Epson R2880 suddenly prints with reddish color cast

[Frans Waterlander]

To you.

I've looked at nozzle checks a few times and I can tell you, these are identical. What more do you want? A hi-res scan? Of all the inks or cyan only?

[digitaldog]

What more do you want? A hi-res scan? Of all the inks or cyan only?

Yes and yes; can you actually do that?

[Frans Waterlander]

Yes and yes; can you actually do that?

Yes, I can actually do that. One image is of the nozzle check print with the old cyan cartridge (magenta cast in B&W print) and the other is with the new cartridge (no color cast in B&W print). Just to make it interesting I don't tell which is which. Fair enough? The prints have been scanned at 800dpi; I couldn't go higher res giving the size of the area to be scanned.

[digitaldog]

Yes, I can actually do that. One image is of the nozzle check print with the old cyan cartridge (magenta cast in B&W print) and the other is with the new cartridge (no color cast in B&W print). Just to make it interesting I don't tell which is which. Fair enough? The prints have been scanned at 800dpi; I couldn't go higher res giving the size of the area to be scanned.

So you can't even produce two scans with the same density? Sad.
Take both, place em in Photoshop on separate layers. Set blend mode to Difference, line em up. Switch back to Normal mode. Scan A is lighter. And if you zoom in you can see what appears to be less magenta ink due in part either to an issue at the time it was made or due to your inability to scan two images identically.
Either way, nothing you've provided here, especially the scans (your only piece of 'evidence') prove the predetermined concept in A which is wrong. So we are back to:

A is wrong and explained why.
B is possible but since your testing methodology and scanning methodology (lack of nozzle checks, trying suggested fixes) was so poorly done, we'll never know. Nor do I suspect does anyone care. Your MO in posting is far more clear than your scanning samples. 

[Farmer]

The resolution is a touch low and there does appear to be a slight variance in the scan dynamics, but it's probably OK.

B shows more scatter and less definition than A in the cyan nozzle check.  A has more definition and generally higher density.  During the B print one of the nozzles had a partial deflection.

Undoubtedly neither check is showing an empty or virtually empty card - ink supply is consistent.

A has more consistent dot gain but it's slight.

Overall they are indeed very similar, I agree.  However, A is better than B and I would suspect that was your "bad" cart and it appears to be less consistent.  This could be separation, although generally the smaller carts are much less likely to suffer this, so it could also be that you had a slightly bad seal and a touch of air getting into it.  If it happens again with the new cart at some point, you may want to look to see if there's any dried ink or other detritus affecting the seating of the cart in the holder or affecting the deal.

A first hand examination would be better, though, particularly given the slightly inconsistent scan.  Also, hold on to your "good" check for future reference as that is really very useful.  In this instance, with the scans available, it's inconclusive but my suspicion is as above.

[digitaldog]

The resolution is a touch low and there does appear to be a slight variance in the scan dynamics, but it's probably OK.

If we examine just paper white, using a 5x5 sampling in PS and extract Lab values, we see that the two scans are a deltaE*ab 2000 of 4.58; pretty bad!
So just an attempt to make two consecutive scans correctly and matching were a struggle and the data provided for 'analysis' is off the bat, flawed.

The prints have been scanned at 800dpi; I couldn't go higher res giving the size of the area to be scanned.

That statement makes absolutely no sense as written.

[Doug Gray]

One of the differences between a nozzle check and a print is that the overall flow rate is lower for nozzle checks. However, the flow rate at each nozzle is higher.

If there was a flow rate restriction at the nozzle the difference should be apparent because more ink per sec. goes through the nozzle.

If there was an overall flow rate restriction, but not a per nozzle restriction, then there would be little or no difference since the overall flow rate is much lower.

[Frans Waterlander]

The resolution is a touch low and there does appear to be a slight variance in the scan dynamics, but it's probably OK.

B shows more scatter and less definition than A in the cyan nozzle check.  A has more definition and generally higher density.  During the B print one of the nozzles had a partial deflection.

Undoubtedly neither check is showing an empty or virtually empty card - ink supply is consistent.

A has more consistent dot gain but it's slight.

Overall they are indeed very similar, I agree.  However, A is better than B and I would suspect that was your "bad" cart and it appears to be less consistent.  This could be separation, although generally the smaller carts are much less likely to suffer this, so it could also be that you had a slightly bad seal and a touch of air getting into it.  If it happens again with the new cart at some point, you may want to look to see if there's any dried ink or other detritus affecting the seating of the cart in the holder or affecting the deal.

A first hand examination would be better, though, particularly given the slightly inconsistent scan.  Also, hold on to your "good" check for future reference as that is really very useful.  In this instance, with the scans available, it's inconclusive but my suspicion is as above.

Phil, thank you so much for your constructive feedback. You are quite right: A is the new cart and B the old one. I'll certainly inspect the orifice area where the cart connects if this color cast issue happens again. I've attached higher res and better matching scans of the cyan nozzle check areas.

[digitaldog]

Phil, thank you so much for your constructive feedback. You are quite right: A is the new cart and B the old one.

And pretty easy to see IF YOU LOOK, that there are issues with scan B, by again realigning them up as layers using difference then not, there IS an issue and alignment of dots within the nozzle check with Scan B!!!!
Congratulations on figuring out how to work your scanner....

[Farmer]

One of the differences between a nozzle check and a print is that the overall flow rate is lower for nozzle checks. However, the flow rate at each nozzle is higher.

If there was a flow rate restriction at the nozzle the difference should be apparent because more ink per sec. goes through the nozzle.

If there was an overall flow rate restriction, but not a per nozzle restriction, then there would be little or no difference since the overall flow rate is much lower.

Absolutely right, Doug.  This is where checks taken before and immediately after a print can help (not always, but often) as an ink supply issue that occurs during regular printing may still be evident in a nozzle check done immediately following the print.