Key to making your film sound like it had a budget

Checker bordering is putting alternate clips/ regions on alternating tracks, as opposed to having them all on one track. L cuts and J cuts are mostly picture editing terms and are when you cut the picture and sound at different places so one overlaps the other.
 
Noiz2 said:
Ted,

Mix stages are calibrated to the "X Curve", and yes that has a high roll off. This originally had started off as the "academy curve" and had to do with optical track limitations and speaker playback systems in theatres. There is an interesting article here.

Cheers
SK
Click to expand...

Given the number of features and docs I've mixed the music scores for, I'm quite embarrassed to say I didn't know this. But it does help explain why some of them came out duller sounding in the end than I had intended (although quite a few didn't). I could blame the post-mix guys in those cases for not going ahead brightening things up, but ownership is usually the better path...

The article was very interesting, and linked to another that was even more so. Many thanks again for the enlightenment!
 
Great write-up Noiz2! You just don't realize there's so much going on with sound on set all the way to the mixing boards.

/j
 
Edgen said:
You just don't realize there's so much going on with sound on set all the way to the mixing boards.
Click to expand...
I think part of the problem lies in the fact that most people consider film to be a visual medium, not an audio-visual medium. During the Silent Film era the larger theatres would have a full orchestra but, what most people don't know is that there would be up to a dozen people behind the screen doing live sound FX in real (reel?! lol...) time. In fact, todays drum kits evolved from percussionists who added all kinds of odd "trappings" to their kit - hence the term "trap kit" - in order to get the movie gigs in smaller movie houses. A lot of our sound FX cliches, especially comedic sound FX, owe their origins to this era.

When movies went "talkie" they were filmed on sound stages, sound was king and the production sound mixer was allowed the privilege of calling cut if the sound was not correct.
 
This is a great thread! I have already learned so much, thank you all.

I was wondering if you experienced guys could help straighten something out for me.

I am new to doing post for video but not new to film or audio. I have the one man post show for 3 web episodes, two 1 hour indy movies, a few shorts and a couple of trailers. On all of these projects I was the only one for post audio therefore going through all of the aspects. I did most of this in adobe audition and STP 2, as I stated in other posts I am now moving into PT LE and loving it.

My issue is with levels. Im so bloody confused here. In school we were simply taught get as close to 0 dBfs without clipping. Clearly this is based towards recording. I've been reading about leq(A) and dolby surround for a bit now and am still not sure what applies to me. Most of the stuff I have been working on ends up being viewed online or at festivals (dvd and/or digibeta) I am not yet set up for surround and am working with two KRK rokit 6 properly aligned for stereo nearfield monitoring.

I did one project for the Canadian National Film Board and CBC and they asked for dialogue to not peak over -12 dbfs, that threw me way off, so I kept the whole mix under -12 and they were cool. Phew!

Am I correct with my conclusion that I should be playing the dolby pink back from my DAW and increase monitor level until i reach 79 dbC on the spl meter and then mix to taste? Am I supposed to slap all the dialogue to -27 leq(A) and mix everything by ear around that? I chose 79dbc because I read this is better for DVD, internet home viewing to ensure dialogue is loud enough in an uncontrolled environment is this right?

I did this procedure before working on a project in STP 2 (its almost done so Im going to finish it in there then mix in PT) just to test things out I set the pink playing from STP to 79-80 dbc and then played back my project. I adjusted the dialogue bus to a comfortable listening level based on that calibration. Right before I did this I also watched a scene of Master and Commander at the same monitor level, to try and match dialogue levels by ear. The dialogue ended up sitting around -29 to -15 dbfs. Is this right? I mean I really don't know what to think at this point so Im putting this out there and maybe someone can set me straight!

If somebody could make sense of this for me It would rock my socks!:dankk2:
 
Last edited:
Awesome thread, Scott.

Montrealaudio, what you are doing as far as calibrating your monitors with pink noise and then mixing to taste is absolutely correct. As my boss, who has been mixing for 30 years, always says, you don't mix by numbers. Mixing is an art, not a science, so at some point you have to trust your ears.
 
Noiz2 said:
There is this misconception that the BIG difference (sound-wise) between "Hollywood" big budget films is better production recording. Don't get me wrong that IS a big difference, but the BIGGEST difference is how those tracks are handled in post. As a rule almost nothing but dialog is used from the production tracks, and well air FOR the dialog track. NOTHING else. The rest of the sound isn't even recorded on location for the most part. Things like footsteps and anything people touch, handle,etc will be Foleyed (or covered by FX or WIFO). FX will fill in all the backgrounds sounds and all the specifics (hard FX) like door closes, gunshots, doorbells, door opens and closes, car sounds, birds etc. For the most part these sounds will not come off of CD libraries you can buy, though some will. Most will be or have been recorded by the editors, facility, studio etc. Some will be recorded specifically for this film others will have been recorded for other films and then put in the "library". The biggest asset a facility or editor has, other than their skill, is their "library". And editors trade and "loan" sounds to each other all the time and even from time to time sell them. Usually if you need a LOT of something that you know so and so has a LOT of then the show will buy use/ access to that library. But if I need a musket sound (this is an actual case) I posted to the post section on digi Users conference (though there are a number of other lists that would also work) and in less than an hour I had an email from an editor in Canada that just done a bunch of revolutionary reenactment recordings and sent me a link to 20 or so long files of musket shots and voleys and gobs of other good stuff. It goes both ways I send a (now) friend in New Zealand some protools sessions that I had used for a Jackie Chan fight sequence. He was working on a no budget Nija flick. He sent me two CD's of recordings around New Zealand.

Any way the point is commercial CD's are often limiting and if your spending less than $100/ disk are probably not very good. There are better options, like recording your own. Once you have some recordings you are in a much better trading position. Every city sounds different, if you live in one go out and record it. Professional editors are always looking for stuff that was not recorded in NYC, LA or SanFrancisco.

Now that you have some sounds and your hot to replace things it's important to remember it's almost never a one to one deal. You should probably have two to three stereo tracks that are broad backgrounds (wind, "city roar", rain, backyard "air") and then you should have some specifics. You in an apartment say. You should throw in an occasional car by, an alarm chirp, maybe some voices in the street, a garbage truck etc. The idea is to make it a living place where life goes on. So now you have maybe 7-8 tracks and you haven't even gotten to the hard FX yet! "Hard FX" are specifics that are linked to something you see or know is there. So for instance, door opens and closes (even if the door isn't in the shot). This is a very "see a dog hear a dog" stuff. If you see it you should hear it. That is going to take a bunch of tracks. For one thing a lot of hard FX are layered. A Kung Foo hit say is going to be three or four tracks on it's own. That door close might be two or three tracks. Lets take a peek at those and you will see how just about everything can get "layered up". The KF hit: you want to have usually three "hits" one in the high register one in the mid and one low to add weight. You often also want a touch of whoosh. The high one adds "snap" the mid is your main sound and has a lot to do with the kind of hit. If things are wet then the mid hit is a really wet hit. The low one is just to add beef. Playing with the different levels and slight timing changes will give you a lot of variety in how the combo will sound (but don't keep using the same combo - keep 10 to 15 hits between repeats, because even though you can make them sound different the audience will pick up on the repeat if they are too close together). Often a little whoosh of the arm/ foot swinging into the hit will give you a MUCH harder hit. Whooshes are like salt a little enhances a lot tastes bad.
The Door: The door could be a single sound. But often it good to add some texture. Maybe a little hinge squeak, or security chain rattle, maybe the doorknob creeks a bit??? It depends on what kind of door it is and how important it is. Most doors are a one or two layer deal but some especially "high tech" or mechanical doors will have a stack of layers.

What are we up to now 15- 25 tracks or so.

Foley. Foley is not FX. Though it is similar and if you use Foley recorded for one show in a second show poof, now it's FX. It's kind of confusing but technically Foley is recording sounds IN SYNC to picture. If your not watching your movie and recording in sync your not doing Foley. Side note on a pet peeve, it's Foley not foley. It's named after the guy who invented it and his name was Foley, so even though it's used as a name of a process it's also a proper name and should be capitalized.
Anyway Foley is one thing often not done at all on no budget films. Truthfully you probably don't have the time or equipment to do a real Foley session with full coverage. What you can do that will work almost as well is a combo of more FX and WIFO. WIFO stands for WIld FOley. And if you look at my rant your saying THAT"S NOT FOLEY. yah I know but it's the movie biz where even cloths pins have strange names. Wild Foley is pretty much recording in sync with your memory of the film. Traditionally this was because it's hard to project 35mm in the field. Now it's often done to record something that will fit better if recorded in the "real world". I've use my house to record all kinds of WIFO. And the great part is it often just drops right in because it sounds like it happened in the space in the movie. So if you have a dinner scene and the camera made production unusable. Set up you dinning room table with similar items (you only need one place setting) and record the actions, either your best guess from memory or play the scene on a laptop etc. Do a few takes and do different takes for each character. You don't need to do everything but more is generally better. At least get silverware sounds and glass downs etc. This will add a huge amount of texture and believability to your track.

So now your at 20-30 tracks.

Anything that is clean AND useful in the production tracks, use. The biggest time savers are if you can get footsteps and cloth sounds. Cloth is usually done in the Foley session and is important if you do any ADR. It is simple some sound for all the soft stuff people are wearing, handling, sitting on etc. Usually it's just a loose sync track that is going to play very low when it's needed so that people actually sound like they are wearing clothes. It's often in production, if you can use it. The norm is to pull the "production FX" (PFX) on to there own tracks so the dialog tracks are JUST dialog. That isn't always possible even in big budget films and in no budget I usually only separate some. It gets tricky is the dialog needs a lot of tweaking because you often don't want to alter the PFX.

If you have done all of the above then if you need ADR your most of the way to making it fit since all the sounds you lost with the production you have replaced.

Dialog. Big films checkerboard characters and often put characters on their own track so they can be tweaked individually. I tend to even though in a no budget it's usually a wast of time. What you do get out of it is a familiarity with the tracks and that's good, also if you do need to work on the tracks they are already separated. You want to separate them out by problems anyway and you my want/ need to checkerboard cuts so you can smooth the edits. But in the end it usually doesn't make any sonic difference if you don't spend a few days pulling apart the dialog.

Now pretend there won't be any music and mix that baby. A good track shouldn't "need" any music. When you add the music it will be the special sauce that really polishes the track. And when you add the music decide in each case who is going to lead the scene, music or sound and have the other one get out of the way. Sometimes, well often it will change in the scene so I should probably say for every moment. But a touch of FX under a heave score moment keeps the score in the world of the movie. BG's are more likely to go away under music and Foley is more likely to stay but?...

The above will get you a track that does not sound like a low budget film, that echoey empty "low budget" sound. Your film will have that full nuanced textured "Hollywood" sound. It takes time but it's not nec expensive. For a full on track the rule of thumb is 10-15 man-hours per screen minute plus a day per reel (10 min) to mix. You can cheat a little and the more you work the more you know how to cheat but it's still a lot of time. That 90 min no budget epic should take a single editor around 90-135 days (10 hour days) to finish. Assuming you cut a LOT of corners and get lucky your only going to cut that in half or so. After that your cutting into what is possible. Truthfully a lot of no budget films don't really need a full on treatment and it may make them "look" worse, if the visuals don't live up to the sound by a wide gap. But the closer you get to the above as a rule the better the film will come out "looking". A good sound track will look better through some magic in the brain. A great looking film with a bad track will not look as good as it really is to an audience.

One way to cheat is for scenes you know are going to be BIG score, don't do much coverage. Pick a few things that should be heard and then just skate the rest.

This turned out to kind of a novel but I didn't think it had been really covered here before.
Click to expand...
thank you so very much. this taught me a lot =]
 
TV and film are different worlds in terms of the mix. Film is mixed by ear on a calibrated system. TV may also be mixed that way BUT you have delivery specs to hit. As a very general rule TV doesn't want anything above -12dB. But it get a lot more complicated. Discovery channel, and a bunch of others now, want you to hit a specific "dialog normal" number. The only way to get the number is to run your mix through Dolby a hardware box that runs ~$5k (there is a software option now also that is a bit cheaper). Film gets complicated if your going to an old school optical track like 16mm.

Nobody in film is aiming at the hottest poss (well producers maybe but if you calibrate your system then getting close to zero is going to hurt. Even in music there are some very good reasons to leave atleast 3dB of headroom, though you don't see it that often on pop music CD's.

You should head over to the DUC. There is a lot of information about calibrating and mixing levels and delivery specs.
 
Thank you everyone once again. I have been reading all morning! I will most probably be reading this stuff for quite sometime before I am comfortable with my workflow. Heres a couple of additional questions.

Situation / Question. When calibrating my monitor levels to 79 - 80 dBc and then testing out some movies / videos I get mixed results. For example Master & Commander sounds nice loud and comfortable at that level setting, ok seems like im doing well. But, I then put in Buffy the vampire slayer (love that show) season two and it was unbearably loud at the same monitoring level. I was wondering what the reason for this was and how it would effect my monitoring settings since most of the things I will be mixing are for internet and dvd release. Was 79 dbc too loud?! Isnt that the lowest level one should be mixing at?

I will be mixing by ear as much as possible but I would like to "plant" my dialogue somewhere safe by the numbers and mix around it. Am I safe to assume using phasescope on my dialogue bus and keeping it around -30/-29 LEQ(A) is a safe bet for internet/dvd?

Of course I will be asking each client for in depth specs as to what they want, but most of them so far had no idea what they wanted in terms of mix or final print levels as we are dealing with smaller independent things more often.

I think im starting to get my brain wrapped around this! For now, back to gearslutz and the duc!

cheers
-MTLA
 
How to record Foley, WIFO, etc.

How to record Foley, WIFO, etc.

Thank you Noiz2 for all the great information. I am new to audio and would ask you what equipment you would use (recording device, mic, etc.) to record your Foley, WIFO, etc. Thanks again for the help. JDB.
 
MontrealAudio said:
For example Master & Commander sounds nice loud and comfortable at that level setting, ok seems like im doing well. But, I then put in Buffy the vampire slayer (love that show) season two and it was unbearably loud at the same monitoring level.
Click to expand...
As Noiz already mentioned, it's the difference between Film and TV.

My biggest challenge these days is figuring out how to mix stuff for the 'net. There's everything from iPhones to earbuds to crappy laptop speakers to really nice computer speakers. Trying to find a "standard mix" that works on all those platforms is quite a challenge.
 
So, I summarize for personal clarity:

- Understanding basic standards with the different media venues,
- Proper monitoring calibrations (79 - 85 dBC) and testing on different monitoring systems @ different levels
- Trustworthy ears

All of this combined makes for a proper mix in todays confusing world!

Hey Bob, are you mixing for web broadcast as well? What kinda mixes are you going for in terms of levels for an embedded web video? I am thinking a DVD style mix would be the most appropriate?

-MTLA
 
That's a question I would ask on the DUC. Internet is the wild west, I have no idea if there are any standards at all other that loud as possible. TV tends to be pretty compressed, compared to film so part of what your hearing is the average being kept higher. My guess is that when they take TV shows and master them for DVD they boost the level since TV peaks would be generally lower than film and dialog probably also. If the show was heavily compressed that would make the show quite a bit louder. But I have no idea what is going on with the DVD's of shows. Especially a broadcast show like Buffy. The distributer might have shopped out the DVD "remaster" to just about anyone. The TV specs come from two things. One TV generally plays in a noisy environment so you need to keep the levels up or the quiet stuff gets missed and Two stations can get their licenses pulled if they "over modulate". That's also why broadcast video has no real black or white. If the white is too bright it bleeds into the sound info (that buzzing you hear sometimes with on screen titles. If they go over their allowed spectrum (over modulate) they can bleed into another stations spectrum. So broadcast has big ass limiters that they slap on all the signals going to the transmitter and they insist that you give them material that won't hit the limiters too hard. Today this is a bit antiquated since a digital signal doesn't "over modulate" and a lot of the stations with draconian standards are cable! The new mantra is that all shows should sound the same loudness so viewers never have to touch the remote (hey they might just change the channel). This dialog normal metadata is used by a processor to alter your mix as it gets sent out the cable so that when the viewer switches from the Monster truck rally to the symphony they won't have to adjust the volume. Doesn't quite work since different networks have different standards, but it keeps post jumping...

DVD's don't have the same restrictions so Buffy may have been remixed for DVD and who knows what standards were used for that.

The short story is while it's pretty safe to use movie references TV shows are not predictable. Also with movies only the original format is really reliable. If the film was mixed in DD 5.1 there was probably some time making the DVD mix, but the "stereo" mix may be either just an inbox crash down or something that not much time was spent on. Few directors get too worked up about the LtRt version of the mix and it's often done in close to real time with minimal adjustments. Level wise it should be OK but it isn't nec what you heard in the theatre. By the same token old films that have been "remastered" for DVD and made into "surround" mixes may not be the mix stage levels either.
Your best bet is to go to a film mix stage if there are any around you and talk to someone about sitting in on a mixing session. Depending on the facility you can probably get in to sit through a few hours (it gets real boring if your not part of it). But you will get a feel for what the levels should sound like. So even movies on DVD may not be great level references, but they should be good balance references.
 
Very insightful thank you Noiz2. I will go drop that Q over at the DUC. I went to a post house as a "field trip" at the audio school I went to, and am now doing a little teaching at. Gonna find out who the contact was and take you up on that advice.

cheers

-MTLA
 
Noiz2 said:
Even in music there are some very good reasons to leave atleast 3dB of headroom, though you don't see it that often on pop music CD's.
Click to expand...

Hardly! I do quite a bit of album mastering, and the "volume wars" in pop CDs have pretty much forced all of us to cram everything into the top 6 db or so of the dynamic range, typically with a lot of clipping. There have been countless discussions of this in audio professional groups, mastering engineers' websites, internet forums and the like. Simply put, it's a tragedy. Thankfully, movie sound is completely unaffected by the trend.
 
The "loudness war" in pop CDs has nothing whatsoever to do with how loud it's played. It has only to do with the relative volume of one CD versus another when the playback (amp) gain is the same. The theory (generally correct) is that "louder sounds better", so program directors in radio stations will tend to favor them, and the louder CD on the changer will sound better at home. Again, this is statistically proveable.

The unfortunate consequence is that in order to be "level-competitive" nowadays, mastering engineers (like me) are forced to use compression and particularly fast peak limiters (aka "maximizers") to get the level up to match the competition, and some of those competitors evidently care only about volume. Clariity be damned. This has literally destroyed most of the fidelity of modern pop CDs.

For proof, take any 10+ year old pop CD from your collection, play it and set a comfortable listening level, whatever that might be for you. Then take any CD made in the last couple of years and play it. It will blow you out of your chair and have you diving for the volume control. I guarantee it. Then, if you can, set up a situation where the two discs are matched for apparent level. I also guarantee you the newer disc will have far, far more distortion (of an undesirable kind).

For movies, lucky for us where this is concerned, the volume decision is out of the audience's hands, so the "absolute recorded level" of the soundtrack is almost irrelevant. The theatre will play it at whatever level its technical staff deems appropriate. Even in home theatre, there isn't exactly the equivalent of a CD changer (or iPod playlist) so movie listening levels are usually readjusted each time (and often during, which is another subject - perhaps relating to your comments).

So a movie soundtrack doesn't benefit from "maximizing" or other digital audio level tricks, and consequently its original fidelity is largely uninjured. A much wider dynamic range is called for and generally used compared to music CDs. If a movie score never hits full digital level, or does so only briefly, it's no biggie. In pop CDs, that would sound vastly quieter than the next guy's CD (unless, of course, you simply turned up the playback). Much, much cleaner too, all else being equal, but sadly "clean" is a casualty of this particularly unfortunate "war".
 
Last edited:
You must log in or register to reply here.
Back
Top