|November 10th, 2013, 07:49 AM||#1|
Join Date: Dec 2002
Location: Augusta Georgia
Odyssey7Q with Canon C500 - 4K Raw
I just spend two days at an event in Atlanta, Georgia with a production model of the Odyssey7Q, using our initial firmware release.
I was at Showcase, for their in-store event.
Sony, Canon, Panasonic, Tiffen with Steadicam, and others were there.
The camera that we had available for this event was a Canon C500. I would have also connected the
Odyssey7Q to an FS700, but one was not available for this event.
Our team has devoted countless hours, in development, to ensuring that the image on the Odyssey7Q is as perfect as we can make it. We purchased a $15,000 color calibrator to ensure a very high degree of accuracy.
We have built in, three separate color calibrations, one for High Brightness, one for normal, and one for low.
The low brightness one is for when working, on set, in a dark environment, where the light from the monitor would interfere with the filming.
In each setting, the colors are accurate, just the brightness of the image changes.
At this event, with the Odyssey7Q connected to the C500, I tested our color calibrations and the Odyssey7Q's overall ability to display colors accurately.
I setup the Odyssey7Q for Canon 4K Raw, and set up the C500 for 4K Raw at 24.00 fps.
I initially pointed the camera at some colorful grip items on a pegboard, adjacent to a colorful, dark red wall.
The colors were gorgeous!
There were grip clamps with bright orange handles, and these displayed perfectly on the Odyssey7Q.
The red wall displayed accurately. A bright, lime green roll of gaffer tape also looked perfect.
There was a black bag, with the company's name printed in bright white, and a blue logo. All of these were also perfect, including the subtle shades of black of the case created by light shining on the black case.
In another area, there were rolls of very colorful backdrop papers. All of the colors displayed accurately.
I found a very colorful package of 1" wide tape. This included a bright pink, an orange, a lime green and another color. Again the colors were perfect.
I found a laminated page with color chips of all of the colors available for the background rolls.
We shot this, and had independent people check our Odyssey7Q's color reproduction against the actual clock chips about 4 feet in front of the lens. Everyone agreed that all colors displayed accurately.
The accurate color rendition is attributable to the Canon C500 capturing the colors properly, including the very difficult colors, and the Odyssey7Q's ability to accurately display these colors.
Here are the technical details:
We were under a mix of lights, some overhead florescent, some 5600K softlights, plus other various sources, as this was a regular video store environment.
We set the camera to 5500K. We also used 5600K for some tests.
I used our focus tools to focus on the subject, and used the Odyssey7Q's waveform monitor to set the exposure levels.
I set the general peaks to 85% (IRE), which worked very well.
If I had a specific hotspot, I did allow it to go over 85%, but I did not allow many peaks to go over 85%. This seemed to ensure that the colors were accurate.
We also had an unusual greenscreen cloth (with a velour texture). We had trouble with this one cloth, and spent a great deal of time trying to determine the cause of the problem.
When we used other, better, greenscreen samples, the color was perfect. But the original greenscreen cloth had too much blue.
This was obvious using our color histogram, as well as the RGB Parade Waveform, or using the Red, Green, and Blue individual Waveform displays.
We used a different greenscreen backdrop, which solved the problem.
I selected the Green waveform display, in full screen mode, which allowed me to carefully analyze the uniformity of the greenscreen (checking each spot, from left to right across the image). I repositioned the lighting to get a very uniform level on the Green Waveform Monitor.
I also used our Green histogram, which cleared showed the variations in the greenscreen image.
Our RGB Histrogram, which shows three separate Histrograms,, one Red, one Green, and one Blue simultaneously, was another tool that I used.
Our initial firmware release does not have our Color Vectorscope enabled. We omitted it since we did not finish the graticle (color boxes) in time for the initial release. We will add this feature, for free, as soon as possible.
We also used the Multi-Stream Monitoring feature to set the exposure of two cameras, using a quick, easy and highly accurate procedure.
I enabled the Multi-Steam Monitoring, then took an HD feed from both cameras.
This enables Camera 1 to be displayed in the upper left quadrant of the Odyssey7Q, with Camera 2 being displayed on the upper right quadrant. This leaves Quadrants 3 and 4, the lower ones, empty.
Then I enabled our Waveform Monitor in Luma Mode, full screen.
Thus this Waveform Monitor is below the two upper images, with Camera 1's waveform, being directly under the Camera 1 image, and the same for Camera 2.
Then we just adjusted Camera 1 for the proper exposure, then adjusted Camera 2 for the same, using the dual camera waveform, on one screen, to ensure accuracy.
This is a great technique that takes advantage of all of the advanced features of the Odyssey7Q, enabling two cameras to be setup and adjusted perfectly, in a minimum amount of time and effort.
I hope someone who was at this event will post their impressions.
|Thread Tools||Search this Thread|