There Are Different Kinds of Pan Pots?

[Samulis]

Sound travels 1.2 feet (37 cm) in 1 millisecond. Typical distance between human ears is something like 7-9 inches I think? As such, you want the aforementioned 0.6 ms or so to be the maximum (for something hard left/right like basses, tuba, harp, etc. depending on layout) and lesser values for most items. Like trombones or the such could be 0.3-0.4 ms delayed on the left (to push them moderately right).

For attenuation, consider that the cardioid pattern attenuates at right angles by about 3 dB at 1000 Hz (it generally attenuates higher frequencies more and low frequencies less, coloring off-axis sounds). Thus a reduction in the area of 3-6 dB for the more distant channel when fully panned should combine nicely if you're "modeling" ORTF/NOS/DIN/XY/Mid-Side. Attenuation due to distance is fairly minor for main/room/distant mics, as the ratio of distance is small (and the contributions of the far field, which is a whole different matter).

[Nayrb]

(01-05-2021, 01:44 AM)Samulis Wrote: Sound travels 1.2 feet (37 cm) in 1 millisecond. Typical distance between human ears is something like 7-9 inches I think? As such, you want the aforementioned 0.6 ms or so to be the maximum (for something hard left/right like basses, tuba, harp, etc. depending on layout) and lesser values for most items. Like trombones or the such could be 0.3-0.4 ms delayed on the left (to push them moderately right).

For attenuation, consider that the cardioid pattern attenuates at right angles by about 3 dB at 1000 Hz (it generally attenuates higher frequencies more and low frequencies less, coloring off-axis sounds). Thus a reduction in the area of 3-6 dB for the more distant channel when fully panned should combine nicely if you're "modeling" ORTF/NOS/DIN/XY/Mid-Side. Attenuation due to distance is fairly minor for main/room/distant mics, as the ratio of distance is small (and the contributions of the far field, which is a whole different matter).

Excellent! Thanks for the numbers, Sam! I will experiment. By "attenuation," you mean the volume reduction caused by balance panning?

[Mattias Westlund]

(01-04-2021, 10:41 PM)Nayrb Wrote: Out of curiosity, do you slap that plugin on everything or just certain instruments? I've been trying to be deliberate and to choose wisely which instruments I make these changes to. But I'm not really sure yet of the best way to go about it.

Right now, only on strings and brass (i.e. on every section channel). I've tried it on woodwinds, but considering the moderate panning required there, I can't hear a difference between the plugin and Reaper's standard stereo panner. I have not tried it on percussion or anything else yet. To avoid unnecessary clutter, I'm sort of listening for problematic instruments rather than putting it on every channel in the mix, just because.

[Nayrb]

(01-07-2021, 03:11 AM)Mattias Westlund Wrote: Right now, only on strings and brass (i.e. on every section channel). I've tried it on woodwinds, but considering the moderate panning required there, I can't hear a difference between the plugin and Reaper's standard stereo panner. I have not tried it on percussion or anything else yet. To avoid unnecessary clutter, I'm sort of listening for problematic instruments rather than putting it on every channel in the mix, just because.

Pretty much my approach as well. My percussion is just so huge sometimes that I thought I'd try it out along with some side EQ. So far I'm liking the results. I agree, a stereo pan for winds and things nearer the center is probably all you need.

I've come across a new, related issue: All my VO life, I've often encountered certain notes of certain sample libraries that seemed to just veer off in one direction or another, making positioning difficult. I've since discovered that, for whatever reason, it's because there is literally more volume in one channel or another in the sample itself. Like, the right channel is louder than the left, for instance. I can see this visually under the hood in Kontakt. It's a surprisingly common issue in my older libraries (like, anything recorded pre-2010).

For the longest time, I just assumed it had to do with some aural/proximity phenomenon, but the wave editor indicates that it is actually just something up with the sample itself--It's very annoying! My approach going forward will be to pan each offending sample (or "zone," as they're called in Kontakt) individually and cautiously in an effort to mitigate some of the issues that it can introduce into a mix. It's the only thing I can think to do at the moment

[Mattias Westlund]

I definitely know the problem you describe, and yes, it's relatively common in older libraries. It still happens, though: IIRC there are some samples in Miroslav 2 that sound completely wonky in places, like the players suddenly moved off to the side for one or two notes. It's incredibly annoying and it's hard to say what the cause may be. I guess it might be a recording issue. Placing mics and players EXACTLY in the same spot between two different sessions is virtually impossible, and even slight discrepancies become much more pronounced when recording in stereo. Maybe they went back to re-record some bad notes and weren't able to set things up exactly right the second time, perhaps due to time or budget constraints. But who knows.

I'm not super familiar with the editing capabilities of Kontakt yet, but if it's possible to perform edits to the samples, maybe you could just go in and attenuate the offending channel until the sample matches up with the rest?

[Nayrb]

(01-07-2021, 05:47 AM)Mattias Westlund Wrote: I definitely know the problem you describe, and yes, it's relatively common in older libraries. It still happens, though: IIRC there are some samples in Miroslav 2 that sound completely wonky in places, like the players suddenly moved off to the side for one or two notes. It's incredibly annoying and it's hard to say what the cause may be. I guess it might be a recording issue. Placing mics and players EXACTLY in the same spot between two different sessions is virtually impossible, and even slight discrepancies become much more pronounced when recording in stereo. Maybe they went back to re-record some bad notes and weren't able to set things up exactly right the second time, perhaps due to time or budget constraints. But who knows.

I'm not super familiar with the editing capabilities of Kontakt yet, but if it's possible to perform edits to the samples, maybe you could just go in and attenuate the offending channel until the sample matches up with the rest?

All of which makes perfect sense, really. I'd like to say, "Yeah, but how come they didn't fix it before releasing it!?" But I know that sometimes deadlines and whatnot simply make that impossible. It's a bummer for people who don't have the editing capabilities, though (Kontakt Player users, for instance). Can you edit Miro?

I think that's the best way to go about it, r.e., attenuating the offending channel. Kontakt's individual zone pan function appears to be a stereo balance (don't quote me on that yet). So it seems to be just what one would need to address the issue. It's just that it isn't always a matter of one or two samples; I was surprised by how all-over-the-place the Westgate clarinet is. I spent some time trying to fix it up, but I realized it's probably better to wait for it to manifest in practice and then dive in to correct it. Now that I know how to rectify it, it's not weighing on my mind as much. My theory is that, over time, I'll eventually weed out and correct all the issues.

[Samulis]

(01-07-2021, 04:51 AM)Nayrb Wrote:

(01-07-2021, 03:11 AM)Mattias Westlund Wrote: Right now, only on strings and brass (i.e. on every section channel). I've tried it on woodwinds, but considering the moderate panning required there, I can't hear a difference between the plugin and Reaper's standard stereo panner. I have not tried it on percussion or anything else yet. To avoid unnecessary clutter, I'm sort of listening for problematic instruments rather than putting it on every channel in the mix, just because.

Pretty much my approach as well. My percussion is just so huge sometimes that I thought I'd try it out along with some side EQ. So far I'm liking the results. I agree, a stereo pan for winds and things nearer the center is probably all you need.

I've come across a new, related issue: All my VO life, I've often encountered certain notes of certain sample libraries that seemed to just veer off in one direction or another, making positioning difficult. I've since discovered that, for whatever reason, it's because there is literally more volume in one channel or another in the sample itself. Like, the right channel is louder than the left, for instance. I can see this visually under the hood in Kontakt. It's a surprisingly common issue in my older libraries (like, anything recorded pre-2010).

For the longest time, I just assumed it had to do with some aural/proximity phenomenon, but the wave editor indicates that it is actually just something up with the sample itself--It's very annoying! My approach going forward will be to pan each offending sample (or "zone," as they're called in Kontakt) individually and cautiously in an effort to mitigate some of the issues that it can introduce into a mix. It's the only thing I can think to do at the moment

Do you notice this panning issue with woodwinds in particular?

I ask because woodwinds are inconsistent emitters; as the holes are opened, the sound is emitted in different directions and quantities, meaning the instrument will shift to the left and right channel in volume. This is particularly problematic in small, reflective spaces where interference patterns and standing waves within the audible range can easily form, causing one mic to be at a point of constructive or destructive interference, or when the mics are placed very close so that certain holes are much closer than others.

There is also of course the natural movement of musicians. Almost all musicians move somewhat while playing, especially in the course of a many-hour sampling session. I often rely on a visual sightline (e.g. 'this mic should line up with this floor tile' or 'these two mic stands should be in line'), but there are several times I have failed, or misaligned mics in the case of multi-day sessions (worst I can think of is one of the upright basses in VSCO 2).

One of the likely reasons older sample libraries are more subject to this issue is that they are often closer mic'd than newer libraries. If the mics are 6 feet (1.8m) away from the source, then they will pick up much less of the motion than if they were 2 feet (0.6 m), as the angle of the motion is less the further you move from the source. If a stereo mic array is very very close to the source it will be subject to major position shifts if the player so much as leans back/forward in their chair.

This is still a problem if the recording is mono too, just instead of having inconsistent panning, you can end up with inconsistent volume even when a note was played at the correct intensity.

[Nayrb]

(01-08-2021, 04:24 AM)Samulis Wrote: Do you notice this panning issue with woodwinds in particular?

I ask because woodwinds are inconsistent emitters; as the holes are opened, the sound is emitted in different directions and quantities, meaning the instrument will shift to the left and right channel in volume. This is particularly problematic in small, reflective spaces where interference patterns and standing waves within the audible range can easily form, causing one mic to be at a point of constructive or destructive interference, or when the mics are placed very close so that certain holes are much closer than others.

There is also of course the natural movement of musicians. Almost all musicians move somewhat while playing, especially in the course of a many-hour sampling session. I often rely on a visual sightline (e.g. 'this mic should line up with this floor tile' or 'these two mic stands should be in line'), but there are several times I have failed, or misaligned mics in the case of multi-day sessions (worst I can think of is one of the upright basses in VSCO 2).

One of the likely reasons older sample libraries are more subject to this issue is that they are often closer mic'd than newer libraries. If the mics are 6 feet (1.8m) away from the source, then they will pick up much less of the motion than if they were 2 feet (0.6 m), as the angle of the motion is less the further you move from the source. If a stereo mic array is very very close to the source it will be subject to major position shifts if the player so much as leans back/forward in their chair.

This is still a problem if the recording is mono too, just instead of having inconsistent panning, you can end up with inconsistent volume even when a note was played at the correct intensity.

Yep, it is usually woodwinds, but it happens on other instruments (notably that solo horn from OBC). I can see what you mean, though; it makes sense. I guess it's just the nature of the beast. My winds are definitely old and definitely closely miked.