Near-field room correction measurements: With or without chair? Occupied or not?

[echopraxia]

I've never seen this question addressed anywhere, so forgive me if it's redundant. I'm trying to determine how my room should be configured when I'm making room correction measurements for my home office setup (in a small room, with speakers positioned for relatively near-field listening).

For a small room in particular, I imagine the position and presence of a primary listening chair (especially with a high backrest) could affect the in-room frequency response of the primary listening position significantly. The speakers are supposed to be pointed at it, after all, so I assume its presence or absence would have a big impact on the overall behavior of room reflections. But then similarly, so could the presence of me sitting in the chair vs. not sitting in the chair while measurements are being taken.

So I am wondering: which would lead to the most preferable and/or most accurate results after running room correction:

1. Measure and equalize to an empty room.
2. Measure and equalize to the room with office chair in its usual position, with without me in it.
3. Measure and equalize to the room with office chair in its usual position and with me in it.

Is there a best-practice guidance for this?

 

[DJBonoBobo]

I don´t know about a best-pratice guidance, but for practical reasons i push the chair back a bit (for placing the mic stand or to gain some space for the RTA MMM method) and go out of the way, but stay in the room. If i´m doing RTA MMM i have to be quite close, of course.

Personally i only do room corrections below Schroeder frequency and use narrow filters only up to 150 Hz or so. Maybe some broader filters (Q around 2 or so) up to 500hz. For this i guess neither the chair nor my body make much difference.

"Best" option would be probably number 3 with measurement microphones in your ears, i think... But that´s not very practical for home use. (edit: This was not really meant seriously, therefore crossed out to avoid confusion.)

 

[QMuse]

I've never seen this question addressed anywhere, so forgive me if it's redundant. I'm trying to determine how my room should be configured when I'm making room correction measurements for my home office setup (in a small room, with speakers positioned for relatively near-field listening).

For a small room in particular, I imagine the position and presence of a primary listening chair (especially with a high backrest) could affect the in-room frequency response of the primary listening position significantly. The speakers are supposed to be pointed at it, after all, so I assume its presence or absence would have a big impact on the overall behavior of room reflections. But then similarly, so could the presence of me sitting in the chair vs. not sitting in the chair while measurements are being taken.

So I am wondering: which would lead to the most preferable and/or most accurate results after running room correction:

1. Measure and equalize to an empty room.
2. Measure and equalize to the room with office chair in its usual position, with without me in it.
3. Measure and equalize to the room with office chair in its usual position and with me in it.

Is there a best-practice guidance for this?

I'm not sure which is the best way but I'm doing it as per 2.

 

[gr-e]

"Best" option would be probably number 3 with measurement microphones in your ears, i think... But that´s not very practical for home use.

You definitely don't wanna do this. Your ears as well as your body are a part of your hearing system, every sound you hear is affected by them.

I personally do the measurements without chair in place

 

[echopraxia]

You definitely don't wanna do this. Your ears as well as your body are a part of your hearing system, every sound you hear is affected by them.

I personally do the measurements without chair in place

Yeah that’s why it’s confusing what’s the best for perception. You definitely don’t want to measure in the ears, but you also don’t want to measure an anechoic chamber because that would be pointless. So where is the ideal point in between? So far I’m going with chair in place but not myself, as well.

 

[Hipper]

I'm sure I posted something like this before but can't find it.

Using REW I did some tests on various aspects to see if there was much difference:

What Difference does a Person Sitting in the Chair Make?


1/24 smoothing, full range, purple being with me in the chair (I place the microphone where one of my ears would be then carefully got in the chair without knocking it it).

Oh! The right speaker had broadly the same effect, notably the increase from 600-1500Hz

What Difference Where You Stand When Taking Measurements?

I always stand behind my equipment in the back left hand corner so I compared that with standing on the right back corner, and standing behind the chair.



No smoothing. Purple is left, green is centre, brown is right.

It doesn’t seem to make a massive amount of difference

What Difference Where You Point the Microphone?





Brown - at my ear pointing to speaker
Green - in the middle pointing at speaker
Purple - in the middle pointing down the middle of the room
Red - in the middle pointing upwards
Turquoise - in the middle pointing downwards

The first FR is up to 500Hz, no smoothing. The second is full range at 1/24 smoothing.

The makers of my Earthworks microphone say to point it to the source, but I’ve seen other recommendations to point mics up or down. It seems to make a bit of difference in actual measurements.

I suppose all we can do is be consistent.

My current attitude is to use microphone and software as a guide, not an exact solution.

I put everything in place in the room, including the chair where it will be placed when listening (mine isn't a desk top but a dedicated listening room with lots of room treatment and the freedom to position where I want).

My first step is to try and find the best locations for the speakers. I first find that for the left speaker alone. I place the microphone where my left ear will be and point it horizontally towards the speaker, and take measurements using only sound from the left speaker and concentrating on the 0-500Hz range. When I've found the speaker location that has the smoothest FR for the left speaker I place the right speaker in its equivalent position and measure that. Hopefully it too will be more or less as smooth. Although it is said that the bass sounds are omni-directional I'm not convinced (how is it you can hear some bass sounds from one speaker only?).

Once you have the best speaker locations, measure them again (perhaps full range if you want that) individually (and together if you wish - microphone then in a central position between your ears [you can compare the combined FR with the two separate ones by averaging the latter. Hopefully they should be pretty much the same]). Make yourself the curve you want with your software or hardware but the final judgement should be in the listening (try not to get obsessed with how a curve looks on paper).

For my current set up all I did was my best at flattening the 0-200Hz range (using the 'no smoothing' option for the FR graph). I got REW to describe the filters I needed then applied the most important ones manually to the ten Parametric EQ settings of my Behringer DEQ2496. I could have done with more PEQ settings but I can still use its graphic equaliser (GEQ) if I want. However I've enjoyed listening since then and felt no reason to tinker with it further. There are still issues according to the frequency response but I'm OK with that. One day maybe.....

 

[ernestcarl]

If it's an office chair type with a mesh back, I don't think it would make much of a difference at all. For quick single point measurements while I'm sitting on the chair, sometimes I just move backwards and hold the mic in front. Using the MMM, of course, I'd definitely get off the chair -- but keep it in place as a guide as where my head is approximately going to be moving around.

EVERY 172E Series Task Chair

 

[Alexanderc]

I'm sure I posted something like this before but can't find it.

Using REW I did some tests on various aspects to see if there was much difference:

What Difference does a Person Sitting in the Chair Make?

View attachment 65502

1/24 smoothing, full range, purple being with me in the chair (I place the microphone where one of my ears would be then carefully got in the chair without knocking it it).

Oh! The right speaker had broadly the same effect, notably the increase from 600-1500Hz

What Difference Where You Stand When Taking Measurements?

I always stand behind my equipment in the back left hand corner so I compared that with standing on the right back corner, and standing behind the chair.

View attachment 65503

No smoothing. Purple is left, green is centre, brown is right.

It doesn’t seem to make a massive amount of difference

What Difference Where You Point the Microphone?

View attachment 65504


View attachment 65505

Brown - at my ear pointing to speaker
Green - in the middle pointing at speaker
Purple - in the middle pointing down the middle of the room
Red - in the middle pointing upwards
Turquoise - in the middle pointing downwards

The first FR is up to 500Hz, no smoothing. The second is full range at 1/24 smoothing.

The makers of my Earthworks microphone say to point it to the source, but I’ve seen other recommendations to point mics up or down. It seems to make a bit of difference in actual measurements.

I suppose all we can do is be consistent.

My current attitude is to use microphone and software as a guide, not an exact solution.

I put everything in place in the room, including the chair where it will be placed when listening (mine isn't a desk top but a dedicated listening room with lots of room treatment and the freedom to position where I want).

My first step is to try and find the best locations for the speakers. I first find that for the left speaker alone. I place the microphone where my left ear will be and point it horizontally towards the speaker, and take measurements using only sound from the left speaker and concentrating on the 0-500Hz range. When I've found the speaker location that has the smoothest FR for the left speaker I place the right speaker in its equivalent position and measure that. Hopefully it too will be more or less as smooth. Although it is said that the bass sounds are omni-directional I'm not convinced (how is it you can hear some bass sounds from one speaker only?).

Once you have the best speaker locations, measure them again (perhaps full range if you want that) individually (and together if you wish - microphone then in a central position between your ears [you can compare the combined FR with the two separate ones by averaging the latter. Hopefully they should be pretty much the same]). Make yourself the curve you want with your software or hardware but the final judgement should be in the listening (try not to get obsessed with how a curve looks on paper).

For my current set up all I did was my best at flattening the 0-200Hz range (using the 'no smoothing' option for the FR graph). I got REW to describe the filters I needed then applied the most important ones manually to the ten Parametric EQ settings of my Behringer DEQ2496. I could have done with more PEQ settings but I can still use its graphic equaliser (GEQ) if I want. However I've enjoyed listening since then and felt no reason to tinker with it further. There are still issues according to the frequency response but I'm OK with that. One day maybe.....

This is excellent. I wonder how this would affect room correction results. It seems that sitting in the listening chair makes a difference that would be audible.

 

[watchnerd]

I thought the whole point of sitting in the near field (< 1 m) was to sit in the area where direct, on-axis sounds dominate and reverberant, room-sounds are reduced.

 

[Alexanderc]

I thought the whole point of sitting in the near field (< 1 m) was to sit in the area where direct, on-axis sounds dominate and reverberant, room-sounds are reduced.

Fair point, and seems like what the OP was asking about. The thread veered off a little bit I suppose, but I find the measurements very interesting.

 

[Daverz]

I put some blankets and a pillow in my listening chair to simulate my meatsack. However, I have to recline the chair all the way to allow for the USB cable coming out the back of the UMIK-1, so measurements miss that the leather seatback is behind my ears.

If you are only doing correction below 200 Hz, you don't have to worry about it.

 

[watchnerd]

I put some blankets and a pillow in my listening chair to simulate my meatsack.

Frozen turkeys are better.

More flesh-like.

 

[echopraxia]

What I was getting at can be explained with this thought experiment: Consider a continuum between anechoic chamber, room fully populated minus the human, and at the extreme to having the measurement mic in a model of a human head and ear canal etc.

I’m pretty sure that both extremes are not what you want for EQ: an anechoic chamber is useful maybe for EQing the speaker with no consideration of room, but room effects are usually the whole point, right? But we also definitely don’t want to measure inside a reproduced human head because we don’t want a flat response there — we want whatever the natural response is that reaches your ears when the sound field in a room is flat.

So it raises the question, where along this spectrum is ideal? It’s kind of the last detail to resolve in the “circle of confusion” in a small room, where chair and/or body seems reasonable to influence the sound audibly. Yet, it’s not clear what exactly is ideal.

In larger rooms I think it’s much simpler, since room effects should dominate the effects of the human being present or not.

 

[echopraxia]

Frozen turkeys are better.

More flesh-like.

I find that calibrating using ham (thawed, but not cooked) with a chicken head glued on top widens the soundstage and has a huge impact on pace, rhythm, and timing.


(Kidding, of course.)

 

[watchnerd]

What I was getting at can be explained with this thought experiment: Consider a continuum between anechoic chamber, room fully populated minus the human, and at the extreme to having the measurement mic in a model of a human head and ear canal etc.

I’m pretty sure that both extremes are not what you want for EQ: an anechoic chamber is useful maybe for EQing the speaker with no consideration of room, but room effects are usually the whole point, right? But we also definitely don’t want to measure inside a reproduced human head because we don’t want a flat response there — we want whatever the natural response is that reaches your ears when the sound field in a room is flat.

So it raises the question, where along this spectrum is ideal? It’s kind of the last detail to resolve in the “circle of confusion” in a small room, where chair and/or body seems reasonable to influence the sound audibly. Yet, it’s not clear what exactly is ideal.

In larger rooms I think it’s much simpler, since room effects should dominate the effects of the human being present or not.

Are you doing this for playback or as part of DAW setup?

I've never met a recording engineer who EQ'ed a near-field DAW with listening distance <1 m for seat-occupied vs not.

Because it's near-field.

And people have different shapes / sizes / masses / resonant frequencies / ear heights, anyway.

 

[echopraxia]

Are you doing this for playback or as part of DAW setup?

I've never met a recording engineer who EQ'ed a near-field DAW with listening distance <1 m for seat-occupied vs not.

Because it's near-field.

And people have different shapes / sizes / masses / resonant frequencies / ear heights, anyway.

For me it’s for playback. But I’m just curious. I realize it probably doesn’t matter, but was just curious if there was any conventional standard procedure here.

Are you saying that near field EQ is only EQing the direct sound? What about desk bounce and other effects — are those not a part of measurements? If things like desk bounce is measured, then at what point do reflections stop being measured, if at all?

 

[watchnerd]

For me it’s for playback. But I’m just curious. I realize it probably doesn’t matter, but was just curious if there was any conventional standard procedure here.

Are you saying that near field EQ is only EQing the direct sound? What about desk bounce and other effects — are those not a part of measurements? If things like desk bounce counts and is measured, at what point do reflections stop being measured?

If you're serious about understanding, all of your questions are extensively addressed in most of the common recording engineering text books.

Huber's "Modern Recording Techniques", part of the AES presents series, is a good one.

"Mixing Secrets for the Small Studio" is also good.

 

[Eurasian]

Frozen turkeys are better.

More flesh-like.

I'm really having to bite my tongue, guys

 

[watchnerd]

I'm really having to bite my tongue, guys

Well, I was going to say pig carcasses.

Those are common for firearms testing.

But harder to buy.

 

[Hipper]

I thought the whole point of sitting in the near field (< 1 m) was to sit in the area where direct, on-axis sounds dominate and reverberant, room-sounds are reduced.

The OP asked about a 'relatively near field' arrangement, whatever that is in his case. The chair and body will still surely have an impact wherever the sounds come from. In fact one thing not mentioned is the type of chair. Leather chairs can reflect for example. My chair is a cloth covered type.

The main point still stands - be consistent with your measurement set up using them as a guide, not an exact solution.