Understanding the State of the Art of Digital Room Correction

[Chromatischism]

Just ideas in general about how to implement it, like possibly running everything through a PC but not using it as the source, for example.

Thinking more about this, we would probably need to run all fully-decoded channels through a PC via the AVR's RCA preouts, then send the processed channels to amps. You could leave the PC volume at max and still use the AVR preamps and remote. Right? Then the question is, how to get 10+ RCA channels into and out of a PC?

AVR (room correction OFF) -> RCA - > PC -> RCA - > Amps

The main limitation is that the AVR can not also be used as the amplification because it (usually, these days) does not have multi-channel analog inputs. And, even if it did, how would you set the AVR to use those inputs, and also the input of your Blu-ray player or other source? You can't. So now we're talking about purchasing 10+ channels of amplification just to make this work.

One other potential problem I see is regarding multi-sub integration. Denon AVRs can adjust the delay of two subs independently, so no additional hardware is needed for time alignment. However, you will have two sub outputs going to the PC. I am unsure how you would handle that, so I think you would just use one channel and wait until after the PC does the EQ to split the sub outputs with a MiniDSP. It begs the question though: how is multi-sub normally handled with "SOTA" PC-based room correction, if not for a MiniDSP 2x4?

And finally, we would be performing room correction after bass management. @markus , remind me, should RC be done before or after bass management? Intuitively I would think before but I could be wrong.

 

[j_j]

Let me propose another link. In this thread folks quickly agreed that when comparing different mics you have to do it in the same position as even few cm of difference in position makes a difference in measurement.

What better argument than that would you need that even in the single seat scenario multipoint measurement is needed in order to get reliable spatially averaged input for room correction?

This is why smoothing is required. Goodness.

 

[thorvat]

This is why smoothing is required. Goodness.

Goes without saying that smoothing is absolutely requried once you average the multipoint measurement. But with smoothing you can't compensate for what you might (and probably will) miss with single point measurement. Goodness.

 

[markus]

This is why smoothing is required. Goodness.

Some kind of smoothing is certainly required along the way – for psychoacoustical reasons. But if one wants to learn about the soundfield in a room smoothing and the loss of information that goes along with it could hide important aspects. So does using just a single mic location.
mitchco for example suggested somewhere to remove furniture like a coffee table before doing measurements because he could hear "phase issues" when moving his head after running his one point optimization with the coffee table in place. In my mind that's a classic example of overcorrection when the optimization is based on just a single mic position.

 

[markus]

And finally, we would be performing room correction after bass management. @markus , remind me, should RC be done before or after bass management? Intuitively I would think before but I could be wrong.

Speaker/room correction filters ALWAYS need to be applied downstream (after) bass management.
More details: https://mehlau.net/audio/bass-management/

 

[thorvat]

mitchco for example suggested somewhere to remove furniture like a coffee table before doing measurements because he could hear "phase issues" when moving his head after running his one point optimization with the coffee table in place. In my mind that's a classic example of overcorrection when the optimization is based on just a single mic position.

IMHO you are right with this.

Removing furniture (like coffee table) before taking measurements is simply wrong because they are part of the room, hence part of the room influence on the speaker's response, and will be there when you will be listening. If you are to make room correction in a recording studio would you remove mixing table before measurement and then be surprised when the response is not what you expected once you repeat the measurement with the mixing table in place? I don't think so.

One more thing: you don't have "phase issues" when person is speaking to you from the other side of the coffee table and nothing in the person's voice is changing much when you are moving your head while listening to that person. If some strange effects are happening when you're moving your head when listening to your speakers it's not coffee table to blame, it's inaccurate correction, very often because it was based on a single point measurement.

 

[j_j]

Goes without saying that smoothing is absolutely requried once you average the multipoint measurement. But with smoothing you can't compensate for what you might (and probably will) miss with single point measurement. Goodness.

Really now, this is only for psychoacoustic reasons? No, it's not. Consider wavelength vs. head position.

 

[markus]

Really now, this is only for psychoacoustic reasons? No, it's not. Consider wavelength vs. head position.

Please elaborate. For example an analysis of low frequency room modes requires high resolution measurements. Smoothing means loss of resolution.

 

[Chromatischism]

Speaker/room correction filters ALWAYS need to be applied downstream (after) bass management.

Then that is good news, if anyone wants to try to wire up such a setup as described above...

Any takers?

 

[thorvat]

Really now, this is only for psychoacoustic reasons? No, it's not. Consider wavelength vs. head position.

No, not only for psychoacoustic reasons. I am really not such expert to name all the reasons that can cause such effect, but I know for sure inaccurate room correction is one of them.

Regarding your remark about wavelength vs. head position - same ratio between wavelength vs. head position is there when a live person is speaking in our rooms from the other side of the coffee table but no such artifacts occur when we move our head as the ones caused by bad correction.

 

[thorvat]

Some kind of smoothing is certainly required along the way – for psychoacoustical reasons. But if one wants to learn about the soundfield in a room smoothing and the loss of information that goes along with it could hide important aspects. So does using just a single mic location.

High-res (unsmoothed) measurement, combined with phase and other derived data, is used to identify what can and what cannot be corrected. Smoothing is adjusting resolution of the emasurement to the resolution/perception of our ears(brain) otherwise we would obsessively be correcting choppiness of the response to no effect.

So smoothing shouldn't be compared with single point measurement as with single point measurement we are failing to collect all the relevant data and with smoothing of multipoint measurement we are simply adjusting the resolution of the relevant data.

 

[fluid]

Here is an example from some of my own measurements. I measured about 25 different points left to right covering the area of a large couch with some further out and a few at different elevations. Above 1k they converge well enough. The Bright Purple trace is a Vector average of all those measurements. It weights the dips more highly and changes the spectral balance as it rolls off more quickly above that 1K range. A basic frequency Average in REW lands in the mess of curves and is very similar to a MMM.

Here is the single centre point vs Vector average less smoothing but the same data.

And Vector vs Frequency Average

 

[thorvat]

Here is an example from some of my own measurements. I measured about 25 different points left to right covering the area of a large couch with some further out and a few at different elevations. Above 1k they converge well enough. The Bright Purple trace is a Vector average of all those measurements. It weights the dips more highly and changes the spectral balance as it rolls off more quickly above that 1K range. A basic frequency Average in REW lands in the mess of curves and is very similar to a MMM.

View attachment 162355

Here is the single centre point vs Vector average less smoothing but the same data.

View attachment 162357
And Vector vs Frequency Average

View attachment 162356

Nice!

Btw, vector average is not something you really need unless you want to average phase response for the purpose of phase correction. For the purpose of FR correction "standard" average is what you want to use. And yes, standard average is pretty much equal to what you get with MMM if you do it the right way.

 

[fluid]

Btw, vector average is not something you really need unless you want to average phase response for the purpose of phase correction.

I agree. I have had good results using both which led to what I said earlier in the thread. Mitch's position as I understand it has phase/time correction as part of the magic where single position measurements make sense to me.

A manual alternative is average or MMM frequency correction combined with a secondary single position (frequency corrected) measurement for phase correction purposes.

 

[markus]

I agree. I have had good results using both which led to what I said earlier in the thread. Mitch's position as I understand it has phase/time correction as part of the magic where single position measurements make sense to me.

Especially when you want to do "phase/time correction" a single mic location makes no sense to me as it is blind to certain aspects of the room response.

Measurements need to be able to provide information about which effects are caused by the source and which are cause by the room. In my mind state of the art room correction needs to consider which reflection is coming from where. It is crucial to know whether a peak is caused by the speaker or if it is caused by a high seat back for example.

 

[fluid]

Especially when you want to do "phase/time correction" a single mic location makes no sense to me as it is blind to certain aspects of the room response.

I don't disagree with you and my comment was made within the confines of programs that only take a single impulse as a base measurement. I didn't add it explicitly as there are very few programs available than do use the information from multiple positions to do anything with the measurements other than average them somehow.

Even with that said the existing single impulse programs give a result that I like and to me is an improvement over not using it.

 

[thorvat]

Even with that said the existing single impulse programs give a result that I like and to me is an improvement over not using it.

While this is true in any case, I suspect that reason for your satisifaction is also the fact that your central point sweep correlates very well with the average of multipoint sweeps/MMM. Can you please put them together on a graph with Psy smoothing?

 

[dasdoing]

changes the spectral balance

that makes it useless

can you share the mdat, please?

 

[Le Concombre]

Please read the collection of Dr Toole's quotes collected by @thewas : https://www.audiosciencereview.com/...ut-room-curve-targets-room-eq-and-more.10950/ . Meditate on : "The stated or implied sales pitch is: give me any loudspeaker in any room and my process will make it "perfect". A moment of thought tells you that this cannot be true." When I was young and naive and had a computeraudiophile account, I thought @mitchco was a good guy sharing insights on a forum where people help each other. Fishing for customers is not bad per se but I've lost hours misguided by his half truths and twisted presentations of serious research, downward slopes only good for boom boom etc etc . If you have seriously engineered speakers, unless you're a guy (any lady here?) confusing the map and the territory, forget about eQ above the transition frequency, full range targets and other nasties and then the debate about FDW etc will become mostly moot and the messing with time domain conter productive since the right modal frequency eQ will be essentially coupled with the right time domain mod : https://www.hometheatershack.com/threads/waterfalls.7135/ .

 

[ernestcarl]

Then that is good news, if anyone wants to try to wire up such a setup as described above...

Any takers?

I prefer to equalize the sub, at least partially to a workable shape, before bass management.

Although, one could always apply some correction in the area beyond the xo interaction zone beforehand. It is not going to affect the bass management outcome anyway.