Upgrading from HPZ3100 to iPF6400

[Robert Ardill]

Hi Robert,

Have you looked carefully at the LEDs used during measurement when performing calibration? Green is used to measure paper white, and the blue LED is used for the first set of patches, and then red for the second set of patches - I may have gotten that mixed up slightly.

I'm still a bit concerned about the OBA side of things.  I would have thought that unless the LEDs have no UV or that the densitometer has a UV filter, that the OBAs will affect the calibration: for example, the white will be bluer and brighter than it would be without UV.  Perhaps that doesn't matter if we are profiling the papers, but it could make quite a difference if we use manufacturer-supplied profiles with common calibration (if we use manufacturer supplied profiles I would think that we should do a unique calibration on the paper).

I'm thinking that this whole thing is potentially quite messy, and the best thing is to use a common calibration using a paper that has minimal or no OBAs and always profile the papers.

Robert

Robert

[samueljohnchia]

I used an .am1 file from Canson, which seemed to have wrong head height/vacuum settings (but it could be the paper path/start of roll problem you mention)...
I take it the .am1 should have things like head height and vacuum settings?  Seems strange then that Canson would have the settings at Auto(lowest) and Auto(weak) for a thick curly paper like the Platine (must try the Baryta and see what the settings are for that)

Canson uses the "Auto" setting for head height and vacuum setting, which is possibly just a safe bet on their side. They certainly don't want to invite complaints that a head strike destroyed a user's print head because of their settings. Leaving it to auto allows the blame to fall to Canon. Unfortunately our dumb machines still cannot figure this out on their own so the settings that were determined are suboptimal.

The .am1 files will contain the head height and vacuum settings. If you import the .am1 file made using the MCT of another iPFX4XX printer, all settings return to their default values, except for the base media setting and inking setting. I learned that the hard way after I imported .am1 files I created with the iPF6460 and SU-21 for my iPF8410.

So it would seem that with existing .am1 files, we should do an edit, skip the paper feed and set the other parameters like head height and vacuum, then go back and to the paper feed adjustment before doing a calibration, profiling.  Is that correct?

Yes, whatever you do, determine the head height and vacuum settings first, then do the paper feed adjustment last.

That's a problem with rolls of these heavy papers!

Canson's papers are actually quite well behaved in that respect. Platine comes off the roll almost flat - very nice. It has got to do with the paper path design of the iPFX6XX printers, which bend the paper counter to the direction of the paper curl. My iPF8410 printer has a curved paper path, curved in the same way as the paper is. It's is a fantastic design, better than Epson's in my opinion, unless one is needing to print on metal or stiff card that requires a straight paper path.

I would be interested in how you guide the paper with your fingers.  I don't suppose there's any risk of getting the fingers cut off is there? Smiley

Nope, not if you are quick and know what you are doing. Don't do a Mr Bean! The exit path of the paper is large enough. Look for the silver pizza wheels on the top side. Before the paper reaches that point, just guide it with your fingers by holding it down a little. After enough of the paper has been fed through, it will bend downwards naturally and you can let go.

Yes, that's certainly a good point, especially if one uses many different papers (or 15m rolls of expensive papers like the Canson Platine/Baryta/Photo HiGloss papers ... by the time you've done a paper feed adjustment and a calibration you would need to put in a new roll!).

Yup, it uses an insane amount of paper for the adjustments. Notice how much paper gets used for a simple calibration print out? Crazy. Not to mention all the wasted time too.

That's worrying and would indicate that the calibration process is not accurate.

Depends on your definition of accurate. I should mention I am unbelievably fussy about this, so perhaps re-calibration is good enough for you. I suggest to test it for yourself. Calibrate once, make a print, then calibrate again, make a print. Can you see a difference? Or can you measure a difference? How big (or small) is it?

So then really, from what you say, the best solution is to do a common calibration (for all papers)

Yes, I think this is the best for the majority of users. It's the least painful method.

If we add a new paper then we should run the common calibration again before profiling the new paper.

No!! The common calibration only needs to be performed once. It is applied to all other media .am1 files and any new .am1 files that you install even after calibration. You only run common calibration again if your printer drifts, or you swap a print head.

If you run common calibration on every new paper you try all your previous profiles will be invalidated.

I assume the calibration of the 6450 would be better (than the 6400/8400)

I don't understand what you mean. The on-board densitometers of all these printers are the same and give equal calibration quality.

I think you refer to the SU-21 unit, which can be placed on both the 6400 and 6450 and their S cousins. The 6450 cannot make profiles on its own.

I'm still a bit concerned about the OBA side of things.

Anyone interested in print permanence needs to be anyway. Why would you select a paper loaded with OBAs, say like Epson Exhibition Fiber? If you print on high quality papers like the Canson Platine you use, I don't see any reason to worry about this. Besides, it works great all the same. I have calibrated on papers with moderate amounts of OBAs in the paper core (not ink layer) with excellent results.

Indeed profiling does help equalize things a lot. Calibration helps more to prevent things likeblocking up in the shadows, which a profile cannot help in.  And of course a sort of standard to return to on a regular basis, or after print head change.

[Robert Ardill]

Nope, not if you are quick and know what you are doing. Don't do a Mr Bean! The exit path of the paper is large enough. Look for the silver pizza wheels on the top side. Before the paper reaches that point, just guide it with your fingers by holding it down a little. After enough of the paper has been fed through, it will bend downwards naturally and you can let go.

I take it you need to have the top cover open to do that (or else have very long, thin, bendy fingers  ). 

Yup, it uses an insane amount of paper for the adjustments. Notice how much paper gets used for a simple calibration print out? Crazy. Not to mention all the wasted time too.

It's something that has always annoyed me with HP also: instead of going across the paper it goes along it, wasting masses amount needlessly.  One thing I found with the Z3100 is that you can feed A4 sheets into it for calibrations, which saves a whole lot of paper ... maybe the same with the Canon?  I'll give it a try.

Depends on your definition of accurate. I should mention I am unbelievably fussy about this, so perhaps re-calibration is good enough for you. I suggest to test it for yourself. Calibrate once, make a print, then calibrate again, make a print. Can you see a difference? Or can you measure a difference? How big (or small) is it?

Oh, I'm an obsessive-compulsive perfectionist!  If it's not perfect it goes in the bin.  If recalibration requires re-profiling then that's what I'll have to do. I'll do some checks and see what sort of dE differences there are between calibrations.

No!! The common calibration only needs to be performed once. It is applied to all other media .am1 files and any new .am1 files that you install even after calibration. You only run common calibration again if your printer drifts, or you swap a print head.

If you run common calibration on every new paper you try all your previous profiles will be invalidated.

Well, they will be invalidated to the extent that the calibrations are inconsistent.  If the calibration was accurate and you always used the same media for your common calibration then it shouldn't matter how often you recalibrate.

How do you know when your printer has drifted?  Do you print a test target every now and then and do a check?  Or do you rely on the printer to tell you when it's time?

I don't understand what you mean. The on-board densitometers of all these printers are the same and give equal calibration quality.

I think you refer to the SU-21 unit, which can be placed on both the 6400 and 6450 and their S cousins. The 6450 cannot make profiles on its own.

Yes ... I meant the printer with the spectro (the 6460, I guess).

Anyone interested in print permanence needs to be anyway. Why would you select a paper loaded with OBAs, say like Epson Exhibition Fiber? If you print on high quality papers like the Canson Platine you use, I don't see any reason to worry about this. Besides, it works great all the same. I have calibrated on papers with moderate amounts of OBAs in the paper core (not ink layer) with excellent results.

Indeed profiling does help equalize things a lot. Calibration helps more to prevent things like blocking up in the shadows, which a profile cannot help in.  And of course a sort of standard to return to on a regular basis, or after print head change.

The reason I'm concerned about OBAs is because I also use the Canson Photo HiGloss 315gsm (for prints that are mounted behind acrylic, because I need a thick paper with a very smooth surface, so that the paper sticks to the acrylic perfectly).  This is the only paper I use that has OBAs.  So do I common calibrate with a paper that has OBAs, or one that does not, seeing as how I'm using papers that have OBAs and papers that do not?

I suppose the profile would correct for any differences, so I'm probably being a bit neurotic  .  So I'm thinking now of using a roll of Canon Photo Glossy 170gsm for the common profiling.  Any advice to use a different paper?

Robert

[Robert Ardill]

I have one or two (or maybe three) more questions ... I hope I'm not pushing my luck!

• Is 16-bit printing on Windows only possible using the Canon Print Plug-in?  I don't see any settings in the driver that would allow 16-bit printing (whereas there seems to be on the Mac).
• For profiling papers, do you have a recommended number of patches? I have an i1Pro2 and I've used both ArgyllCMS, and also i1Profiler. They both seem to give pretty good results ... I usually use 2584 patches, which just fits onto 4 A4 sheets (or 1 A2 sheet).  I haven't done any comparisons with less/more patches so I'm just flying blind.
• When you're checking your print quality (for example after a recalibration), do you use the Data Analysis workflow in i1Profiler?  I haven't used it but I guess it would make sense to print an A4 target, recalibrate, print the target again, scan both and compare them using the Data Analysis.  This would be OK to check color drift, but I don't see a way with i1Profiler to validate the target print against the chart Lab values.  Any suggestions?  Maybe ArgyllCMS would be an option if it can't be done with i1Profiler.
  

Robert

[samueljohnchia]

I take it you need to have the top cover open to do that (or else have very long, thin, bendy fingers

Absolutely not. I wrote "exit path". You reach in from the front, and only a little ways in. Find the pizza wheels first! That's your marker. I don't have long thin bendy fingers. I'm not about to go back to my dealer's at this point and make a video of all this, and you don't seem to have any trouble with the pizza wheels. This was one of the more annoying issues I had with the smaller format Canons, coming from an iPF5000 more than 8 years ago. The paper path design is not the best.

instead of going across the paper it goes along it, wasting masses amount needlessly.

If you purchase the SU-21 for your printer, it distributes the patches across the width of a 17 inch roll to minimize paper wastage. But I don't think the unit is worth the price.

Oh, I'm an obsessive-compulsive perfectionist!  If it's not perfect it goes in the bin.

Haha I understand that feeling. Would it irk you to know that your i1 Pro 2 is not as consistent as we would like it to be? I tore my hair out for years over this.

If the calibration was accurate and you always used the same media for your common calibration then it shouldn't matter how often you recalibrate.

I dug up my notes of one of the tests I did on recalibration at the start of this year, and I noted a max 1dE2k difference, but I was measuring with an i1 Pro 2 then. Same roll of paper. I'm not sure how much of that error is user error. Then I just did a recalibration on Epson Hot Press Natural, two different rolls, different lot numbers. I measured with my iSis for this one. Max 1.2 dE2k. So I should say it can be reasonably consistent, but bordering on the need for re-profiling for me. It would be good to know what kind of numbers you are seeing too.

How do you know when your printer has drifted?  Do you print a test target every now and then and do a check?  Or do you rely on the printer to tell you when it's time?

Scott Martin recommends every 6 months for low volume users, and every 3 months for high volume users. Yes, I actually print out my profiling target to do a check, since I am going to measure it anyway, it might as well serve two purposes. If it is different, the profile would still be useful for the next print job, being able to correct for most of the drift. I have not known about the usefulness of calibration long enough and used different machines in different atmospheric environments to provide useful numbers besides this. The Canon Color Calibration Management Console also has a status timer which tells you when to recalibrate. I'm not sure how long the countdown is and I don't use that.

Yes ... I meant the printer with the spectro (the 6460, I guess).

The 6460 is the Southeast Asian name of the 6450. Canon adds 10 to all the X4XX printer names. I believe they modify the ink cartridge keyed slots for these printers so folks in Europe and USA don't buy our inks, which are somewhat cheaper.

The spectro might offer slightly more accurate calibrations, but that kind of differences are worked out with excellent profiles. After profiling, I considered it to make no difference at all to the final output quality, after testing carefully. The consistency of the calibration is much more important. I am about to run more tests on this now that I have my iSis and write up an extensive article on what we have discussed so far. I also rate the printer's performance as far more important. I have uncovered multiple issues leading to less than optimal quality for these newer Canon printers. When all the stars align, the dot pattern is ridiculously fine, better in some color regions than Epson OEM driver, and more consistent across the color range.

So do I common calibrate with a paper that has OBAs, or one that does not, seeing as how I'm using papers that have OBAs and papers that do not?

Does not matter at all to me. It is about consistency - pick one and stick with it. Especially if you choose media with OBAs, keep the same roll around for future recalibrations.

So I'm thinking now of using a roll of Canon Photo Glossy 170gsm for the common profiling.  Any advice to use a different paper?

That paper is very thin and translucent and the black plastic of the platen would affect calibration to some extent, although in a consistent fashion. I'm not sure how it would affect calibration as I don't have any around to test against - I saw it only twice when Canon engineers brought it down for my printer repair. Actually, it would be great if you could test it against your Canson Photo Higloss paper and see what kind of differences you can measure.

My preference is to calibrate on a glossy paper that can take the highest ink loads for the greatest dmax. Hahnemuhle Photo Rag Bartya, Harman by Hahnemuhle Gloss Baryta are two I have used successfully before. I noted a gamut increase of just over 4% for matte media when calibrating to a high quality glossy paper, instead of matte paper. I used the same matte paper for the profiling gamut test - instead of favoring calibration on the same paper, it was actually better when I calibrated to glossy.

[samueljohnchia]

I have one or two (or maybe three) more questions ... I hope I'm not pushing my luck!

• Is 16-bit printing on Windows only possible using the Canon Print Plug-in?  I don't see any settings in the driver that would allow 16-bit printing (whereas there seems to be on the Mac).
• For profiling papers, do you have a recommended number of patches? I have an i1Pro2 and I've used both ArgyllCMS, and also i1Profiler. They both seem to give pretty good results ... I usually use 2584 patches, which just fits onto 4 A4 sheets (or 1 A2 sheet).  I haven't done any comparisons with less/more patches so I'm just flying blind.
• When you're checking your print quality (for example after a recalibration), do you use the Data Analysis workflow in i1Profiler?  I haven't used it but I guess it would make sense to print an A4 target, recalibrate, print the target again, scan both and compare them using the Data Analysis.  This would be OK to check color drift, but I don't see a way with i1Profiler to validate the target print against the chart Lab values.  Any suggestions?  Maybe ArgyllCMS would be an option if it can't be done with i1Profiler.
  

Robert

No problem Robert. I am happy to share what I know. It would be fantastic if you could run some of the tests yourself and report your findings. They might be different from my observations and I think we would both want to know why - and understand these printers better.

1. The plug-in supports 16 bits, but the on-board hardware may or may not quantize the bits to 12. Scott Martin said they did on the X1XX printers. I don't know about the X4XX printers.
2. We could talk about this for weeks, it is such an interesting topic. Generally speaking, 2000+ patches (well distributed or well chosen sampling) is at the point of diminishing returns. You are safe here. With active calibration, the X4XX printers are very linear and actually need less patches to profile well. Like you, my preference is also to fill my chosen paper size with patches to reduce paper wastage.
3. I used ColorThink Pro for comparing my data. I just did a quick look at i1Profiler's Data Analysis function. It performs just as well. Being able to parse its own mxf measurement file is convenient. I might switch over!

[Robert Ardill]

I also rate the printer's performance as far more important. I have uncovered multiple issues leading to less than optimal quality for these newer Canon printers. When all the stars align, the dot pattern is ridiculously fine, better in some color regions than Epson OEM driver, and more consistent across the color range.

That sounds both good and bad: I take it what you mean is that it's really important to get everything set up right, and then the results are very good; but anything wrong could give poor (or at least sub-optimal) results.

I would certainly be interested in what you've found to be the key things that need to be done.  Off the top of my head, from what I've learnt so far,  I would have thought:
1. Paper with .am1 with spectral data
2. Head height and vacuum set correctly for this paper
3. Do head alignment on Canon HW Coated Genuine Paper (or good glossy paper) from time to time
4. Do a nozzle check on glossy paper (maybe not necessary?).
5. Do paper feed adjustment.
6. Calibrate paper using Unique calibration preferably, otherwise Common calibration is OK (do Common Calibration if calibration test shows that calibration has drifted beyond acceptable limit.
7. Profile paper with 2000+ patches (but possibly 1200 patches would be OK ... 2 A4 sheets would do this)

Anything else you would consider important?

Robert

[Robert Ardill]

No problem Robert. I am happy to share what I know. It would be fantastic if you could run some of the tests yourself and report your findings. They might be different from my observations and I think we would both want to know why - and understand these printers better.

1. The plug-in supports 16 bits, but the on-board hardware may or may not quantize the bits to 12. Scott Martin said they did on the X1XX printers. I don't know about the X4XX printers.
2. We could talk about this for weeks, it is such an interesting topic. Generally speaking, 2000+ patches (well distributed or well chosen sampling) is at the point of diminishing returns. You are safe here. With active calibration, the X4XX printers are very linear and actually need less patches to profile well. Like you, my preference is also to fill my chosen paper size with patches to reduce paper wastage.
3. I used ColorThink Pro for comparing my data. I just did a quick look at i1Profiler's Data Analysis function. It performs just as well. Being able to parse its own mxf measurement file is convenient. I might switch over!

1. So 16-bit is only available from the plug-in?  Not if using the printer driver (on Windows, that is)?  If the hardware knocks it down to 12 bits I wouldn't be too concerned, it would still be a lot better than 8 bits.
2. As I mentioned in my previous post, perhaps 1274 patches would be enough (2xA4 sheets for the i1Pro), seeing as how the printer is quite linear.  Still, it would be leaving double the interpolation distance (I guess), so if we want very smooth gradients it might be best to go for the higher spot count.
3. Yes, i1Profiler looks OK.  I haven't tried it yet, but what I have tried is a test using ArgyllCMS.  It produces the Lab values for the two targets, delta E values (plain old dE, dE94 or dE2000), the worst, best, average, 10% and 90% best/worse and a vrml file that shows the differences visually (quite nice).  What was interesting is that I got 3 values with dE2000 values of over 1, with an average of 0.33. And that's from prints done one after the other (could be some drying-time differences, I'll need to repeat the test).

If you're interested, here are the Argyll commands you would need:

targen -v -d2 -G -e8 -B8 -f400 chart1
printtarg -v -ii1 -a1.0 -T300 -M6 -pA4 chart1

copy chart1.ti2 chart2.ti2
copy chart1.tif chart2.tif

Pause print chart1.tif
chartread chart1

Pause print chart2.tif
chartread chart2

colverify -v2 -k -s  -w  -W  -x chart1.ti3 chart2.ti3 > printerdrift.txt

You can then read the printerdrift.txt file into Excel with delimiter set to Space.

You can also use Argyll to check profile accuracy:

profcheck -v2 -k -w -x -p -Ir chart1.ti3 profile.icm

In this case the chart needs to be printed with the correct profile (so replace 'profile.icm').

Robert

[Robert Ardill]

That paper is very thin and translucent and the black plastic of the platen would affect calibration to some extent, although in a consistent fashion. I'm not sure how it would affect calibration as I don't have any around to test against - I saw it only twice when Canon engineers brought it down for my printer repair. Actually, it would be great if you could test it against your Canson Photo Higloss paper and see what kind of differences you can measure.

Actually, I'm still wobbly about using a non-genuine Canon paper for common calibrations. For example, if I use the Canson HiGloss paper then I would have to put it in as GlossyPhotoHG255 (probably?), but the papers won't have the same white points, OBAs, spectral profiles, ink absorption etc.  So it would seem to me that the chances are that the calibration wouldn't be the best.  What do you think?

Your point about the thinness of the Glossy 170g is a good one ... I'll get one of the heavier papers.

Robert

[Robert Ardill]

About reducing paper usage during calibration etc.

The printer will calibrate on one sheet of A4.  For feed adjustment it requires 2 sheets of A4.  Nozzle check 1 sheet of A4.

The feed adjustment and the calibration only uses part of the sheets, so they can be reused for ink level set/head height/vacuum strength adjustments.

So I reckon that for a new paper that I would need:
Calibration:          1 sheet
Feed adjustment: 2 sheets
Level settings:     Re-use above
Profiling:             2-4 sheets (1292 - 2584 patches for i1Pro)

Assuming only 1292 patches this would require 630mm of 24" (leaving one spare A4) or 295mm of 44" papers.  For 2584 patches we would need 840mm of 24" (again leaving 1 spare A4), or 590mm of 44" papers (leaving 3 spare A4 sheets).

I don't know how that compares to just using the rolls, but I image there is a good saving there (a calibration on 24" roll used 260mm of paper with most of the paper unused).

Robert

[JRSmit]

How do you see the head height in mm? I only see Super Low, Lowest, Low, Standard, High and Highest.

samuel , it is the paper thickness i am referring to. I assume the head height is relative to the specified paper thickness. Like it is in the epson stylus pro printers.

[samueljohnchia]

samuel , it is the paper thickness i am referring to. I assume the head height is relative to the specified paper thickness. Like it is in the epson stylus pro printers.

Hi JRSmit, thanks for clarifying. But I'm pretty sure I saw the head height names correspond to exact mm values somewhere...

Haha I found it! Service manual of the iPF8300, page 222 and 250.

Super Low - 1.2mm
Lowest - 1.4mm
Low - 1.8mm
Standard - 2.0mm
High - 2.2mm
Highest 2.6mm

[Ernst Dinkla]

I'm still a bit concerned about the OBA side of things.  I would have thought that unless the LEDs have no UV or that the densitometer has a UV filter, that the OBAs will affect the calibration: for example, the white will be bluer and brighter than it would be without UV.  Perhaps that doesn't matter if we are profiling the papers, but it could make quite a difference if we use manufacturer-supplied profiles with common calibration (if we use manufacturer supplied profiles I would think that we should do a unique calibration on the paper).

I'm thinking that this whole thing is potentially quite messy, and the best thing is to use a common calibration using a paper that has minimal or no OBAs and always profile the papers.

Robert

Robert

It is unlikely that the iPF densitometer triggers a fluorescent effect from the OBAs, it would not correspond well with the common UV cut spectrometers that should create the iPFs profiles. The spectrometer of the HP Zs is UV cut as well in both calibration and profiling.


--
Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
September 2014, 650+ inkjet media white spectral plots.

[samueljohnchia]

That sounds both good and bad: I take it what you mean is that it's really important to get everything set up right, and then the results are very good; but anything wrong could give poor (or at least sub-optimal) results.

I would certainly be interested in what you've found to be the key things that need to be done.  Off the top of my head, from what I've learnt so far,  I would have thought:
1. Paper with .am1 with spectral data
2. Head height and vacuum set correctly for this paper
3. Do head alignment on Canon HW Coated Genuine Paper (or good glossy paper) from time to time
4. Do a nozzle check on glossy paper (maybe not necessary?).
5. Do paper feed adjustment.
6. Calibrate paper using Unique calibration preferably, otherwise Common calibration is OK (do Common Calibration if calibration test shows that calibration has drifted beyond acceptable limit.
7. Profile paper with 2000+ patches (but possibly 1200 patches would be OK ... 2 A4 sheets would do this)

Anything else you would consider important?

Robert

Yes, that is right. It is also extremely helpful to know the optimal gamut behavior of the printer, so when something goes wrong, it is easy to catch it. Keep notes of your printer - it should be pretty ideal as it is new. Look very closely for micro banding issues too. Sometimes the  auto adjustments are not perfect and needs to be done again. This happens when the paper is wavy or curled - varying the space between it and the print head.

1. Spectral data is not important. Don't forget that until the X4XX printers, this was never available, and people could still make excellent prints, as they can do now. I have not seen any evidence of spectral data providing any significant benefits to print output quality.
3. Head alignment should be done on your favourite paper - the distance from the paper to the print head is important. Thin matte papers are not ideal at all. If your favourite paper is a textured or matte paper, I would suggest aligning on a smooth glossy paper of equal thickness. I only run the alignment again when I notice issues with the dot placement precision, otherwise only when changing the print head or the entire carriage assembly, belt drive etc. I usually run the initial alignment if I have problems. I find it far more effective than the auto or manual alignments.
4. This is usually not necessary until you notice a problem like density loss, color drop outs, misfiring, color channel specific banding parallel to carriage travel etc. You can print this on any standard laser 80gsm paper.
5. I strongly recommend doing this on roll paper, if you are printing from a roll. It keeps the paper under a consistent tension, and the paper is flat across the width of the roll because the rest of itself is wrapped around. The portion over the platen will always lie at the same height to the print head. Cut sheets will always be somewhat wavy across their length and width, and this will affect the alignment. I have wasted a lot of paper to learn this!
6. Actually, calibrate using common calibration always. Unique calibration is an advanced user option, IMHO.
7. That is a bit simplistic, but ok for now.

[samueljohnchia]

So 16-bit is only available from the plug-in?  Not if using the printer driver (on Windows, that is)?  If the hardware knocks it down to 12 bits I wouldn't be too concerned, it would still be a lot better than 8 bits.

Yes, 16 bits only with the plug-in. True, 12 bits is much better than 8 bits. But I would not mind having 16 bits.

I haven't tried it yet, but what I have tried is a test using ArgyllCMS.  It produces the Lab values for the two targets

No recalibration in between?

What was interesting is that I got 3 values with dE2000 values of over 1, with an average of 0.33.

That is unnaturally high. I typically see average dE2k values of less than 0.1, and max of no more than 0.5. Measuring by hand with the i1 Pro 2 is not the most consistent method I'm afraid, but good technique should yield results of less than 1 de2k for the maximum deviation. If they were printed one right after the other and left to dry in the same location that should not play a significant role in the measurement difference.

Thanks for the argyllcms arguments. I'll look into it in more detail when I have the time. It's an excellent and lovely thing.

[samueljohnchia]

So it would seem to me that the chances are that the calibration wouldn't be the best.  What do you think?

Remember, the calibration routine is just a linearizing routine. It serves the nature of calibration only if you use the same paper, same setup and run the calibration function in the future. Calibrate on your favourite glossy media. I have done so with great success, and I am very fussy about my results.

EDIT: I'm with Ernst on this! I doubt the OBAs matter at all.

[Robert Ardill]

No recalibration in between?

That is unnaturally high. I typically see average dE2k values of less than 0.1, and max of no more than 0.5. Measuring by hand with the i1 Pro 2 is not the most consistent method I'm afraid, but good technique should yield results of less than 1 de2k for the maximum deviation. If they were printed one right after the other and left to dry in the same location that should not play a significant role in the measurement difference.

No calibration in between ... so not too good!  I'll scan the targets again and see if I messed up somewhere.  But a dE2K of less than 0.1 I'm not expecting!

[samueljohnchia]

About reducing paper usage during calibration etc.

Excellent, thanks for sharing. I will still stick with working with roll paper, for reasons I mention earlier. Much better material structural stability and height consistency. Also, unless you are very careful in handling and selecting cut sheets, you may potentially have scuffing in the printed areas. My own limited experience with cut sheets is that the smaller sheets have potentially more surface damage (less area to grab when packing them into boxes). These papers are very fragile. I notice that when calibrating from a roll, the printer will feed a great deal of paper out first before printing - but that ensures that it prints on the cleanest part of the paper that was never touched.

Feed adjustment should best be done on roll media. If you primarily print on sheets, use a single (very flat) sheet for doing the adjustment I feel is better.

[Robert Ardill]

Remember, the calibration routine is just a linearizing routine. It serves the nature of calibration only if you use the same paper, same setup and run the calibration function in the future. Calibrate on your favourite glossy media. I have done so with great success, and I am very fussy about my results.

Yes ... I agree.  The only thing is though that if you only do a common calibration then you really have to profile all the papers (can't use manufacturer profiles, for example).  But as Geraldo says the Canson profiles are quite poor, reprofiling seems necessary anyway.

I'm coming at this from the HPZ3100 angle, where one always calibrates each paper ... and quite regularly too.  Seems like the Canon printheads must be much more stable.

Robert

[Robert Ardill]

Excellent, thanks for sharing. I will still stick with working with roll paper, for reasons I mention earlier. Much better material structural stability and height consistency. Also, unless you are very careful in handling and selecting cut sheets, you may potentially have scuffing in the printed areas. My own limited experience with cut sheets is that the smaller sheets have potentially more surface damage (less area to grab when packing them into boxes). These papers are very fragile. I notice that when calibrating from a roll, the printer will feed a great deal of paper out first before printing - but that ensures that it prints on the cleanest part of the paper that was never touched.

Feed adjustment should best be done on roll media. If you primarily print on sheets, use a single (very flat) sheet for doing the adjustment I feel is better.

As calibration seems to be something that's not needed very often, there's really not much advantage to using sheets.  Multiple targets can be printed at the same time for profiling, of course, so again there's nothing to be gained from using sheet paper.  For feed adjustment the printer uses 2 A4 sheets ... so might as well stick to roll again!

Things are getting a bit simpler in my head as I go along ... with a lot of help from you guys!

Robert