LaserSoft Imaging

Using latest update AI 6.4.3r9

I don't use this tool often enough to be very proficient with it but I don't remember getting such massive changes when using the White point tool before.

See before and after here:
http://home.jtan.com/~reader/exp/web_ready/disp.cgi

I've reset histogram and Graduation curves I've also removed any automated settings under Options/auto

Then scan around in my picture with densomiter on the arm of the couch on left which appears to be the brightest/whitest spot I get this reading:
R 192 => 196
G 196 => 200
B 204 => 224

And click that spot with white point tool. I get a hefty dramatic change when I expected to improve the general brightness.

The after pic is unusable as you see. So whats wrong here? am I clear off the mark with usage?

Harry,

It looks like you picked a specula highlight to use as the white point. Try using the white wall behind your subject and see if that doesn't produce better results.

RAG wrote:Harry,
It looks like you picked a specula highlight to use as the white point. Try using the white wall behind your subject and see if that doesn't produce better results.

First off what is a `specula' highlight? I didn't google just out of lack of time.

I tried doing what you said I think and have posted the results with a red x showing where I clicked for white point. I started with everthing neutralized. reset etc.

Before/After

Its hard to imagine how anywhere I might click would cause this drastic change.... I mean it looks like I ran it thru some kind of psychodelic distortion filter...

Do you think it is something more serious than choice of where to click?
Is there some trick settting somewhere that would invoke this mess?

That was a typo, meant specular (The blown out white that is created from the light reflecting off of a shinny surface). When viewed with the densitometer it will read Red=255 Green=255 Blue=255.

I noticed the extreme blue cast in the second image. Try removing color cast prior to setting the white point to see if that helps any.

I just tried setting the white point in one of my images and did not have the same problem. I used the "Show brightest or darkest point of the frame" button to find the brightest point.
<img src="/img/forum/light_dark.gif">

RAG wrote:That was a typo, meant specular (The blown out white that is created from the light reflecting off of a shinny surface). When viewed with the densitometer it will read Red=255 Green=255 Blue=255.

I'd be about the last one to call attention to a typo... Had you spelled it correctly I still would not have recognized it.... thanks for the definition.

I noticed the extreme blue cast in the second image. Try removing color cast prior to setting the white point to see if that helps any.

That helped immensely. It made the difference between using whitepoint and moving on or creating a useless adjustment that has to be backed out and something else tried.

I just tried setting the white point in one of my images and did not have the same problem. I used the "Show brightest or darkest point of the frame" button to find the brightest point.

Thats something I was sort of confused about. Does user usually want to use whitepoint on the brightest area? I find that often will darken the image beyond usability. My problem here is poor picture quality in many ways.

These are someones home shots probably many of them taken with those 3 dollar cameras from walmart. And shot by homebodies with little or no knowledge of what to do. That is, I'm working with what are probably on most working class peoples mantels and dressers, in collages and etc.

So a common problem with peoples pictures is they are lit with a little flash that got the center but falls away to quick making it hard to adjust so that the darker 40% is also visible.

I was hoping I could use whitepoint carefully to get the best level between the over lighted and underlighted.

I posted another set with filenames [Before/After]White2.png where I used the little guys white tee at his neck... it made the pic pretty useless.
This was after removing cast 100%. Also notice the odd sort of neon effect on the edge of the book pages. How does one dampen down what whitepoint does? Or is it a matter of lots of experience.
http://www.jtan.com/~reader/exp/web_ready/disp.cgi

Probably getting a little confusing on my site ... ignore those other two with names like Auto*.png, those are related to another thread I've started but probably also related to this.
OR usa a QUERY STRING of White2 to just get the newes ones:

http://www.jtan.com/~reader/exp/web_ready/disp.cgi?White2

Am I just using the wrong tool. Expecting it to do something it will not do.

Harry,

What type of scanner are you using and did you do any color management at all?

RAG wrote:Harry,

What type of scanner are you using and did you do any color management at all?

Sorry I missed your reply `Rag':

Scanner is Epson 4870

I didn't do anything special about color... I had hopes of things just working by default since my Silverfast is supposed to be for this model of scanner.

I selected that one when I bought it as I suppose everyone does.

I'm somwhat confused by the manual where they talk about using some upgrade to get calibrated. I don't remember now what all I bought.

The quick start manual that came with my boxed set talks about using
IT8 calibration... it appears to be some more software that is sold separately...

But ... silverfast is supposed to be somewhat intelligent so I'm guessing it would have sensible defaults and have decent scanning ability out of the box. Since it was supposed to be matched to this scanner. Is that not true?

Harry,

It is matched to the scanner from a functionality standpoint, but color management is not something that it is currently possible to set by default. Take a look at the color management documentation and at least set your "CMS" options accordingly. Doing these settings at a minimum will improve the quality of your scans.

The 4870 is a good scanner, I used to have one until I upgraded to the 4990 and I got excellent results.