Multicam Audio

[Philip Howells]

AQlthough I've been 30+ years in this business, we now mainly shoot weddings using three EX1Rs, typically each with one or two channels of radio mic and the others with short guns on the cameras. We also use a Zoom to record any special sound eg organ, string quartet etc, so our normal sound mix down involves at least eight tracks.

I'm at the stage of leaving Liquid (and rather embarassed to find how much earlier many other Liquid users "saw the light") and I'm jumping in with a real project on the demo version of Edius. Actually the main multicam sections of that project have already been done in Liquid but I'm tempted to re-do them in Edius to get some learning in with its Multicam feature.

I have two specific concerns and a more general question.

1 In Liquid I sync'd using the sound track of the clips, dropping the vision and displaying the waveform in the preview monitor which I note isn't possible in Edius. Is Plural Eyes a practical solution?

2 Is it possible to do my multicam vision edit with the audio following and then drag all the audio tracks in to the timeline and do a sound mix?

3 I am very excited about the possiblity of using a Behringer BCF2000 as I've always felt that audio is best mixed with fingers on faders - or perhaps it's my age! Is it as simple as the text suggests or have I been unfairly seduced. At £150 for the surface it sounds like a snip.

I have searched on these topics but without much success so many thanks for indulging a learner

[zvit]

First, why aren't you using a visual sync mark at the wedding? Like all cameras filming a flash together or someone using their fingers as a movie slate? It's much easier to sync that way. (if you don't use timecode etc...)

You can also sync by the visual waveform after stacking up the audio tracks. (Make the height of the tracks max to see better and set the waveform to linear (in settings).

Yes, you can make audio mixes.

I think I read that Plural eyes has a plugin for Edius but never tried it. I know they have an external engine that syncs audio pretty good.

[Philip Howells]

I know a number of people think this but I can give you three reasons for not using visual markers

1) they're are very hit and miss - if they're short enough to be useful (eg 1/1000th second flash) they can easily be missed off the vision, ie simply not recorded on some or all cameras;

2) with three camera each with widely separated eyelines a single flash often doesn't cover them all and certainly not to the same intesity;

3) flash isn't very visible out of doors or in brightly lit churches.

After years of experimentation (we did a lot of bands before starting in weddings, I found audio is by far the best way to sync tracks but that's just my opinion). I believe Plural Eyes relies on audio and won't actually work with mute tracks.

[John Hooper]

There is a 30 day free trial of Plural Eyes here:

http://www.singularsoftware.com/pluraleyes.html

To add extra audio after the vision syncing, just match timecodes from the other sources. I assume you lock your cameras timecode together?

If you don't have timecode references, why not try a cheap version of the Digislate? You can get an app for the iPad or Android tablet, see here:

http://apps.mechnology.com/my-apps/sl-digislate/

I have not used the Behringer, but it is a large item to have on your desktop, just to use as a mixer.

I bought a brand new Fadermaster 4/100 on eBay for a good price, but have hardly used it. I think it would be useful when you have long sections to do that involve ducking music behind voice overs etc.

[Philip Howells]

John, Many thanks - I've downloaded the trial and am waiting for it to make it's first synchronisation. Probably complicated because two of the three tracks are .m2t with separate audio - the result of fusing after colour correction in Liquid (I mentioned that I'm halfway through an edit didn't I?).

With two 24inch monitors and a Spirit F1 routing sound to my monitor amp and the Betacam-SP(long story!) I'm not stuck for room so the BCF2000 sounds like a luxury come true. AFAIK Liquid could only be used with very expensive external control surfaces.

[GrassValley_PS]

To answer your question, Manual Audio sync works pretty easily, it is just done down in the timeline vs. the preview monitor.

[Philip Howells]

Thanks for that assurance mod; in truth I never had difficulty in syncing manually in Liquid and since PE simply went nowhere, my guess is that it's one purchase I won't be bothering to make.

[Harro]

I have used PluralEyes to sync a concert performance shot with a Sony EX1 and Canon HV20 without any problems, worked perfectly.

[antonsvideo]

To answer your question, Manual Audio sync works pretty easily, it is just done down in the timeline vs. the preview monitor.

I agree, I have never had a problem syncing with the waveforms of the clips on the timeline, especially if you change the audio track height and use linear waveforms and not log

[Philip Howells]

I have used PluralEyes to sync a concert performance shot with a Sony EX1 and Canon HV20 without any problems, worked perfectly.

In case I gave the impression I felt PE was at fault, clearly, although I've followed the brief instructions and allowed Edius to "prepare" the timeline for me, I'm doing something wrong.

I followed the instructions this way:

I created two more V tracks (ignoring 1 which is VA evidently by default).
This also created to additional A tracks for each V. I'm used to working with separated A tracks, in fact Stereo tracks would complicate things since we record different mics on different tracks.

I dragged the three clips to the V tracks and the appropriate A tracks were filled by the relevant audio. I say relevant because there seemed to be no choice. It appears that this allocation of tracks acts as if the 1Title track is the centre line and V tracks number 1,2,3 etc upward from this track whilst the A tracks number downward.

Since Edius worked in this way, I assumed that PE identified the relevant tracks according to the names not the number of the tracks.

I tried both selecting all the tracks and leaving them (as per the instructions) unselected and also had the timeline in and out of Multicam mode.

I activated PE and when the files had loaded, selected Sync.

And that's all that happened.

As I say I have been in this game for long enough to recognise that when this solrt of situation arises it's invariably if not always operator error so I'm blaming no-one other than myself.

I also recognise that I need to some manual reading because presumably it's possible to change tracks from VA to V and A. It's just not been exceptionally intuitive in this area. And please good colleagues (who've been very polite and helpful to date), if it's a matter of reading the appropriate page in the manual, don't feel you have to do anything other than perhaps give me the page or chapter number - at over 1000 pages it's quite a tome!

Thanks in advance.

[Philip Howells]

(snipped)

To add extra audio after the vision syncing, just match timecodes from the other sources. I assume you lock your cameras timecode together?

If you don't have timecode references, why not try a cheap version of the Digislate? You can get an app for the iPad or Android tablet, see here:

http://apps.mechnology.com/my-apps/sl-digislate/

(snipped)

John, many thanks for your full answer.

Unfortunately coming from Liquid (and Z1's with MRC1s where timecode is another land) I don't bother with the timecode. I did once, but since the Z1s cant be genlocked and in any case at weddings it's impractical, setting ToD with the remote commander simultaneously by pointing at all three cameras still leaves a couple of frames inaccuracy.

I'm not an iPad or Android user (age probably, sorry - I bought a phone to make phone calls not take pictures etc probably sounds like Luddites Anonymous though I'm not really!) the app isn't very applicable but I'll look into it and see if it justifies my investment. At first glance it would at least show me the offset for each camera from the Remote Commander method.

Many thanks

Philip

[plasma_video]

Hi Philip, I can't give you a specific set of "instructions" for doing the synching and track layouts, as there are many different ways to do it.

I will say that I routinely edit 3 camera church productions with Edius and have found the multicam mode to be very good for my needs.

You will have one of those "ohh - that's pretty cool" moments shortly, once you find what works for you. I've always found the A/V track thing a bit mysterious, but once I had one of those "ohh" moments, things fell into place.

I sync using visual cues and the audio waveform, although sometimes it's more difficult with the audio, as the time delay and room acoustics sometimes throw the waveforms into something hard to match.

Typically, I start all cameras out on the same shot, and sync them to, for instance, a movement where a hand touches an object and quickly pulls away.

There is a "nudge" feature in Edius. For Multicam, you will find it to be your best friend. Once I get the gross sync set, I use the nudge to fine tune the positions.

I also do something a little different than others here, and place each camera in a different sequence, and then put the sequences into a master sequence for the multicam portion of the edit. I do this for several reasons. For me, it's easier to make color corrections to each sequence, and then the color corrections carry over into the master edit. There are other ways of doing this, but that's what I do. This has its pluses and minuses. For instance, the waveform display doesn't carry over to the master sequence.

Also, take time to see how transitions work with the final edits once you have made your cuts. there are several different ways of adding transitions and the y interact differently, depending on the mode you are in.

Once you get used to it, you really should find it to be a very good multicam editor.

[Philip Howells]

Thanks for the encouragement plasma; you seem to have grasped exactly the point I'm at. I have been additionally frustrated by first trying to import clips that were colour corrected in Liquid which means I'm using separate audio (.wav) and video (.mt2) files which evidently Plural Eyes doesn't like and second by inadvertently corrupting one of my cards which means I can't start over and bring the XDCAM EX clips in natively. The instructions (such as I could find) mentioning nothing about waiting until each clip had been "processed" and got its blue dot in the corner of the thumbnail. And I still don't understand what value there might be in Adding to Bin and Adding and Transferring to bin - without the clips in the bin how can one proceed?

The best I can do now is to re-import my .mxf clips which, for some unfathomable reason (for I'm following the instructions religiuously) Plural Eyes won't use. Because I've never seen a demo of PE I don't know whether I should expect some sort of action immediately I start the sync or whether I should expect the activity bar to show no movement for a while - I think I've rules out the last because after leaving it for 6 hours nothing happens, but that's just an assumption.

I appreciate your workflow because colour correction is always a feature of our post production but I thought it would be something I could do after the multicam edit. Can not the same correction be copied to each shot which needs it in the final multicam edit?

As far as syncing is concerned, the other thing I considered was with all three clips in the same sequence but on different tracks, to cut each clip at the sync point line them up and then drag the earlier part of each clip out backwards so I could work on the section pre-the sync point. I haven't tried it so it's just a theory.

You might wonder why I've decided to change NLE mid edit - it's principally because since my last clean re-install of Liquid I have had trouble with the latency of the Liquid/Windows XP installation (I have two C disks in a caddy so I can swtch between operating systems). I'm at the point of abandoning Edius and soldiering on with Liquid but I sense that if I give it up now I'll not bother coming back and "soldier on" again. I'm not a quitter but you're absolutely right about needing a Eureka moment. Fingers crossed.

[Ron Evans]

I do my colour correction after multicam edit. Multicam will run a lot faster/smoother if there are no filters on the clips. When done I correct a clip from each track and save filter set as a named preset. Clear all filters from the tracks and apply these named filters.

Ctrl+A to select the track. ctrl+shift+alt +F will delete all filters from the selected track. Ctrl+A to select the track then drag the named filter over the track to apply to all clips on the track.

It's easy then to check through and see if each clip needs the same correction or modify the filter set on the particular clip.

Ron Evans

[Philip Howells]

Ron, thanks for that advice which of course makes logical sense.

So often logic seems to be the answer. One of the things which Liquid failed to mention in the early days of multicamming was that unless one had a really powerful video card (which in those days was expensive and hence rare), you had to render the multicammed track before it would play or edit. It's a bit like my error with the importation of my native XDCAM EX clips. I simply missed it (if it was stated rather than being thought "logical") - so I'm not trying to shift responsibility - I cocked up.

In that respect, I've just watched the Event DV video on Proxy editing and I wonder how much use people here typically make of proxy editing. Going back to my old days that was, of course standard practice using say a VHS offline to produce an EDL which was then actioned and tweaked in the online, only in those days it wasn't called proxy.

Edius seems set up to encourage that sort of workflow with its filters being able to be applied easily later as you do. In contrast, in Liquid if you didn't do your correction to the original clip, it meant selecting every clip that required a particular correction and applying it individually - albeit by pasting.

Incidentally I was interested to read an early comment of the moderator on the samples thread about this not being a social forum. I happen to agree with the philosophy in so far that very few samples show off specific examples of what that particular NLE can do so it's really a bragging channel. However, that's OK for those who want it and so I make no further comment. What I do want to add is how civil and helpful this forum seems to be - notably lacking the "this is my way and it's the only way" syndrome so prevalent on other forums. Just a glance through this thread alone will find so many caveats like "this is my way but there are others". Very grown up. Thanks.