Vision Mixing

NG

noggin Founding member 8 August 2008 3:01pm

william posted:

What do people think about the rule that sound follows vision and vice-versa?

I'm not talking about a vision mixer where the moment you cut up a different source its sound is automatically selected, but the artistic merits of the order you do it in.

Sound follows vision seems standard for live productions, news etc. - but I've always thought shows that have been through post-production - e.g. wildlife documentaries, benefit from a few seconds overlap of the audio from the next shot over the previous one (say you're about to cut from a rainforest to a waterfall or something.)

Is there a standard method people are taught?

Often sounds leads vision a tad on live shows - as sound faders can be open prior to a source being run and cut/mixed to.

Don't think there are hard and fast rules - but when mixing from a VT to studio I think sound lags (where the sound continues as it fades away) work quite well at softening transitions.

Whilst cuts are routine in vision mixing, sound almost universally requires fades, which means you always end up with sound leading or lagging.

NG

noggin Founding member 8 August 2008 3:09pm

Spencer For Hire posted:

There's a great temptation to sit and admire your finished product on the TX monitor whilst you're working. Obviously you need to know what's going out, but if you lose concentration on the preview monitors, you're doomed!

Yep - transmission is the last monitor a vision mixer should be concentrating on - you know which source is on air from the preview stack (it has a red light underneath it if you've forgotten...) The next source - if you know what it is - is the one you should be concentrating on, or all of them that are likely candidates.

It also helps to look at your script now and again if you have one

Quote:

Other things that I was taught that have stuck with me... Always 'cut on the action' - so if you're going from a wide-shot to a closer shot where someone's, say, about to sit down, aim to cut as they land on the chair. Don't cut before or after, or it'll look messy.

Yep - though cutting early to allow action to start in-vision can also be useful, as can cutting once the action has finished.

One of the most annoying cuts is when you are on a wideshot, and see something about to happen, and then cut exactly as it happens. Often you want to see it happen in the shot that you've seen it about to happen in.

Quote:

If you're vision-mixing for an interview, look for the facial movements which indicate when a subject is about to start talking so you can cut to them just as they start speaking, rather than just after.

The most important thing is to look at those listening for reactions - are they violently disagreeing in their body language? Are they nodding in agreement?

It is also vital to listen. Is the speaker talking about someone else present in the studio - is there a shot of them? Are they reacting?

Watch out for two-shots where you can see how people relate to each other. Don't feel you have to stay on an MCU of the person speaking.

Quote:

Also avoid cutting whilst the camera is moving - so if you're cutting away from a tracking, panning or zooming shot, wait til the movement has finished before cutting... although sometimes you can get away with a mix out of a moving shot.

Hmm - that is very much a decision the director will make when deciding the style of the show. Some shows employ lots of tracking wideshots or jib shots, where the style is very much to cut to and from the camera on the move, rather than waiting for it to steady. Equally on a music show the move should ideally happen just before you cut to the shot, rather than just after, as you don't want to see the move start in-vision, particularly on music numbers where the style is for constant movement.

Quote:

If you're using music too, aim to cut your shots with the beat.

Hmm - cutting on the down or off-beat can also be very useful if you are trying to keep things interesting.

NG

noggin Founding member 8 August 2008 3:15pm

GaryC posted:

all of noggins comments are top notch.

However the kit you are using is quite low grade (but IS used for broadcast - 'live roulette' and the 'babe' style channels use the panny MX series a lot) and if the cameras are not fully pro and genlocked together outside of the mixer (which has a frame sync to fix the problem of using low grade equipment) then DO NOT hot cut the show by puching programme row only!

These type of mixers need to sync the video in a frame buffer before switching, so always try and load the next source to the preview row. If you do not the MX series has an odd freeze frame effect from 4-10 frames of the current source on-air while it hunts for sync. Sometimes not noticable, but it can become a problem.

As to fade/cut argument; all i would add is things that are at the end of a sequence like a VT should make you ask the question 'is this next bit related?' If its a news report and you are moving on to the next story, then while you cut into the report, fast fade mix out.

If you have someone on cameras, then movement is GOOD! but only if cutting/mixing to the source (avoid starting to move a shot that is static and already on air) A slow zoom or track in at the start of something welcoming is pleasing. As is the reverse at the end. Don't ever go in the opposite direction (ie zoom out on talent at the start of something)

The only major point I would add is try and use establishing shots (all guests and interviewer in frame) to give a context to the viewer if you are cutting between question/answer/reaction more than 3 or so times.

Yes - I had assumed that the system would be genlocked (not everyone who uses the mixers in question relies on the internal synchronisers to time things)

Not being able to hot cut on the programme bus is a major limitation - you simply can't do a good job of cutting a live interview in a reactive manner if you have to preset every camera and then cut to it. You'll always be late - or end up keeping things to safe.

Additionally, when directing interviews, try not to repeatedly cut between the same shots. If you are able to, try and reframe your cameras each time you use them, apart from guest close-ups (you can have more fun with the presenter close-up camera) This is not always possible - but in some situations it is. In simple 1+2, 2+2 situations etc. this isn't always possible if you have only a small number of cameras - but if you are doing a 1+1 with 4 or 5 cameras with cameras on peds that are able to re-position quickly then this can be a good way of keeping things interesting. Obviously it depends on the programme style as well.

Always consider if anything you are doing is improving things for the viewer, or is likely to annoy, irritate or distract...

Another standard rule is to never make your presenter bigger in frame than your guest - it makes the guest appear less important.

A personal bug bear of mine is pulling the guest close-up camera out to two-shot and then tightening on-shot. (Zooms are not a natural shot change as the eye doesn't zoom - tracking is so much nicer. When watching drama from the 60s when zoom lenses weren't widespread you see far more interesting camera moves - as re-frames in-vision have to be done entirely by camera movement. The only other way was to rotate the lens turret to a different lens angle - but would have to do this out of vision)

(August 2008)