What's new

What makes a mix sound wide if its bad practice to pan orchestral samples?

OP
thevisi0nary

thevisi0nary

Active Member
What soundtrack is your reference? in a traditional orchestral style, the feeling of width often comes from recording and orchestration more than mixing techniques.

For example, spaced mic pairs like "outriggers" or Decca tree will be good for a 3d spatial image compared to coincident mic techniques. And capturing a bit of the room in those mic positions also add more feeling of width in the recording. So if you can use Decca mic positions or similar in your template over close mics, that could be worth a try.

But even more important could be orchestration. Layering too much can actually smooth out the high frequencies that would have given clarity and power to each instrument, and the result becomes more muddy and vaguely centered. Instead, assigning separate notes of a chord to individual wind instruments, or individual string sections, will spread the music in stereo, and it will also allow the brightness/clarity of each instrument to shine on its own even when the entire orchestra is playing.

Great examples of these points are the Andy Blaney demos for spitfire symphonic brass. I've heard those were done using outriggers only, but I don't really know.

Do you feel like these points might explain some width that you're hearing in your reference track, or do you feel it's something else?
In this track it feels like there is a fair bit of distance between the dueling string lines. Also when the horns come in they feel reasonably to the left.
 

shawnsingh

Active Member
Yeah, in my opinion, the example video you've posted feels wide for both of the reasons I've mentioned earlier - the orchestration and the localized positioninings of each instrument (i'm not sure how much virtual and real this has?). The strings are actually chamber-like in size, at least some of the time, which gives it a more detailed feeling that also helps - it enhances the feeling of positioning and clarity even more. This mix also did *not* leave the center empty, which is probably a good thing. (I don't have too much experience with extreme panning, but I do expect a hole in the middle would actually detract from the feeling of width more than it would help.)

The room/reverb also, is recorded/mixed quite transparently, in my opinion. I can't tell for sure but seems like there may be a few different recordings or instruments that have slightly different room/reverb spaces, but it doesn't really matter. For the most part, this amount of room/reverb almost melts away into the background and keeps things feeling clear, but still provides the undeniable feeling of distance and space. I've always been partial to that kind of mix. As a side note I feel like Teldex recordings usually have that property too, something about the room that has slightly weaker first/second early reflections, but still has nice room and long tail.
 

Dietz

Space Explorer
I do expect a hole in the middle would actually detract from the feeling of width more than it would help
Exactly. Mixing is always about relations: If everything is loud, nothing is loud. And this is true for any aspect of sound, of course: If everything is dry and close, nothing is dry and close. If everything is bright, nothings sounds bright any more. And so on.

Or put simply: The basic question of mixing is "Louder than WHAT?" :)

So the sensation of "width" will be much bigger when there are narrow sounds in the center, by comparison. ... this is a common trick in pop music, BTW: Have a verse in (almost) mono, and overwhelm the listener with a super-wide chorus.
 

MartinH.

Senior Member
In this track it feels like there is a fair bit of distance between the dueling string lines. Also when the horns come in they feel reasonably to the left.
Iirc Penka said in the GDC talk on the Bloodborne OST that they recorded in different locations for the DLC. The base game was recorded in Air Lyndhurst I think:



I think she said some aleatoric stuff was taken from Symphobia, but otherwise it should mostly be real players.

But don't take my word for it, it's been a while that I watched it. Here's the talk:

 

ProfoundSilence

Senior Member
two main things, 1.) microphone choices. If you don't pick the wider mics it wont sound as wide and 2.) you can still pan close mics freely.

bonus tip would be using a HAAS delay on the close mics as well, which in laymans terms is delaying the weaker side by a few milliseconds.

i.e. pan your violins I close mic more to the left, then delay the right signal by 2-8 or so milliseconds(honestly just use your ears and a pair of headphones) and often times I mix both pan and the delay at the same time to approximate where I actually want the instrument to "feel" like it is
 

Per Boysen

New Member
A fun and useful exercise is to route a ms-matrix in your DAW:

1. Split the stereo mix into three busses.
2. On the first bus reverse left and right.
3. On the second bus, shift the phase.
4. On the fourth bus, make it mono.

Now, the two first busses together produce an extremely wide field and also a hole in the middle. Use the mono bus to fill that hole. By altering the levels of these three channels you have full control over the stereo width and sound definition. The last step would be to use another L-R shifter to get it back to the original.

I might be wrong, but I guess this is basically what these kind of stereo widening plugins are doing. But setting it up manually and experimenting can be quite educational.
 
OP
thevisi0nary

thevisi0nary

Active Member
Iirc Penka said in the GDC talk on the Bloodborne OST that they recorded in different locations for the DLC. The base game was recorded in Air Lyndhurst I think:



I think she said some aleatoric stuff was taken from Symphobia, but otherwise it should mostly be real players.

But don't take my word for it, it's been a while that I watched it. Here's the talk:

I've watched this video before and I absolutely love it. I wish so bad that I could study the sheet music. There is really incredible music on this soundtrack.
 
OP
thevisi0nary

thevisi0nary

Active Member
A fun and useful exercise is to route a ms-matrix in your DAW:

1. Split the stereo mix into three busses.
2. On the first bus reverse left and right.
3. On the second bus, shift the phase.
4. On the fourth bus, make it mono.

Now, the two first busses together produce an extremely wide field and also a hole in the middle. Use the mono bus to fill that hole. By altering the levels of these three channels you have full control over the stereo width and sound definition. The last step would be to use another L-R shifter to get it back to the original.

I might be wrong, but I guess this is basically what these kind of stereo widening plugins are doing. But setting it up manually and experimenting can be quite educational.
That is very interesting, will have to give this a try. Do you think it's possible to run into phasing issues doing this?
 

pkm

Senior Member
That is very interesting, will have to give this a try. Do you think it's possible to run into phasing issues doing this?
Phasing issues are on purpose and are what make the stereo effect work. The stuff in the center is out of phase and cancels out, leaving only the stuff on the sides. When collapsed to mono, the first two busses completely cancel out and the 3rd mono bus is all that remains, phasing-free.
 

shawnsingh

Active Member
I think @thevisi0nary was asking whether the phase differences you're configuring would cause unintended issues like comb filtering, or some kind of "chorusy" effect, or mono compatibility issues, due to the phase differences.

Also something else that is keeping me awake at night... Whatever happened to the third bus? :grin:
 

pkm

Senior Member
I think @thevisi0nary was asking whether the phase differences you're configuring would cause unintended issues like comb filtering, or some kind of "chorusy" effect, or mono compatibility issues, due to the phase differences.

Also something else that is keeping me awake at night... Whatever happened to the third bus? :grin:
Yeah, there can definitely be those types of phase issues, it’s inherent in the technique. Kinda like when you hear “a capellas” or instrumentals that have a weird phasey ghost of other instruments. It’s not perfect but can be very effective.

But mono compatibility is almost perfect because the sides completely 100% cancel out, leaving only the mono bus by itself. The only possible problems are overall volume.

And we don’t talk about the third bus.
 

Per Boysen

New Member
That is very interesting, will have to give this a try. Do you think it's possible to run into phasing issues doing this?
It's mainly a mastering technic, so given you have fixed phasing issues in the mixing phase this should not add any. On the contrary actually, you may reach a better mono compatibility (even for a wide stereo mix)
 

Andrew Souter

Active Member
Sounds fantastic, will definitely keep this on my watch list. What is the difference between something like this and Virtual Sound Stage by paralax?

I am not completely familiar with Virtual Sound Stage, and as I know how much work goes into making products like this, I don't like to directly point out any perceived weaknesses of competitive products. Indeed these guys seem to do nice work, and they even recommend pairing their product with our algo verbs such as Aether/B2 to supply tails, so I have a mutual respect for their work. I believe Virtual Sound Stage is primarily a gain panner with a built-in Early Reflections engine, but I am not 100% sure, so don't quote me.

Our system changes the direct sound itself providing instantaneous audio source width, uses several psychoacoustic techniques related to those discussed in this thread to achieve positioning, offers as much mono-compatabilty as you like/need via three different algorithm modes and control over various positioning rules, and is modulated to give additional life and organic feeling to the result.

Our system is furthermore capable of inter-plugin communication between Precedence and Breeze 2.5, where position information in Precedence is communicated to a linked instance of the reverb engine, and the entire DSP settings of the reverb engine updates in response to position. This creates something like an algorithmic Multiple Impulse Response system. There is infinite variation in Precedence and Breeze 2.5 depending on position, and both are modulated. Combined they create an incredible sense of depth and positioning. It's truly next level!

It's almost like when Spitfire or other library company records in Air studio and offeres 20 different mic positions or similar, and sometimes in situ positional variations. Our system can take a completely dry library, or physically modeled instrument like Sample Modeling, or real recording from your studio, and do the same, but not with 20 or so positions -- literally infinite.

And it also has the ability to work well with and compliment libraries that are already room-y, as we are very well aware some great libraries are recorded with lots of room-sound. We have various input modes to address this and help blend wet libraries with dry libraries.

Furthermore, the new Precedence 1.5, offers Multi-Instnace Editing, and Edit Groups! Not only can you see 10, 50, 100, 200 instances within a single plug-in GUI, but you can also EDIT them! Changing instance selection within a shared GUI is MUCH, MUCH faster than constantly having to switch between the DAW mixer and many plug-in instances! The linked Reverb Engine can be have instance selection controlled by Precedence as well, so you can keep one GUI editor open for both a and quickly control many instances with the same connivence as controlling one!

Finally parameters in both Precedence and Breeze 2.5 can be changed en masse for the entire Edit Group! So you can load preset changes for the entire group with a single click! Position information is retained! You can metaphorically simply transport your mix from Air, to Boston hall or whatever else you like by changing the preset in Breeze for the entire group. While retaining the relative in-situ positions! Or you can change the Alg Mode in Precedence between Beta and Mu and export two different mixes, the later with enhanced mono-comatabiliy if that is a critical concern. Or change the Delta and Loss Parameters to change the positioning rules for the entire group, and create macro-changes to "spatial contrast" for the entire group.

etc etc. but I will stop bc this starts to sound salesman-y. ;)

we hope to have videos ready shortly to explain this all better, but the manual is arleady online with full details. Hope it helps.
 
Top Bottom