than use my SPL meter and individual test tones, I use the sweep pattern generated by the SMS-1 and adjust the sub level until the two portions of the line (the sub below 80Hz and the mains above 80Hz) are at the same level.