Canon Pro 1000: large color drift after ink running out warning

[Doug Gray]

Large changes in delta E occurred printing with my Canon Pro 1000 in the last month. In the worst case the change was (dE1976:11 or dE2000: 5). I have never seen this amount of change on my Epson 9800.

Back at the start of Dec., I'd been printing on the same Pro1000 cartridges since about May and they were about 1/3 empty except for two that were near empty. GY and PBK. Around the 4th they issued the low ink warning but there's still quite a bit of ink left so I've been running them until they quit not thinking too much of it. As of now (Jan 3) they still haven't run out.

However, I've been developing a tool to evaluate profile patch sets for accuracy that's specifically designed to minimize printer/ink variations. Along with that are mechanisms to track changes over time. Both to evaluate ink color stability on a print and such things as drying time impact. This involves printing a linear ramp, RGB(0,0,0),(2,2,2),(4,4,4)...(30,30,30) as well as a set of all colors using 0, 85, 170, 255 in each RGB channel. This is a total of 79 colors. These are then repeated to fill a full page and duplicates each color 12 to 13 times. Averaging these gives a pretty precise estimate at the time the page was printed then measured.

The unexpected happened. The colors shifted radically after the ink warning light came on but while the printer still continued to print. In the worst case the color associated with RGB(255,0,85) a slightly orange-ish red, shifted 9 deltaE's after 20 days to 11 deltaE's after 30 days. In the latter case the original Lab value was (50.05, 77.34, 42.14) and 30 days later it printed as (48.76, 77.08, 53.19). More saturated and darker. This might be consistent with ink in cartridges gradually drying and becoming denser. I would expect, from the principles of physics, this would occur at higher rates as the remaining ink got lower. But this seems rather much.

Needless to say, this sort of error greatly exceeds any inaccuracy caused by using a small patch count, say 400 v 2000 when making profiles.

The printer has passed nozzle checks in each case prior to printing.

[Czornyj]

This is very weird, never saw anything like that in any iPF. I would recalibrate the printer with internal spectrodensitometer using Canon Pro Luster. If it won’t help it there may be some issue with ink pressure, so I’d try to reinstall the print head to purge ink lines.

[Doug Gray]

This is very weird, never saw anything like that in any iPF. I would recalibrate the printer with internal spectrodensitometer using Canon Pro Luster. If it won’t help it there may be some issue with ink pressure, so I’d try to reinstall the print head to purge ink lines.

I'm going to finish running the low ink down and take samples along the way then replace all the inks. If it does or does not resolve will be an indicator of the problem nature. Right now it's most consistent with an increase in ink density due to evaporation. It would be great to have more technical details on the Pro1000 design.

If it doesn't resolve with new inks, I'll recalibrate. Currently I have the printer using the factory defaults since calibration affects all profiles and I like to make profiles for each paper/settings combo. calibration would then shift all my same type profiles. And I'm uncertain how consistent recalibration is. In the best case it will produce very consistent results each time which would mean I could make profiles immediately after calibration. Then just re-calibrate as needed with occasional checks for drift.

As an aside, humidity here has been low at 30%.

If new inks continue to exhibit this strong drift, I'll re-install the head.

It's an interesting problem even though the print differences are hardly noticable. Even prints side by side 6" apart. Have to line up the same colors next to each other to see it. And it's only more saturated colors that are impacted. Neutrals are very slightly changed which is odd since those are the low levels.

[jmichael]

This is interesting since we just saw something similar with a Pro 300. However, we also moved to an M1 Mac from an Intel iMac at the same time the ink changes occurred but saw the color shift to slightly warmer and lower contrast.

[aaron125]

What was the end result/cause of these inconsistencies Doug? I’ve the same printer and have always replaced the inks only once the printer refuses to print. I’ve never noticed any differences but you did say that you had to have the same colour patches next to each other to masked the difference noticeable. Therefore, I guess it’s not surprising I’ve never noticed this, that is, if it’s a systemic problem affecting all Pro1000 printers.

[TheNinth]

If it doesn't resolve with new inks, I'll recalibrate. Currently I have the printer using the factory defaults since calibration affects all profiles and I like to make profiles for each paper/settings combo. calibration would then shift all my same type profiles.

Although the post is old, I'd like to chip in on this. Calibration does not necessarily affect all profiles. There is Unique calibration affecting only one media type and Common calibration, affecting all media types that have not had a Unique calibration.

If you have a custom media type - and I assume that's what you are refering to with „profile“ - you can create a calibration target and then run so-called Unique calibrations specific for this media type. Ideally you create the target at the same time as your ICC profile and then use calibration to get back to that state when needed. Once you have run this Unique calibration for a media type, it will not be affected anymore by Common calibrations. Also the built-in media types partly support Unique calibration.

Here you can find some more information on this: A Guide to the Internal Color Calibration of Canon Printers.