Interpretation of impulse response delay

[mib]

Hi, I have a question about the interpretation of the time delay measured using impulse response for a passive speaker. Do we expect the time delay measured that way to always correspond to the distance between the speaker and the listening position or are there other acoustical influences?

Background of my question is that I have speakers where the IR delay reflects the speaker distance (as measured with a Laser or using Audyssey) very well and others where the IR is delayed more than a ms further (I have excluded that the addl delay is in the electronics by swapping channels). Typically the speakers with higher IR delay are closer to the LP and close to a back wall.

So I am wondering if the delay in the IR in such cases can represent something else than just the distance to the front of the speaker?

 

[John Mulcahy]

It should be pretty close to the measurement, unless the impulse response is from a limited bandwidth speaker which broadens and delays the peak. A very strong early reflection could also confuse things.

 

[mib]

Thanks for the response. Well it is different speakers (though they all count as full range) and the back wall is certainly contributing early reflections. However, the impulse response is pretty concise and there are no major peaks before the first impulse.

If that helps I attach the original mdat file. The front speaker FL is an example where it is very close, the rear speaker SBL is an example of the additional delay.

Are there any other pitfalls in the measurement process?
Without having a firm grasp of these concepts - could it be something like the difference beween time delay and group delay?

The strange thing is that configuring the “incorrect” delays estimated with REW sounds actually pretty good. Though it is hard to compare as it takes a while to switch things back and forth, I noticed that I kept the REW distances.

 

[John Mulcahy]

No issues evident there. Seems like a very lively environment though, with some very sharp resonances. Lot of bare surfaces?

 

[mib]

Yes, my guess is that this is due to all the glass. I will probably try some curtains to tame that at least in the evening a bit. Especially the center has issues I still need to address. Could the resonances influence the distance measurements?

Now the difficult question: If there are no obvious issues with the delay measurements - would you use these numbers? Or would you recommend to better use the actual physical speaker distance / Audyssey number?

 

[John Mulcahy]

I would use the physical distances unless there were some reason not to, such as an equaliser in the path for some channels but not others. The differences were small though, so shouldn't make much difference either way.