Today's rantlet (a comedic piece of sound advice in the style of Jeremy Clarkson) is aimed at video editors and producers:
Number one to tick off on any editors todo list should be chat to your dubbing mixer. We are often nice and know the latest jokes from all the super talented vo comedians we regularly record! We can suggest workflow and solutions - we want to make it work smoothly for you so that you come back for more. Communication is vital. Even more so if you are editing a drama.....
Please understand the term "locked off" to mean what it says - and understand the implications otherwise. I totally understand the need for changes due to late arrival of GFX and VFX etc There are workflow solutions to this such as reconfirming software like conformalizer (by maggot software) and the use xml files from your edit (in fcp). These solutions help speed the process up dramatically, but it still takes time and you need to both allow for this and budget for it. If there is a possibility that you may need to alter your cut, flag this up at the start and plan for it. Throwing in a change at the last minute can mean horrendous complications - not to say that it is not do-able, but it can have many ramifications.
Consider your audio tracks layout. Number one of any dubbing mixer's pet hate is the randomness of audio tracks from an edit and the mixing of mono and stereo tracks on the same audio tracks. Please see this example of naughtiness:
It slows us down inordinately and makes copying tracks from your omf import to our pro tools tracks very time consuming . this example took a while to sort out:
to something nice like this:
So the rules are:
- Stereo tracks must start with an odd number i.e 1&2, 3&4 etc and not 2&3, 4&5 etc
- do not mix stereo and mono on the same track
good guidelines for practice are:
1) keep your talking heads on separate tracks to music and/or fx - see my example omf above where I have separated out the dialogue (red), music (green) and FX (blue).
2) keep same interviewees on the same track - it really speeds up sound mixing if 'man a' is on track 1, 'man b' on track 2 etc especially if they occur multiple times in a show.
3) keep cut-aways if they are non dialogue based on different tracks
4) if a track is mute - discard it. The golden rule of post is "do not pass sh1t on!"
5) keep vo (or guide vo) separate on its own track so that we can mute it whilst recoding new vo or mixing the M&E
I am sure handling vast numbers of tracks in any edit program is not as easy as in sound, but going some way towards these goals will help your production sound better as more time is spent doing sweetening - rather than rescuing a train smash of random bits!
Talk to your dubbing mixer about video codecs before you hand him a low bit rate h264 formatted for ipod (as in the nicest way) - it may not best please him.
Certain video codecs like h264, whilst being small and dandy for internet transfer, do not play so nicely with pro tools - as they are very processor intensive. These codecs operate by working out how a frame has changed compared to certain key frames and calculate the difference - this calculation can lead to laggy video performance.
Personally, I have an avid mojo for guaranteed frame accurate sync (unlike rubbery quicktime). Anything that is sync critical (or for long form work) I transcode on my avid media composer whilst I import the omf and sort the audio tracklay.
In my studio I spend my cash on the best plug-ins and software for sound that money can buy, and not on HD video infrastructure/playback. SD video is good for me to see lip sync well and for me to work nicely to picture. You can enjoy your HD pictures at the layback.
My picture preference is for avid dv pal mxf media at 720x576 (or widescreen equivalent)
For fcp users I like a dv pal quicktime at 720x576 (or widescreen equivalent)
If internet video transfers are required, I will accept H264 pictures, but I like every frame to be a key frame. I have 100mb fibre internet connection, so if you have a fast server, I can download your video very quickly, so major league compression may not be needed!
Consider your sequence.
1) It is common UK practice to start your picture/programme at 10:00:00:00. Deviation is by consultation (i.e if your programme is longer than 90 minutes as it can cause some versions of pro tools to be a pig at 11:30:00:00 timecode and beyond.
2) have you considered part breaks? Please start them at logical whole minutes.
3) agree final out time for your parts and final duration time for the entire show with your dubbing mixer. Reverb tails chopped off by networks are very ugly sounding.
Consider synchronisation - belt and braces is the sensible option here.
a) A good old fashioned 2 plop and sync flash at the beginning and end of the film are hard to beat when used in tandem with
b) burnt in timecode (bitc). Please avoid putting your bitc over name supers and subtitles etc - it makes it tricky to see what is going on (or can make lip sync a chore)!
Current UK digital production partnership specs require
- The sync plop must be between timecode 09:59:57:06 and 09:59:57:08
- The audio plop must be 1kHz tone on all tracks at -18dB (standard zero level)
- The duration of the vision flash must be 2 frames to allow it to pass through standards conversion successfully
- The audio plop must be synchronous across all audio PCM audio tracks and with the video flash (within +/- 5 ms)
- If an end sync plop is used it must be no closer than 10 seconds to the end of the programme and comply with the points above.
Delivery requirements - who are you mixing for and what do they want?
Suddenly asking your dubbing mixer to deliver a 5.1 version of your programme by tomorrow (because you forgot) may cause him to choke on his food.
Making a 5.1 mix is not just a simple thing. Ask your editor to turn a shot into 3D and watch him fall off his perch.
If your programme is for different clients, then they might have different loudness criteria to mix to and different delivery requirements. Discovery Channel have very different mixes to Sky or Nat Geo. DVDs often have a different mix to broadcast.
How is your stereo mix derived? from a fold down of the 5.1 mix in the satellite box (Sky) or from a separately made stereo mix?
I try and avoid “scope creep” on a project by asking the right questions at the beginning of the project and making sure that each party’s hopes and expectations are agreed, along with the implications of any changes.
Payment terms should also be agreed up front so there are no nasty surprises for either party. In over 15 years of busy freelancing I have lost only one weeks work due to non payment.
Forwarding the broadcaster’s delivery requirements really help those nasty little “quality control itches” that can occur. Here is the Digital Production Partnership version for the for the majority of the UK.
Please bear in mind that it is often a job requirement of many qc departments to do silly things such as in these examples (they pay peanuts - they get monkeys):
- 1) fail a period drama for having nothing in the subwoofer channel (not really required - as no space ships are landing or atom bombs have gone off)
- 2) complain that there is voiceover on the M&E - when they are in fact listening to the M&E tacks and the full mix tracks all together (strange they didn’t notice that it was phasing or that the mix was sounding just plain wrong)
- 3) fail a programme for digital clicks when there are clearly footballs being kicked in the back of shot
Consider the export options.
Before you export have you checked your sequence that it doesn’t contain mixed sample rates or mixed bit rates?
Is your sequence locked - no really, is it locked or “locked”....
Are you exporting an AAF or an OMF?
Have you discussed handle lengths? This really can save your bacon. Ignore at your peril. Long is good!
Is your export embedded or referenced? - this can have a massive effect on the metadata.
Have you made an XML file or vision edl before you go ahead and make changes to your “locked” picture?
Is your AAF/OMF embedded file less than the 2Gb file size limit. If not, split the omf into parts out vertically into tracks (i.e tracks 1-8 and tracks 9-16 etc).
How are you doing the Layback?
The cheapest option as I am sure you are aware is to have the mix completed while you grade, and then have the dubbing mixer send you files for you to layback sound and graded picture concurrently.
Some dubbing mixers are aghast at this and worry - in my view unnecessarily. I work with many very good editors and have never suffered a problem yet in many years of working this way. I include line up tones on my mix files and stems as per the delivery requirements. The editor just has to light blue touch paper and listen to my mix, smile and collect the BAFTA! The wonders of SDI and an agreed synchronisation workflow are the key here.