Comments

farss wrote on 6/30/2006, 4:27 PM
Why downconvert the HDV footage in the first place?

Surely it would make more sense to upconvert the SP footage and bring that into a HDV project?

Given that your SP footage is most likely 4:3 if possible leave it as such unless it's truly intercut. You can still use 4:3 in a 16:9 frame by filling the frame with a letterboxed and blurred copy of the 4:3 footage, I've seen this done a lot for broadcast, piece of cake in Vegas.

Bob.
williamk wrote on 7/1/2006, 8:19 AM
Thanks for the reply Farss

That would be ideal but that brings up another problem. cost

We have about 9 hours of HDV footage use maybe 45-1hr
about 15 hours of beta Sp footage use30-45
and about 2-3 hours of DV footage use 10-20

But we cannot see the Beta Sp to decide what to use in the timeline until it is transfered to our harddrive (outsourced). To upconvert all that footage to HD would be too expensive as to transfering 1-1.5 hours after editing

We are a low budget production.

Can you downconvrt and then upconvert

Thanks

Spot|DSE wrote on 7/1/2006, 8:24 AM
You could use GearShift to create proxies of the HDV files, bring those into an SD timeline along with your Beta footage, then shift out the SD proxies for the HD files via GearShift, without affecting the Beta.
Then, I'd recommend keeping those Beta files as 4:3 and underlaying a cool graphic as Bob suggests, or other generated media. If you use black as your underlayment, I'd recommend staying w/in legal limits, but a slow moving graphic or at least some kind of graded background looks better, IMO.
farss wrote on 7/1/2006, 8:42 AM
Well yes one CAN downconvert and then upconvert. Except (big except), you're reducing everything to the lowest common denominator, i.e. standard definition.

Now if you're in PAL land starting with say 16:9 Digibetacam upconverting that to 1080 doesn't look too shabby, particularly if you do all the graphics in HD you just might fool the viewers.

I don't know if you're working in PAL or NTSC so it's a little difficult to know what the best advice.

However the ideal is to keep everything at the best possible quality, that means ingesting the SP as 4:2:2 and that means a lot of disk space. Then I'd use Vegas tp upscale it, same with the DV material, maybe add a little Unsharpen mask to the SP and leave the DV alone, even a little chroma blur might help the DV. If you cannot afford the copious amount of disk space for the SP at 4:2:2 then ingest it as DV. You can do this yourself using the Sony J30 deck however it lacks a few features of the real SP decks like dynamic tracking, so dodgy tapes can go real bad. You can use a regular SP deck with the SD Connect to get a half decent ingest from SP as DV25.

Given the large 'shooting' ratio I'd be inclined to look at doing all of this inhouse, a post house will charge a fair bit for 15 hours of SP and you'll need enough disk space as well. If you were to rent the gear you can just ingest what you need or even dump it all to DV or DVCAM and ingest what you need.

Anyway, what's your delivery format, HDCAM?

The ONE advantage of downconverting everything and the upconverting is it might be easier to match it all, cutting real HD with SD might be rather obvious. Particularly if you're going to convert the SD to 16:9.

Bob.
williamk wrote on 7/1/2006, 9:33 AM
Thanks for the answers

I shot HDV with the FX1, NTSC in Canada.

We are planning to do everything with the best quality and disk space is not a problem. It is just the quality if I downconvert and then upconvert that was worrying me. I think it would be easier to line everything up together as SD as well.

So basically I should Downcovert HDV-DV
Downcovert SP to DV 4:2:2 (outsourced to post house as we do not have Beta capabilities) ($300)
Play with masks and other to get best visual quality
Master and upconvert

Does this sound about right?

Thanks again

Williamk


fwtep wrote on 7/1/2006, 9:46 AM
> But we cannot see the Beta Sp to decide what to use in the
> timeline until it is transfered to our harddrive (outsourced).
> To upconvert all that footage to HD would be too expensive
> as to transfering 1-1.5 hours after editing

Capture and use the Beta SP just as it is, then upconvert only what's in the final edit.
Coursedesign wrote on 7/1/2006, 9:58 AM
Downcovert HDV-DV

That's a sad thing to do. When you go from 4:2:0 to 4:1:1 you end up with 4:1:0.

With a bit of luck, you can recover some of that with a chroma smoothing tool, but it won't be the same.

If you're gonna have any text at all in the picture, it will look sucky when upconverted.

There are various possibilities depending on how you want to see the SP footage, but IMHO all of them will look best with doing the edit and post in the final HD format or at least resolution (i.e. probably not HDV).

If it was me, I'd do it with one of the Cineform codecs.

Farss' suggestion to put the Beta SP footage in a window is good.

If the window is SD res (720x486) out of a total 1920x1080 frame, the SD footage will look just as sharp as the rest.

Even if you carefully uprezz the SP footage to nearly full frame (and Beta SP can look very respectable uprezzed, if captured 4:2:2), you'll come out ahead.

Just my $0.015.
farss wrote on 7/1/2006, 4:20 PM
NO!

For the very best result ASSUMING you are going out as true HD i.e. HDCAM and / or DVCPro HD then:

Get the SP captured to the BMD QT 4:2:2 codec, pretty well industry standard but you'll need plenty of disk space.

Using Vegas convert that to say the Cineform DI at 1080, leave it as 4:3 in a 16:9 frame.
Convert all the HDV to the same CF DI at 1080.

Do the same thing to the DV, using Vegas upscale to CF DI, if it's 16:9 great, if not leave it 4:3 as you did for the SP.

Edit etc, render out Sony YUV at 1080 or BMD QT codec at 1080. Send to post house to put to tape as HDCAM or DVCPro HD.

Do all graphics, text etc at HD res.

This will keep your external costs to a minimum.

You do need to decide what your output will be, and find a post house that can output to that format. Neither HDCAM or DVCPro HD decks are exactly cheap! HDCAM is the best but expensive but getting cheaper as it's now kind of superceded in the Sony lineup.

Needless to say given the amount of work your planning on doing, TEST EVERYTHING, TRUST NO ONE.
Make a 5 minutes segment using bits from all source, render it out and have a post house do a HD transfer, go there and look at it on a true HD monitor, by that I mean a full res broadcast monitor of at least 24". If they don't have one RUN, go find a post house that takes their work seriously.

Trust me on the imprtance of testing. The number of times I've been asked if I can help after the fact is truly scary and I'm talking productions that have blown very serious money with paid talent and months of location shooting. Having it 'in the can' is useless if nothing can open 'the can'.


Bob.