ATMOS NEARFIELD ROOM TUNING

[Demetri Evdox]

I'm setting up a small nearfield 7.1.4 system and want to do the tuning myself. All the large studios I have worked at we hired a Dolby tech to come and tune the systems so I learned a little about it and aquired the tools (smaart 8, sonarworks XFRE 20 mic), and an interface atmos controler combo: metric halo LIO 8 with DSP for room correction.

I want to have different tunings for ATMOS MUSIC, ATMOS HOME THEATER TV, and setero music .. want bass managed and non bass managed versions.

The main thing I dont understand is how much smoothing, averaging or banding to apply to the magnitude trace.. obviously the more banding and smoothing applied the less it shows the problems in the magnitude response of your speakers.. so it will affect how much you attempt to correct.

I have aquired the Dolby recommended target curve for atmos music but couldnt find the proper curve for atmos tv.
so then also need to decide on a curve for standard stereo music.

Any help on this topic would be much appreciated.

 

[Jeff Babcock]

Hi,
When using Smaart, for this type of situation, I normally go with 1/12 octave, 2sec,. However, don't chase down every little peak or dip, keep the number of corrections to a minimum and think in broader strokes unless you are chasing down a very specific issue.. Obviously you want to shoot for somewhere in the ballpark of Dolby's target, but don't worry about matching it exactly. Keep in mind the effects and limitations of the room make a big difference. Also keep in mind Atmos has a very wide range of sub crossover points, as high as 250Hz. There is a big difference in how the system will sound and be perceived in terms of imaging with an 80Hz crossover vs 250Hz, so keep this in mind (lower is strongly preferable IMO)

If this is for your own room, feel free to tweak slightly based on what sounds good to your ears, not what looks best on the screen.

 

[Demetri Evdox]

Hi,
When using Smaart, for this type of situation, I normally go with 1/12 octave, 2sec,. However, don't chase down every little peak or dip, keep the number of corrections to a minimum and think in broader strokes unless you are chasing down a very specific issue.. Obviously you want to shoot for somewhere in the ballpark of Dolby's target, but don't worry about matching it exactly.

If this is for your own room, feel free to tweak slightly based on what sounds good to your ears, not what looks best on the screen.

thanks for the reply. do you think it's best to utilize allpass filters to handle some phase linearity correction?

 

[Jeff Babcock]

Allpass filters are useful if there is a major phase issue, but be careful to ensure your Smaart timing calibration is correct, and only make one change at a time and listen to whether it improves things. Keep in mind with so many speakers, room reflections and points of interaction you might just end up chasing your tail. Things may look good on screen with one individual speaker, but the result of everything together could be different in the room. Again, make sure you use your ears at least as much as your tools.

 

[Demetri Evdox]

greatly appreciate the advice. so as far as timing goes is it as simple as adjusting all speakers except for the 1 with the longest delay time to the mix position (measurement position)? ie.. use smart to calculate delay with a TF measuring pink loopback against mic.. (find delay) .. then manually do the math and subtract the differences ?

i did a test another way playing 1 frame of 1k tone from pro tools to each speaker and recorded it at the mix position 1 at a time.. slightly re-aiming the mic at each speaker.. then measured the offset from the ref track in samples .. then adjusted the delay to the 1 speaker with the most delay .. re recorded the pop again to all speakers and the start of the tone lined up, but then the waves would all drift out of phase on the recorded tracks that were starting off in sync.. does that mean I have a phase problem that an allpass filter would correct?

 

[Chris Hindle]

You can't beat or cheat Physics.
Adjust delays and phase all you want. It will only be "correct" in ONE spot.
Less speakers is more accurate. A true point source being the best you could expect.
More than that is always a compromise.
Chris.