RED testing by David Stump, ASC. - Files, Quotes, Mirrors, etc...

[mikkowilson]

Last Tuesday David Stump ASC was over at the RED facility and did a series of test shots last week with the RED prototype; "Frankie".
Tests included latuitude testing and a greenscreen shot.

As David posts results over on his site: [URL="http://www.cinematography.net ,we"]http://www.cinematography.net ,we[/URL] here at DVXuser will be passing on his reports and files through mirror servers to keep the load down on his server.
Relevant quotes will be updated in the 2nd post of this thread.

Considering that a single 4K frame .tif is 74MB, that's a lot of data!


Tip for examining image sets: Don't download every image, just the ones you want to look at. For example each extreme and then a few in between gives a lot of info!

Sites for downloads:
David's exposure test results: http://www.cinematography.net/red-exposure.html
David's green-screen test results: http://www.cinematography.net/Red/red-greenscreen.html
Matrix & Redcode Greenscreen images: http://www.cinematography.net/Red/comp-matrix.html
My mirror server, hosting copies of all the images as fast as I can: http://red.mikkowilson.com
A mirror thanks to Ramjet1979:http://www.red.ramjetfilms.com/


Do you want to mirror?
If anyone else has any mirrors that they are setting up, please PM me with info. Also let me know if you server is going to go down and I can remove the link from here before you run out of bandwidth! These things eat up a lot of bandwidth fast, so keep that in mind. These releases are not for feable servers, but any help with load distribution is appriciated.

Do you have your own files to add to the mix?
As these files come out, many people like to try post processing them. Color correction, scaling, etc, are all very common.
If you have files that you have changed, for now you will need to host them yourself somewhere. (Your 10mb of online storage included with your DVXuser account in the top toolbar is a good place to start.) I may start accepting those files for temporary hosting too if there is demand.
PLEASE POST ALL RE-PROCESSED IMAGES AND DISCUSSION IN THE "RED IMAGES" SECTION OF THE FORUM: http://www.dvxuser.com/V6/forumdisplay.php?f=99

- Mikko

 

[mikkowilson]

Assorted comments and quotes from David Stump ASC regarding his RED testing...

Assorted comments and quotes from David Stump ASC regarding his RED testing...

Comments and details on the exposure tests:

David Stump ASC RED Color Chart Over Under Frames
Details of 11/07/2006 Shoot

Subjects were DSC Chroma Du Monde, DSC Focus Star, MacBeth Color Checker
Lens was a Cooke S-4 65mm S/N 65-0937. Focus Distance was 9 feet. Charts were cross front lit with 2K tungsten from 45 deg. both sides. Average color temp was 3100 deg. Kelvin. I conducted a preliminary exposure test to determine how to rate the camera’s ASA which I determined to assume at ASA 200. Ther results of that test can be observed in the series of greenscreen wedges seen elsewhere in these materials.) Charts were lit for f 16 @ ASA 200, 180 shutter at 24 fps, 1/48th sec. exp. time. Exposure was metered with Spectra Cine Pro IV meter s/n 11752 calibrated 5/2006 by Nassar Zaidi of Spectra Cine. Color temp was metered with Minolta Color meter II s/n105677 calibrated 5/2006 by Nassar Zaidi of Spectra Cine. Deep underexposures above f22 were achieved through the use of a full set of new Tiffen 4x5 neutral density glass filters, no attempt was made to characterize any color shift created by the use of these filters. Camera Gain setting was 5447 (standard neutral setting) and at the camera’s native 5000K color balance. The native 12 bit bayer images were debayered and converted to uncompressed RGB 16bit tiffs in a prototype RedCine application developed by Graeme Nattress. The Redcine application added 2.2 gamma, black offset correction, some small white balance correction, and added a small contrast curve. No dead pixel correction was applied. No color correction has been attempted yet, I will color correct the images in the next week and post some examples then.

Commenting on artifacts in some of the exposure test images:

Congratulations! You won the "find the artifact Dave Stump caused by running over a cable" contest! I had the choice of restarting the tests or carrying on. Check the other series and that artifact is not there....

And:

I must also state that in the course of dollying "Frankie" from one setup to another I ran over a cable which caused a sync issue in the signal for part of the exercise. We restarted the camera and the issue went away, but I am using some frames from the series of grayscale and color chart brackets we did before the reboot.

http://www.dvxuser.com/V6/showthread.php?p=767847


Commenting on image softness:

As I said in the posts, the sensor is not square to the lens in the prototype, so back focal distance is not good side to side, it would be wise to eval intrinsic camera focus quality from some of the material done at deeper stops, in order to take the back focus issues out of the equation.

Comments and details on the green-screen shots:

Subjects were DSC Chroma Du Monde, DSC Focus Star, MacBeth Color Checker, and David Stump ASC “One Eyed Jack” HD exposure ball. Lens was a Cooke S-4 65mm S/N 65-0937. Focus distance to subject was 15 feet, greenscreen distance was 28 feet. Greenscreen was lit with 6 x 4 globe Kinoflo units at 45 degrees to screen at approximately 8 foot distance, using chroma green globes, no gels. The greenscreen used was a Composite Components Spandex Green 12’ x 12’ lit to EV 9.3 @ ASA 200.
Subjects were cross front lit with chimera 2K blond tungsten from 45 deg. left side f4.8. Fill from 1K tungsten on camera right to f 3.5. Cross backlight from 600w tungsten tweenie on camera left to f 2.8 and on camera right from tungsten 1K to f 3.5. Average color temp was 3100 deg. Kelvin, no color gels on any of the lamps.
I conducted a preliminary exposure test to determine how to rate the camera’s ASA which I determined from the test to assume at ASA 200. Greenscreen wedge test was evaluated in Adobe CS2 Photoshop which indicated that a green exposure of 58% was achieved at f4.8 while holding red level in the screen at about 8% and blue level in the screen at 26% (see attached chart) which I determined to be the optimum level of exposure and color difference for pulling mattes from the material.
Subjects were lit for f 4.8 @ ASA 200, 180 shutter at 24 fps, 1/48th sec. exp. time. Foreground exposure was metered with Spectra Cine Pro IV meter s/n 11752 calibrated 5/2006 by Nassar Zaidi of Spectra Cine. Foreground color temperature was metered with Minolta Color meter II s/n105677 calibrated 5/2006 by Nassar Zaidi of Spectra Cine. Greenscreen was spot read with Pentax Digital Spot meter s/n 103772 also calibrated 5/2006 by Nassar Zaidi of Spectra Cine.
Camera Gain setting was 5447 (standard neutral setting) and camera was left at it’s native 5000 Kelvin color balance. The native 12 bit bayer images were debayered and converted to uncompressed RGB 16bit tiffs in a prototype RedCine application developed by Graeme Nattress. The Redcine application added 2.2 gamma, black offset correction, some small white balance correction, and added a small contrast curve. No dead pixel correction was applied.

Other Misc Quotes:

"The RED Mysterium sensor has finally pushed digital acquisition past the timeless 35mm film barrier in resolution and clean color fidelity". David Stump, ASC

I will post frames from all of the work, but I have just completed my
first series of tests on the Red camera. Shooting went very well, the
prototype was extraordinarily well behaved. I am very pleased with the
keys that I am pulling from the material, and I can't wait to see a
green screen comp on the big screen. The image from the sensor has a
pleasing look to it, not noisy, not grainy, not sharpened by any
sharpening kernel, and it color corrects easily, yielding very pleasing
and true colors.

More comments from David will be posted as they become available and we can process them. Stay tuned...


- Mikko

 

[mikkowilson]

Update: Greenscreen info & link added & mirrors are updating.

- Mikko