Color Changes: Drying Time v. Warm Up

[Doug Gray]

I've been looking at the impact of these two variables the last several days as I try to squeeze out the best technique for color profiling my main two printers, a Pro1000 and Epson 9800.

These two exhibit the above but differently.

In general, drying time beyond an hour produces little change on Pro1000 prints but significant changes on the 9800. However the changes on the 9800 are almost entirely in the neutral and very near neutral patches with little change in color patches.

The largest change in the 9800 is on the black patch where L* starts off after an hour at 3.8 but gradually increases to about 4.6 over the next 3 days. Color patches, OTOH, change little, around dE 0.1 in the same time frame. The Pro1000 changes little, also about 0.1 dE over the 3 days on all colors. And the black patch starts off at L*=2.9 and increases to about 3.0.

However, the Pro1000 collectively, is very sensitive to what I call warm up. The first page is significantly different from an identically printed second page in succession. The average dE print of defined colors using Abs. Col. increases from about .35 dE after warm up to about .55 dE on the initial page.  And the differences remain after 3 days of drying. The 9800, OTOH, exhibits no detectable warm up difference.

That said, the Pro1000 requires fewer patches to obtain the same level of accuracy as the 9800 so long as one accounts for the warm up effect.

The slow 9800 changes over three days on the K inks can be essentially eliminated by first allowing the print to dry flat for 60 minutes then "baking" them at 120F for an hour. Doing this stabilizes the colors and produces the same results as just drying at room temp for 3 days. I found that it's important to first let the prints dry for 60 minutes rather than just putting them directly into 120F. Significant differences show up if the prints are put into 120F before at least the water component of the inks have evaporated. Inks mostly have a water base with a smaller glycol component that evaporates very slowly over many days.

The main purpose of this was to identify a protocol that could be used to experiment more quickly and get around the need to wait days to evaluate experiments. So my protocol is now:

1. For the Pro1000, printing an initial page of color patches and tossing it.
2. For both printers letting the print dry at room temp for an hour then another hour at 120F.

[Alan Goldhammer]

Doug - thanks for the suggestion about the Pro-1000.  I did a lot of profiling after I got my printer set up.  One profile was a little bit off and I think it was the first set of patches that I printed that day.  I reprinted the patches the other day after running 4-5 prints so the first sheet of patches should be OK now.  I'm going to read them later today and then compare that profile with the one I did previously.

Your observations are not surprising in that the chemical composition of Epson and Canon pigment inks is bound to be different because of the ink delivery systems.

[nirpat89]

Hi, Doug. 

I wonder if the warming up requirement in the Canon printer have something to do with the fact that it works on a thermal printhead technology as opposed Epson's piezoelectric.

Does this mean, for Canon, one also has to toss out the first print or may be print a bunch of junk before printing the real thing on a good paper to get the correct color as intended by the profile?

Do MK vs PK make a difference?  Or perhaps the paper, when it comes to drying time.  May be the standard 1 hr. @RT + 1 hr @ 120 F protocol will take care all different combinations.


:Niranjan.

[Doug Gray]

Hi, Doug. 

I wonder if the warming up requirement in the Canon printer have something to do with the fact that it works on a thermal printhead technology as opposed Epson's piezoelectric.

Could well be. It's still a subtle difference and probably not observable in any normal sense.

Does this mean, for Canon, one also has to toss out the first print or may be print a bunch of junk before printing the real thing on a good paper to get the correct color as intended by the profile?

Do MK vs PK make a difference?  Or perhaps the paper, when it comes to drying time.  May be the standard 1 hr. @RT + 1 hr @ 120 F protocol will take care all different combinations.


:Niranjan.

Good question. I haven't tested MK as much but my earlier testing suggests that matte prints have much less variation at low L* than glossy prints.

[Ernst Dinkla]

Hi, Doug. 

I wonder if the warming up requirement in the Canon printer have something to do with the fact that it works on a thermal printhead technology as opposed Epson's piezoelectric.

Does this mean, for Canon, one also has to toss out the first print or may be print a bunch of junk before printing the real thing on a good paper to get the correct color as intended by the profile?

Do MK vs PK make a difference?  Or perhaps the paper, when it comes to drying time.  May be the standard 1 hr. @RT + 1 hr @ 120 F protocol will take care all different combinations.


:Niranjan.

Yes, some idea about the papers used would help. Doug's "warm up" should not be taken too literally I think, other effects may cause it as well.  I wonder whether the density per patch decreased in the second print or increased. Pump efficiency going down or up.

On drying at 50 Celsius, yes I kept an eye on that effect if done directly or when the inks have settled more on the coating. I expected that the pigments stay more on top with heated drying right away and by that also are less bonded and more prone to scratches. So I do the same Doug does but for another reason. The drying cabinet I have has frames with mesh stretched, prints can dry reversed too. Have to experiment what effect that has.


Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots

[Ernst Dinkla]

Could well be. It's still a subtle difference and probably not observable in any normal sense.

An 0.20 dE difference. I wondered about that. If all the patches together increased or decreased in density makes that a sub 1dE shift more visual or not. Or are individual changes of patches adding up to max differences in contrast of say 0.80 dE more visible?  Tossing a target print is something different as tossing the first print of every print run.  On the HP Zs the targets have a bar up front of the patches that has wide patches of Black and CMYRGB at 100%.

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots

[Doug Gray]

Yes, some idea about the papers used would help. Doug's "warm up" should not be taken too literally I think, other effects may cause it as well.  I wonder whether the density per patch decreased in the second print or increased. Pump efficiency going down or up.

On drying at 50 Celsius, yes I kept an eye on that effect if done directly or when the inks have settled more on the coating. I expected that the pigments stay more on top with heated drying right away and by that also are less bonded and more prone to scratches. So I do the same Doug does but for another reason. The drying cabinet I have has frames with mesh stretched, prints can dry reversed too. Have to experiment what effect that has.


Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots

I've seen the same effect using Canon Plat Pro, Epson Ultra Prem, and Costco Glossy. They all exhibit similar characteristics even though the UP and Costco have tons of OBAs.

To determine the drying time, I use a thermal imager which picks up tiny differences in temperature. As water evaporates, the surface is initially significantly cooler and over half an hour or so one can see the temp. differential drop as less and less evaporation occurs. The imager only picks up the water evaporation. Glycols evaporate much slower by several orders of magnitude as the vapor pressure is quite low. The 9800 k inks change the most as the glycol slowly evaporates or gets stably adsorbed by the paper.

[Doug Gray]

An 0.20 dE difference. I wondered about that. If all the patches together increased or decreased in density makes that a sub 1dE shift more visual or not. Or are individual changes of patches adding up to max differences in contrast of say 0.80 dE more visible?  Tossing a target print is something different as tossing the first print of every print run.  On the HP Zs the targets have a bar up front of the patches that has wide patches of Black and CMYRGB at 100%.

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots

I certainly don't toss the first prints normally. It's just something I do when critically evaluating patch sets for making profiles. These small differences can impact how one evaluates patch sets.

And you are right about other possible variations. At these low levels I've seen small changes after weeks up and down that may be due to variations in humidity since it's occurred when RH varied from 30% to 60%. Also, the printers do things on their own schedules and it's not possible to get exactly the same starting conditions each time. But tossing the first chart print seems to produce the least variation on the Pro1000.