Dummy head for measuring room response

[Tim Link]

Dummy heads are used to measure headphones against a target curve. The target curve as I understand it was derived from measuring dummy head response of a good speaker in a good room. So why not measure room / speaker response using a dummy head and once again shoot for the good target response at the ear?

I'm thinking of this becasue it's been my experience that sometimes over-deadened rooms can be magically brought back to life with some eq in the mids and highs. It's hard to know what to do though, making it hit and miss becaue the equation of what's happening to the ear cannot be measured with a standard microphone. We need a microphone that will show the response our ear will have to the total of sounds hitting from all directions in that particular space with those particular speakers and acoustic treatments. If we can adjust for absorption effects it could open up a lot of options perhaps to maximize imagining and clarity without adversly affecting tonal perception, or just make the most of a difficult set of speakers and room.

The last system I heard that surprised me was a big pair of Dunlavy speakers in a very heavily damped room. I was expecting it to sound like over damped rooms sound but it didn't at all. It was very bright and clear so I suspect that possibilities might be there that just need some clear method of measuring to attain reliably. I was thinking about getting some simple in-ear microphones to start and EQing my system to give a good target response. I don't know how good those are for that purpose, or how much I need to spend to get a decent set.

Any thoughts or experience with this sort of thing would be highly appreciated.

 

[staticV3]

With speakers, there's no worthwhile advantage measuring them with a HATS vs a regular measurement microphone.

It'd be like measuring a DAC using actual music signals. Technically possible, but with extra steps required and with results that'd be more difficult to understand and interpret.

Synthetic test tones (=traditional mic) do just fine there.

 

[Tim Link]

With speakers, there's no worthwhile advantage measuring them with a HATS vs a regular measurement microphone.

It'd be like measuring a DAC using actual music signals. Technically possible, but with extra steps required and with results that'd be more difficult to understand and interpret.

Synthetic test tones (=traditional mic) do just fine there.

Well I don't think it would be quite like that. We're still using the same synthetic test tone. We're just shooting for a different target curve because the microphone is different. I guess a simple test would be to first achieve the smooth gently sloping target curve measured with an omni mic., and then use a HATS to see if the target curve for that has also been achieved. It should have been, right? Or no? The HATS target curve is what we're really after, and we're assuming that the smooth room curve will make it happen, so why not just measure directly what we really want to achieve rather than assume a surrogate curve from an omindirectional mic. will always get us there - unless the two do correlate so highly that you can really count on it with a high degree of certainty? I'm wondering about it because after hearing a lot of speakers that follow that Harman room curve I get a frequent sense that they're often a bit soft in the midrange compared to good headphones or even the same speakers in a different room. If I try to fix it with EQ it can be difficult because I'm trying to fix what's at my eardrum and I don't know how to do it for a particular room when I'm measuring with an omni mic. I just recently put some absorbers at first reflections points to improve the imaging. It did that but it also dulled the sound. The room curve as measured the normal way still looked almost identical. Maybe 0.1 decimal differences here and there. So it seems to me that depending on the specifics of the room the omni measurement may need to deviate from flat a bit to sound correct at the eardrum.
I guess a related question would be why do people say it sounds bad to listen to a speaker in an anechoic chamber? Can you make a speaker sound good in an anechoic chamber by giving it the proper response? I saw a show where some people tested musical instruments in an anechoic chamber and most sounded really bad because they rely on room reverberation. But the one that defeated the chamber was the gong. It had so much of it's own reverberation that it didn't need any help. A speaker can play a recorded gong so that should sound fine too. Or an instrument recorded in an ambient space.

 

[hwest]

Dummy heads are used to measure headphones against a target curve. The target curve as I understand it was derived from measuring dummy head response of a good speaker in a good room. So why not measure room / speaker response using a dummy head and once again shoot for the good target response at the ear?

I'm thinking of this becasue it's been my experience that sometimes over-deadened rooms can be magically brought back to life with some eq in the mids and highs. It's hard to know what to do though, making it hit and miss becaue the equation of what's happening to the ear cannot be measured with a standard microphone. We need a microphone that will show the response our ear will have to the total of sounds hitting from all directions in that particular space with those particular speakers and acoustic treatments. If we can adjust for absorption effects it could open up a lot of options perhaps to maximize imagining and clarity without adversly affecting tonal perception, or just make the most of a difficult set of speakers and room.

The last system I heard that surprised me was a big pair of Dunlavy speakers in a very heavily damped room. I was expecting it to sound like over damped rooms sound but it didn't at all. It was very bright and clear so I suspect that possibilities might be there that just need some clear method of measuring to attain reliably. I was thinking about getting some simple in-ear microphones to start and EQing my system to give a good target response. I don't know how good those are for that purpose, or how much I need to spend to get a decent set.

Any thoughts or experience with this sort of thing would be highly appreciated.

How much does those dummy heads cost?

 

[kemmler3D]

We're just shooting for a different target curve because the microphone is different. I guess a simple test would be to first achieve the smooth gently sloping target curve measured with an omni mic., and then use a HATS to see if the target curve for that has also been achieved. It should have been, right?

Dummy heads are really only necessary for making binaural recordings and things of that nature. They compensate for not having your actual head present in the listening space.

When you listen to speakers, your head is always present. So a dummy head (when doing room EQ) would just add an additional approximation, when all you really need is for a flat measurement mic to show you that you've hit your original target curve. Your own actual head takes care of the rest.

Now, I think where using a dummy head could be interesting for room correction is measuring stereo effects. Not that we have agreement on a standard test suite for stereo imaging and spatial effects, but if you were going to do measurements for those things, I think a dummy head could be a better place to start than omni mics.

 

[NTK]

Using dummy head for measurement is often done in the study of building acoustics to get the so called binaural room impulse response (BRIR). For example:

360° BRIR & FOAIR - APL

apl-hud.com

The problem with this type of measurement is that the room response is in general non-minimum phase. It is therefore not easy to use the acquired data for the purpose of EQ.

 

[DVDdoug]

I'd expect a calibrated measurement mic and a calibrated dummy head to give the exact same results. ("Exact" within tolerances.)