[John-michael Trojan]

To simply answer your question; yes – film has more latitude than a conventional video master and so a DI master will as well… Typical CPD log range is 64-685 where 685 is 90% white on a grey card. So there is latitude above 685 for super-whites, etc that film can handle. Spectral highlights can go much higher at times, but one needs to be wary of going too high as whites above 940 or so can overload the laser on the film-out.

Generally speaking a projection master (4:4:4 or 4:2:2) that simply has the viewing LUT baked in, can have levels from 0 – 1024 and needs an additional video color pass to create a broadcast legal master honed for a television rather than projection or film-out.

For more information on Cineon Print Density, you can see the official PDF with lots of graphs of curves!

https://www.dotcsw.com/doc/cineon1.pdf

[John-michael Trojan]

FYI – made it slightly better by switching the field dominance to upper from none. I think this is as good as it gets but still less than satisfactory in my book.

Anyone care to share a good reason to shoot 30p?

[John-michael Trojan]

done that – same ugly SD footage via the aja – just not as bad…

[John-michael Trojan]

yes – it is getting rather ridiculous.

I was mistaken in my original post. The footage is not 29.97i, it is 30p running at 29.97.

I really just want to get it into a 720×486 29.97 NTSC sequence – or at least master out that way using the aja kona 3 to add pull-down.

Since I can put the 30p footage into a interlaced timeline nicely I thought I would use a high-end coverter to get the footage to 23.98 (which looks pretty great) then add the pull down back in during the down-convert (still looks ugly).

I think I may just ask the client to check with post before he shoots P2 again for the correct settings…

[John-michael Trojan]

not sure what camera it was… wish I did as that would help the google searches! All I can say is the movement goes to jello…

I’m trying to go to uncompressed 8-bit.

[John-michael Trojan]

OK – I think I’ve got this straight.

Since it is really recording 59.94 at all times what I have is two matching fields creating a 1080 30p file that runs native at 29.97. (When brought into FCP it says its 1080p30 codec with a frame rate of 29.97).

Then – since it is 30p adding pull-down doesn’t work since it is already at 30 – so some other odd math is being applied and most likely losing frames somewhere along the line.

What has provided limited success is using a separate piece of software to convert the file to 23.98pN then adding pull-down. The footage ends up rather smeary, but doesn’t display the same ugly large interlaced looking artifacts.

[John-michael Trojan]

I also just noticed there is some pretty bad rolling shutter artifacting going on – has anyone seen this on the 30 p setting before? I have a man’s face distorting rather severely.

[John-michael Trojan]

As an update – the footage is actually 1080 30p (called pN I think?)

Also – I have looked at the aja down convert and it displays the same problem.

thanks!

[John-michael Trojan]

thanks – thought that and will do a test shortly using my Aja Kona 3 card. Still, this is unusually bad. I could send you a still to look at if that helps.