Re: Measurement Geometry
Re: Measurement Geometry
- Subject: Re: Measurement Geometry
- From: Ken Fleisher <email@hidden>
- Date: Thu, 16 Feb 2006 14:42:30 -0500
On Feb 16, 2006, at 1:05 PM, Steve Kale wrote:
Thanks for that. I'm not sure, though, how it addressed the question
of why
an integrating sphere spectrophotometer is better for measuring colour
samples on canvas or other course materials.
Think about what happens when you take a light and skim it across the
surface of an oil painting. The light exaggerates the texture in the
painting by creating shadows from one side. If you've ever done copy
photography, you will know that you usually try to set up two banks of
lights, one on each side, that are about the same intensity. The idea
is to eliminate the effect of shadows created from texture.
So this relates to measurement geometry because a 45/0 geometry is
similar in concept to skimming a single light across an oil painting
from one side only. So you can see that if your light/sensor are placed
slightly differently, the sensor may have a different response because
what is being illuminated varies due to the changing texture.
With d/0, you have diffuse illumination, so the sample is evenly
illuminated from all sides, similar to the banks of light in copy
photography. the effect of shadows is cancelled out because of the ring
of illumination. In addition, the integrating sphere works in a way
that your sensor is not pointed directly at the sample, but rather the
light reflecting off of your sample is bounced all around the inside of
the sphere thus averaging, or integrating, the color. The sensor
measures off of an inside surface of the sphere. This is how effects of
texture are minimized when measuring with an integrating sphere.
At this point, I feel I must state more clearly that I'm not actually
suggesting using integrating spheres for measuring targets printed on
canvas or other textured materials. For one thing, I don't think there
are any automated integrating spheres that can do the job like a
spectroscan or eye-one with the new iO. Instead, you can use these 45/0
devices if you make many measurements and average them. This should
produce reasonable results for these types of materials. As I stated
before, make sure that after you read the target in once, you "remove"
it from the table, then replace it before taking another measurement.
You might also try to not center the registration exactly the same way
each time on purpose to be sure you are taking readings from multiple
locations of each patch.
I've been typing this in a bit of a rush, so I hope I've been clear.
Ken Fleisher
_______________________________________________
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