DV blanking issues?
I read this statement on http://www.screeneditors.com website and it made me worried about onlining a couple of TVC's shot on Dvcam. the online will be 8bit uncompressed thru AJA component input. Do I have to blow the picts up? I didn't think I had to, but this has me worried. It's for PAL as well. Output to Digibeta.
"MiniDV has a very wide blanking that is outside broadcast specs. Fixing these kinds of errors lowers the quality of the pictures. On DV and DVCPro programs the quality drops away quite fast as the picture is blown up to reach the edges. This not only reduces the markets for your program but is also going to limit the life of your program."
Thanks for any clarification.
docofilms PAL 25fps
Bad horizontal blanking? Well, that's a new one on me! I can't get very enthused about DV, but for other reasons: the main one being that it doesn't hold up to re-encoding very well. There is a BIG generational loss... and wasn't digital video supposed to eliminate that problem? Guess not...
I see that screeneditors.com is in Australia, and I also see that you're working in PAL, so I bet you live Down Under. I don't know how Australian TV is organized, but I'll give you my two cents' worth from the perspective of an American broadcaster: it depends on the size of the market.
If I were to shoot and edit a spot on DV with expectations of getting it played coast-to-coast on the NBC television network, I'd be smoking the fibers in the drapes. However, if I were to send the same spot to a local station -- let's say it's the mythical KCUD-TV in Cutbank, Montana -- I don't think I'd have a problem. The reason: small-market stations don't have the time, resources or manpower to scrutinize every single thing that comes through the door, especially for something as trivial as horizontal blanking which might be quite within specs. Besides that, it's a spot for a paying customer.
So: If you're making something for air on the Austraian Broadcasting Company, you'll probably have a problem. If it's going to air only on the locally-owned ABC affiliate in Alice Springs, the chances are pretty good you'll be okay. Besides, this screeneditors.com web site is for guys in film, and if you're going to all the expense of shooting on film, you're probably not going to transfer your precious footage to a prosumer tape format.
The best advice: ask! Find out if they'll accept stuff shot on DV and mastered on digibeta. If they don't, go rent a digibeta camera and spare yourself the grief.
I think they are all just spreading misinformation?
What I believe they're talking about:
Fixing these kinds of errors lowers the quality of the pictures. On DV and DVCPro programs the quality drops away quite fast as the picture is blown up to reach the edges. This not only reduces the markets for your program but is also going to limit the life of your program.
DV frame size is 720X480.
Many editing systems that follow the ITU 601 spec use a frame size of 720X486. (They can probably do 720X480 too.)
They may be trying to scale the video up so the 720X480 picture fills the 720X486 dimensions. There's really no reason to do this, although some editing systems may do this by default. What you should be doing is moving the image down 2 or 4 pixels (but not 3), with no scaling/resizing.
Or they can be talking about something else entirely... there's such thing as a blanking interval, but shooting DV doesn't affect that.
Read about analog video at http://www.danalee.ca/ttt/analog_video.htm
*Some of the little details may be wrong, but I don't know of any better sources on the internet that are free, as easy to read, and as correct. Poynton's books and SMPTE publications would be more technically-accurate sources.
**Those figures are for NTSC, not PAL.
2- DV generation loss is pretty low in a properly configured editing system using a decent DV codec. see http://codecs.onerivermedia.com/articles/adam_dv.htm
You can however use lossless codecs while editing, so generation loss isn't really an issue.
Also, firewire/ieee transfer is lossless (while sdi is not). SDI/digibeta dubbing to SDI/digibeta over and over will actually pickup generation loss, while firewire/DV to firewire/DV won't (SDI/DV to SDI/DV will however).
3- Adam Wilt (video engineer who writes for dv mag and has the excellent DV FAQ) rates DV as 9.0/10 for picture quality, and betaSP 8.9/10 as far as the FORMATS go.
In practice however, you'll probably get better results with betaSP for non-technical reasons. The budgets are generally bigger, the cameras better (tens of thousands versus thousands), there's a greater knowledge/experience base for betaSP, etc.
And of course, shooting digibeta will yield better technical quality. And 35mm is likely better yet.
4- the online will be 8bit uncompressed thru AJA component input.
You might get slightly better quality if you maintain a digital workflow by capturing through SDI. Going to analog will add a generation.
If you want to be assinine, you kind of incur a generation if you capture through SDI. see http://www.lafcpug.org/feature_capture_card.html
Disclaimer: Take what I say with a grain of salt! I'm still learning about this stuff and I've been wrong many times. I have tried to point out references to sources that are much more well-informed.
Disclaimer 2: Even experts are sometimes wrong (although this is a lot rarer).
If you read that article, the author points out that many textbooks on digital video are kind of wrong.