HD footage w/TC on audio channel 2

Hey all --

Well, I've got something I have not seen before, thought I would toss it out there.

I have a HD feature coming, which they shot 23.98. The audio configuration on the HD camera masters is scratch track on Channel 1 and TC on Channel 2. The TC came from the audio recording device (some sort of hard disk recorder, not DAT). And I believe the audio TC is 29.97, audio is 24-bit. So there is "proper audio" as files, and HD picture with temp audio and TC on channel 2. There is a visual TC slate. They are planning on downconverting to DVCAM for the offline and editing in 29.97.

My first response to this was that I would be hand-synching the audio to the picture, shot by shot, using TC, claps and/or waveforms as reference. I would also throw away the Channel 2 timecode because I'm not using it to lock to anything. Or, the pic and audio would need to be synchronised in telecine, just as if they had shot on 35MM. But that may be outside the budget.

But the more I think of it, I wonder if there is a way that the pic and audio can somehow be put together in a more automated sense. What way is that? No idea, but I thought I would ask the forum, see if anyone had ever heard of this and might have some experience/ideas. Or is this something better suited for Adrenaline?

I have FCP 5.1, DSR-11, and usually log/digitize using Firewire. I also have a KonaLH card -- perhaps the Kona card + some kind of deck rental might be useful here? Not sure.

Anyway, I'm ready to "adjust the edit schedule" for hand-synching 30-plus hours of dailies, but before I do... anyone?

thanks so much
Karl

First, buy the production people something expensive for remembering to put the clapboards in.

By definition, certain portions of the audio timecode must line up--in time--with the top of the video frame. If you shot 23.98, then 29.97 time code will never line up and you may be creating more problems that you solve.

I wonder what everyone else has to say.

Koz

Hi Karl,

You can use Time:Tools to automate this procedure, but this will reset your video timecode to the timecode of the audio 2 and video master TC is lost. So if you use it you should work with an independent copy of your video files and then type in an AUX TC for each clip by hand.
To use the timecode and play rate of the audio HDR you can use FCP 5.1.2 as long as the audio recording is not 29.97 DF - this TC setting is not supported by this/any version of FCP. Additionally FCP has to be launched with 23.976 NDF if you want to import audio at this TC setting.
You can use the free SebskyTools or my expensive BWF2XML to convert the audio to QTs with TC.
You also can use Convert It to change TC and play rate virtually for use with FCP.
With any of those last three options, you'll have have to synch AV manually.

Hope this helps to get a better understanding.

Regards
Andreas

Quote
Andreas Kiel
my expensive BWF2XML

...only if you are misusing it.

For a show that uses it every day to process two 4 track recorders on an improv heavy set it was the best cheapest thing we bought.
ak

Here's an idea that may be useful.
Being that you are transferring the footage to DVCAM at 29.97 and then capturing into FCP from that, you could get a hold of the HDD recorder that the audio was captured to (or a similar system) and treat it like a playback deck.
Take the audio Ch.2 (TC out) from your DVCAM deck and send it to the TC input of the HDD recorder and put it in slave/chase mode. When you press play on the DVCAM deck the HDD should lock to the incoming timecode and play.
Now you take the AES audio outputs from the HDD recorder and plug them into the AES inputs on the Kona card and the video from the DVCAM deck to the Kona card as well. Send the scratch track on CH. 1 of the DVCAM to it as well via the analog XLR inputs so you can then check sync in FCP against that.

Hey guys --

WOW. I really appreciate all the thought you gentlemen put into this one.

Andreas, I did read some sort of tutorial on your BWF software before posting. What worries me about that is that this show is only offlining here. When editing is done, I have to deliver to an online facility, and deliver OMF (or similar) to post sound. I am VERY nervous about changing any timecode properties of any video. And, if I read the article correctly, the BWF software creates what seem to be reference files only of the audio -- small 4k files -- not true media files. I'm just not quite sure what that would mean in an OMF situation. Laying off from my studio, of course it would work, but sending things off to other people... I dunno. Any thoughts on that?

Koz -- I have mixed several HD shows in 29.97 and then laid back to a 23.98 master with no sync problems whatsoever. I was told once that that's because of "math," I have seen it work, and I choose not to question it beyond that. It's alien technology to me, but I've seen it match up perfectly. That said, my experience in 29.97 matching up with 23.98 is only in post-production, NOT PRODUCTION. But the recordist does not sound like a louse at all, and conceptually I think it can work. So, I asked the producer to provide me with some test material before I start, mainly to make sure that WE'RE IN SYNC TO BEGIN WITH.

Frank, I think your idea is the most intriguing, thank you so much. Your idea puts the Ch 2 TC signal to use -- this is what I think I was getting at -- WHY is the Ch 2 TC there, and can we use it to our advantage instead of throwing it away in the downconversion process because I don't usually use it? Perhaps this is worth a test, maybe even a day's rental of a HDD (I guess I'll be finding out what that is in due time). I will consult the sound recordist, see what he says about your idea.

Anyway, guys, thanks again for your thoughts. I'll post again when I come up with a conclusion.

Now, can anyone tell me why the Quicktime Conv-to-Mpeg-2 function disappeared? Grr

thanks again
Karl

Andrew Kines Wrote:
-------------------------------------------------------
> my expensive BWF2XML
> ...only if you are misusing it.
>
> For a show that uses it every day to process two 4
> track recorders on an improv heavy set it was the
> best cheapest thing we bought.
> ak

Thanks Andrew :-)

Karl,
BWF2XML does not change any source timecodes. The software Andreas mentioned that does that is called Time:tools. The major task of BWF2XML is to convert BWF files to FCP readable files. You can set it to make those 4k reference files that you mentioned or it can be set to create complete .mov versions of the BWF files. This is how we work and we have sent 4 episodes of OMF files off for mixing with no problem. Tons of multitrack recordings all imported, synced by timecode and multicam'ed with 3 cameras.

You can (and have aparently) resolved 23.976 audio to a 29.97 sequence with the right tools. as long as no Drop Frame has been introduced anywhere in the process you'll notice that the seconds roll over at the same time. The XDCAM HD decks can demonstrate this when you can get both 30 and 24 TC rates up on the LCD display for 23.976 footage with pulldown. You of course cannot mix 23.976 and 30 or 29.97 and 24 without some rate changes.

How many AES inputs does the Kona card have? Will it be enough for your HDD recorder.

It would be nice if that Time:tools package put the audio TC into one of the AuxTC slots instead of wiping the LTC.

ak

ak
Sleeplings, AWAKE!

<<<Now, can anyone tell me why the Quicktime Conv-to-Mpeg-2 function disappeared? Grr >>>

That would be the Curse Of Compressor. Someone at Apple fixed the minor shortcomings of MPEG2 Export by producing a bloated software package that produced the same or worse work and took an insanely long time to do it with a steep learning curve.

Then, Marketing compounded the error by deleting all the original tools.

Koz

If Time:Tools is changing the video-TC, what to do if you have to deliver an EDL.

Thanks

stefan

Stefan,

skampie Wrote:
-------------------------------------------------------
> If Time:Tools is changing the video-TC, what to do
> if you have to deliver an EDL.
>
It currently won't work if you did not edit the video TC for all clips by hand.

The classic method Frank described above at least will give you the option to have AV synched during capture and KEEP the video TC intact (I never did it with a 23.976 to 29.97 conversion - the audio square-waves might change somehow-non-square-waves), but it won't allow to keep the audio TC since the audio HDR runs in slave mode, also metadata can't be transported thru the AES connection and the amount of channels which can be connected to capture cards is somehow limited. But the audio TC/metadata stuff isn't needed if you just work within FCP and don't have need to interchange with DAWs other then with self-contained OMFs.

Conclusion:
As Koz said: a clapper is a good and cheap investment, TC clappers are a better one. (And having a color table on the clapper is helpful as well.)
Having a camera which allows a TC setup/synch is helpful also - just calculate the money and time you spent matching/synching clips.
Using separate A and V devices while recording can be a problem with working in NTSC (with FCP) as FCP never supported DF TC/playrate for audio, so some helper apps maybe worth the googling time and/or the money as well.
The final and most important thing is to test the setup in advance.

Kind Regards
Andreas

P.S.
Who is interested in an audio C2 TC to AUXTC conversion, maybe I can make some money on that ;-)

What are the quality tolerances for TC audio? Our current audio ops send a scratch mix wirelessly to camera and while we have never had to rely on it, when ever I have listened to it it's crap quality. Barely useable for sync. It's pretty low on a long list of things we need to fix.

Would audioTC HAVE to be a hard patch?

Quote

Who is interested in an audio C2 TC to AUXTC conversion, maybe I can make some money on that ;-)

Might be too niche to make alot of money on.

For example:
Set of editors.
Subset of those people using FCP.
Subset of those people shooting two system.
Subset of those people using Ch2 TC Audio.
Subset of those people with producers willing to give up one visit to the "shiatsu therapist" for some software.

Just a theory. Maybe you are living like Karl Lagerfeld on your BWF2XML income. I hope you are but I doubt it. I am the worst person I know to take financial advice from though.

ak ak a dak

ak
Sleeplings, AWAKE!

Quote

What are the quality tolerances for TC audio? Our current audio ops send a scratch mix wirelessly to camera and while we have never had to rely on it, when ever I have listened to it it's crap quality. Barely useable for sync. It's pretty low on a long list of things we need to fix.

I have no idea currently what will happen to an audio TC when transfering from 23.976 to 29.97.
If audio will not be affected by the pull procedure (which should be standard procedure), you will only have a valid timecode on those frames which will match the original.
Maybe you should do it the other way round send a wireless video to the audio device and connect Audio, Audio TC (or Serial RS-232) and Video to an nNovia A2D, I know some sound recordists who did follow my advice to do it that way and they seem to be happy with it - even if the setup needs to include some additional hardware.

Quote

Would audioTC HAVE to be a hard patch?

Shouldn't be that difficult, but it'll take some time though.

Quote

? Subset of those people with producers willing to give up one visit to the "shiatsu therapist" for some software.

That sounds like a good marketing idea, I will put that on my website like: Instead of paying one visit to the "shiatsu therapist" buy my software and you'll be more relaxed - and you don't need the next visit at the "shiatsu therapist" as well. ;-)

Quote

?Maybe you are living like Karl Lagerfeld on your BWF2XML income?

That would be cool to be the Karl Lagerfeld of FCP (audio) utilities. There is a world famous painter living here in town and he makes money (a real lot) on putting his signature beside the signature of the person who created the artwork - meaning this artwork has been reviewed and found as good by me. So that's what I would love to see for me - some like Apple pays me 2 bucks on every FCP licence, because I will state BWF support now works (which currently is not the case).

Quote

I am the worst person I know to take financial advice from though.

Do you know a good one?

Back to serious.
I did sent out a mailing to users some time ago about what people would like to see when dealing with BWF. There hadn't been such a big response.
Maybe "non-users" have some ideas - like making BWAV 29.97 DF audio work with FCP

Regards
Andreas