Shooting an entire sequences at 40fps with sound. (3 replies and 2 comments)
Is it possible at all? I've been wondering if I shoot certain sequences at a higher frame rate (but not yet considered slo-mo) just to have a dreamy, smooth motion to the scene but with dialogue. Not an expert with sound so I hope some of you would know if it can be done or may have already tried it. Any input on the matter will be greatly appreciated. Thanks!!
OK, so the Fairy Godmother has to be slowed down to give her a magical quality, but still has to be able to speak. I recommend 60fps at least. You will find that some shots are better at one speed and others at another. Just squeeze and stretch in the NLE to suit the action. Some shots can be slowed right down: others can be more or less in real-time.
As for lip-syncing - don't worry. ADR the whole thing to give the Fairy GM role an intimate sound, i.e. lips fairly close to the mic. You obviously can't do that on camera! Make sure that there is no 'popping' of the mic, so double pop-filters - a foam and a shield and try different mics to suit the speaker.
You may find that the perfect Fairy to look at may not have the perfect voice anyway, so someone else can be used for the ADR.
Most NLR software packages are not very useful at this task and I recommend Reaper for all ADR tasks, as you can put so-called 'Stretcher-Marks' anywhere on the audio and even assign the placement of the SM to a single key-stroke. This speeds things up enormously - if we remember that there are some five to ten thousand audio events in a movie and each and every single one has to be tweaked and moved, saving a minute on one event can end up saving several days work!
To sync' the voice, simply place the audio next to the bit of video you want it to match. Then find the consonant where the lips close (B, M, P etc.) and place a Stretcher Mark on it and move it to where the lips close on the video. You will need to practice this a few times until you get the hang of it.
Reaper plays and can edit video as well, so it is ideal for the task. TBH it is my go-to Swiss Army Knife for all things A-for-V, as you can throw any media file onto the time-line without any conversion.
When you are recording the audio for ADR, show the ADR artist the video as it will appear on-screen and get them to more or less match the new speeds, but thanks to 'Elastique' you do not have to get them to lip-sync - the horrible days of endless 'looping' and following cue bars on-screen are over! (Other squeeze-and-stretch plug-ins for audio like 'Elastic-Audio' tend to introduce warbling, so the audio-post editor has to spend ages cutting and pasting to match the lips.)
I feel that you are going to learn a great deal if you get this sequence right!
Not sure what you are really trying to achieve here!
Vision and sound are two entirely different subjects and have to be treated as such.
You are recording an image at around 40 frames per second as you want to slow down the movement to achieve a “dreamy” sequence. I assume you are using a film camera and a separate recording machine or are you using a hand held video camera with synchronised recording. Your concern is that lip sync will not line up with the dialogue. If you slow down the image then the dialogue will naturally be out of synce, you then will have to edit the recorded dialogue that will ‘skip’ parts of the script so certain words will line up with some of the spoken dialogue. This is going to be very tricky.
It all depends on the script and the positioning of the words. How many words are allowed to be cut before you lose the overall meaning of the scene. How important is this scene? There are a number of ways you can do this. One way is not to film the actors face full on, hide or disguise theirs mouths so the dialogue is heard but the audience does not see lips moving. This has been common practice for years. The voice must match the reverberation of the location shot otherwise it will sound artificial. Always record a “wild Track” of the location sound without actors, this will make it easier to blend in the dialogue during post. Most high speed filming is usually to highlight an alarmed situation, a woman screaming, man shouting for help etc, etc, but is accompanied by a long scream or shout which does not need lipsync being an hysterical re-action.
Probably better if you advised us how you envisage this scene.
I was just considering for a project in which a character will be speaking to her "idol/fairygodmother" to shoot the idol in a faster frame rate again just to give a slightly different sense of motion to the "apparition". Yes, I understand the difficulties. I will also consult a live sound recordist if he has ever tried this or what, if any, are the possible techniques to achieve this. Thank you!
You'll need a camera that does 40fps - though I'd go 60fps or more and then make the frame rate variable to choice and taste in whatever editing package you're using. But Yes! Frame rates are speeded up (e.g. bits of fight scenes) and slowed down very, very often.
Sound - That's why the gods invented ADR! If you slow down audio, it will just sound silly. Nowadays it must not drop in pitch, thanks to a plug-in called Elastique from ZPlane in Berlin, but it still will sound out of whack.
So you will have to ADR any dialog and probably do incidental sounds as Foleys.
If the sequence is otherwise using live sound, remember to use the same mic and at the same rough distance and in the same or matching acoustic environment. That is why all post-production rooms have a Sennheiser MKH416 lurking in the mic locker and of course have an anechoic vocal booth and either a good reverb machine like a Lexicon 960 or at least a healthy selection of reverb plug-ins!
Tip - Elastique comes free with DaVinci-Resolve and with Reaper and both packages have a good selection of reverbs - but in this instance, you will just have to redo the audio as ADR and Foley tracks.
Thank you for this!!