I am presently preparing a "Introductory Night on the Digital Intermediate" for the Australian Cinematographer Society along with fellow member Ben Allan ACS.
We want to discuss the "Politics " and changing relationship a DOP has on a film with a DI. Would love to hear from DP's with direct experience of DI on a commercial project. To what extent will DP's maintain control of the look of a film? Why do a project using DI?
We also want to cover the process itself. We are not going to do a "shoot out" or a comparison of approaches but we are interested in discussing the KEY issues and processes that a DOP should now understand.
We are also interested in any reference material people can point us to. Is there a list of Feature projects that have had a DI finish ? Any morsel of information or experience would be valuable. I hope I have asked not too many questions but we believe feature production is facing one of the largest single revolution in 100 years of making movies.
It is a very good overall picture of the DI Process. It uses Quantel equipment to illustrate the process but of course many other manufacturers equipment is being used.
Regards, Bill Hogan
Summary : The Quantel guide to Digital Intermediate
Tom Gleeson wrote :
>Is there a list of Feature projects that have had a DI finish ?
There is a table of DI films on the Telecine Internet Group's TWiki web site.
Scroll down to the Digital Intermediate Process for Film Releases heading
The origins of the table are :
"...Wade Odlum, Colourist at Cutting Edge and myself (Warren Eagles) set about making a basic DI Database. We used American Cinematographer, IMDB, Kodak Fuji and other publications as reference. Rob and Dave (of the TIG) then worked hard to get the document into TWiki form. It was only ever meant to be a START..."
The TIG TWiki pages are editable (although the structure of a table is awkward to deal with.) I recommend that anyone with information to add place a text comment to the bottom of the page -- I will collect those comments and incorporate them into the next version of the table on the TIG TWiki pages.
Digital Intermediate is a challenge for cinematographers and will certainly change the way of our work.
To know more about DI and to see what is possible, / impossible, more easy / difficult doing a DI in comparison to colour grading the conventional way in a lab I did extensive testing with the support of leading postproduction facilities in Munich, Germany. DAS WERK MUNICH (scanning, digital grading, online, recording on Arri Laser), ARRI lab (negative development), BAVARIA lab (color grading and doing the 35mm prints)
Basically what we did was :
1) developing a 5 min. short film that contained all the typical problems and situations cinematographers and colourists have to face in real shooting situations and later in color grading. Like :
- high contrast ratios from 7 stops over to 4 stops under
- low contrast ratios at about 2-3 stops under
- mixed lighting: daylight, tungsten, candle light, green fluorescent lights mixed in different combinations
- cars interiors moving in bright sunlight, face against windows
- available light night shots etc.
2) We shot the film 5 times under exactly the same conditions with 5 cameras 35 mm, Super 16, HDCam, Digital Beta, Mini DV. (These are the formats we mainly use in Germany for TV and the screen). Thus the project got the name “5eyes“.
3) We made the complete postproduction process with all 5 formats for SDTV (Pal) and for the screen in 35 mm prints.
Main interest was :
A) to have a comparison of 35mm, HDCam, Super 16 in SDTV and 35 mm prints.
B) to compare conventional prints with prints from digital intermediate and prints from HDCam.
C) to get a show-reel that is able to demonstrate in an easy and visual way to producers and directors what and how much they can get from each format. (Because in Germany it is sort of a fashion to ask cinematographers and post-facilities : “Can’t we shoot Digital Beta or Mini DV and make it look like 35?“)
I admit the whole project was crazy and very, very German! Without great support form the industry it wouldn’t have been possible. But the results have been very interesting and worth the time, money and effort.
Regarding DI :
My experience having printed the same film from a 35mm negative on 35mm via lab and via digital internegative(C-reality, 2k Da Vinci, Arri Laser film recorder) is :
The difference is not as big as one would think regarding a normal color grading. Using power windows, secondary color correction and all the nice things digital colour correction can provide us with we thought it would be quite hard for the lab to achieve similar results.
Fact was the print from the lab looked as good as the print done via DI. In some scenes the print from the lab could solve problems even better. There was especially one scene in a subway station full of green fluorescent lights. I lit the faces of the actors with daylight kinoflos without adjusting the „correct“ light from the kinoflos to the green fluorescent lights by plusgreen as I would have done normally. I wanted to see what happens when you try to get the green out of the image in color correction. I thought the faces would go magenta because of the light of the kinoflos.
Actually we had quite problems in digital color correction (Da Vinci) with the faces going magenta when we took out the green. Regarding this scene we had the biggest problems in digital color correction with :
>1) Super 16, 2) Digital Beta, 3) HDCam.
35 mm was the easiest to handle apart HDCam. But we couldn’t get the touch of magenta out of the faces in the DI.
The surprise was that the 35mm print from the lab looked in this scene way better than the 35mm print done via DI. Almost no magenta in the faces while the green light was corrected to a nice almost white light.
So DI has its limits and in the lab is possible a lot more than we thought.
Clear advantages of DI are :
DI can get the full contrast ratio out of the negative. Prints from the lab have in this point clear limits because of the more narrow contrast ratio of the positive stock. Regarding contrast ratio lab prints still look a bit better than HD.
Biggest advantage of DI is the possibility to create „looks“ by changing certain colors of an image completely or manipulating the gamma curve.
Example : Oh brother where are...“ from the Coen Brothers : DoP Deakins made all the green trees and meadows look brown.
DI for 35mm negative is an interesting tool, but conventional color grading in the labs is still very powerful, and sometimes can give you even better results in a still more easy way.
On the other hand: Our short film shot with HDCam and recorded to 35 on the
same Arri Laser as the 35mm DI was in many scenes very close to the 35mm DI print. Only the 35 mm print directly from the neg looked way better. And recently I had a presentation in Hamburg where screened the 35mm prints and the HDCam tape and the HDCam projection looked even better an a mid sized art house theatre screen. The projector has been an 8.000 USD JVC-projector.
There are a lot of things to cope with in the near future: e-cinema, HDCam SR, the first digital camera from ARRI, the new Vision 2 stocks from Kodak. The art of cinematography will surely be interesting and very demanding regarding all this new developments.
Oh Ben Allen! I never succeeded to load your files on the Sony HDCam F-900. However, I used your graphs to show what is possible working on the menus. It would be interesting to see the real thing.
All the best!
German based Cinematographer
From Munich, Bavaria