How do you like to record ambiances, and backgrounds?
Stereo? Mono? ORTF? MS?
How do you like to layer them in a project? Mono? Some Stereo? Lots of Stereo files?
That depends... In Sweden we call ambiances "atmospheres" which I think is a better word for describing the function of ambiances.
An ambience tells us where we are and what we should feel like.
When I started out working with post sound I used to think ambiances was all about putting a lot of stereo recordings on top of each other to create a fill... Now I know it's a lot more complex. So I do lots of stuff.
Most of the times I record a 2 minute "ambiance" = a raw "this is what this location sounds like" stereo in ORTF (Imho, I think ORTF sounds the most natural and spatious) and then I fill that up with single mono things that's panned. The FX are important part of the ambiances as well. It's really easy to just throw a lot of stuff into the AMB-tracks, but it should be handled more gracefully than that.
The ambiances are by far the most important part of the overall feeling of sound. It's also the hardest part of sound design. Imo.
So to answer your question (I took a trip to preaching to the choir-land) I deliver one or two base stereo (or 5.1) files and then lots of mono or stereo FX-type sounds bundled in the Ambiance tracks to make a nice, fluffy airy mix.
I could write a book on this subject - I love working with ambiences, no matter the size of the team ambiences is usually something I prefer to cut myself, I totally agree ambiences play a crucial role in many different ways... Here is my rant for now - suspect it will become a blog post/chapter in a book etc etc..
In terms of recording ambiences, it depends on the location & what ambience I'm recording as to the best way to capture it, or elements of it. It also depends on whether it is a location I have plenty of time at, or its a one take scenario. All of the recent films I have done have involved me revisiting locations from the shoot to capture ambiences, and those by far have been the most satisfying ambience tracks to build & edit eg a week spent in Samoa recording ambiences achieved a far better result than trying to composite layers from the library, without any real world reference for how ambiences there actually sound (which is a very important part of recording your own ambiences ie the memory). Most ambiences I record 5 minutes minimum and eg some night ambiences I recorded in Samoa I rolled from 10pm through to and including dawn. If you are prepared (large hard drive, plenty of batteries) it is fascinating to later be able to skip through time and hear how night changes. When I listened to 'night' in Samoa in real time it didn't appear to change much, but replaying the recording and skipping forward an hour the changes are very significant...
I prefer using a combination of spaced omni, stereo, and discrete stereo/dual mono and mono recordings. For any location I seperate the tasks of (a) capturing a general ambience, and then (b) capturing individual elements intended for specific placement. No one recording technique can achieve both of these aims, simply because most ambiences change over time, and when dealing with film we are not making a real time documentary of a location - we are aiming to create and evoke the world of the film, scene, moment as best serves the story.
With respect to editing ambiences & preparing them for predubs, there are many considerations.... Each project has to be considered seperately, general rules are only that i.e. general, and films require specificity.... As I am editing ambiences I determine how many predub stems the ambiences will need, so as to allow enough rebalancing during the final dub & to not have any debateable elements locked together. As a minimum I would want to predub to 4 x 5.0 stems (main ambiences) and 4 x LCR stems (spot/split elements) but thats a minimum...
First & most important imho is consideration of the centre channel. The main reason film exists is to convey stories and that is primarily achieved via dialogue, which is mostly coming from the centre speaker & mostly was recorded on location/set etc with varying backgrounds & time of day... On a shot by shot, and scene by scene basis I think it is important to listen to the background behind the production audio and provide enough material in the centre channel to (a) match & mask any inherited ambiences and (b) help make these appear interesting, characterful, appropriate & of good fidelity. For most scenes/shots I would layer up to 4 mono tracks, solely to meet these requirements... Then depending on the set and setting, I'd layer 2-6 stereo (& offset stereo/dual mono) ambiences for use in LCR and an additional 2-4 stereo ambiences for surround LR per stem to build each stem of the 'ambient world of the film'. If I have quad ambiences I will consider using them, but the chances of them matching the perspective and picture cuts within a scene are less likely than cutting layered stereo (& offset stereo) ambiences.. But if I have them I will use them... Experience and specific context dictates how much offlaying/cutting is required for perspective cuts, which is why it is so important to attend the predubs of material you have edited.
So that set of tracks (Cx4+LRx6+sLRx4) are my source elements for each predub stem. The content & number of elements depends on what the scene/moment needs. There are no prizes for using every track - its a matter of making it only as complex/simple as it needs to be. Sometimes stems are split between foreground, mid ground & background/distant, other times they are divided by constituent elements eg Ambience stems A/B = winds C/D = wildlife E/F/G/H = spot ambience elements... it totally depends on the context....
[NOTE: this is all subjective -there is no right or wrong way to do anything, experience reinforces what works etc etc....]
I personally stay away from MS for recordings intended for BGFX. dolby decoding can sometimes react strangely when you pan decoded MS recordings into the surrounds.
For recording I like ORTF, but I'll roll XY if I have to and mono if there's no alternatives and I need to capture the moment.
Olle spelled out the standard way to approach things in post. layer a few washy tracks and then spot one offs to taste and with respect to the story.
I've started using a Soundfield mic (SPS200) recording to 4 track (R-44), that way I can decode to stereo, 5.1 or even 7.1 in Pro Tools HD. To date we have used stereo, 5.1 and 7.1, but not mono.