Testing Dynamic Range of my5D: Need Help

Hello,

Just got a 5DMk2 and want to test to determine it's dynamic range as well as create a zone chart and have unexpected results. Here is my methodology.
1) Tripod mounted, fill frame with an 18% gray card, camera set for neutral picture style, manual focus and manual exposure using partial metering and also spot. First in open shade then in bright sunlight.
2) Custom white balanced on grey card, set meter to get a correct exposure (open shade f8 @ 1/30, ISO 100 - bright sun f16 @ 1/125 also iso100), took a series of 14 exposures (7 in each direction), changing shutter speed in 1-stop increments, first faster shutter speeds to reduce exposure towards zone 0. Next progrssively slower shutter speeds to increase exposure towards zone 9.
3) After loading in to Aperture, I used it's densitometer to read each frame. There were no color cast issues throughout the study as my RGB readings were identical to no more than 2 points difference.
4) Working towards zone 0, my results were as expected, ie. achieved nice steps of progressively darker gray to black with no detail and a reading of 0 across the RGB spectrum. So far so good
5) Working towards zone 9, here is where my results get kooky. My results showed a nice gray step change to white from zone 5 to zone 6 and appropriate increase in RGB readings, again equal, indicating an RGB readings of 100 across the board, so no color shift, but here is the kicker, the readings for each consecutive frame & zone also read 100 and there is no visible change in density from zone 6 on out and never reach a reading of 255 which I should.

I was beginning to think my camera, although new, may have a problem, so I borrowed a Nikon D300 and put it on a copy stand and while the lighting is different (tungsten) vs direct sunlight and open shade, everything else in the study remained the same and got the same problem where I reach readings of 100 RGB at zone 6 and progress no further and not reaching 255.

Any suggestions, ideas would be appreciated.

I have no experience with aperture, but the thing that comes to mind is that the RGB reading may be a percentile (0-100).

TheBurningCrown wrote in post #12162941
I have no experience with aperture, but the thing that comes to mind is that the RGB reading may be a percentile (0-100).

Thanks for your thought and reply, I wondered about that to, but a couple things. I did not mention it but I also did my readings using Mac's digital meter and achieved the same results. Secondly, I only see a visual difference in frames from zone 5 (18% gray) to zone 6 ( a lighter gray, almost white) and then it remains the same. I would think I would see a progression of gray to white and then to paper white, as I did going the other direction.

So you're saying that the photographs don't look...normal? Did you shoot in RAW or JPEG?

Would you be willing to post some samples?

To measure DR you have to take WB out of the equation. WB is the application of multipliers to the red and blue channels for two purposes; first to bring the channels to parity because the sensor array is most sensitive to green light and less so to red and blue, and second to compensate for the lighting. A typical WB for daylight might be R X 2.0 and B X 1.4. Suppose you overexposed your grey card by two stops so that the RAW capture was 155/238/183. After WB that would become 255/238/255, which would clearly, although falsely, seem to be clipped.

Any RAW converter will automatically apply some WB or another, so what can you do? One of two things; either disregard the red and blue channels or use a UniWB custom WB setting in the camera. UniWB is a WB in which the multipliers are 1.0, in other words, the RAW values are not changed, You can fool the camera into making a CWB like that by taking a shot with the lens cap on and then using it as the image from which the CWB is made. Be sure to use the "As Shot" setting in Aperture.

My own experience with the 5D2 is that it captures 3.67 stops above medium grey in native space or when converted to ProPhoto RGB. For DR below medium grey the floor is defined as the point at which noise obscures detail, a definition that is inexact and somewhat subjective, but most testers give the 5D2 around 7.5 stops below medium at ISO 100.

Thanks for the replies, I believe I have found the answer. It seems my initial hunch and Burning Crown's suggestion, that it may be a percentage may be correct. I found that paper base white reads at 100 in LR, A3 & Mac's digital meter, but in Photoshop it reads at 255. So I believe that the 100 represents 100% and was probably done for simplicity sake. So I now have the results I was seeking. By the way I did shoot this study in Raw, the only format I use.