Dirac Unison

[rhbblb1]

In that case you are correct and I stand corrected... in general I personally prefer a larger volume as in the sofa setting, I'm more into Stereo than Home Theater and follow a conservative approach to avoid any risk of eventual overcorrection but in the case that you mention where speakers are blocked then a smaller variation as you mention would be preferable.

Flavio

Flavio,

I am also now confused. Are you saying that for stereo measurements, the vertical spacing should be 36 inches (18 above the ear and 18 inches below the ear)? If so, the lower measurements will be taken well below the top of the back of the chair. If so, how far away from the back of the chair should the microphone be placed?
Thanks for your input.

 

[Mark Seaton]

Flavio,

I am also now confused. Are you saying that for stereo measurements, the vertical spacing should be 36 inches (18 above the ear and 18 inches below the ear)? If so, the lower measurements will be taken well below the top of the back of the chair. If so, how far away from the back of the chair should the microphone be placed?
Thanks for your input.

Any set of measurements need a quick sanity check before you start. A great example is with tall ribbon or AMT drivers having very narrow vertical dispersion. The same can be said of very tall speakers that have a narrow intended listening window. Starting with a simple check by ear and pink noise will alert you to the effective boundaries you might want to limit your measurements to. Quick measurements (something I wish Dirac would add between Levels & Measurement steps of the calibration) can then be used to better understand what is going on, and even guide what upper frequency correction limits might be worth experimenting with. If you are well outside the listening window of the speaker, expect that to muck with the average.

The key understanding is to give the software enough information and a large enough sample to see what is affecting the entire listening area and not just point-specific effects. This requires a few measurements offset in all 3 dimensions. It is still an average, so the more measurements you take in one axis or dimension, the more aggressively issues at those locations will be corrected. A good example is taking 7 measurements at the exact same height, and 1 set higher by 6-12". The result will be aggressive correction of issues common to the height at the 7 locations, where the response might quickly change at other elevations.

 

[kach22i]

UPDATE: 01/24/2020

This is an interesting thread, but the very first link from 2014 no loner works.

I found this link doing a search and it has some pretty good graphics in it.

https://www.dirac.com/unison