Re: DataColor SpyderCube
Re: DataColor SpyderCube
- Subject: Re: DataColor SpyderCube
- From: Robin Myers <email@hidden>
- Date: Thu, 19 Mar 2009 07:55:26 -0700
Hello Andrew,
My tests with fine art reproduction and using neutral balanced raw
capture show that the N8 patch you recommend on the ColorChecker is
not as good for neutral balancing as the N6.5 patch next to it.
Titanium dioxide, the white colorant used for the White patch, has a
falloff in reflectance in the deep blue to violet portion of the
spectrum which results in the camera "perceiving" the patch as
slightly yellow. The N8 patch next to the White patch still has too
much influence from the titanium dioxide and also exhibits a slight
yellow cast. The titanium dioxide influence is low enough with the
N6.5 patch as to be effectively imperceptable and the reflectance is
high enough to give a good signal to noise ratio. As you noted,
depending on the exposure, the N5 patch can be too low in reflectance,
resulting in a poor signal to noise ratio, especially if the
photographer is utilizing exposure compensation to correct for the
over exposure common with automatic exposure systems in most digital
cameras.
Robin Myers
On Mar 19, 2009, at 6:55 AM, Andrew Rodney wrote:
On Mar 18, 2009, at 9:29 PM, Cdtobie wrote:
Well, there is a user's guide on the website that you apparently
missed, which includes a spectral graph that shows the Cube to be
more spectrally neutral than the gray Munsel chips in a
ColorChecker, or a Kodak gray card, or even a WhiBal card, though
the WhiBal comes in second.
I'd prefer to measure it myself as I've done with all other
products. The BableColor Tile measured nearly identically on two
samples using an EyeOne as stated on their site. Maybe yours does
too. I'll believe it when I see it.
You are welcome to white balance on the white face if you prefer,
but it's certainly more foolproof to use the gray, since it avoids
spectral clipping. When unclipped the two produce very similar
results.
For Raw capture, the ideal tone is something close to the 2nd white
(#20) of the Macbeth color checker, certainly with the Adobe Raw
engine. You don't want that so called "18% gray" you call your gray
in your product or you could produce some casts due to where that
falls within the linear capture. In fact, when Thomas Knoll designed
the color profiles in ACR, he sampled this #20 patch under two
illuminants to build them and, should you clip channels, ACR will
pop an error. Further:
"Trying to adjust white balance using a gray that is too dark is not
advised. Since you are working primarily with a linear capture where
white balancing Raw captures, a middle gray is too far down the tone
scale to provide an optimal amount of image data to evaluate. A
gray card that has traditionally been used by photographers in the
past should be avoided because its too dark and gray cards aren't
noted for their neutrality. Third party white balance cards are
available whose spectral neutrality have been tested and found to be
particularly useful when used as a white balance sample point for
Camera Raw." (Page 116 of Real World Camera Raw by Schewe and Fraser).
Lastly, the video might be more useful if you teach the behavior of
the Color Sample Tool which provide 5x5 sample based on the zoom
ratio. Zoom in closer to the target, you're more likely to get an
actual 25 pixel sample using the sample tools or WB tool in LR. In
LR, its 5x5 regardless of zoom.
Andrew Rodney
http://www.digitaldog.net/_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden