To crossfeed or not to crossfeed? That is the question...

[pinnahertz]

Stereo mics that are more than about 10 inches away from each other produce too much ITD and the directivity + set up of mics easily produce too much ILD.

And yet many feel that recordings made with widely spaced stereo pairs and mono spot mics sound just fine on headphones!

There is SO much more to creating an aesthetically pleasing recording than microphone configuration, and to evaluate recordings in any way using only that information (usually unknown and only presumed anyway) would simply be the equivalent of viewing the night sky while looking through a toilet paper tube.

Many orchestras are recorded using combinations of X-Y pairs (with "correct" ITD by your definition), widely spaced omnis, spot mics, possibly even a coincedent pair or two, all mixed together. The goal of nearly every recording is not to clinically replicate an event, since that is impractical even using binaural techniques. It is to present the sense of a performance that represents the intent of the creator given the limitations of stereo reproduction. A great many of these performances never existed acoustically at all! So how can you claim what is "correct"? Even recording with panned mono sources (which is hardly ever done exclusively) they can be placed spatially with delay and reverb.

Since you cannot measure ILD and ITD built into a recording, especially if there are multiple variations of each occurring simultaneously, and specific mic and mixing parameters are never published, coupled with the lack of full knowledge of the creators intent for his work in every playback situation and system, the need for, desired amount and type of crossfeed must.... that's MUST.... be the resul of subjective preference...only. There cannot be any rules. There cannot be any hard specifications, because of everything we don't know.

And around we go again, with more circularity in discussion than sound in 5.1 surround.

 

[jgazal]

Since you cannot measure ILD and ITD built into a recording, especially if there are multiple variations of each occurring simultaneously, and specific mic and mixing parameters are never published, coupled with the lack of full knowledge of the creators intent for his work in every playback situation and system, the need for, desired amount and type of crossfeed must...

Thank you for that information!!!

 

[pinnahertz]

Thank you for that information!!!

You'd think this would be completely obvious by now, but...guess not. Perhaps it's inertia from all the spin in the other direction.

 

[jgazal]

You'd think this would be completely obvious by now, but...guess not. Perhaps it's inertia from all the spin in the other direction.

Well, I could only have imagined that if one previously had a RIR in the exact location where the microphones were in the recording.

But as Claude Lévi-Strauss once said: "The scientist is not a person who gives the right answers, he is one who asks the right questions."

 

[pinnahertz]

Well, I could only have imagined that if one previously had a RIR in the exact location where the microphones were in the recording.

Measurements of any kind are not often part of recording sessions.

 

[bigshot]

Tracking is a creative and evolutionary process. You don't know when you start what you are going to end up with. All you have is a demo of the song with the bare bones of it. Then you start building and tailoring tracks based on what you've recorded before.

 

[71 dB]

I see it.

So do you digitally analize the ILD and ITD parameters in the recording using some software the allows to do that and then you are able to know how the recording was made and how much ILD and ITD it has?

And then do you adjust your algorithm accordingly?

Do you mind telling me what sofware I can use to discover ILD and ITD from a given recording?

I beleive the Realiser with a crossfeed free PRIR would allow to spatially perceive those ITD differences, but it would be nice to have a sotware that allows to numerically/quantitatively confirm such perceptions...

Not really. I listen to the sound and set the crossfeed level so that spatial distortion just disappears. I have writen analyser for ILD, but simulating human hearing isn't easy, at least not to me.

 

[pinnahertz]

Not really. I listen to the sound and set the crossfeed level so that spatial distortion just disappears. I have writen analyser for ILD, but simulating human hearing isn't easy, at least not to me.

What is your indication that spatial distortion has disappeared?

An analyzer that measured ILD would still not tell you how much cross-feed to apply to correct for spatial distortions anyway.

 

[71 dB]

And yet many feel that recordings made with widely spaced stereo pairs and mono spot mics sound just fine on headphones!

Yes, even I did feel that way, but then again I was spatially ignorant back then...

Many orchestras are recorded using combinations of X-Y pairs (with "correct" ITD by your definition), widely spaced omnis, spot mics, possibly even a coincedent pair or two, all mixed together. The goal of nearly every recording is not to clinically replicate an event, since that is impractical even using binaural techniques. It is to present the sense of a performance that represents the intent of the creator given the limitations of stereo reproduction. A great many of these performances never existed acoustically at all! So how can you claim what is "correct"? Even recording with panned mono sources (which is hardly ever done exclusively) they can be placed spatially with delay and reverb.

XY has zero ITD which is not optimal, because in real life we have ITD up to about 640 µs for sounds that aren't coming from very near. For speakers XY creates a narrow sound image because of the zero ITD information.

The purpose of music is enjoyment, and if for example use of crossfeed increases that enjoyment, it's hard for me to deny it is the right thing to do. I don't believe that most recording engineers understand (some do understand, e.g. Jürg Jecklin) that well human hearing. They understand how to get balanced sound. .

Nothing wrong with panned mono tracks as long as you do it in a way that makes sense (ITD + ILD + perhaps some spectral stuff). I make computer music with very little recorded elements. I have to create the spatial information and out of principle I use only free software, that is Garage Band and Audacity. I export raw tracks from Garage band to be processed in Audacity using the plugins I have written. So, I am creating spatial information from scratch all the time and that's why I think I understand that stuff pretty well. If you use expensive professional software and effects you don't know what those effects are made of and you don't learn about human hearing.

Since you cannot measure ILD and ITD built into a recording, especially if there are multiple variations of each occurring simultaneously, and specific mic and mixing parameters are never published, coupled with the lack of full knowledge of the creators intent for his work in every playback situation and system, the need for, desired amount and type of crossfeed must.... that's MUST.... be the resul of subjective preference...only. There cannot be any rules. There cannot be any hard specifications, because of everything we don't know.

What counts is how good it sounds. Our ears are amazing ITD/ILD detectors. That's why spatially enlightened listeners get annoyed by excessive ITD and ILD. We can have guidelines such as "limit ILD below 500 Hz to 6 dB." or "avoid ITD more than 640 µs below 1 kHz." ILD and ITD is easiest to control while recording and creating the music. We can estimate the theoretical maximum ILD and ITD values of a given mic set up for example. Say we have an AB pair with 1 m (40 inches) distance between the mics. The mics are at a distance where the orchestra to be recorded covers ±45° angle for the mics. This means the sound from the far left instruments arrive to left mic sin (45°) * 1 / 345 s = 2 ms earlier than to the right mic. Omnidiractional mics create ILD due to distance difference. If the mics are 5 meters away from the closest players, the theoretical maximum ILD is 20*log10 (6/5) = 1.6 dB. So, we see ILD won't be a problem, but ITD will be. For a XY pair it's the other way around etc.

 

[71 dB]

What is your indication that spatial distortion has disappeared?

It sounds spatial distortion free. No different from say harmonic distortion. If you can reduce it, at a certain point the distortion goes below hearing threshold.

An analyzer that measured ILD would still not tell you how much cross-feed to apply to correct for spatial distortions anyway.

Actually it does if we know the threshold (target) value. I have wrote analyser plugin for "D" value:

D = S/(S+M),

where S is absolute value of L-R and M is absolute value of L+R.

D=0 means mono sound. D=1 means L and R are out of phase (antimono). If the target value for ILD is 3 dB, it means R = 0.7 * L.

M = L + R = L + 0.7*L = 1.7*L
S = L - R = L -0.7*L = 0.3*L
D = S / (S + M) = 0.3*L / 2*L = 0.15

Now, if the analysed D value in 0.4 for example, we can calculate the needed correction:

x = 20*log10 ((0.4-0.15)/(0.4+0.1-2*0.5*0.1)) = 20*log10 (0.25/0.4) = -4.1 dB.

 

[pinnahertz]

Yes, even I did feel that way, but then again I was spatially ignorant back then...

You have just defined anyone who doesn't share your opinion as "ignorant".

Again.

Your arrogance is offensive, even if someone partially agrees.

XY has zero ITD which is not optimal, because in real life we have ITD up to about 640 µs for sounds that aren't coming from very near. Also, the cardioid mics of XY create excessive ILD at low frequencies.

No, you need to examine the LF pattern of a few cardioid mics. Besides, X-Y is only one of many stereo mic arrays.

For speakers XY creates a narrow sound image because of the zero ITD information.

Again, incorrect. The image location of sound between speakers does not depend on ITD alone, or even primarily. If it did the ubiquitous "pan pot" wouldn't pan.

XY isn't that good for much anything, but it's not very bad either.

Its great for producing stereo recordings that are highly mono-compatible.

The purpose of music is enjoyment, and if for example use of crossfeed increases that enjoyment, it's hard for me to deny it is the right thing to do.

Not always. And not for everyone. This is, again, arrogance.

I don't believe that most recording engineers understand (some do understand, e.g. Jürg Jecklin) that well human hearing. They understand how to get balanced sound.

Arrogance! To think that those that have a career in making commercial recordings are not as "enlightened" as you, a rank amateur! Just incredible!

Nothing wrong with panned mono tracks as long as you do it in a way that makes sense (ITD + ILD + perhaps some spectral stuff).
No pan pot works that way. You know that, so what you're doing here is dissing the entire recording industry relative to your own enlightenment.

I make computer music with very little recorded elements. I have to create the spatial information and out of principle I use only free software, that is Garage Band and Audacity. I export raw tracks from Garage band to be processed in Audacity using the plugins I have written. So, I am creating spatial information from scratch all the time and that's why I think I understand that stuff pretty well. If you use expensive professional software and effects you don't know what those effects are made of and you don't learn about human hearing.

So... let me make sure I understand this.... professionals with life long careers in recording who invest in professional tools don't understand them, but an amateur who spends nothing on free tools does?

You see nothing even slightly wrong with that statement?

What counts is how good it sounds. Our ears are amazing ITD/ILD detectors.

Detectors, yes. Objective evaluators, no.

That's why spatially enlightened listeners get annoyed by excessive ITD and ILD. We can have guidelines such as "limit ILD below 500 Hz to 6 dB." or "avoid ITD more than 640 µs below 1 kHz." ILD and ITD is easiest to control while recording and creating the music. We can estimate the theoretical maximum ILD and ITD values of a given mic set up for example. Say we have an AB pair with 1 m (40 inches) distance between the mics. The mics are at a distance where the orchestra to be recorded covers ±45° angle for the mics. This means the sound from the far left instruments arrive to left mic sin (45°) * 1 / 345 s = 2 ms earlier than to the right mic. Omnidiractional mics create ILD due to distance difference. If the mics are 5 meters away from the closest players, the theoretical maximum ILD is 20*log10 (6/5) = 1.6 dB. So, we see ILD won't be a problem, but ITD will be. For a XY pair it's the other way around etc.

You still don't seem to understand the reality of recording. The number of recordings made with a single stereo mic pair is about as small as the number of recordings made binaurally. The vast bulk of stereo material had been recorded with many Mocs in many positions and configurations all mixed together. There is no way to extract or even to anticipate ILD or ITD because there are multiple values of each within the mix, and unknown artistic intentions. Any presumption that these values could be known or compensated for is the very panicle of arrogance and superciliousness.

But arrogance and superciliousness have become the theme of this thread.

 

[gregorio]

Science is on my side. All you have is "artistic intent". Please.

Huh? How many commercial music recordings do you have which have been made by scientists and/or to purely scientific principles? No, commercial music recordings are made by artists and "artistic intent" is always the fundamental and overriding concern. Your statement is therefore complete nonsense and pretty much backwards! And that's even if science were even on your side to begin with!

[1] I sense that you almost fear stereo sound with natural ILD and ITD [2] but that's not a limitation really because there's so much more you can do in music, so much other possibilities for artistic intent.

1. I can't answer for pinnahertz but I certainly do! It would in effect mean dumping all the progress made in the last 50 years or so in music recording techniques, mixing and production, along with all the modern popular music genres, plus narrative film and TV sound. That's pretty scary to me!
2. Such as? I can't think of anything more limiting to artistic intent than dumping all the technical and artistic progress of the last 50 years!

[1] XY has zero ITD [2] which is not optimal, because in real life we have ITD up to about 640 µs for sounds that aren't coming from very near. [3] Also, the cardioid mics of XY create excessive ILD at low frequencies. [4] For speakers XY creates a narrow sound image because of the zero ITD information. [5] XY isn't that good for much anything ...

1. No it doesn't but pretty close. 2. The opposite is true! As near to zero as possible is in fact optimal, to avoid phase artefacts.
3. No they don't! 4. No it doesn't! 5. Yes it is! In fact it's probably the most commonly used stereo mic technique.
3 and 4 depend entirely on: What you're recording, where you're recording it, how the XY pair is positioned relative to what you're recording and artistic intent!

I'm not sure where you've got all this nonsense from, whether you've zero practical experience of mic'ing techniques and are just misinterpreting what you've read or if you're deliberately misrepresenting the facts just to support your opinion?

[1] What counts is how good it sounds. [2] Our ears are amazing ITD/ILD detectors. [2a] That's why spatially enlightened listeners get annoyed by excessive ITD and ILD.

1. Absolutely!
2. They are only potentially amazing given an exact HRTF and even then they can be pretty easily fooled.
2a. That's presumably why you've managed to convince yourself that you're spatially enlightened when in fact your statements indicate the exact opposite! For example:

Open headphones leak some sound and there is very minor acoustic crosstalk happening, but unless you do something to the signal entering your headphones, none of the violet stuff is happening. That's why some people including me find headphone listening without crossfeed unnatural, annoying, spatially broken and tiring.

With extremely few exceptions, whether you're listening to headphones with crossfeed, listening to them without crossfeed or even listening on speakers in an acoustically excellent room, the spatial information of what you're listening to (commercial music, TV/films) is ALWAYS "unnatural" and "spatially broken". The only explanation I can think of for why you seem completely unaware/oblivious/ignorant of this basic fact is that you are spatially ignorant/un-enlightened.

[1] Nothing wrong with panned mono tracks as long as you do it in a way that makes sense [2] (ITD + ILD + perhaps some spectral stuff).

1. Again, absolutely agreed. 2. Nonsense, virtually all commercial music/audio comprises at least partly, if not mostly or entirely of panned mono tracks with little or no ITD and a fair amount of it with no ILD either and it makes perfect sense! Or more precisely, our brains can make perfect sense of it.

[1] I listen to the sound and set the crossfeed level so that spatial distortion just disappears.
[2] I don't believe that most recording engineers understand (some do understand, e.g. Jürg Jecklin) that well human hearing.
[2a] So, I am creating spatial information from scratch all the time and that's why I think I understand that stuff pretty well.
[2b] If you use expensive professional software and effects you don't know what those effects are made of and you don't learn about human hearing.

1. Spatial distortion never disappears, unless you remove ALL the spatial information and that would sound pretty ridiculous. Maybe it just seems to disappear to you personally because you are unaware of it/spatially ignorant?
2. That's clearly complete nonsense. You somehow don't seem to realise that being a sound engineer is an extremely competitive career or even that the two requisites which determine their success is the quality of their work and their efficiency. Neither of which they could achieve without excellent judgement of human hearing perception.
2a. You mean that's how you've managed to convince yourself that you know more than anyone else.
2b. And this is why this is all "clearly complete nonsense"! You're making a correlation where there is none. Is an F1 car mechanic a better driver than a professional F1 driver because he has a far better understanding of how the car works? What about a fighter plane designer being a better pilot than a fighter pilot, a tennis racket maker, a designer of surgical equipment, a piano manufacturer, was Stradivarius a legendary violinist? The best mixes are made by the best sound engineers/artists (typically using professional software), not by those who design/code that software. Isn't this obvious, do you really believe what you're saying or just making it up to justify your belief?

You've made it abundantly clear that you have a preference for the "unnatural, broken spatially" reproduction of HPs with crossfeed. I generally prefer the "unnatural, broken spatially" reproduction of HPs without crossfeed and better still, the "unnatural, broken spatially" reproduction of speakers in a good acoustic. So what all this comes down to is how each of us personally prefers their "unnatural, broken spatially" reproduction. What I object to is you coming out with all kinds of nonsense to try and contradict the basic facts and prove that your reproduction preference is in fact not "unnatural, broken spatially", simply because you're apparently unable to hear it and then calling everyone who can hear it and/or do know the basic facts "spatially ignorant"! Please, enough already!

G

 

[bigshot]

Yes it is!

No it isn't!

You are wrong!

No I'm not!

Yes it is!

No it isn't!

You are wrong!

No I'm not!

Yes it is!

No it isn't!

You are wrong!

No I'm not!

 

[pinnahertz]

^^^^How is this helpful?

 

[bigshot]

That's exactly what the recent posts look like to me. Go through and read just the replies without reading all the quotes. That's what most of us out here in forum land do. The amount of actual content has dropped precipitously. But if you're having fun, feel free to continue.