Datacolor User to User Support Group.

--- In colorvision_group@yahoogroups.com, "davedoughman" <davedoughman@...> 
wrote:
>
> I understand the choice of paper type selected for printing &
> profileing changes the ink flow onto the paper but I would like to
> have some information on what this really means. Just how much a
> change is there between selections, what inks are affected, etc.
> Dave

Hi Dave,

You could make custom profiles for all media settings on one paper and compare the 
results either by eye or using CDTobie's famous "LabDifferenceCalculator.xls" in the Files 
section.

I applied a much simpler route...

I was never happy how Epson's driver always block the shadows, so I did some testing last 
month to find the media setting for the best shadow detail before I made a profile. 

Here's a repost of what I found: 

Try exploring media settings to open shadows...that's what I'd recomend over changing 
Color Density as Color Density is more of a gamma correction so every part of the image 
takes a hit instead of just the shadows.

I recently tested the "best" media setting for William Turner paper on my Epson 4800.

Epson's black usually comes in hard and fast, less so with the newest printers, and I 
wanted to find the media setting that has smoothest black from 100% to 90%.

I printed a ten patch gray ramp from 100% to 90% with five media settings.

My Excel results are below for William Turner on an Epson 4800:

http://homepage.mac.com/johnvito/WilliamTurnerMediaSettings.png

From 100% to 90% only two media settings show expected linear density loss albeit 
stepped at times. While the other three actually have spiked increases in density "called 
reversals" on some patches. For example Epson's Enhanced media setting has the same 
density at the following patches: 100%, 99% and 97%. but has a reversal at the 98% patch 
and another reversal at 94%.

My media setting choice for William Turner was UltraSmooth as it has the most linear path.
In the end the custom paper profile will have less work to handle.