Is that your Final Answer?
 

Ok,

No more guessing, no personal opinions, someone with experience has to know the answer.

We are shooting a show in HDV. Our delivery is standard def.

Would we get better quality by:

A) Downconverting during capture, editing standard def, and exporting

or

B) Capturing and editing in HDV, and downconverting during the export process.

I am looking for a definitive answer.

Thanks,

-Andy

I think that you'll have to tell us what camera, what kind of computer and what software you have before anyone could attempt to answer this question....

Thanks for your reply,

3 Sony Z1U's and a Sony HC-3.

Running Final Cut Studio 5.1 on a G5 Tower and a new Mac Pro tower, both computers loaded.

I guess you are shooting in HDV for futureproofing?
I would down convert in camera while digitising (that's got an 's' in it ;-)) The Z1 downconverts very well.

Editing in SD will be much snappier, you'll get more RT effects/mixes etc in your edit system. Less rendering etc.

You can reconform the HDV version at a later date.
But there are issues with HDV reconforming. Make sure your recordings are clean, with solid timecode, usual thing really.

The HDV GOP can screw up redigitising, if the FCP needs to capture anywhere near or over camera TC jumps.

Also my personal preference for multicamera editing is lay the cameras onto layers in the timeline and delete down what you don't want. I'll put the wide on layer one, the two main CU cams on 2 and 3, the two handheld on 4 and 5.

In the last two years I've done about 5 5xZ1 concert shoots. We always get a full multitrack recording made at gigs, this is given a rough mix and put to DVD as a 90min+ AIFF.

Drop this onto the timeline, and use it against the cameras sound, to check they are all in sync. They various picture layers often need to be cut and moved around a frame or two during the night. (Camera flashes from the audience help to sync up precisely)

Once the cut is done the sound guy uses the pictures as a reference for tweaking the room atmos and applause etc. Then gives me another full length AIFF, which reconnects to the original file.

After this I'll make edits to the content of the show, dropping or reordering songs etc. Then a final sound tweak if there are areas where my edits messed up the sound beyond what can be repaired in FCP (usually one or two places where applause needs to be lengthened or remixed)

I knocked up a HTML page for a particular gig last year. I've dug it out and uploaded it to my site here:

http://www.darkhorsegraphics.co.uk/P.../Paradise.html

The artist is Joe Longthorne (.com) I don't work with them anymore, or I would still have video clips too.

Short answer: you get better quality shooting and cutting HDV, and downconverting to format better than DV at the end. Also providing you downconvert with care.

I've done both a lot of times, the above is better.

Thanks guys for your repsonses.

I am going to do some tests this week to see which would work best for our application.

Cheers,

-Andy

One of the advantages of editing HDV material in a SD project is the ability to zoom back out to 100% which appears to the user as a zoom in, without loss of quality.

Now, this method is certainly not available if you convert in camera, or if you edit HDV in an HDV project, but I find it quite handy for certain purposes.

I've done it both ways.
My experience is that shooting HDV, editing in Cineform CFHD codec, transcoding the edited 1080i CFHD directly to SD DVD .m2v using 3rd party encoder results in somewhat better images.
On the other hand, when I have a larger project (10+ hours of raw footage, 40-60 min. final cut), the system hassels, storage requirements, render times, etc. outweigh the quality advantage for me.
I have been very pleased with the in camera downconvert from the Z1 (actually, I use the Sony HDV M-15 deck), edit in 16:9 DV, and use third party transcode to .m2v. The result is definitely better than 1/3" chip standard DV acquisition, and almost as good as the HDV workflow
My producing partner, who shoots, but is not a techie, cannot see the difference b/t the two workflows on the final DVD. I think that might be typical for the average viewer.
So, for me it comes down to the practical considerations re my system, and how much time I want to put into the project.

Thanks Robert,

We are going to have approximately 40-50 hours of footage to deal with each episode, with a 46 minute final show time.

If it's not a definite "You have to do it this way!" it may not be worth our hastles. We too will be using capture decks, so the downconversion will be done before the capture.

Thanks again for the replies.

-Andy Peterson

Andy
Your project is definitely big, and ongoing.
It would probably be worth your while to put together a 10 min test segment that includes a variety of lighting, effects, color correction, etc. that will be used in the production. Using the same footage and editing, try both workflows all the way out to DVD and see what you think. It's fun to put both workflows on the same DVD and show it to a variety of viewers. If hardly anyone can see the difference it's food for thought.
A difference without a distinction???

Yeah, my advice is providing cutting an HDV project is feasible for you. On FCP (my work), it's workable enough to do as I mention.

The criteria in the end is what device is doing the downconversion, and to what format. The Sony cameras do a pretty good downconversion, but they don't add any edge enhancement backinto the signal so sometimes things can wind up looking soft.

Also, downconverting to DV as the cameras do is not ideal. You can get an end result MUCH better than the HDV format of 4:2:0 if you downconvert to some sort of 8 bit uncompressed, or DV50.

I have no personal experience, but the next frontier is surely to convert the raw M2T, or even edited Cineform CFHD into one of the high end HD formats, and use commercial grade encoding to get to DVD. From what I've heard, that does make a quantum difference.
On the other hand, if you are going to end up in 480 DVD, I wonder if DV50 wouldn't look about as good?

If he is future proofing with HD and doing a down convert to SD, or even if he doesn't, does the camera transmit the actual time code into the editing suite?

And I've heard that the HDV codec actually re-generates the time code in editing, sometimes. Was this ever a huge issue that you guys have noticed?