Orchestra + Soloist

[fifthcircle]

So 13 years of making this work is wrong? Comeon folks...

And Roland "In the real word?"

Yes, in the real world. I do this on a regular basis and I don't give a care as to what the books should say is possible. Yes, it does work. Yes, you need to get your measurement right. Yes, it helps put stuff in perspective by not making the solo feel so "close." and in many occasions, it has removed comb filtering issues because of the 5-8 ms of delay difference between the mains and a solo spot.

Of course nothing is perfect- I've heard comb filtering on an ORTF pair. Sure, whenever you have 2 mics with any distance and a single point of source, you can have issues that range from timing to image problems. This is part of the reason why I almost never use A-B setups. Too easy to have problems. But to outright discount a delay on a spot microphone is IMO a pretty poor idea.

--Ben

[Roland]

I've mentioned this before, but I will state it again, about 15 years ago with the roll out of the first Yamaha high end digital console which had the ability to delay any channel, a series of listening tests were done. The panel included some very illustious enginners and industry personel. The results were that universally the undelayed spots were preffered. If you really think about what is happening, (draw a picture if it helps), there are good reasons why this might be the case. The notion that delaying a spot mic by the distance away from the main mic stands somehow puts everything to right is also nonsense. 1 ms is equilelent to approx 1ft in distance, the wavelength of 1ft is around 1.2khz, half that distance would be 2.4khz, not compensatable for as it's a half ms, bang in the middle of the audio spectrum, a horrible frequency if you are considering vocals (a very frequent spot miking job). I'm not even going to go into the issues of a vocalist that may move around as they sing.

Comb filtering as a result of ortf or AB techniques can only be on playback.


Regards


Roland

[ghellquist]

I agree on delaying the spot mic.

My personal experiences of this is that it sometimes helps, sometimes not, but it is worth trying.

The effect I hear in mixing is that when i delay the spot mic I hear the solo instrument better in the mix. This means that I can turn down the level on the solo mic a little, having less bad effect on the rest of the sounds.

I do a rough calculation for the time to delay as about 3ms per meter from main pairs to solo mic and then add a little extra.

There is a, more or less theoretical but still, explanation for this in what is called the Haas effect. Now, when recording music, theory and practice is often not the same thing.
Haas effect - Wikipedia, the free encyclopedia

Gunnar

[ghellquist]

Originally Posted by Roland

... The results were that universally the undelayed spots were preffered. ...

Just curious, is there any writeup of this experiment on the net? As always I would like to learn more.

// Gunnar

[Roland]

Originally Posted by ghellquist

Just curious, is there any writeup of this experiment on the net? As always I would like to learn more.

// Gunnar

The article was in Studio Sound, unfortunately now long defunct.

Roland

[d_fu]

Roland,

Originally Posted by Roland

The results were that universally the undelayed spots were preffered.

Would you happen to remember why they were preferred?

1 ms is equilelent to approx 1ft in distance, the wavelength of 1ft is around 1.2khz, half that distance would be 2.4khz, not compensatable for as it's a half ms, bang in the middle of the audio spectrum, a horrible frequency if you are considering vocals (a very frequent spot miking job).

How do these numbers alone explain potential (comb filtering) issues? Without delaying spots, considering the distance of soloists from the mic etc., you will not have the signals from the main and spot mic totally phase aligned, either (even if you nail the soloist to the floor). The good thing about the delay is that you can yary the exact time, it is not fixed. If a certain delay time will cause an audible phase issue, you can alter it, can you not?

I use delay most of the time. Sometimes not on soloists up front, but mostly on choirs, winds, etc.


Daniel

[Roland]

Originally Posted by d_fu

Roland,

Would you happen to remember why they were preferred?

It's a good number of years ago and I don't think that any particular reasons were given except the sounds more solid, intergrated, 3D generic conclusions. Possibly what was more interesting was that most were surprised at there conclusions as they expected to prefer the delayed version.

How do these numbers alone explain potential (comb filtering issues)? Without delaying spots, considering the distance of soloists from the mic etc., you will not have the signals from the main and spot mic totally phase aligned, either (even if you nail the soloist to the floor). The good thing about the delay is that you can yary the exact time, it is not fixed. If a certain delay time will cause an audible phase issue, you can alter it, can you not?

I've use delay most of the time. Sometimes not on soloists up front, but mostly on choirs, winds, etc.


Daniel

They don't, however they do point out the frequencies where you may potentially get problems. Of course there are other factors and variables to take into account, but to blanket say that you can deal with these issues using delays is incorrect. It is quite possible that using delays may, on some occasions, subjectively be better, it may also be considerably worse, but the most important factor is to keep an open mind and (IMHO) understand why this may very well be the case.

Regards


Roland

[fifthcircle]

Just trying to put another viewpoint out there... Did similar tests years ago in 2 different locations with two different sets of engineers. One used a Yamaha O2R, the other an O3D.

The biggest issue in those tests was a set of engineers that was preconditioned to listen without *any* delays due to the fact that they had always worked in an analog paradigm where delays are nearly impossible to deal with.

The question of the comb filtering is not just in relation to the wavelength (although that tells you the frequencies that are going to have that sound), but in relation to time. The reason I have found more comb filtering is because of a time-domain problem. ie- the same wave is hitting two points at different times, yet are being presented in the mix at the exact same time. The whole purpose of time alignment (delaying the spots) is to keep this from happening.

Places I don't delay- when I have a ton of mics out there- either a scoring-style session or when mixing live sound. Sometimes when the spots are far enough away tha I don't have to worry about the comb filtering- in those cases, the gain of the mic and careful mix of ambience can make the close feeling of the mic go away.

I almost always *do* delay on a soloist- the fact that the mic is usually close and there are issues with the main pair makes it important. As I said before, though, you need to make sure you have the delay time right. You can cause as many problems as you can fix.

--Ben

[dtf]

Originally Posted by fifthcircle

The question of the comb filtering is not just in relation to the wavelength (although that tells you the frequencies that are going to have that sound), but in relation to time. The reason I have found more comb filtering is because of a time-domain problem. ie- the same wave is hitting two points at different times, yet are being presented in the mix at the exact same time. The whole purpose of time alignment (delaying the spots) is to keep this from happening.

I am not sure if I understand what you are trying to say. To begin with, I would be surprised if the sound of e.g. a first violin section reaching the spot-mic and later the mains is identical / more or less "the same". I could only imagine this with one instrument with a very directional sound radiation pattern towards the mics. If it really were like this, I would only expect problems when using a coincident main setup, as there are a whole lot of other things happening with any kind of spaced array.

Secondly, in the mix balance the two signals are not added "at the same", because the actual time relationship still exists. The signal of the undelayed spot comes earlier than the mains and can be heard with much less level than a delayed one (masking mechanism of human hearing), with the benefit that the signals won't influence each other too much. Hence no (or much less) comb-filtering.

Originally Posted by fifthcircle

Places I don't delay- when I have a ton of mics out there- either a scoring-style session or when mixing live sound. Sometimes when the spots are far enough away tha I don't have to worry about the comb filtering- in those cases, the gain of the mic and careful mix of ambience can make the close feeling of the mic go away.

Many engineers successfully use a reverb on undelayed spots to blend them. I couldn't imagine a situation where a spot is so close to the mains that there would be audible comb-filtering. In such a case I would move either the mains or soloist to a preferable location.

Best,
Dirk

EDIT: This doesn't mean, of course, that there can be a situation in which a delay will deliver the best result.

[RobAnderson]

dtf sez:

I couldn't imagine a situation where a spot is so close to the mains that there would be audible comb-filtering. In such a case I would move either the mains or soloist to a preferable location.

I have seen this - most often in the case of flown spot mic's.

For me, if I have a spot mic, I want whatever it is I am spotting to be prevalent (if it's pointing at the violin, I want to hear the violin), but some engineers are going for a more integrated sound. Most of these guys know how to make it work quite well, but I am a bit nervous when the mic is that far away.

d_fu says:

(even if you nail the soloist to the floor)

Now why didn't I think of this?

As for delays:
I have tried using delayed spots on different occasions, but have rarely preferred them to the "undelayed" version. Some people get them to work quite nicely, but perhaps they are more used to using this technique than I.

However, I have often found that simply delaying the spot does not always fix the comb filtering problem - sometimes it can, but many times you are simply stuck with it. Not saying it hasn't worked for you Ben, but for me there were times when it simply was not fixable, which is why I always try to get it right with the microphones at the point of the recording.

It is not always easy to guess what you will or will not be able to fix in post. Personally, I would rather fix it right there and then.

[maestro]

the dpa lapel mic would pick up the sound of the performer's clothing as he/she bows & moves. amazing mic though in the right situation.

[chris319]

At the moment I'm leaning toward using a spaced pair to record this, putting both mics way up high and downstage of orchestra and soloist both. The soloist will play into the left channel mic. Problem solved, hopefully, as the mics will no longer be upstage of the soloist. Now if we can only deal with the tone-deaf second violinist (not the soloist).

[WVUtubadude]

Originally Posted by chris319

Now if we can only deal with the tone-deaf second violinist (not the soloist).

New venture for autotune?

[maestro]

Originally Posted by chris319

At the moment I'm leaning toward using a spaced pair to record this, putting both mics way up high and downstage of orchestra and soloist both. The soloist will play into the left channel mic. Problem solved, hopefully, as the mics will no longer be upstage of the soloist. Now if we can only deal with the tone-deaf second violinist (not the soloist).

Put the spot mic on the soloist. Better to have it and not need the track than . . .

[chris319]

Originally Posted by maestro

Put the spot mic on the soloist. Better to have it and not need the track than . . .

This will be a live mix.

[fifthcircle]

Originally Posted by RobAnderson

dtf sez:
However, I have often found that simply delaying the spot does not always fix the comb filtering problem - sometimes it can, but many times you are simply stuck with it. Not saying it hasn't worked for you Ben, but for me there were times when it simply was not fixable, which is why I always try to get it right with the microphones at the point of the recording.

It is not always easy to guess what you will or will not be able to fix in post. Personally, I would rather fix it right there and then.

I find the greater the space of your main pair, the greater the chance of an issue with comb filtering. I can't say that I've never had an issue between a soloist and a coincident pair, but I can't remember ever having that issue. Then again, I tend not to use particularly widely spaced pairs, even for orchestral work. The biggest I get is my decca tree that is probably only 3 feet or so across the back. If the tree is too far out, though, and I have a vocal soloist, I can have issues that nothing will solve. Put a blumlein pair out front, though, and things will be solid and a small delay on your solo mic makes a world of difference.

The times I use a tree most often are with choral performances with orchestra and piano concertos. Put the tree over the conductor's head and things work out quite well.

And I absolutely agree when it comes to getting things right from the beginning. If I hear an issue, I fix it. Period- I never "fix it in the mix." I learned early on that problems are best fixed by mic choice/position. Not by covering it up with some effect.

--Ben

[audio ergo sum]

Originally Posted by fifthcircle

...
The question of the comb filtering is not just in relation to the wavelength (although that tells you the frequencies that are going to have that sound), but in relation to time. The reason I have found more comb filtering is because of a time-domain problem. ie- the same wave is hitting two points at different times, yet are being presented in the mix at the exact same time. The whole purpose of time alignment (delaying the spots) is to keep this from happening.
...

The echo threshold is level and signal dependent between 30-80 ms.
Anything with less time difference will be perceived by the brain as one sound source yet with coloration (Haas effect).

Originally Posted by Roland

...Of course there are other factors and variables to take into account, but to blanket say that you can deal with these issues using delays is incorrect. It is quite possible that using delays may, on some occasions, subjectively be better, it may also be considerably worse, but the most important factor is to keep an open mind and (IMHO) understand why this may very well be the case.
...

+1 +my € .02
the more the sounds of main and spot resemble each other spectrally and the levels in the mix are in the same range (±6 dB) the more problematic comb filtering can become.

Another factor when thinking about to delay or not to delay is, what we need from the spot. If we need a little transient and presence/brilliance then I would not delay, or maybe delay a little but not the full distance to the main array. But if the stationary sound of the source's tones shapes the sonic impression more and we need amplification with spectral quality intact, I would delay.

It is a delicate issue, not black and white at all. An artistic decision basically that shapes the quality of the sound, no right or wrong here.
As to the listening tests: It takes a while to get used to the new sound. If you never worked with delays it is unlikely you would like them, even if they sound better
I had situations where without a delay the singers would have sounded unbearable. Other mics in the same situation sounded dull if used with delay.

Any digital mixing desk or DAW should have a continuous delay pot as prominent and easy to access as a pan pot. Delay is a powerful tool to shape sound in mixing.
It influences perceived depth, transient response, timbre.

[dtf]

Originally Posted by audio ergo sum

the more the sounds of main and spot resemble each other spectrally and the levels in the mix are in the same range (±6 dB) the more problematic comb filtering can become.

This is exactly why not delaying a spot can be favourable. It will not be needed to be mixed in with a higher level. Apart from that I don't believe that, in a typical situation, the signals are really that equal spectrally. And if they were, this would only become a problem the moment they are time-alligned by delaying.

Originally Posted by audio ergo sum

Another factor when thinking about to delay or not to delay is, what we need from the spot. If we need a little transient and presence/brilliance then I would not delay, or maybe delay a little but not the full distance to the main array. But if the stationary sound of the source's tones shapes the sonic impression more and we need amplification with spectral quality intact, I would delay.

I never quite understood the principle of this reasoning. Either you can hear a spot or not - there really is not such a large grey area here. Many factors dictate the way how you can use a spot microphone (placement of mains, placement of spot, their relation, signal content etc.), so either you do it that way or you change some of the contributing factors.

Originally Posted by audio ergo sum

As to the listening tests: It takes a while to get used to the new sound. If you never worked with delays it is unlikely you would like them, even if they sound better

As much as I know how preconceived ideas can influence the way you hear, I find this a rather weird thing to accuse professional colleagues of. It all depends on the actual listening test, how it was prepared and performed, and on the test subjects. All of which we don't know.

Originally Posted by audio ergo sum

Delay is a powerful tool to shape sound in mixing. It influences perceived depth, transient response, timbre.

Exactly. Mix in a spot mic and then vary the delay - now we're talking comb filters! A delay really works as an EQ as well.

I completely agree with the notion that occasional delaying spot mics can be better, as I have had the same conclusion (though very rarely). In most cases, however, as have been described here as well, I suspect the cause to be not ideal mic placement. Therefore, and given the explanations here, it would be wrong to believe that delaying is a constructive tool in acoustical recording in everyday studio-life.

Dirk

[audio ergo sum]

Originally Posted by dtf

...I don't believe that, in a typical situation, the signals are really that equal spectrally. And if they were, this would only become a problem the moment they are time-alligned by delaying.

Not my experience at all, but I have a suspicion that there also are big differences in how individuals process sound. Some are more sensitive to timbre alterations than others it seems.

It becomes a fascinating subject onces we think about, how much of a sound we actually need to trigger that stored sound in our brain and we literally start to imagine.

I never quite understood the principle of this reasoning. Either you can hear a spot or not - there really is not such a large grey area here. Many factors dictate the way how you can use a spot microphone (placement of mains, placement of spot, their relation, signal content etc.), so either you do it that way or you change some of the contributing factors.

Yes, and time difference between spot and main is one contributing factor that can be changed
Certainly easier than asking in an orchestra live recording the harp to come closer to the main mic.

Except in the analog days. There it would have been easier to ask the harp to come closer.

... Therefore, and given the explanations here, it would be wrong to believe that delaying is a constructive tool in acoustical recording in everyday studio-life.

That is certainly wrong. Delaying is a powerful tool that can be used constructively or destructively. It depends on the operator.

Saying that delaying is a priori destructive does not make sense.
Saying that always time differences between mics have to be compensated is also nonsense. (and impossible).