Mike Chaney's Tech Corner

Alright, done as in the videos.

Image comes out a bit red and over-saturated.  This is disabling ICM in the Printer and letting QImage handle the color through setting the proper profiles.  There is no double-profiling, I've checked.

Monitor is profiled.
Z3100 has a profiled paper in it (new profile and calibration)

However, turning on the Printer to do the Color Management in the Driver seems to give much better, though not perfect, results.

What, if anything, am I doing wrong?

Thanks

I wish I had a copy of the driver installed, but so far I haven't very successful installing HP drivers unless one has the printer connected.
Nevertheless, unless otherwise instructed, 99% of printing using properly made printer profiles for the paper, ink, and printer, require that the Driver be set to No Color Management.
The phrasing is not standard, so it might be called something else such as Color Matching OFF, APPLICATION MANAGED COLOR, No Color Correction... and perhaps others.

I will make another attempt to get a Z3100 driver installed.
In the meantime, maybe Ernst will check in here and he is expert at the HP printers and especially the Z series.

Fred

I finally squeaked by and got the Z3100 driver installed.
The Z 3200 would not go through, but you have a Z3100.

Mind you, I don't have a manual nor the printer, but the driver seems very straightforward.
I would DISABLE ICM in the ADVANCED TAB, and then in the color Tab, place a dot in APPLICATION MANAGED COLOR.
See screen snaps.

Fred

Done.

Just as you have shown. 

Also, just profiled the Monitor.
Also, just calibrated and Profiled the Paper in the Printer.

Color is off - too Red, too Saturated.

Only option is that the Color Management Engine in Qimage U is at fault.

Printing with - Printer Managed Colors works well.  Can't figure out why.  Image is an Adobe RGB Tiff from Lightroom.

Thoughts?

Not an option if your profile is correct and made for the printer and the paper.

Any chance I can get the image here?

wathree.ssz@verizon.net.

If the image is too large for emailing, you can send it for free using https://www.wetransfer.com/ to the same email address.

Fred

One more test!
Please set the driver to the color setting as shown in *this* screen snap, and Qimage Prtr profile to the setting shown in this post too.
Make a print and tell me if you color problem is still there, or gone?

Fred

Greetings! I have been running an HP Z3100 44" in a production shop with Qimage since 2007 and I'm getting excellent results. Perhaps I can help.

With some reds on some papers, including canvas, there can be a noticeable and (as far as I know) unavoidable gamut problem. However, if you can get better results with different settings, then it is likely not the paper gamut. A couple of questions:

- what firmware are you running in the printer? There was a known issue with reds a while ago that was addressed with a firmware upgrade. The latest is 7.0.0.3-7.

- what driver version are you running? A new driver v1.14.1.9 was released last summer and I would recommend installing it if you haven't already.

- are you on a Mac or Windows system?

- did you use the HP Utility Color Center with both the calibration and profile steps with the built-in spectrophotometer? Or some other method (e.g. Colormunki)?

- what paper are you printing on?

- have you done soft-proofing in Photoshop to select the best rendering intent? This can make a dramatic difference. In many cases P (Perceptual) is not the best choice, whereas RC (Relative Colorimetric) is. However, it is good to soft proof both, with and without BPC (Black Point Compensation) to determine visually which combination introduces the least noticeable shifts.

I have posted two screenshots from my system (Windows 7 64-bit)

Cheers!

Brad

Using HP paper?   Image has a color space assigned and which one?
Try the following; print with the HP driver's CM on with the media preset for that paper and a test image either sRGB or AdobeRGB assigned, the driver's CM set to sRGB or AdobeRGB depending on what is assigned to the image, Qimage's CM set to ¨Let Printer/Driver manage color".
The other print with the HP driver set to "Let application manage color", same test image with the same color space assigned as above, media preset as above, Qimage CM with the HP OEM profile for that paper not a custom profile.
That should deliver about equal color in the two prints, the driver CM is based on internal LUTs for that paper but the OEM profile used in Qimage should be close. Rendering choice in Qimage CM could make some difference.
If no color space is assigned to the image or another color space than the driver CM expects you will get different results.

Z drivers have a habit of falling back to the default settings so make absolutely sure that your CM settings in the driver stick by going back to the driver settings after the print is made and see that they are as intended. There is also the Advanced Color setting in the driver, all settings on zero there?

The use of a custom profile comes after you made similar prints with the conditions sketched above.


met vriendelijke groeten, Ernst
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

I definitely agree that if there is no colour space in the image file, then different printing setups (driver vs. application) can and often do produce vastly different colour results. It is recommended to have warnings enable in Photoshop so that when you open the image you'll be told that it doesn't have a colour space assigned. If that is the case, assign one (either sRGB or Adobe 1998), save the file (ensure the colour space tag option is checked when saving), then try printing it. You should at least get more predictably consistent results that way.

Cheers!

Brad

Thank you Brad and Ernst.
I knew that Ernst has Z series, but Brad too? Wonderful.
I was hoping someone would pop who actually ran the printer.

Fred

Fred,

I am out of town at the moment and will send the image next week when back.  The image is 24 mpixel and a TIFF with embedded RGB soft proofed in Photoshop prior to output.  When set as you show above (what I set to for the printer to manage the colors) everything prints perfectly, in fact I am very impressed with the match.

It is wen I try to have QImage handle the color that it all falls apart.

I have used the Color Management in order to have the printer calibrate and profile the paper (Calumet Briliant Satin) and have the new Eye One Pro to calibrate the monitor.  I've been doing this for years, so I know when it is spot on, and it is.

What I am trying to determine is why the printer can get it right, but Qimage can not.

Thanks

I will be happy to see the image and have a look at the colorspace if any.
But more important, I think we are closing in on the cause.

What the evidence points to is not a defect in Qimage, but a defect in your profile. (Assuming your driver settings are correct) (And as Ernst said, recheck the driver before printing as sometimes, the HP drivers decide to lose a setting)
First let me point out a couple of anomalies.
As long as you mentioned the paper you were using, I checked the paper from Calumet Brilliant Satin.
They make printer profiles for their paper, but none for your printer.... see screen snap!

Next we ask why no profile for your printer which is a class above the B series, so I am told?
Some papers defy profiling for certain inks or drivers....
Next you seem to put a lot of reliance on the profile that the printer is making for the paper.
It may not be working right!

We know the printer is OK, and we know Qimage is sending the proper data. The prints are good WITHOUT THE PROFILE.

I don't know if you have any HP PAPER that comes with the HP profile  (Some HP Glossy or HP matte; any paper that matches a premade HP profile that came with the printer driver) but if youi can make a test print with a proper match of paper and an HP profile made for the paper, I would bet you get an excellent print!

That would convince you that the Calumet Satin profile you made is not working.

Fred

Peter, is it possible that the profiles on the printer are out of sync with the profiles on the computer? Because when you choose Prtr ICC = "Let printer/driver manage color" in QU, you're using the profile on the printer. When you choose Prtr ICC = "some profile" in QU, you're using the copy of the profile on the system. It can't hurt to re-sync them using the HP Print Utility Color Center.

The other issue could be the rendering intent chosen in QU. Which HP printer driver "ICM Intent" option gives you the best result when you "Let printer/driver manage color"?

When you get back and if you are willing, I would be willing to test print your file on my Z3100 with QU using one of my similar papers and we can compare steps taken and settings used. I use Dropbox and I also have a submit link on my web site.

Cheers!

Brad

Brad,
Excellent idea!!
You actually have the printer...
That's what I wanted him to do.... print to a known HP paper with a known HP matching profile.

Super:
You read that he is using an off brand paper for which there is no profile supplied for that specialized paper.

Fred

I will happily send the image next week when I return.

However, I have synced the profiles.  It is also a new profile so I am sure it is synced and properly showing in QU with the correct build date.  I have checked the settings in the HP driver and they seem to have stuck.

I am also pretty sure i am using the latest driver and FW, but can not confirm until next week.  Please send me a link to both if you get a chance and i will check when i get back.

Thanks for all the assist.  Appreciate it.

Peters has created a custom profile using the built-in spectrophotometer of the Z3100. I'm running a couple of dozen Z3100-generated custom profiles myself with awesome (according to my clients) results. So this is not an inherent problem unless the medium is one of a very few that are so brightened with OBAs or fluorescing agents that the built-in system can't get a proper reading. I have encountered only one such case in my time and it was one of HP's own canvasses! HP posted a bulletin and I found myself another canvas!

Even so, one would expect even a bad profile to produce the same results from different workflows if they use the  same profile with proper and consistent settings. I think that's the issue here.

My contact info and a submission link are on my website at http://www.turningpointarts.com (http://www.turningpointarts.com). If you prefer to use Dropbox, just drop me an email and I'll invite to a shared folder.

Cheers!

Brad

Brad,

My understanding is that it is not that simple, when printer driver CM is used the color management is based on internal LUTs (so not the OEM ICC printer profile) and the driver expects either sRGB or AdobeRGB assigned images. There is not a complete color engine in that driver that will use the OEM profile and can work with more than the two color spaces mentioned. Nevertheless the print results are very close to what Qimage will do with an OEM driver and the same image.

I agree that the integrated spectrometer and Color Center do a decent job in profile creation though not better than the HP profiles for HP media are (they are created with other software). The optional APS software does a better job on the Z3100 in my opinion. On the Z3200 both are equal and better than Color Center on the Z3100, HP improved its own profile creator for the Z3200 while little has been done by X-Rite on APS.

I agree with Fred that Peters should first make prints in as much an identical path as possible with both driver CM and Qimage CM before blaming Qimage CM.


met vriendelijke groeten, Ernst
Shareware too:
330+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

Brad,
As I understand it, Peter is getting decent prints using Let Printer manage color.
That uses a wide gamut color feed from Qimage and lets the DRIVER invoke whatever setting you use, ICM or Color, Argb or sRGB.
It makes no difference.
If you use the same settings in Photo Shop as in Qimage, same driver settings and paper selection etc, you will get the same color either way.
I have been doing this a long time. I have had to prove, over and over again that there is no such thing as PS prints fine and Qimage with a color shift if all other parameters are set the same.
People just refuse to believe that Photo Shop can fool them. It does. With color  space, with sizing, with saving.... 
People refuse to believe that a profile could be less than perfect too, sometimes corrupt.

We will find the discrepancy, but you have to maintain logic as you make tests.
If I can see the image and ascertain the profile embedded in the image, and also have Peter make a few other prints using HP paper and an HP matching printer profile, we will have the answer.
We might have a bad profile. We might have some kind of brand of paper that doesn't like his inks. We might have a situation where he made a profile with the driver set to color, and prints using the profile with the driver set to Printer OFF.
I don't know the answer.
I do know that all settings equal, and same profile in use, Photo Shop and Qimage will produce the same color.

Brad,
MOST OF THE TIME, (I capitalized for effect to get your attention) I have found in an overwhelming number of issues like this, the person trying to get the prints to come out right, ARE NOT ACTUALLY PRINTING. THEY ARE USING PRINT PREVIEW instead.

That is worthless. As Terry said earlier, there is no color management incorporated in the preview.
I have also been stymied by being told about misguided printer profiles and prints, (Being told: "I get better color using the XYZ profile from RED River paper on my Canon paper than the Canon profile")
After probing I find out that the person with the question found it wasting too much ink or paper to actually make prints so he found it more expedient to use SOFT PROOF instead.
This may be cheaper but not definitive and almost of no value.
So forgive me if I dig my heels in. Experience talking!
Brad, I get, on average, a dozen phone calls a week with similar questions.... I make them print real prints, or don't waste my time!
The results of that are eye openers.... Problems are solved.

Fred

Well let's hope that's not the case with Peters!  :)

I'm back.

According to the Printer the FW is TR12-TR_7.0.0.3  OS TR_2.1.2

Driver is version 61.101.366.41 (from Color Control Panel)  But this is a 2010 driver (HPL3).

Are there later versions of either I should download?

Thanks

The driver I have is the same version, it is called PCL3 to my knowledge.
Firmware is also the same:  TR. 12-TR_7.0.0.3

I am not aware of a later version and it has not been that long ago that I upgraded the firmware.


met vriendelijke groeten, Ernst
Shareware too:
330+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

ALRIGHT!

It is now printing perfectly with the program handling the color management.

What was the difference you ask?

I was using a Calumet Brilliant Luster paper.  No matter how many times I profiled it, the color management in the program would oversaturate the colors.  The Printer Managed was MUCH better.

Ok, so I figured why fight it.  Went and bought a roll of HP Professional Photo Satin.  Perfect.  In fact the color match when using QImage is MUCH better than the Printer Managed.  Now I know, don't experiment.

So, I am asking this group now,

"Which Satin Papers?"
"Which Matte Papers?"

I really did like the look of the Calumet, it was also about $100 per 100ft. x 24" roll, which was nice.

Please let me know what you are successfully using.

Also, began to get the drive belt breakup issue with the Z3100.  Found a REAL CHEAP solution there and simply put some double-stick tap inside the printer on the metal plate which runs under the drive belt.  Now the droppings, when there are any, are caught by the plate and don't end up on the print.  Eventually I will have to get a new drive belt, but not looking forward to doing the service myself (not going to spend $1,500 that HP wants).

Thanks

Peter

Peter,

There are no Calumet papers in my SpectrumViz spectral plot collection but given the "Brilliant" I expect it is one of the papers with a very high load of Optical Brightening Agents. Your Z3100 spectrometer does not measure into UV light spectrally and extrapolates the numbers below 400 NM from the numbers between 400 and 700 NM. That may work nicely with low OBA or non OBA papers but not with high OBA paper content. The OBA compensation or what is more often called Fluorescence Whitening Agent compensation build in the profile creator is more suited for the average OBA content.

I am sorry to say that while HP Prof Photo Satin is a very nice paper to print on it also shifts its paper white quite fast in time. The OBA content is high and the OBA degrades fast due to light or oxygen. See www.Aardenburg-Imaging.com tests. In general avoid high OBA content papers and check both the OBA content in my SpectrumViz application and the paper white fading at Aardenburg. That is if you are interested in long term stability of the prints. Aardenburg has a very good test for the Canon Heavyweight Satin Photographic RC paper, Called Glacier Photo Quality in Europe. There are more there that tested quite good but not as good, HP and Epson brands. Some Lasal RC qualities show a similar spectral plot to the Canon ones so may be related but are not tested at Aardenburg.

If you check the spectral plots of matte papers against the HP paper spectral plots in the same matte category you will get an idea what should fit. HP's HM Smooth Fine Art is similar to HM Photorag, the Texture version is similar to HM William Turner. HP's Matte Litho-Realistic is a nice, cheap, heavy paper in itself. not rag though. Tests for all at Aardenburg. There are good and lasting papers ranging from the cheap Red River Aurora Natural rag (dual sided) to expensive Canson qualities with low OBA content that will print good on the HP Z3100 and will profile nicely too. The Aurora will not have the gamut of the Canson Rag Photographique and it has a slightly warmer white but at that price one shouldn't complain.

I will measure the spectral reflectance of papers that are not yet in my SpectrumViz catalog, preferably from two A4 sheets as I also want to measure Dmax from one print too later on. So I can measure the Calumet paper or more types you have that are not in my list. Mail address is at my site.


met vriendelijke groeten, Ernst
Shareware too:
330+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm