In the previous post, I mentioned that having multiple gray ink tanks can potentially help us improve the quality of our black and white prints. Long-time printers know that Epson have the R-series of great printers (R1900, R2000, etc) that have been touted as the best desktop inkjets in the world. There are specially made monochrome inks like Piezography inks (which I first read about on the Northlight Images website*) that one can use to replace the color inksets for perfectly neutral prints. A custom software, called Quadtone Rip can also be used to prepare excellent ICC profiles. Finally, custom drivers have been written to manage the amounts of inks of different density to spray onto each dot, thereby achieving fine gradations of gray. As a Canon owner, no such luck for me. By the way, Epson printers are expensive!
Printer Calibration with Spyder3Print
Nonetheless, all these sounded great to me, so a couple of years ago, I decided that it will be a good idea to try something out. I purchased a printer calibration device, the Datacolor Spyder3Print SR to calibrate my Canon iP4600 (color printer). I found that the images looked great after calibration, so I was fairly satisfied with the purchase. It works similarly to the screen calibrator mentioned in the last post (and it’s made by the same company). Basically, colored squares are printed out and scanned using the spectrometer. The colors are then compared to stored values and corrected.
In theory, scanning can be done very quickly. The scanning frame guides the spectrometer as you pull it smoothly across a row. Repeat till finished. Simple enough. However, the execution is terrible – the flimsy and warped frame will not sit flush against the printout, and the rail has significant friction so instead of a smooth motion it is more like stick-and-slip (which the program cannot handle!!!!!!). The software is poorly written, and extremely slow. For more than 200 USD, I really expected more than this. In the end I had to press the stupid button at each square, and there were hundreds of them. Urgh!! Nonetheless, the printer works much better after calibration, so I guess that is alright.
Unfortunately, after calibration it still didn’t give great black and white images, as they often contain tints. The grays looked muddy and brownish, almost, instead of neutral. I became convinced that the profiler is able to map the color very well, and it should be possible to do the same with gray inks. After all, my primary goal was to get good monochrome printouts. I knew that the iP4600 I had already had color ink run through the print head, so I cannot put gray inks in there without cleaning it (I have heard horror stories about cleaning print heads, so I decided against it). Besides, I needed to print color images sometimes (mainly class notes), so when I saw a deal for an iP4700 for 60 bucks, I jumped at it. The ink that came with the printer would probably be worth that much, so the printer was pretty much free!
To select the proper inkset, I first had to consider the number of tanks that I had. There was no pre-packaged inkset that I could use, so I had to go out and try a few different dye-based inks. I know that the Canon Pixma MG6000 series of printers had a grey dye ink, so I went to Inksupply.com and got 2 bottles of that. I read online that some old HP printers had 3 different gray inks (dark, medium and light) in a single cartridge, and I got the light and dark gray versions (I cannot recall why I did not get he medium gray from them too). By brushing these on a piece of paper, I knew that in decreasing density, we have dark gray, light black, and finally light gray. I also purchased empty, refillable ink cartridges for this purpose.
So, being the genius that I am, I decided that if I used only gray inks, and used the printer profiler to create a mapping profile, then I should be able to force the printer to print in perfect black and white! Since I will be performing a map, I figured I didn’t need to care about where the inks go. All I had to do is to print many squares of every shade, and do a straight mapping!
But to help the system achieve the colors more easily, I printed a nozzle check pattern and converted it to black and white. That way, I can better assign the density to the cartridges. I believe I assigned the dark gray to cyan, light black to magenta, and light gray to yellow.
After performing the calibration, the software started giving me errors. I pushed through, reasoning that the software was just confused about the colors. I was right; I just didn’t know how right I was. I printed an image after ‘calibration’ and it was basically black. There were no shades of gray at all. Since gray was generated by stacking C, Y, M and K, what I ended up with was patches of photo black topped off with dark gray, light black, and light gray – the density was so high that no shades were visible! What I had neglected to take note of is that the calibrator was expecting color squares, not gray ones. The mismatch in the values was so great that it wasn’t really sure what it was doing. Therefore, the printer was simply unable to get the colors right, no matter how I tried to adjust it. I suppose you cannot use color management to perform color conversion after all.
I wasted 20-30 sheets of photo paper and lots of ink in this futile exercise. I also wasted a perfectly good printer, and was honestly pretty upset. Thankfully, I decided to try one last thing out before quitting, which I shall talk more about next time, in the third and concluding part of this journey!
*The Northlight Images website is a great resource for printing your own images. In particular, this was immensely helpful for what I described in this post.