Question about stereo image of samples.

[Fealow]

So i'm having a bit of an inner conflict about how to deal with close mic'd stereo samples in the context of a full orchestra. 

I'm working with Miroslav philharmonik 1 and 2 and both sample sets are miked in a studio setting at a reasonably close distance. The issue is mainly with Miroslav 2 as all of the instruments (strings, woodwinds and brass) are recorded in a central position of the room in stereo as far as I can tell where as miro 1 has all the ensemble samples recorded in their actual seating position within in the orchestra. Now the miro 2 samples all sound lovely, but obviously orchestra players do not sit on top of each other in a straight line so obviously I would need to pan these out to their respective positions and that is where the biggest problem occurs.

Panning/narrowing stereo tracks never sounds that great. You lose detail, fidelity all while introducing phasing. Any tips on how to deal with this? should I just pan them out while leaving the stereo image at it's original width or should I narrow the image to mono and then pan or do you personally have other ways of dealing with this?

Thanks.

[kmlandre]

I recently discovered this and I'm finding I really like the pan "quality" it produces.  It seems to work happily with both mono and stereo samples.  If I understood it right, it's mostly based around the power pan principal ( pretty sure that's the wrong term) and the Haas Effect.

Dotec Audio "DeePanpot"
http://dotec-audio.com/deepanpot.html#main


-- Kurt

[Fealow]

Ooo nice, that seems like it would help a lot. I'll give it a go when I get a chance. Thanks for sharing!

[Nayrb]

This is something I'd like to know more about as well. I use a variety of sample libraries and so I need to fiddle with stereo imaging to get things to sit right, but I'm still training my ear to really figure out the best way to treat my samples. In one of Mattias' articles (and I paraphrase here) he gives the general rule of thumb that narrowing is usually safe while widening is something to be a bit more careful with. I interpret "safe" as "narrowing probably won't introduce phenomena that I don't understand and which might create problems in my mix that I have a hard time readily identifying"; but as yet I can't say I'm entirely certain I'm not creating any issues from an objective technical standpoint. It comes down to experimentation in the end, I guess.

As far as losing fidelity when narrowing goes, I think that's usually a problem when one is admiring the quality of a sample(s) on its own rather than in context. I wouldn't say that is a rule, of course, as I've found plenty of samples that lose their effect when narrowed too much. But it all depends; context is key.

[Samulis]

Basically the Haas Effect is that when you turn your head, the signal reaches one ear slightly faster than the other. This is how we identify where sounds are coming from- by detecting that tiny millisecond delay, and is the basis of most stereo recording techniques. Basically, by delaying one signal a little bit, you can produce a more realistic pan than a traditional pan (i.e. the pan knob in your DAW), which uses just volume to make one louder than the other (a very primitive approach dating back to early multi-tracking). In fact, you likely already have a stereo width/pan tool as part of your DAW's default plugin set- if not, a channel delay tool (not a delay effect) set to a few ms can do the trick manually (I used to always do panning and phasing via manual delays, hehe).

As a rule of thumb, typically it is not safe to narrow an orchestral sample much, because they are typically recorded with what is called a non-coincident array. What this means is that the mics are placed quite far apart, such as in a Spaced Pair setup. Decca Trees theoretically combat this by having a third mic, placed in the middle and slightly forward (therefore, it, theoretically, is possible to narrow a decca tree- being a glorified spaced pair with an extra mic in the center, narrowing it will, like a Mid-Side, if phasing is present, cancel out the two sides and have just the center mic left). If you have samples with XY or ORTF/NOS, the former a coincident array, the latter two, semi-coincident, then it is probably safe to narrow. If you have Mid-Side, unfortunately somewhat rare, then you can most definitely narrow.

Typically you shouldn't need to narrow things much. Just pan with a stereo imager.

[Nayrb]

(07-20-2016, 07:05 PM)Samulis Wrote: Basically the Haas Effect is that when you turn your head, the signal reaches one ear slightly faster than the other. This is how we identify where sounds are coming from- by detecting that tiny millisecond delay, and is the basis of most stereo recording techniques. Basically, by delaying one signal a little bit, you can produce a more realistic pan than a traditional pan (i.e. the pan knob in your DAW), which uses just volume to make one louder than the other (a very primitive approach dating back to early multi-tracking). In fact, you likely already have a stereo width/pan tool as part of your DAW's default plugin set- if not, a channel delay tool (not a delay effect) set to a few ms can do the trick manually (I used to always do panning and phasing via manual delays, hehe).

As a rule of thumb, typically it is not safe to narrow an orchestral sample much, because they are typically recorded with what is called a non-coincident array. What this means is that the mics are placed quite far apart, such as in a Spaced Pair setup. Decca Trees theoretically combat this by having a third mic, placed in the middle and slightly forward (therefore, it, theoretically, is possible to narrow a decca tree- being a glorified spaced pair with an extra mic in the center, narrowing it will, like a Mid-Side, if phasing is present, cancel out the two sides and have just the center mic left). If you have samples with XY or ORTF/NOS, the former a coincident array, the latter two, semi-coincident, then it is probably safe to narrow. If you have Mid-Side, unfortunately somewhat rare, then you can most definitely narrow.

Typically you shouldn't need to narrow things much. Just pan with a stereo imager.

So the stereo width control on the mixer in my DAW isn't quite the same as a dedicated plugin?

[Samulis]

(07-21-2016, 01:07 AM)Nayrb Wrote: So the stereo width control on the mixer in my DAW isn't quite the same as a dedicated plugin?

Your DAW's mixer has a stereo width slider? Never seen that before...

Maybe it has two knobs, one for left pan and the other for right pan. This is no different from a normal pan, in that it uses volume. In this case, instead of using delays to bring the channels into coincidence, it just is essentially a volume balance knob for each channel (like on an old fashioned mixer). A dedicated stereo pan and stereo width plugin will likely only work with delay, or a mixture of delay and volume based pan.

[Mattias Westlund]

(07-21-2016, 02:18 AM)Samulis Wrote: Your DAW's mixer has a stereo width slider? Never seen that before...

REAPER has it, and IIRC Cubase has had it since forever...

[Samulis]

(07-21-2016, 09:02 PM)Mattias Westlund Wrote:

(07-21-2016, 02:18 AM)Samulis Wrote: Your DAW's mixer has a stereo width slider? Never seen that before...

REAPER has it, and IIRC Cubase has had it since forever...

Interesting, yes, I see that now in Reaper's mixer. Thank you for the correction, Mattias.

I do wonder however, if it actually is a true stereo width control or just a dual channel pan function. Probably in some documentation somewhere.

[Nayrb]

(07-22-2016, 02:15 AM)Samulis Wrote:

(07-21-2016, 09:02 PM)Mattias Westlund Wrote:

(07-21-2016, 02:18 AM)Samulis Wrote: Your DAW's mixer has a stereo width slider? Never seen that before...

REAPER has it, and IIRC Cubase has had it since forever...

Interesting, yes, I see that now in Reaper's mixer. Thank you for the correction, Mattias.

I do wonder however, if it actually is a true stereo width control or just a dual channel pan function. Probably in some documentation somewhere.

I'll have to check the manual for specifics. I use FL Studio and there is definitely a stereo width knobby on each mixer track below the typical L/R pan control. FL also comes with at least two dedicated stereo enhancement plugins I can think of off the top of my head that I think will also warrant further exploration. This has all been good info, though. Thanks!