Pages: [1]   Go Down

Author Topic: Two Canon Pro-6000 printers, same paper, different output?  (Read 798 times)

drgonzo

  • Newbie
  • *
  • Offline Offline
  • Posts: 17
Two Canon Pro-6000 printers, same paper, different output?
« on: August 13, 2021, 05:42:58 pm »

I'm searching for ideas on why my two Canon Pro 6000 printers have different output when using the same paper and same icc profile.

It's most notable on Epson Premium Luster paper. Sometimes it is the hue of a blue sky that is different, other times it's areas with a lot of solid colors.

Some notes on the printers:
- Both printers are the same model (Pro- 6000)
- Using OEM ink
- Both printers have Canon color calibration performed regularly
- Color calibration/nozzle cleaning and nozzle checks performed before printing any profiling test charts
- Targets for each machine printed from the same roll of paper
- I use the same profiling test chart file and profile creation settings in i1Profiler to create profiles for each printer
- Measured the targets using the same ISIS XL after 48 hours dry-down
- Test charts were printed through Mirage RIP using the "Profile Creation" setting to turn off color management.
- I am always printing images through Mirage with Perceptual intent.

When I built my icc profiles, I made one profile for each printer/paper combo, specific for each machine. The Canon tech advised me to profile each paper/machine combo as a best practice -  thinking that a profile for each paper/machine combo would reduce differences in output between machines (vs using the same Premium Luster paper profile on both printers).

A Canon tech came out to troubleshoot shortly after the printheads on both matches were replaced (so the heads are the same age and approximately the same amount of use) but couldn't ever determine why there is a mismatch in output. They chalked it up to "each machine has a different electrical fingerprint".  So I'm wondering if there is some other variable here that I'm missing.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #1 on: August 13, 2021, 05:57:24 pm »

Did you save the measurement data for each profile? Be useful to check them out.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

drgonzo

  • Newbie
  • *
  • Offline Offline
  • Posts: 17
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #2 on: August 15, 2021, 04:15:34 pm »

Thanks for the reply, Andrew - I did save the measurement data, and just ran a report in Colorthink Pro comparing device measurements from when I read the profiling charts with the ISIS.

Looking at the Color Worksheet, it looks like the highest dE between the two machines are in blues and greens. Would it make sense to measure the charts for each printer multiple times and then average each machine's measurements before creating a profile for each machine?

This is what the dE report said:
--------------------------------------------------
dE Report
Number of Samples: 905
Delta-E Formula dE2000
Overall - (905 colors)
--------------------------------------------------
Average dE:   0.78
    Max dE:   1.92
    Min dE:   0.04
 StdDev dE:   0.38

Best 90% - (814 colors)
--------------------------------------------------
Average dE:   0.71
    Max dE:   1.34
    Min dE:   0.04
 StdDev dE:   0.31

Worst 10% - (91 colors)
--------------------------------------------------
Average dE:   1.50
    Max dE:   1.92
    Min dE:   1.34
 StdDev dE:   0.13

--------------------------------------------------
--------------------------------------------------
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #3 on: August 15, 2021, 04:26:50 pm »

Those values are enough that I'd go no farther measuring those targets and averaging. But the big difference you report seem to fall in colors you initially reported on the prints so that seems to point to there being an issue. I do have to say the difference in two of the same printers seems high, at least based on doing similar comparisons from Epson when building profiles for exhibition fiber for several models. 10% max dE of nearly 2 is not great. Why? Dont know but if you go to Canon for help, you have good data to show them and ask: is this expected?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #4 on: August 16, 2021, 12:07:14 pm »

The largest difference (deltaE 1976)  I see seems associated with the yellow channel. I'm using dE76 because it's more aligned with ink density while dE2k changes that to human perceptual differences. Note that the dE2k for this was 1.08.

Largest dE76 was 4.2,  Lab1: (91.52, -10.76, 69.90)  Lab2: (91.69, -10.62, 65.66) RGB:255,255,95

The shift is also in the greens which uses a lot of yellow ink but produces higher dE2k though lower dE76.

Might be some process difference. I would run at least 5 letter size pages of profile charts through each printer then print profiling charts and make profiles from that. I've seen differences like that on my Epson 9800 when the printer hasn't been used a lot for a week or so. I have a Canon Pro1000 and it also exhibits shifts over about 2 pages when not used for a week. And nozzle checks won't show this.

This can result in an apparent dot gain. Here's the full list of the Yellow channel desaturated over the range of full to none(unprinted)

  95.06   -1.05   -2.20       95.11   -1.02   -2.08     255 255 255
  94.08   -4.01   11.25       93.97   -3.96   10.91     255 255 223
  93.25   -6.85   26.81       93.16   -6.63   25.23     255 255 191
  92.62   -9.02   41.94       92.72   -8.74   39.80     255 255 159
  92.11  -10.07   55.57       92.18  -10.00   53.13     255 255 127
  91.52  -10.76   69.90       91.69  -10.62   65.66     255 255  95
  90.87  -10.94   83.56       91.11  -11.08   80.43     255 255  63
  90.09  -10.20   97.82       90.17  -10.60   95.67     255 255  31
  88.49   -4.89  109.13       88.63   -5.53  108.83     255 255   0


« Last Edit: August 16, 2021, 02:03:06 pm by Doug Gray »
Logged

drgonzo

  • Newbie
  • *
  • Offline Offline
  • Posts: 17
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #5 on: August 16, 2021, 03:54:24 pm »

Thanks Andrew and Doug for the insights - the Canon tech said that creating a profile for each paper type on each machine would minimize output differences, but this report does make that more clear and it's something more quantitative to show the Canon support people. 

Doug, I'll try reprinting the targets after running some prints through the machines first. Neither one of the printers ever sits idle for more than a weekend, and we are usually printing 10-30 linear feet of prints on 44" or 60" rolls per day, sometimes a lot more. That said,  I don't remember exactly how many prints were done right before the targets were printed. It's likely the targets were printed first thing in the morning after running the Canon color calibration process. The calibration process does print a small color test chart after doing a nozzle check, but it's a lot less ink volume than running 5 letter size prints through.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #6 on: August 16, 2021, 04:04:27 pm »

the Canon tech said that creating a profile for each paper type on each machine would minimize output differences,
File that under 'duh' <g>.
Are these deltas expected? I'm not sure this tech can answer that based on the previous conversations reported.
I now have to wonder how consistent output is over a number of days. Is it 'drifting'? That would not be good. Are the inks or papers consistent? They should be.
Again, much of my experience doing these kinds of tests are with Pro Epsons (or a device that can be all over the planet; differing digital presses).
Maybe someone with a similar printer as yours would be willing to run some targets for you to measure.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #7 on: August 16, 2021, 05:23:50 pm »

Something I've noticed on my Canon Pro1000, is there is a short term warmup effect. This causes a shift between the first page and subsequent pages. That effect doesn't exist with the Epson. It might be the different head tech. between the 9800 and Pro1000. But what I do is always print the first page of a set of charts twice and spool them all. Then toss the first page. That said, the most shift I've seen on the first page is about half of what you are seeing. But you might give that a shot.

Also, these being 2 different printers, you should expect some difference between them printing charts. The custom profiles for each printer should eliminate these differences printing using color management. Also, the Canon can self calibrate. I disable that feature.  It will improve the machine's accuracy with their OEM profiles but can be counterproductive when you make your own.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Two Canon Pro-6000 printers, same paper, different output?
« Reply #8 on: August 16, 2021, 06:37:57 pm »

Another thing is you really can't compare profile patch readings from two different printers. They won't give you any idea how accurate the printers are when each printer has their own profile.

Here's a chart similar to one I use. It has a 16 bit tiff file in ProPhoto RGB that is set up to be read by an i1isis. Also included is the RGB file associated with the chart. Load the RGB file into i1Profiler, select USA letter size profile with defaults. The image should match the tif file.

Print the tiff file from Photoshop or any app using Absolute Colorimetric.

The patches are all at Lab spaces of 10 so when you read the measured data you can easily see how far off the printer is. There's also a CGATs file with the Lab values.

These patches should all be in gamut for any glossy type media with uVcut M2 profiles on a Pro1000 and up series.
Logged
Pages: [1]   Go Up