Print setup for a new fine art paper
Print setup for a new fine art paper
Setup and profiling a new paper for colour and Black & White printing
Keith gets to test a lot of new papers, but these usually come directly from suppliers. This time it’s a box of unbranded paper.
This short article looks at how Keith profiles and uses the paper for his Canon large format printer, although the process would be virtually the same with a good desktop printer.
Correction: May 2020 – The B&W print section has been corrected to the proper sequence of profile adjustments, after an error was introduced in a 2017 edit.
A new ‘mystery’ paper
I’ve a lot of boxes and rolls of paper in our print room, many from testing printers over the years, but only some I’ll regularly use for printing my work.
In this instance I’ve several boxes of A3+ (13″ x 19″) matte art paper that are all identical, but unbranded.
It so happens that there are not that many companies producing high quality art papers, and a detailed look at this one suggests that it’s a (very) smooth coated art paper – maybe a cotton rag paper, whilst the slightly yellowish tone and lack of bright glow from my blue laser pointer suggests that it has relatively little optical brightening agent.
This photo shows a sheet of bright white photo paper on top of two sheets of the ‘mystery’ paper. Quickly scanning the beam of a a blue laser pointer shows how the OBA in the top sheet is glowing much more.
As ever, take care with lasers. The deep blue colour of this one means that we don’t see it as bright as it really is (in terms of power output)
Using the damp finger test, shows that it’s single sided.
Weighing a sheet of it (241mm x 329mm) gives a weight of around 21gm.
Dividing 1 square metre by the area of the sample and multiplying this by 21 gives a rough weight of 265gm/m2 [gsm].
That’s not a weight I immediately know, so a search for ‘fine art paper 265’ comes up with an HP branded Hahnemuhle smooth fine art paper. My digital micrometer gives a thickness around 0.42mm (it’s a nice stiff paper).
A bit of hunting round suggests that this paper might be discontinued, but given I’ve several hundred sheets, I’m strongly inclined to look at it for some of our print work. It seems like a lighter weight version of the Innova IFA11 Smooth Cotton Natural White that I regularly use (and have several rolls of in different widths).
For desktop printers, it’s a matter of picking a media setting that works best – drivers normally have a range of settings and my choice is first to go with a similar type (lustre, Fine Art etc.) and then maybe test similar options if there is no really good match. I might make an ICC profile for several settings and evaluate test prints, but that’s really something to experiment with when I’m using someone else’s printer, paper and ink for a review ;-)
Media settings – I’ve an article about choosing media settings I wrote, some time ago , about using third party inks and paper in my old Epson 1160. There is also a page with media test images and links to their sources.
For our Canon iPF 8300 printer, I’ll usually create a custom setting for any paper I’m making. This involves running the Canon ‘Media Configuration Tool’ software and letting the printer do a few test prints.
The procedure is outlined in (much) more detail in part two my recent review of the Canon iPF 6400 along with many other aspects of profiling and setup. See also my review of the Epson P5000 for Epson custom media settings.
I end up with a setting that I’ve decided to call ‘Han SFA 265 fahw’ – this reminds me what (I’m fairly sure) it is and that the setting is based on the Canon standard ‘Fine Art Heavyweight Photo’ setting.
If I was less sure of the paper, I might have made test prints using different media settings and see how they responded to different ink amounts.
Note that I base the custom setting on an existing setting rather than one of the ‘Special’ settings, since they prohibit use of the printer Monochrome printing setup. Something I know works well on similar papers.
Here are the test sheets I used in setting up printing for this paper.
The large one is for creating a custom ICC profile using X-Rite’s i1 Profiler software, the smallest one is for linearising the printer’s black and white print mode, whilst the other two were created by the printer during the initial media configuration stage.
The printer profiling target I use is a custom 2938 patch one I created for reading with the X-Rite i1 iSis scanning spectrophotometer.
I could create targets and measure them by hand with my i1Pro 2 spectrophotometer (or i1iO motorised unit), but loading the paper into the i1iSis is so easy, that I almost always use it (remembering to clear the desk space behind it).
Here’s the i1Profiler software during the scanning process.
The i1iSis scans back and forth. Its light source is looking decidedly blue here, since the scene is lit by tungsten lighting (no CFL or LED lighting in any of our working areas).
With little OBA in the paper, I’m not using the dual scan mode with UV – this sort of paper profiles perfectly well with a single M2 scan.
I don’t profile every day (it’s not a service I offer via our business) but the results are good enough that I almost never use paper manufacturers profiles.
If you can find a profile though, then by all means try it, since I’m well aware that the kit you see above is rather expensive for casual use. The cheaper ColorMunki and its successor, the i1Studio are aimed at a broader market.
I’ve included the gamut comparison shot, captured from i1Profiler just after creating my custom ICC profile to show how much smaller the gamut (range of colours) is with a matt art paper (blue shape) compared to my ‘everyday’ lustre paper (red).
I know from experience that the number of colour images that look great on such paper is fewer than a lustre paper.
As ever it’s a matter of trusting your own taste and experimenting.
There is more about such papers in the review of Hahnemuhle 310gsm Photo Rag Bright White I carried out when testing the Canon iPF6400.
There are two other observations I’d make about the blue shape above.
It’s smooth, no ‘holes’ or dents, which suggest that the profiling process has gone well and the paper is a good match for the printer.
I often get asked for suggestions about learning more about the nuts and bolts of Colour Management.
My usual suggestion is Bruce Fraser's Real World Colour Management. My own copy is well thumbed. It's my first port of call if I'm asked a question and I feel I don't quite understand an issue well enough to be absolutely sure of an answer.
Check latest price/availability from AmzonRWCM 2nd Edition RWCM
See some other books Keith has on the shelf, on our Books Page
Note how its base doesn’t reach down to the lines.
This reminds me that the blacks on this paper are nowhere as black as the lustre paper, and that I should be sure to use black point compensation (BPC), if available, when printing.
I’ll come back to this in the B&W section below, but it’s something that I know can cause problems when people who are used to printing on gloss/lustre papers first try out matt art papers and find they get very washed out looking prints.
With the colour profile created, I have a number of test images I’ll print, but actually, looking at that target print gives me a very good feel for how the paper handles colour.
I’ve printed a 51-step target using the Monochrome print mode of my printer driver. My aim is to be able to linearise the print output using a QTR profile (note that the target is also included in our specialist B&W test image shown at the foot of this article).
The test image is always printed as you would a normal B&W photo. It should never be printed as a profiling target.
Depending on the printer, the B&W print mode can produce a superior quality black and white output – this is available for many of the smaller Epson and Canon printers I’ve reviewed, as well as larger ones. It’s something I always try and cover in all our different printer reviews.
I’ve printed the image towards the top of an A3+ sheet and chopped it off to put in the paper holder for manual scanning with my i1Pro2 spectrophotometer (right).
I’ve looked at using both X-Rite’s free ColorPort software for this, and using i1Profiler. Both produce the measurements needed. The process can use the X-Rite ColorMunki for measurements too.
There are more details in these articles
- ColorPort B/W
- i1Profiler B/W
- ColorMunki B/W
- i1iSis B/W
- SpyderPrint B/W
- Using a scanner for B&W adjustment
I’m using the i1Pro 2, but any measurement device that can give you a set of Lab measurements will do here (even if you have to manually enter them into a data file).
I scan across the target and the print is measured one strip at a time.
My target is available in ColorPort – here it is waiting for me to connect my i1Pro2
There is a lot of data that could be exported from these measurements, but all I need here is a Tab delimited text file of ‘Lab’ values.
The data file is dropped onto the QTR-Create-ICC-RGB application in the QuadToneRIP package.
This software is actually intended as a specialist B&W (Epson) printing driver, but I’m just using the profiling part.
The program produces this file, showing how the density of printed patches change as you step in 2% increments from 0% (white) to 100% (black). You could use a 21 step version of the test print, which has 5% steps.
Once again, there is more about this in the printer reviews and articles…
The relatively smooth ‘a’ and ‘b’ lines tell me that there are few variations in the colour of the output, and there is no sign of the ‘b’ line veering off at the top (a good indicator of OBA’s in the paper if it does). My laser pointer test is confirmed.
The ‘L’ line is straight, so we have a smooth progression of tones, except above 94% where the line flattens off a lot.
This tells me that any detail in the darkest 10% of my print output is liable to be crunched up. So, if I edit an image that has a full range of tones from black to white, I’ll get crunched shadows in my print.
Note that if you look at the 50% row, the ‘L’ is somewhat to the right of the midline. This suggests that a general darkening of midones may be needed in addition to the deep shadows opening up.
The program produces a specialised ICC profile that I can use as an ‘adjustment curve’ to correct this. It’s important to note that even though it’s a printer profile, I’m not using it for that, just as an adjustment curve to linearise print output (if needed).
The shape of the curve is better shown in this view, taken from looking at the profile in Apple’s ColorSync utility, showing how to get a more linear output, I need to lighten shadows and slightly darken midtones (mouse over the image to see more).
Note too that it only shows just the blue channel in this screen grab, but I’m using the profile for monochrome adjustments so it’s not important here.
One other thing is that the numbers I’ve added to the image above match Luminance, from zero (black) to 100 (white). Whenever looking at such data do check what the numbers mean, since you may well see 0% referring to white (0% black ink) and 100% as solid black, as you’d get in CMYK values, or even that QTR graph above.
I’m a photographer and almost never use CMYK, so I live in the RGB world where higher numbers are nearer to white. There is no ‘correct’ way of looking at this, so I remember to double check such scales wherever I see them.
Perhaps it’s clearer if I show some examples with a real print, based on one of my photos of the beach at Shingle Street in Suffolk (this is from the negative I used as an example in my review of the Epson V850 film scanner).
Using a linearising profile [Corrected 2020]
I use the linearising profile to ‘correct’ the image before printing.
In the example below, I take my monochrome (greyscale – profile Grey gamma 2.2) print-ready image, convert to the QTR profile and then assign the original profile for printing. This works similarly if my B&W image was still in a colour space, such as Adobe 98, where I’d convert to the QTR profile and Assign back to Adobe 98.
For the printing I just send it to my printer driver and use the B&W print mode – such modes often include the capacity to introduce toning/tinting if you desire. In this instance I used the Canon Photoshop print plugin for my iPF8300. For Epson it would be the ABW print mode.
For the conversion, there is no change to what you see. The data in the image is being changed, but since we are now using the QTR profile, our colour management system is doing its job and the visual appearance is not changing.
I’m using the perceptual rendering intent since it already incorporates black point correction (BPC). The checkbox would do this for Relative Colorimetric – it does no harm to leave it checked.
BPC in this case is matching the black of our paper to the black in our image file – if this wasn’t applied, blacks in the image file would be mapped to dark greys.
I now need to assign the original colour profile to the image. Since the QTR profile is actually a colour one (RGB three channels), I can use my current working colour profile. This is Adobe98, which is a gamma 2.2 space, just like the original B&W image.
For printing there’s no difference between my B&W image in RGB format or greyscale – just the RGB files are bigger. The Canon and Epson B&W print modes expect files in a gamma 2.2 space
At this point I have my original image with its pixel values changed by the conversion, made visible by assigning the original profile.
The QTR profile is only being used here as a specialist adjustment curve not a printer profile
I find that thinking of it this way makes its use potentially less confusing. This ‘curve’ is actually the one shown as the blue line in the look at the profile earlier. As printers and drivers have improved, the amount of adjustment needed for linearising has reduced, meaning that I often  find a simple adjustment curve achieves smoother results. See my specialist test print version adapted for measurement with the SpyderPrint article for more of a discussion about this.
I now have an image that has had the ‘correction’ of the profile applied.
Here’s the original image…
…and here is the corrected one to send to the printer (Convert then Assign)
The slight darkening of the mid-tones is most obvious – the changes in the deeper black areas are less easy to see on the screen.
Click on the images to see larger versions.
The ‘before’ is what I’d edited and how I wanted the tonality distributed in the image.
The ‘after’ version has been altered by the profile/curve to correct for unevenness in the way the printer reproduces this.
It may be difficult to see on your screen, but the dark corner area is spread out more and the mid tones darkened to compensate for the lighter mid tones in printing.
The annotated profile curve from earlier shows the shape of curve needed if I’d carried out the correction with a simple Photoshop adjustment curve.
It’s a relatively subtle adjustment. If you find the curve is applying a lot of adjustment, take it as a hint that either you’ve made an error in the process somewhere, or you’ve the incorrect media setting for printing. However it’s always possible that the paper just isn’t a good match to the inks/printer you’ve got and no amount of adjustment is going to give acceptable results – it happens.
Remembering the process – My own way is: it’s Convert to the Correction profile and then Assign back to working profile.
However, it’s important to realise that a print is never what you see on a screen, so depending on paper and viewing conditions, I may decide to lighten or darken the print – the simplest way is with a curve, pushing the mid point up or down.
You might wonder why I go to all the trouble of linearising the output, if I sometimes go back to editing after a print?
Well, it’s about predictability. With the linearisation in place the printer isn’t going to crunch shadows or darken midtones, that’s my choice…
Here’s the final print coming out of the printer
Here’s a detail of the top corner – the part where the ‘shadow’ crunching would be very evident (and yes I did use a light orange filter when shooting the film back in 1986).
This article updated and corrected in May 2020 – See the SpyderPrint article for more about making/using curves/profiles in this way.
This (free for non-commercial use) test image was created specifically to test black and white printing.
There is a detailed article explaining all its features, including the step ramp for linearising.
Just remember, that it is only ever printed as a normal B&W photo – never as a profiling target.
Never miss a new article or review - Sign up for our Newsletter (2-4 a month max.)
Enjoyed this article?
More print related information
For information about other printers, paper reviews and profiling (colour management) see the Printing section of the main Articles and Reviews page, or use the search box at the top of any page. There are also specific index pages for any articles connected with the following topics:
- Digital Black and White
- Tutorials and 'How to' articles
- Colour Management
- Printer test images
- Why do your prints look wrong?
More of Keith's articles/reviews (Google's picks to match this page)