Canon Pro9000 ll colour shift noticeable in greys

[ThrillaMozilla]

The 186 cyan strip, which requires a greater flow of ink, printed immediately after, does not appear to lighten. If there is ink starvation how is the second strip not affected?

If you look closely, it is affected, but the effect is masked by other colors. (You can also measure the effect.) Most of those bars look a bit streaky to me, suggesting generally uneven delivery in all or most of the colors. Compare that with The Hat's example, which is not just perfect, but astonishingly good. EDIT: Emulator reports that the streaks are artifacts.

Thank you for providing this example, though. I have never seen ink starvation without obvious horizontal light streaks. This actually looks more like what I would expect if the ink was diluted, as might happen if the cartridge had some water in it when it was refilled.

 

[stratman]

The two sets of colour bars were printed on one sheet starting at the top. You may be able to see the 255 cyan displays a lightening of colour shade down the strip. The 186 cyan strip, which requires a greater flow of ink, printed immediately after, does not appear to lighten. If there is ink starvation how is the second strip not affected?

I see the color gradation in the Cyan stripe now that you have pointed it out. I missed that when i first looked at the image. I also see the vertical streaking Thrilla mentioned in several 186 stripes. Whether this is artifact from image compression and/or my display or actual representation of ink flow alterations is impossible to tell on my end. Was this printed in Portrait or Landscape mode? You said it was printed "starting at the top." Portrait mode then?

What do you think about The Hat's point about how the printer handles 255 versus 186 colors?

 

[stratman]

This actually looks more like what I would expect if the ink was diluted, as might happen if the cartridge had some water in it when it was refilled.

Interesting. Are you thinking there is "full strength" ink in the round cigarette filter of the ink exit port or undiluted ink still in the print head when he first started printing which is then very quickly replaced with diluted ink from the cartridge?

If this 255 stripe was printed in Portrait mode than it's possible, but not in Landscape mode. The vertical streaking I think I see in the 186 stripes would occur if printing in Landscape mode but maybe not Portrait mode. A larger patch/stripe might demonstrate a repeating pattern in Portrait mode which is not seen in Emulator's image due to the width of those stripes.

This is purely conjecture on my part at this point as I'm not even sure I'm seeing properly (artifact) what Emulater has on the printed paper.

 

[ThrillaMozilla]

Interesting. Are you thinking there is "full strength" ink in the round cigarette filter of the ink exit port or undiluted ink still in the print head when he first started printing which is then very quickly replaced with diluted ink from the cartridge?

Yes, or vice-versa.

EDIT: Looking again, I see that several of the bars are shaded, not just cyan. Gray and magenta in the top row, for example. Looking at the background, I see that some of the shading appears to be in the scanner, but probably not all of it (??).

EDIT again: the streaks are artifacts. Ignore the rest of this post.
But now that you mention it, it's difficult to be sure about anything without knowing which way it was printed. The direction of the streaks is not consistent with a dilution effect. The streaks don't look like typical ringing around edges caused by compression artifacts, but then again, there are also streaks on the white areas that don't look like artifacts, so I don't know. The streaks are especially visible in the 186 M and Y bars. EDIT: I changed my mind about the streaks. I think you could be right that they are compression artifacts, but I'm not sure.

So we really need to know if the printhead ran along the bars or across the bars. (I'm confused by "portrait" vs. "landscape" in this case because that doesn't tell me how the paper was turned.)

I didn't understand your question about what The Hat said about the 186 bars. They're darker, of course, because they have black or other colors mixed in.

 

[PeterBJ]

I still think the cause of the color shift after an inch or so is printed is an ink flow problem in one of the cyan cartridges. As the color shift is much less than in my uploaded image, I think the photo cyan cartridge is the culprit. So I suggest you try first replacing the photo cyan cartridge and see if that makes a difference. If that doesn't help then also replace the normal cyan cartridge. I may be right or wrong, but I think this test is relevant to rule out or confirm an ink flow problem.

 

[ThrillaMozilla]

I forget, Peter. Was it shaded like Emulator's? Try printing it upside down and see if the shading runs the other way.

My experience with ink flow problems is that either it prints a normal-sized drop or it doesn't print anything. I suppose in principle, pressure could affect the drop size somewhat. Maybe I just haven't noticed.

 

[stratman]

I didn't understand your question about what The Hat said about the 186 bars. They're darker, of course, because they have black or other colors mixed in.

I think that's what The Hat meant too. The stripes use different ink or combinations of inks when 186 bars versus 255 bars.

Portrait would mean the page was printed in the same orientation as that of the image we see posted by Emulator, such that when using an 8.5" X 11" piece of paper, the print head moves parallel to the 8.5" edge/dimension. In Landscape mode, the print head moves parallel to the 11" edge/dimension.

This is a puzzler, Thrilla.

 

[Emulator]

We don't have NFL in UK. I now see what you mean about spammers.

I don't have time to follow up the suggestions at the moment, but thank you all. To answer the simple questions, the print was in 'portrait' i.e. printing started at the top. The streaks are artifacts, on the original there are none. The colour shift is only visible in the 255 cyan and 200 grey to my eyes and now the print is several days old it is less visible, in fact you might not notice it. I will carry out more tests shortly. I thought of swapping Lt Cyan and Cyan, unless advised against it. I also have spare refilled carts to test. Has anyone tried printing similar strip on their 9000. The carts were wicked dry and warm dried over night before refilling, I don't think there was any significant water left in the carts.

For the time being, regards to all, Ian

 

[PeterBJ]

Emulator wrote:
I thought of swapping Lt Cyan and Cyan, unless advised against it.

I did not suggest to interchange cyan and light cyan. I might not have been precise enough. I meant to replace the cyan cartridge with another cyan and replace the light cyan cartridge with another light cyan. This would be a perfect use for your spare cartridges.

If you try to interchange the cyan cartridges, you will get an error message "One or more cartridges are in a wrong position" or something like that. That's because the chips are color coded. According to Canon that's to prevent users from making such mistakes. Not a word about making refilling more dificult

ThrillaMozilla wrote:
I forget, Peter. Was it shaded like Emulator's?

I don't quite understand the question, but you find the picture in post #44. I cropped the picture to get less artefacts and less loss of detail. Note very pronounced color shift and banding before failure of cyan. I use IrfanView to resize the picture to 80 kB before uploading. I think the result is fair.

 

[ThrillaMozilla]

...in post #44.... Note very pronounced color shift and banding before failure of cyan.

No, I don't think there is one. Look at the ball, not the wedge. I could easily think that there might be a change in drop size just before you get air, but I don't see any evidence of that in your print.