Color correction for hdv

Jorrit Huisman · May 20, 2006

Hello,

I'm about to shoot my graduation film next week. Now i'm planning to shoot on a XLh1.

But i have some questions about the color correction in post production.

I've been searching the net for hours about this topic and found quite some information, but i hope someone can answer me in a more specific way, hopefully based on his/her own experiences.

-I understand that HDV is a compressed video stream but does that really make color correction difficult because of that compression?

-Does minidv (pal) leave more room for color correction than HDV does?

-If I decide to export my HDV recordings as mini dv, does that result in a double compressed video stream? (both HDV and mini dv compression)

Or do i have a clean minidv stream with compression per frame instead of a GOP?

thanks!

Phil Rhodes · May 20, 2006

Hi,

> -I understand that HDV is a compressed video stream

Very much so, yes.

> but does that really make color correction difficult because of that compression?

Depends on too many factors to give a definitive answer, but in general yes. Highly compressed images can look OK straight off the camera - indeed HDV looks startlingly good for the price - but once you start to process them, particularly if you wish to increase contrast or colour saturation, you'll find that it begins to make compression artifacts visible. There's on real way to elucidate this other than to encourage you to run tests - you'll see what I mean.

-Does minidv (pal) leave more room for color correction than HDV does?

I wouldn't have said so, no. HDV is considerably more compressed than miniDV, but then it's using a cleverer codec and it's higher resolution to start with. This means that the artifacts tend to be smaller, which is less objectionable even viewing it as HD, and especially if you are delivering it as SD.

> -If I decide to export my HDV recordings as mini dv, does that result in a double compressed video stream?

It means it's been through two codecs and been degraded twice, yes. This is very bad.

However if your intention is to deliver an SD stream it would probably be best to post the show in HD (even HDV) and then have it downscaled to a format like Digital Betacam - with any luck many of the compression artifacts will be averaged out by the downscaler. I've seen a couple of shows done like this and they looked pretty respectable.

Phil

May 21, 2006

You could always digitize the raw .m2t footage to the computer then make an uncompressed HD copy of that in another sequence for colour correction and heavy manipulation shots. You can then finalize your film with the above specs then make your standard definition master as a downconvert from the original HD master. This way you have both.

The downside of the above route is that it is a rather resource heavy way to work. You will need a FAST computer to work on and a LOT of memory and hard disk space - and even then you probably won't get true real time performance.

One way to speed things up considerably is to make standard definition copies of the original .m2t files, and edit/finalize your work with them. If you ever want to make a HD master, you can simply replace the standard footage using the original .m2t files (so make a backup of them ).

Or you could take the intermediate route between the two above options by transferring to an intermediate codec like CineForm or DVC Pro HD or something similar and edit and correct in that codec.