I recently printed a test print from http://www.outbackprint.com/printinginsights/pi048/essay.html. I printed it on an Epson R1800 to a half-sheet (9.5" x 13") of Super B Premium Semi Gloss. I used the Photo RPM setting (5760 x 1440 dpi) at a -pixels per inch- setting of 581 PPI. The image dimensions are 5520 x 6960.
I was astounded by the tiny details that resulted. They were too small to see clearly, even when I took off my reading glasses and put myself into macro mode. Even using a 12x loupe, the details were hard to see. So I taped the print on a door, and put my Canon SX1 on a tripod in super macro mode, which allows close focus right up to the objective lens. I needed to light the print with a hand-held white LED flashlight.
The crops shown below, in the right pane, (except the ruler!) are just 1/4" wide. The corresponding original image is shown in the left pane, at various 'magnifications' as noted in the lower-left 'zoom' tray.
Two things surprised me - even at 581 ppi, I see single-pixel details, and the black ink droplets are now tiny. In the past, black droplets were clearly visible by eye, and were much larger than the color droplets. I guess Epson changed the printer driver to allow this? You might also note that the print is more 'photographic' than the original image. I assume this is an effect of droplet dithering. Note also the number '10' above the largest set of black bars. That's not faked, that's the way it printed! When I captured the screen image, I actually had to go back and check to see that the left pane contained the original, downloaded image. I thought I had downsized it by mistake!
The color seems quite off. Otherwise pretty nice detail. Does that level of detail really matter though? I mean if you couldn't even see it at 12x magnification, does it make the print any better than printing it with normal mode considerably quicker?
ModifiedPhoto wrote:
The color seems quite off. Otherwise pretty nice detail. Does that level of detail really matter though? I mean if you couldn't even see it at 12x magnification, does it make the print any better than printing it with normal mode considerably quicker?
As explained, I shot with a white LED flashlight providing light. It's not surprizing the colors don't match.
As for the level of detail, it's typical to try to print at 240 - 300 ppi for photos. But you'd be laughed out of a job if you tried to put a 240 dpi laser printer on your bosses' desk. 2400 dpi lasers are the norm. Something to think about...
I know that -I- can see the difference between 300 ppi and 600 ppi when viewing my prints at 18". It's not a huge difference, but it's obvious side-by-side. Now if I could just afford an 89 MP cam!
Thank you for your work, I usually recommend people to do this to find the optimal workflow with their own printers... :-)
One thing that I'd like to comment though - if you do the same, but with a shift in hue instead (like the res-chart, but in orange-red or blue-green in stead of b/w) the resolution drops to half or one third of the b/w example... But it's still impressive how far a "simple" home workstation printer has come the last ten years!
And Epson has come quite far with their built in RIP-modules since I worked with optimizing soft-rips for them. As you say, the ink-spatter patterns looks very "natural" when viewed up close - not at all like the geometrical patterned raster angles of yesteryear.
Hope everyone lived through yesterday, and may you all have a good 2010!
cgardner wrote:
When output resolution is set in Photoshop at say 300 ppi, that indicates how many pixels in EACH COLOR will be printed per inch. Printer makers express resolution in terms of total dots of all colors. So 300ppi in Photoshop x 8 colors becomes 2400 "dpi" on the printer.
-------------------
I disagree. Each 'color pixel' will index to the same spot on paper. A yellow pixel overlays a green one. Epson's diffusion might blur the edges of each color pixel. The Epson R1800 printer indexes 5760 x 1440 dpi at it's highest resolution. The images ppi setting in the image file doesn't control the ink droplet spray pattern, the Print Quality setting in the printer driver does.
--------------------
cgardner wrote:
Ink jet and lasers printers use different imaging methods so you assumptions based on the numerical comparison isn't correct.
--------------------
The point I was making was that your eye, my eye, her eye can distinguish the difference between 1200dpi and 2400dpi laser text, at a normal reading distance. And I can clearly see the difference between 300ppi and 600ppi color prints on my R1800.
--------------------
cgardner wrote:
So put away the magnifier and find out which printer settings make the your 4 x 6 - 8x10 prints look best overall perceptually from reading distance...
--------------------
I have! (sigh)
Quote: "I know that -I- can see the difference between 300 ppi and 600 ppi when viewing my prints at 18". It's not a huge difference, but it's obvious side-by-side. Now if I could just afford an 89 MP cam!"
In a test I conducted with the help of my good friend Bill Glickman, printing a target he designed, to determine the actual on print resolution of an Epson 9900, I was able to resolve on Pictorico gloss transparency media 9-10 line pairs per millimeter. We were able to see increased real print resolution up to 1440 real image dpi, using the highest resolution settings of the printer. These number are contrary to common Epson "wisdom" and were surprising to both Bill and myself. For Bill, who is a certified stereo viewer nut (and I mean this in the nicest of all possible terms - anyone who stops by his house in Mesquite, Nv will agree) images are indeed viewed not by eye, but with a very high quality 6X magnifier optical set, and we were testing to see if inkjets had evolved enough in their inherent resolution to be considered for an application that would have previously been unacceptable.
I haven't tested all the different media I print on, but I do know that what we think we see and what we actually perceive with our eyes can be two different things. Increased on print detail can have an almost unexplained subtle effect on how we feel about one print compared to another, and there is also the fact that some people just see more detail than others. Just last week, when at the movie theater seeing Crazy Heart (fantastic in every way btw) the commercials beforehand were digitally projected. I could easily make out the scanning lines from the back of the theater while by girlfriend could not see them at all.
Here is a good link that talks about the resolution of the human eye in fairly simple terms and how it relates to printing at higher dpi (ppi). In my real world experiments with my students about half the class could see a difference in 8x10's printed at 300 dpi vs. 360 dpi, when viewed at standard distances. Now that I'm older, 53, I notice that my eyes are not resolving the details they once did and I have to remove my glasses and look at prints much closer to see detail I used to see at more normal viewing distances. So, the bottom line is some will see a difference and some won't.
Jeff Donald wrote:
Here is a good link that talks about the resolution of the human eye in fairly simple terms and how it relates to printing at higher dpi (ppi). In my real world experiments with my students about half the class could see a difference in 8x10's printed at 300 dpi vs. 360 dpi, when viewed at standard distances. Now that I'm older, 53, I notice that my eyes are not resolving the details they once did and I have to remove my glasses and look at prints much closer to see detail I used to see at more normal viewing distances. So, the bottom line is some will see a difference and some won't.
skid00skid00 wrote:
cgardner wrote:
When output resolution is set in Photoshop at say 300 ppi, that indicates how many pixels in EACH COLOR will be printed per inch. Printer makers express resolution in terms of total dots of all colors. So 300ppi in Photoshop x 8 colors becomes 2400 "dpi" on the printer.
-------------------
I disagree. Each 'color pixel' will index to the same spot on paper. A yellow pixel overlays a green one. Epson's diffusion might blur the edges of each color pixel. The Epson R1800 printer indexes 5760 x 1440 dpi at it's highest resolution. The images ppi setting in the image file doesn't control the ink droplet spray pattern, the Print Quality setting in the printer driver does. ...Show more →
I think both of you are not entirely accurate here. The pixel in the file does not correspond to one droplet but rather a group of droplets (or dots) which is called a macro dot. In fact, more dots per pixel means smoother gradients at the cost of lower resolution.
For example, a printer I routinely use (Konica-Minolta Bizhub PRO C6501eP) while having the same dot pitch allows to choose finer and coarser raster. Finer raster gives better detail while coarser raster gives better gradients.
Also there are no green inks and there is absolutely no overlay of dots on paper!! This is particularly important because if you make a solid magenta patch and cover it with a solid yellow patch you're not getting a red patch. To get red color you need to have non-overlapping yellow and magenta dots. The amount of blank white paper between these dots will determine how intensive the red is.
I've looked at microscopy of inkjet prints, and have seen (and expected to see) different color droplets overlaying each other. I think that both cgardner and I are clear on the technique of spraying several to a great many ink droplets, to paint a 'pixel' that will be much larger than any individual ink droplet. The diffusion pattern of those droplets can result in overlaying of different colored ink droplets. At 300 pixels per inch, there -could- be up to about 95 ink droplets sprayed into a 'pixel' at the 5760 x 1440 dpi of ink droplet resolution setting of my Epson.
You mention green inks. I was referring to pixels, not ink droplets. Rereading, I see that I wasn't making myself real clear, but was trying to say that image pixels index to a set pattern where any color pixel could be printed into any of those grids in the set pattern, whereas cgardner seemed to say that the different colors were laid out like a CCD or CMOS imager chip-side-by-side.
I no longer have the micrographs... But I suspect it's just a relic of the dither patterns happening to overlay. I know that Epson can vary the droplet size, so there's probably just too many variations of print pattern for them to worry about overlay. I wonder if mfrs go to extra effort to avoid overlay in lasers because the toner would get too thick?
