Physics of low-mid buildup

[locutus]

This may be either a bit academic or obvious to most of you.

I didn't want to hijack the other thread, but I am thinking about a comment in the Flat Omnis thread. Flat omnis?

Originally Posted by mljung

The sound was not "blown up" as I find easily happens because of low+lo-mid build up from traditional pressure omnis.

OK, I have seen occasional references to this phenomenon, and I have experienced it myself. But I don't completely understand exactly why this happens from a physical acoustics standpoint.

Case in point. I recently did a recording of a solo piano (NY Steinway D) in a decent hall, pair of MK2H, 2m out, high enough to avoid lid reflections, spacing 36cm. I had terrible problems with low-mid buildup.

But why exactly? I know that 36cm is 1/4 wavelength of 238Hz, which is roughly the center frequency of the problem area. It seems to consist of some odd phasiness, not just boosted frequency response. This is consistent with a 90-degree phase shift at 1/4 wavelength.

Now, to experiment I tried changing the spacing. Adding just 4cm changed the buildup frequency noticably. But even at 50cm is was still rather noticable. I didn't have time to compare larger spacings.

I have to conclude that this is a direct result of sidewall reflections. But this is a large hall. The sidewalls are at least 10m away. I experimented with using choir shell panels to divert sidewall reflections, but didn't have enough time to prove anything.

I can't believe it is simply related to the low frequency response of omnis compared to cardis. Most cardiods don't have much of any noticable frequency drop at 238Hz.

Does anyone have a handle on exactly why this happens?

Nathan

[David Spearritt]

I am replying to this post to bump it. This topic is close to my heart and I am hoping for any relevant contributions from the group here. I have always been troubled by the woofly boost of spaced omnis, more correctly described, as in the title, as lower mid boost. It must be related to spacing, but I am not entirely sure of the exact physics of it. Have commenced discussions with some colleagues to hopefully offer some clarity on the subject.

Do we all agree that it exists and is a problem? Where have you found it most prevalent and in what sort of acoustic? I should add that I notice it most when positioned within 3m of a solo grand piano and the spacing is less than 450mm.

[mljung]

It's good know to know that this is a phenomenon that I'm not the only one experiencing!

Unfortunately I can't come up with any scientific explanations here, as I'm simply not capable of it.
But I have ideas for some general questions as a start point for investigating this phenomenon.

[and to hear different suggestions/experiences from all the clever people here on this forum of course]:


1) Are we certain this is a stereo phenomenon only?

2) Does it happen equally much with miniature diaphragms and the usual SDC? [and LDC's for that matter]

3) Is it a phenomenon that only happens with omni-directional microphones?
[If Cardioid and Figure of Eights have the same problem - is it better or worse?]

4) Does the microphone angle have any influence – which?

5) Which influence has stereo spacing on the phenomenon and what are the subjective/objective results using different spacings.
[e.g. 20/40/60/80/100/120 cm]

6) Since it's related to piano recordings does placement of the microphones relative to the piano make a difference? [concert listening position/tail end/ behind the pianist, different heights and distances]

7) Does the piano-lid have an influence on this phenomenon - if so which?

8) Does the room acoustics have a influence [such as sidewall reflections] on this phenomenon - if so which?

***

This is an idea for a list of questions that I find relevant [tried to include locutus and Davids subjects] - further suggestions/comments are very welcome!

::
Mads


[unfortunately I don't have a grand piano + pianist in a fine hall at my disposal]

[Baithak]

Hey glad to see that others have noticed this too ... I use small AB config a lot and noticed this phenomenon. I put it down to the following:

The Omnis become progressively directional as frequency increases. At higher frequencies they are almost cardiod and therefore are only "accepting" these sounds from the front of the mic whereas at lower frequencies, they are accepting sounds from all around. In smaller rooms this is immediately noticeable above and beyond the room modes. So the room reflections are an essential part of this.

In open air venues where there is no side and back reflection my guess is that the spaced omnis will not exhibit this behavior. This is just a guess.

I could be really, really off on all this so please feel free to ignore and move on with your discussion.

Baithak

[mljung]

Originally Posted by Baithak

Hey glad to see that others have noticed this too ... I use small AB config a lot and noticed this phenomenon. I put it down to the following:

The Omnis become progressively directional as frequency increases. At higher frequencies they are almost cardiod and therefore are only "accepting" these sounds from the front of the mic whereas at lower frequencies, they are accepting sounds from all around. In smaller rooms this is immediately noticeable above and beyond the room modes. So the room reflections are an essential part of this.

In open air venues where there is no side and back reflection my guess is that the spaced omnis will not exhibit this behavior. This is just a guess.

I could be really, really off on all this so please feel free to ignore and move on with your discussion.

Baithak

This is actually pretty close to my guess, that I stated in the Flat Omni thread.
It may not be all though, if this is the case it would be just as big a problem in mono as in stereo, and according to Nathan and David, they have an idea that the issue is "stereo-related".
::
Mads

[BrianHanke]

I've noticed this phenomenon on many pianos. I think it's due mostly to the acoustic properties of the instrument. Everything below middle C or so has the full power of the piano's soundboard behind it. Unless the piano is voiced brightly that low-middle range can come across as unbalanced.

I'd also say that there is a certain amount of perception involved. What may sound normal and even "warm" in real life can sound different than we expect when heard in isolation on a recording.

Looking at some of Neumann’s frequency response charts, I notice that the directionality in their omni mics starts happening around 2kHz and doesn't get extreme until between 8 and 16kHz. The piano's highest C comes in at just over 4kHz, so any pitches affected by the change in directionality will mostly be overtones. It's hard for me to believe that missing out on a lot of reflected, soft overtones could affect the sound that much.

Anyway, this is all just speculation off the top of my head. I could be totally wrong!

[RobAnderson]

How big is the hall?

What sort of materials are present in the room?

How far from the back wall was the piano?

Were you over the height of the lid?

[spm_gl]

If I understand the problem correctly, it's probably a stereo thing. Adding to phase-coherent correlated signals makes +6dB, adding non-correlated +3dB. Starting at a certain wavelength, both mics will receive more of the same signal, i.e. the bass signal becomes more similar.
Am I making any sense? Perhaps I should make a sketch.

Edit: (Trying to make more sense) With a low frequency and an A/B setup, we are actually sampling the same wave at two different points. The more similar the pressure level at these points is, the bigger the sum of the two.

[7rojo7]

Originally Posted by spm_gl

If I understand the problem correctly, it's probably a stereo thing. Adding to phase-coherent correlated signals makes +6dB, adding non-correlated +3dB. Starting at a certain wavelength, both mics will receive more of the same signal, i.e. the bass signal becomes more similar.
Am I making any sense? Perhaps I should make a sketch.

Edit: (Trying to make more sense) With a low frequency and an A/B setup, we are actually sampling the same wave at two different points. The more similar the pressure level at these points is, the bigger the sum of the two.

I never put omnis closer than about 6 feet apart, they just don't sound good close, it always becomes more mono the closer you put them. better just one. plusses and minuses. I use an xy with cards over an omni (near coincedent) and I never have a problem with woofy, but they're mixed in a certain manner (by ear, does that sound better? yeah. OK!) The floor is often the most heinous culprit.

[spm_gl]

We actually only do M/S nowadays, seldom XY, hardly ever A/B.

[locutus]

Originally Posted by RobAnderson

How big is the hall?
What sort of materials are present in the room?
How far from the back wall was the piano?
Were you over the height of the lid?

Well, to answer a bunch of questions. I will save a thousand words with a photo. I don't have a good photo of the configuration closer up. Sorry about the bad quality. It was just a handheld, lowlight snapshot with a cheap camera.

My notes say the mic were 238cm high from stage, 108cm to front leg, 232 to back leg. I tried spacings of 36cm, 40, 46, 50cm. I don't know precisely how far to the side/rear walls. The mic height was just barely below the plane of the lid, to eliminate possible lid reflections.

On another day I generally kept a 36cm spacing, and moved the mics to a dozen different positions, moved the piano around the stage, angled it off, etc.

I have been very suspicious of the side wall reflections. You can see they are angled, but flat in the lower part. I played with using choir shells to deflect the side wall reflections, but either I didn't have a large enough panel (6ft) for the wavelength, or it wasn't the issue.

[rumleymusic]

If I understand the problem correctly, it's probably a stereo thing. Adding to phase-coherent correlated signals makes +6dB, adding non-correlated +3dB. Starting at a certain wavelength, both mics will receive more of the same signal, i.e. the bass signal becomes more similar.
Am I making any sense? Perhaps I should make a sketch.

Phase is not really a problem in stereo signals, as long as those channels are not combined to mono.

I never put omnis closer than about 6 feet apart, they just don't sound good close

I would disagree with that. Standard AB spacing is 40 to 60 cm. The idea is to sepparate the mics enough so that the longer audio wavelengths will arrive at each mic sepparately, but not far enought as to create an acoustic gap. More than 3 feet is to much in my experience unless you add a middle mic.

I would surmise that this problem is due to diffuse field recording and the nature on the instrument being recorded. I have recorded pianos more times than I can count and I have come to the conclusion that they just don't register well when recorded with a stereo pair at a distance of more than a few feet. The loss of HF information and the added reflections from the side wall as the OP guesses is probably the culprit. Evey piano recording I have done with the mic spacing the OP described has turned out the same way, muddy, dull, and lacking in brilliance.

[7rojo7]

aah! for solo piano! xy or some kind of ortf, if you need more bass or center add an omni. Or AB but it's difficult to describe the placement without a drawing, depends on the recording space. To avoid a lot of noise i would put up 2 pairs, one for piano and the other for ambience
inverse square vs critical distance your looking at a MF to get it right with one pair
cards on the piano omnis for ambience

[mpdonahue]

I may be completely off base here, but I think that it has nothing to do with the frequency response of the mics or the spacing. My suspicion is that the mid range buildup you are hearing is actually coming off the bottom of the piano. For as long as I've been recording piano I've almost always found that the energy off the bottom of the soundboard reflected off the floor needs to be adjusted to balance the clarity with the warmth of the instrument. A little bit of absorption under the piano goes a long way.
All the best,
-mark

[rumleymusic]

For as long as I've been recording piano I've almost always found that the energy off the bottom of the soundboard reflected off the floor needs to be adjusted to balance the clarity with the warmth of the instrument.

Absolutely. That is why I like getting in close. Avoiding the direct reflections from the walls and most especially from the floor.

[locutus]

Originally Posted by mpdonahue

A little bit of absorption under the piano goes a long way.

I could be the one off base here, but assuming you are right, a "little" absorbtion would not go far enough. Considering the wavelengths involved, we would need some substantial absorbtion to accomplish this. A typical rug solution could not accomplish this.

[locutus]

I have had better luck with positioning the mics more off the tail end of the piano, like Plush's "butt end" solution or more precisely, similar to the Simon Eadon configuration described by RecordingDavid.

Even with the same room/piano/player, I felt I had less mud to deal with. (Unfortunately the pianist insisted on the straight out configuration.) Off the tail end I felt I had less mud and harshness, better presence and cleaner lows. However my conclusion for why this produced less low-mid buildup was that the mics had little distance differential to the side walls. (Mic axis closer to parallel to nearest wall.) Right answer, wrong reason?

Clearly, the distance to hammers, and axis of string vibration is a potential factor. But in relation to Mark's point about sound board reflections, how would tail end positioning affect the low-mid mud? Or why?

I fully acknowledge the great results produced by Plush and others with the multi-mic techniques. There certainly seems to be some advantage in tackling this issue. But something of the purist in me insists that there must be a way to produce a great piano recording with just a single pair. Others have done that, too.

I may border on the academic rather than practical side. But I believe understanding the whys is essential to not falling into a cookbook approach, and always pursuing a better sound.

Nathan

[rumleymusic]

But something of the purist in me insists that there must be a way to produce a great piano recording with just a single pair. Others have done that, too.

Of course. Most of the great piano recordings have used only a stereo pair. Following DPA's guidelines is a great way to start. I think 3 things are needed for a great sounding classical piano recording

a) A great piano obviously
b) Very good, transparent mics with excellent high frequency detail
c) Mic position (my favorite is close AB omni spacing which will give you a good, clear, stereo image of the instrument, and provide enough ambiance without mud or proximity effect.)

on a side note, a good hall is important too.

[David Spearritt]

Originally Posted by spm_gl

If I understand the problem correctly, it's probably a stereo thing. Adding to phase-coherent correlated signals makes +6dB, adding non-correlated +3dB. Starting at a certain wavelength, both mics will receive more of the same signal, i.e. the bass signal becomes more similar.
Am I making any sense? Perhaps I should make a sketch.

Edit: (Trying to make more sense) With a low frequency and an A/B setup, we are actually sampling the same wave at two different points. The more similar the pressure level at these points is, the bigger the sum of the two.

I think this aligns with my thinking and experiences. Observations as follows:

* Widening the spacing tends to reduce the problem, ie moves the coherent frequencies lower

* Moving away from the source, reducing the contributions from coherent additions of direct sound coming from glancing angles.

However, this doesn't gel with the stated phenomenon that two 4060's hanging just over the open piano lid, do not exhibit the problem. I haven't heard this so cannot comment. Perhaps the degree of omni-ness is relevant, although the 4006 is just as omni as a 4060 at the woofly frequencies.

Keep the theories coming in. Interesting thread.

[audio ergo sum]

Originally Posted by locutus

This may be either a bit academic or obvious to most of you.

I didn't want to hijack the other thread, but I am thinking about a comment in the Flat Omnis thread. Flat omnis?


OK, I have seen occasional references to this phenomenon, and I have experienced it myself. But I don't completely understand exactly why this happens from a physical acoustics standpoint.

Case in point. I recently did a recording of a solo piano (NY Steinway D) in a great hall, pair of MK2H, 2m out, high enough to avoid lid reflections, spacing 36cm. I had terrible problems with low-mid buildup.

But why exactly? I know that 36cm is 1/4 wavelength of 238Hz, which is roughly the center frequency of the problem area. It seems to consist of some odd phasiness, not just boosted frequency response. This is consistent with a 90-degree phase shift at 1/4 wavelength.

Now, to experiment I tried changing the spacing. Adding just 4cm changed the buildup frequency noticably. But even at 50cm is was still rather noticable. I didn't have time to compare larger spacings.

I have to conclude that this is a direct result of sidewall reflections. But this is a large hall. The sidewalls are at least 10m away. I experimented with using choir shell panels to divert sidewall reflections, but didn't have enough time to prove anything.

I can't believe it is simply related to the low frequency response of omnis compared to cardis. Most cardiods don't have much of any noticable frequency drop at 238Hz.

Does anyone have a handle on exactly why this happens?

Nathan

My guess:
It is not related to low frequency response of omnis but (lack of) high frequency off axis response in omnis.

If it was "a great hall" which had considerable energy in the diffuse field, you probably heard the roll off of the high frequencies off axis.

Plus it would help to know the frequency reponse / RT60 over frequency for that hall.

If the RT60 in that hall was longer in the range you describe, than in addition to the lack of off axis brilliance in MK2H could result in what you heard.

Plus the random unfortunate positioning relative to specific room modes at your mic position.

[just.sounds]

another theory

With ab highs are more uncorrelated than the lows.

Lows add 6 db hights 3 db when played back on speakers.

Check on headphones if this is the case

[BrianHanke]

Originally Posted by just.sounds

another theory

With ab highs are more uncorrelated than the lows.

Lows add 6 db hights 3 db when played back on speakers.

Check on headphones if this is the case

This is very interesting, and it's something I've noticed as well: on playback my headphones often show very little to no low-mid buildup, whereas my monitors will show a ton of it. Can anyone else confirm that they've experienced this and what the reasons behind the phenomenon might be?

[David Spearritt]

I have heard it in phones and equally with monitors. I have also experienced it with cardioid mics, but they need to be closer to the piano.

Maybe there's another explanation or an aggravating factor. How many of you are using mics with transformers?
PSW Recording Forums: Klaus Heyne's Mic Lab => Can you hear the transformer in some of these mics?

I am only using transformerless setups, but still hearing it.

[locutus]

Originally Posted by BrianHanke

my headphones often show very little to no low-mid buildup, whereas my monitors will show a ton of it.

What headphones? What monitors? What room?

The two primary factors that come to my mind are both acoustics related. First, one of my great annoyances in speaker design is that most smaller monitors leverage a higher Q in the low-mid to upper-bass region to substitute for the real bass they lack. It also gives them a somewhat wonky phase response in this region. Headphones don't have the same physics, so are not subject to these issues in the same way. (They have their own issues.)

Second, is room acoustics. I don't know how well your room acoustics are controlled, but most rooms I have seen are only partially controlled. And the low mid region is highly susceptible to issues here. Also in this frequency range, undamped sheetrock on studs can even resonate and cause issues. Headphones obviously don't have the same issues. (Unless your headphones are way too loud. )

Until I can build a proper/perfect room myself, I will continue to use headphones to crosscheck.

[locutus]

Originally Posted by David Spearritt

How many of you are using mics with transformers? I am only using transformerless setups, but still hearing it.

I also am using transformerless, CMC6 bodies. Certainly some mics are warmer, such as the Neumann samples posted recently. But the Schoeps omnis are rather flat on the bottom half.

Originally Posted by David Spearritt

I have heard it in phones and equally with monitors.

Same here. Consistent issue across multiple monitoring environments.

Originally Posted by David Spearritt

I have also experienced it with cardioid mics, but they need to be closer to the piano.

This continues to point to acoustic factors.

[locutus]

Originally Posted by David Spearritt

I think this aligns with my thinking and experiences. Observations as follows:

* Widening the spacing tends to reduce the problem, ie moves the coherent frequencies lower

This agrees with my observations. As little as a 4cm separation increase caused a noticable change.

Originally Posted by David Spearritt

* Moving away from the source, reducing the contributions from coherent additions of direct sound coming from glancing angles.

I am not sure if I am not coming to the opposite conclusion. Direct sound is largely phase coherent for all frequencies with AB < 1m. It is the indirect sound that is more phase incoherent, as it comes from all directions, and coherency decreases with decreasing frequency. (Or incoherency rises with wavelength.) So as the direct to indirect ratio increases you will notice less of the low-mid buildup. Or the more distant placements will exhibit more of this problem.

As evidence, placements very close or inside the piano do not exhibit the problem. This agrees with your following observation.

Originally Posted by David Spearritt

However, this doesn't gel with the stated phenomenon that two 4060's hanging just over the open piano lid, do not exhibit the problem.

Is this making sense?

[MaTr1x2051]

I've personally found the further away you move, the more LF 'buildup' you get. But its really not buildup... its just a lack of highs to balance it. Highs dissipate sooner.

Right now (and this changes frequently), I love a mix of ortf in the hall and ab closer in. But I use cards in both places. I find that cards with a gradual and pleasant rolloff sound better with the halls I work in than omnis do. With omnis, I've never found a placement that doesn't extremely favor the lows. I always need to eq them later, put a large rolloff in, and it always turns out more of a mess. Cards just give me the sound I'm after right from the start. But they usually aren't more than a few db down even at 20Hz.

There have been a few occassions when the room allowed omnis, and in those cases they sounded great. But I don't think its a miking 'problem' that we have here. I feel that its simply omnis don't work in all places. I've felt that way especially if there's a lot of wood around... just my experience.

In my case, instead of getting bogged down in figuring out the bass problems with omnis after trying them on multiple sessions in multiple configs, I just switched over to cards, despite what everyone will tell you about recording piano. They've sounded great, and I will be hesitant to ever go back to omnis again, unless the room is truely spectacular.

[BrianHanke]

Originally Posted by locutus

What headphones? What monitors? What room?

The two primary factors that come to my mind are both acoustics related. First, one of my great annoyances in speaker design is that most smaller monitors leverage a higher Q in the low-mid to upper-bass region to substitute for the real bass they lack. It also gives them a somewhat wonky phase response in this region. Headphones don't have the same physics, so are not subject to these issues in the same way. (They have their own issues.)

Second, is room acoustics. I don't know how well your room acoustics are controlled, but most rooms I have seen are only partially controlled. And the low mid region is highly susceptible to issues here. Also in this frequency range, undamped sheetrock on studs can even resonate and cause issues. Headphones obviously don't have the same issues. (Unless your headphones are way too loud. )

Until I can build a proper/perfect room myself, I will continue to use headphones to crosscheck.

It's definitely possible that either my monitors or my room have some resonances of their own that are skewing the sound. My bet is on my monitors: they're your typical ported nearfields. I've never understood the wisdom of attaching a small, drum-like cavity to a speaker. I'm going to do some tests in the next few days to see if I can work out exactly what is going on.