My 9800 has always printed the same target consistently except for gradual drifting as ink ages and the tank inks gradually become denser over many months. So my testing has come to depend on that, short time frame, consistency. My 9500 II is less consistent but it was never the printer I focused on optimizing.
The Pro1000 has some excellent traits. A large gamut and good black points but just isn't as consistent as the 9800 even though it's now 15 years old. There is the "warm up" effect I've mentioned but I've also seen small changes even when discarding the first print. It seems to just do some sort of operation every now and then before and sometimes after each print. And I've seen unexplained shifts in Lab values printing the same target. These are small, typically only about .2 or .3 in the Lab channels but when one is trying to evaluate optimal targets it interferes. There are even times a 3k patch set produces a worse profile than a 1.5k patch set. Not usually, but sometimes.
So I've come up with an approach. To evaluate profiles it is necessary to print fixed raw device RGB at the same time as I print known colors using Absolute Colorimetric.
So I've come up with a mechanism to print, on one page, a fixed, Abs. Col. set of colorchecker colors as well as a large set of neutrals for 3 different profiles under test and on the same page a fixed set of device neutrals and colors. These are repeated and scrambled in case there is some change during the printing of the page and provides a way of statistically checking this. The fixed device rgb values automates a way to detect changes that have occurred between the printing of one page and another perhaps as a result of the printer doing whatever it does from time to time. But most importantly, it allows the result of 3 different profiles to be tested at the same time on one printed page with a control set of device RGB values.
Yes, I'm a bit OCD about this.
