Which K metering scale is best for cinematic/movie trailers?
Which K metering scale is best for cinematic/movie trailers?
It depends. K-metering isn’t really how you do things in post. In post you either mix by ear in a calibrated studio or stage, or you go by loudness standards chosen by the production company that buys your product and chosen based on where your content will be shown (i.e. internet, TV of theater).
So, either you ask the production company you’re giving your trailers to what they need, or you figure out yourself what the destination distribution is. If you want a recommendation then we’ll need to know what you’re going to use the stuff for.
In a nutshell, “higher k-value” = more dynamic range.
For use in two scenarios: Cinema and TV - trailer music/underscores/cues
Ok, so really you’re asking about music for post, not actual final mixes, correct?
I would approach this by looking at two different scenarios then.
First I would look at true underscore where you get dialog & fx by the supervising sound editor (or video editor) and then compose and mix for that. Since you’re mixing for that the final delivery medium will dictate what the final levels will be. For TV in the US it’s the CALM act, and in a lot of Europe I think it’s EBU R-128. So you can simply grab something that’s mixed for those standards and play that back in your room and set the levels so it works the way you want. I believe there’s an AES recording somewhere of speech that sits at -24LKFS for example, so simply playing that back and setting the monitor level so it’s comfortable is ballpark.
Now, if you’re really mixing underscore you’ll have the dialog and fx as a reference, and you can mix the music to that, all while your monitors are setup as above. This should put you very close to what’s needed for us post-engineers that take over once you’re done. If the destination is the web then typically the levels go up significantly, but again, if it’s true underscore you can probably score it as I just mentioned. We post engineers will bring it up to where it needs to be.
For theatrical real mixes are done on a Dolby certified (or “compatible”) mix stage. Those engineers would do the exact same thing basically adapting your music in that final mix. If you want to adopt a similar setup at home (or wherever) then you probably want a bit wider dynamic range than what TV would give you. But it also places demands on your space in terms of silence etc.
The other way I would approach it is if you are creating music that is closer to modern ‘popular’ music. It might end up functioning as either source or score, but really the production of it is pretty much like production of any music not going to post-production. A lot of times when I get music composed for a commercial for example, it’s actually produced as if it was a ‘regular’ piece of music. A big reason for that is that composers compete with each other to win these contracts, and loudness counts. You can’t have a track that’s softer than your competition when it’s played back at the ad agency, so you produce it just like regular music. When I get the track I obviously drop that level way down, and I generally don’t like stuff that loud and that compressed, but it is what it is.
Then in addition to that there’s a lot of television and film that pulls pieces of music from libraries where it’s pretty much all produced that way.
So for this type of production I would think your calibration would be closer to K-14 or K-12. Heck, even less dynamic range for some music styles.
Not sure if that helps, but it unfortunately really does depend. Fortunately Cubase and Nuendo has the Control Room and it’s a great help. That together with templates should make it easy to switch if you want to approach it that way. You could essentially setup one reference level with your master output level at 0dB gain (unity) and just adjust speakers, and then choose the actual reference level function in Cubendo to set an offset for ‘the other’ work you do. So if you need less dynamic range you can then set Cubendo’s reference level to -10dB or whatever it ends up being.
Does that make sense?
If you’re going to stick with the Katz system, K-20 is for cinematic stuff. K-14 is “typical music” and K-12 is broadcast radio.
I would still just calibrate monitoring for cinematic or broadcast using actual cinematic/broadcast content, and then switch between that and K-12.
Thank you for the detailed reply. What you wrote is very helpful for me - thank you very much
Is there a specific reason why you specified K-12 and not K-20?
Well, I actually wrote ‘either or’ in my other longer reply, but I do feel that for the most part music that ends up in post tends to be very competitive with modern music in general, which means highly compressed. From what I can tell even K-12 isn’t particularly ‘low on dynamic range’ compared to a lot of contemporary music. Unless things have changed recently that is. Remember that his proposition is over a decade and a half old by now.
So in other words, I would expect that working at K-12 would give you loudness that is competitive, if not softer than the competition, when composing popular-music-sounding tracks specifically for post.
While I don’t disagree with your answer based on the assumptions you made, it has to be asked: is dman planning on compressing a lot? I imagine that cinematic trailers aren’t compressed a lot to allow the dynamics to really shine through (especially since they often have huge percussive components). I could be wrong, however.
No you’re not wrong. But that’s why I made the distinctions I did in the longer reply. I’m basically saying that it’s better to calibrate the room according to either cinematic (not sure that K-20 is ideal, but probably ballpark) or broadcast (which would be -24LKFS average)… Either/both of those would be for more dynamic stuff where you’re actually mixing to picture and to a known ‘standard’, and then the alternate calibration would be K-12 (or K-14) for anything that needs to be compressed and “flat”.
Think about it this way:
one type of trailer music:
So what they have in common is that they’re movie trailers, but they differ in the type of music that’s used. Of course, you can hear (I think) that Interstellar has the music dip as a full mix to make way for some of the dialog (might be music from somewhere else), but the point is that it really plays as if it was scored for the trailer. So scoring orchestral music like that for a trailer is one thing, and on the other hand you have “pop songs” that play under a trailer like in the first example (one of the worst movies I’ve seen this past year btw). So for trailers with “songs” in them I would personally mix them as if they were intended to go on a record, even though they’re going into a movie trailer.
Very well articulated. Thanks!
Makes perfect sense to me. Mattias, that’s one of the best posts I’ve read on the forums and you explained things perfectly.
When I was at the ad agency we’d get tracks from leading music companies and all of them were submitted at commercial song levels even if they were eventually only used at “bed level” in the final mix for release, unless, of course the music was carrying the spot, like they did on the AmTrack spots with Richie. I’d often think how it was really too bad that the public couldn’t hear the great song that was going on so far down in the mix while the V/O announcer is screaming about whatever soft drink or candy bar or gasoline was being sold.
Anyway, I do have a few questions but I don’t have time to post now. Again, thanks for your help and advice. Great post.
If anything, let these comments simply amplify the excellent posts in this thread for which I’m also very grateful. I’d never worked with K-Scales until I met Cubase in May 2015. I’ve found working with the K-Scale very helpful, but, it has to be in conjunction with relating the internal monitoring to the external Audio Monitor levels in the room you’re in. Print out a copy of the Equal Loudness Contour and a frequency chart post those on the wall. Calibrate your external audio monitoring output levels using a sound level meter. Read (and re-read) Bob Katz’s papers on loudness and the scales when time permits, if you have not already done so. Download the references files from his site and bring them into a project and check. Make friends with Cubase’s tone generator. Get a free or paid VU meter and line up your reference levels, of course, you can line up perfectly well with the meters Cubase provides. I just like a VU for that (old fashioned).
K=20 highest dynamic range, Classical Music, Jazz – other wide dynamic range program material
K=14 Pop, Rock, EDM, etc.
K=12 Squashed mixes, little to no dynamic range programs, late-night TV commercials, but can but used if careful or wanted.
All the good resources you need to work this up are on the web so just search for the links and it will fall into place. sound on sound magazine, etc.
Or just buy a plug-in that provides EBU-R128.
After the EBU-R128 standard and LUFS/LUKS even Bob Katz himself are NOT using the K-system anymore (his own words)… The K-system were never a standard, just a proposal on how to have consatant monitoring, which also require calibration of your monitoring system…
And it has nothing what so ever to do with delivery specs.