What's the greatest "resolution" that can be printed from an infinitely detailed digital image?

PaperclipGod

Banned
Apr 7, 2003
2,021
0
0
So, let's say you've got one of those 49805730942752 megapixel photographs... how much of that image's detail/resolution/color-information can be reproduced on a printed medium? What's the smallest size a printer can print to?

Is it possible to take one of those huge photos, print it on an A4 sheet of paper, and use a magnifying glass to reveal the greater detail?

I guess what I'm really getting at, is: How much data can you store on a single sheet of printed paper?
 

Quasmo

Diamond Member
Jul 7, 2004
9,630
1
76
Some printers claim to print at 2400 DPI, No I dea how accurate that is, but thats pretty damn small. The naked eye cant really discern anything past 600 dpi. Magazines are usually printed at 300 DPI. National Geograpics are printed at 450 DPI.
 

Injury

Lifer
Jul 19, 2004
13,066
2
81
Any run of the mill home inkjet that claims to print better than 600DPI is full of shit*... but even with the best inkjet printer you can find, using a magnifying glass will only let you see the CMYK dots instead of insane detail anyway.




*Pretty much any print quality higher than this is going to be completely ruined by the fact that ink spreads when it hits paper and is absorbed. While the paper used is a big factor in actual print resolution, you can only make an ink dot so small with consumer-level equipment.
 

PaperclipGod

Banned
Apr 7, 2003
2,021
0
0
Originally posted by: Injury
Any run of the mill home inkjet that claims to print better than 600DPI is full of shit*... but even with the best inkjet printer you can find, using a magnifying glass will only let you see the CMYK dots instead of insane detail anyway.




*Pretty much any print quality higher than this is going to be completely ruined by the fact that ink spreads when it hits paper and is absorbed. While the paper used is a big factor in actual print resolution, you can only make an ink dot so small with consumer-level equipment.


Are there any other print technologies that overcome this? How about laserjets?

Is there professional-grade hardware that can print at uber-high resolution?
 

iGas

Diamond Member
Feb 7, 2009
6,240
1
0

Depends on the application.

Film slide/transparency/negative lp/mm (line pair per mm) resolution often exceed optical lenses resolving power. High quality 25-50 ISO can often produce 100~150 lp/mm. High contrast technical film can reach as high as 350 lp/mm.

The highest resolution has to be spectroscopic plates at 2000 lp/mm.

Films: (350 lp/mm * 2) * (25.4 mm * 25.4 mm) = 451,612 DPI

Spectroscopic plates: (2000 lp/mm * 2) * (25.4 mm * 25.4 mm) = 2,580,640 DPI
 

Paperdoc

Platinum Member
Aug 17, 2006
2,499
374
126
Originally posted by: Quasmo
Some printers claim to print at 2400 DPI, No I dea how accurate that is, but thats pretty damn small. The naked eye cant really discern anything past 600 dpi. Magazines are usually printed at 300 DPI. National Geograpics are printed at 450 DPI.

Here's one of the cute tricks hidden in the specs. Most higher-quality color printers are using 4 ink colors (some use 6 or 7 for photo work) - cyan, magenta, yellow, and black, the same system commercial printing uses. Each color prints at 600 dpi, but since there are 4 of them the printer box says it does 2400 dpi. The eye sees all the full color range because it integrates the impact of all the adjacent color dots. The actual image resolution on such a printer really is about 600 dpi, which is still pretty impressive to the eye.

There is a fundamental difference in how dots of color are laid out between computer printers and commercial publishing presses. A computer printer prints dots of each color at the specified spacing. The intensity of color at each dot position is set mostly by the volume of ink laid down - more ink makes a slightly larger dot as it spreads out before drying, and the total amount of dye determines color of that dot.

A commercial press, on the other hand, cannot vary the volume of ink in one dot directly. What it does do is vary the size of the dot, and that indirectly changes its ink volume, too. This is a holdover from the original technique of "screening" a photo in preparation of the printing plate. That process breaks up a solid image into a rectangular array of dots which vary in size, depending on image brightness, because of light diffraction going through the screen. The resolution of the plate is determined by the spacing of the lines in the screen filter used, so it's usually labeled in terms of lines per inch (lpi).

In modern digital systems to make printing plates, the relationship is established by by treating the printing image as a rectangular array of "cells" (equivalent to dots). A common way to do this is to create a cell as a square array of 8 x 8 tiny ink dot locations, which makes it possible through selection of which dots to generate 256 different "grey levels" of that color in that cell. Whatever line spacing you want to reproduce (in traditional screening systems), you then need ink dot spacing 8x that. So to make a 100 lpi printer's image you need a printer that will do 800 dpi. Very good quality commercial printing will be done at 300 lpi, requiring a good 2400 dpi printer. But since the commercial plate is generated for each of the colors separately, the plate prep equipment is doing this on only one "color" (black) at a time. Still, if your printing house plans to do some very high quality work, the plate prep stuff has to be able to print that one color at a true 2400 dpi, or better.

The other way to look at it is: if your home printer can do a true 600 dpi of one color (say, a black-and-white laser printer), it can generate a printed photo of 256-level grey scale at a resolution equivalent to75 lpi, which is a bit better than newspapers would do. But in fact many laser printers do better than this. My old HP LaserJet 4 does 600 dpi, but it also has a feature called "Resolution Enhancement Technology" or RET that actually varies (within a small range) the size of each black dot (as opposed to either a full dot or no dot, as "normal" 600 dpi would do). The result is that the 256-grey-level impact can be delivered with a smaller array of dots per cell, so the number of cells per inch is higher, and the resulting grey-scale photo looks much better than it would in a newspaper at 75 dpi screening. Back in color printer land, because the amount of ink laid down at each location with 600 dpi spacing can be varied over a significant range, you can reproduce full-color images much better than the finest commercial print process.
 

PaperclipGod

Banned
Apr 7, 2003
2,021
0
0
Originally posted by: iGas

Depends on the application.

Film slide/transparency/negative lp/mm (line pair per mm) resolution often exceed optical lenses resolving power. High quality 25-50 ISO can often produce 100~150 lp/mm. High contrast technical film can reach as high as 350 lp/mm.

The highest resolution has to be spectroscopic plates at 2000 lp/mm.

Films: (350 lp/mm * 2) * (25.4 mm * 25.4 mm) = 451,612 DPI

Spectroscopic plates: (2000 lp/mm * 2) * (25.4 mm * 25.4 mm) = 2,580,640 DPI


I'm wondering more about transferring digital images onto a physical medium, though. i.e., what's the smallest unique point of digital data that can be resolved to paper?

For example: I take an extremely high-rez photo of 1 sq-in of the Mona Lisa. The quality is such that a single brushstroke looks gigantic. I continue taking similar shots of the entire painting, and then stitch them all together. I'm left with an incredibly detailed digital reproduction of the entire Mona Lisa. The computer can abstract the detail in the image, so that even if I zoom the image down to 5% of it's "true" size, I can still "zoom in" and reveal the greater levels of detail.

However, if I wanted to print a hardcopy of that image, I'd be restricted to the detail available at any given "zoom". i.e., if I print the image on a sheet of A4 paper, "zooming in" with a magnifying glass would only reveal individual dots of ink, not further detail.

Is there any way to print points of color (data) small enough that looking at the printed picture under a magnifying glass reveals further detail, not simply the underlying matrix of paper/ink?

I'm kind of a retard, though, so please go easy on me if I'm asking something stupid. :/
 

mugs

Lifer
Apr 29, 2003
48,920
46
91
Originally posted by: Quasmo
Some printers claim to print at 2400 DPI, No I dea how accurate that is, but thats pretty damn small. The naked eye cant really discern anything past 600 dpi. Magazines are usually printed at 300 DPI. National Geograpics are printed at 450 DPI.

I believe the printers that claim 2400 DPI (and really all printers actually) are using multiple "dots" to make up one "dot" - fewer dots = lighter color. I don't think that 2400 number is equivalent to the 300 DPI that is used when printing photos and what have you.
 

Paperdoc

Platinum Member
Aug 17, 2006
2,499
374
126
The finest resolution you can print on paper right now depends both on the paper and the ink, which is one reason many printer makers insist you use their paper for best results.

An ink jet printer shoots a tiny droplet of water-based liquid ink (actually, very dilute with lots of water) onto the paper surface. When it hits, the water with its dyes included starts to soak into the surface in all directions, so its diameter grows. The chemical details of the ink and the paper surface are set up deliberately for two important things. One is that the paper surface must absorb the water quickly in as small an area as possible so that it does not spread out or, worse, stay liquid long enough to be smeared by physical contact. Attached to that is the requirement that the paper wetted by all this water should not expand locally and end up looking wrinkled. The other thing is that the dyes carried in the water should be preferentially attached to the first paper surface component they encounter so that they are no longer being carried along by the flow of water into the paper. This ensures the resulting dye dot is as small as possible. It also ensures the dyes are concentrated right on the surface to have maximum impact on the light it reflects back to your eye, thus yielding sharper colors and fine detail in the image.

For these purposes there are quite a range of papers available for ink jet printers. The simplest ones basically have a light treatment of starch on the surface, just like almost all office copy papers. That's why you can "get away with" using plain copier papers in an ink jet printer - it sort of meets the basic requirements of the printing process, but not well. The next step up is more sophisticated surface coatings with various pigment systems, although you would not see these coatings and probably don't even know they are there. These are much better than plain copier paper, but they look the same and cause confusion and skepticism among paper buyers - why buy more expensive "ink jet paper" when it looks just the same? Well, it actually is different and performance is noticeably better if you care to look at it. The next step up from there is a much higher-quality surface coating based on silica pigments, and these are the original "ink jet papers". The coating materials and process are expensive but produce great printed results. Again, this paper is not obvious because the coating is not glossy, and that's what most people think of when we say "coated paper".

Above that are a big range of multiply-coated papers for higher-quality results, often sold as photo printing papers. These come in both matte and glossy finishes, and they are optimized to accept lots of ink (with lots of water), to keep the ink dyes on the surface, and to minimize spread or "blotting" of the ink dot for sharp detail. Because optimization depends on the interaction of ink and paper surface, best results are obtained if you get both from the same manufacturer since each uses a different system.

Because the limiting factor on an ink jet printer is the process of liquid droplet spreading on the paper surface before its dyes are immobilized, the current limit appears to be about 1200 dpi per color. I have not seen any consumer-level printer that claims more, although there may be some at the professional level.

With the dry toner systems of laser printers the process and limits are different. The process involves creating an "image" composed of electrical charges on a drum surface, then attracting to that some fine pigment particles from a colored dust, then transferring that array of pigment particles onto the paper surface, and finally heating the particles to melt a fine layer of adhesive on the pigment particles so that, when they cool again, they are bound to each other and to the paper surface. This involves a lot of careful balancing of electrical performance of pigment, drum and paper. The pigment particles have to be small and uniform in size. The paper surface does require some modification with simple coatings to balance its electrical conductivity in the right range. The paper body itself must stand up to the "mistreatment" (as papermakers might call it) of being heated drastically but briefly on one side only, and still stay flat so it can continue to travel through the printer's paper path. To do full color, most such printers will do this process four times in sequence (one pass for each color), so the pigments are being laid down over top of a previous pigment and the paper is being hit hard four times. Moreover, each color's image has to be placed exactly right in relation to the other three for details to come out right.

So limits are imposed by toner particle size, laser beam and scanner resolution, balanced electrical characteristics of ink, drum and paper surface, stability of the paper body, and precision of the paper carrying system. The highest resolution I've seen for ONE color laser printers is about 4800 dpi - don't know if there are any professional systems higher. For 4-color laser printers, I doubt you'll see better than 2400 dpi per color.
 

PaperclipGod

Banned
Apr 7, 2003
2,021
0
0
Wow, fantastic explanation -- thanks!!

Does printer technology follow any sort of "improvement curve" similar to Moore's Law and IC's? I mean, can we expect steadily increasing printed resolutions, or are we already at the physical limit of the known/projected technology?

Do you think it'll be possible at some point to print at a resolution high enough to create images which reveal more detail with a magnifying glass, or is that an impossibility? Wouldn't that require an extremely homogeneous printing surface, else the magnified detail would be distorted by the material it's printed on?
 

iGas

Diamond Member
Feb 7, 2009
6,240
1
0
Originally posted by: PaperclipGod
Originally posted by: iGas

Depends on the application.

Film slide/transparency/negative lp/mm (line pair per mm) resolution often exceed optical lenses resolving power. High quality 25-50 ISO can often produce 100~150 lp/mm. High contrast technical film can reach as high as 350 lp/mm.

The highest resolution has to be spectroscopic plates at 2000 lp/mm.

Films: (350 lp/mm * 2) * (25.4 mm * 25.4 mm) = 451,612 DPI

Spectroscopic plates: (2000 lp/mm * 2) * (25.4 mm * 25.4 mm) = 2,580,640 DPI


I'm wondering more about transferring digital images onto a physical medium, though. i.e., what's the smallest unique point of digital data that can be resolved to paper?

For example: I take an extremely high-rez photo of 1 sq-in of the Mona Lisa. The quality is such that a single brushstroke looks gigantic. I continue taking similar shots of the entire painting, and then stitch them all together. I'm left with an incredibly detailed digital reproduction of the entire Mona Lisa. The computer can abstract the detail in the image, so that even if I zoom the image down to 5% of it's "true" size, I can still "zoom in" and reveal the greater levels of detail.

However, if I wanted to print a hardcopy of that image, I'd be restricted to the detail available at any given "zoom". i.e., if I print the image on a sheet of A4 paper, "zooming in" with a magnifying glass would only reveal individual dots of ink, not further detail.

Is there any way to print points of color (data) small enough that looking at the printed picture under a magnifying glass reveals further detail, not simply the underlying matrix of paper/ink?

I'm kind of a retard, though, so please go easy on me if I'm asking something stupid. :/

Fiber paper limitation is around 10~20 lp/mm, RC (resin coated) paper limitation is 40 lp/mm

Transparency (plastic base) limitation is around 80 lp/mm. Transparency can be had up to 4'x100' ( I have experimented with 30"x40")

Technical sheet films can be had up to 11"x14" (20"x24" is no longer available).

I don't have plates experience, but 5"x7" were very common, and some people have experimented with 11"x14"

20 years ago, I mixed my own silver bromine + resin to make a paste that was good up to 40 lp/mm. And, the largest I have printed was 4'X10' (squirt bottles, janitor bucket & mops, rags, and windows cleaner squeegees).

Largest Polaroid that I experimented with was 20"X24" (camera was a loaner to my school by Polaroid).
 

Paperdoc

Platinum Member
Aug 17, 2006
2,499
374
126
Originally posted by: PaperclipGod
Wow, fantastic explanation -- thanks!!

Does printer technology follow any sort of "improvement curve" similar to Moore's Law and IC's? I mean, can we expect steadily increasing printed resolutions, or are we already at the physical limit of the known/projected technology?

Do you think it'll be possible at some point to print at a resolution high enough to create images which reveal more detail with a magnifying glass, or is that an impossibility? Wouldn't that require an extremely homogeneous printing surface, else the magnified detail would be distorted by the material it's printed on?

I almost never say we have reached the limit and can't make further progress. Anyone who says that is guaranteed to be proven wrong. As far as developing papers for high-resolution printing goes, the potential improvements are most likely to come from the surface modifications through coatings, and not from the paper itself. But that's not new. In the coated fine papers business, it is well recognized that papermakers are selling surface, not paper. We have passed the point where surface roughness irregularities interfere with ink laydown. Use a strong magnifying glass on a magazine photo and you'll see easily the regular arrays of colored ink dots that create the image. You also will see that the underlying glossy surface looks pretty flat compared to the ink dot size. Now, if you go to a photo from a good ink jet printer with that same glass, you will find it hard to see individual ink dots clearly, except where there are only a few dots in an otherwise white area. But using more powerful magnification like a stereo microscope you will get to seeing each ink dot clearly, plus some really tiny surrounding ink bits caused by "splatter" of the wet ink as it hit the paper. On that photo paper you will see very little surface irregularity compared to the ink dot size.

So, the limiting factors on resolution currently are dot gain and ink droplet volume. That is, once the ink drop hits the surface, how does it spread out as it soaks into that surface? Now, it certainly is possible to prevent that soaking and spreading, to keep the ink as just a little spherical drop sitting on the surface. With enough time its solvent (water) will evaporate, leaving a film of ink dyes on the surface with a diameter close to the original droplet size. The problem here, though, is that this situation normally means that the ink components are so completely different from the surface that they refuse to interact chemically, with the result that the dried dyes have no bond to the surface and will simply slide off the first time they are touched by anything. Someone may find a way to solve that.

Then we have the droplet size. Printer makers constantly work to make even smaller droplets so they can produce finer-resolution prints. That involves making even smaller ink jet nozzles, but as you do that you increase then probability that, when not being used, the ink sitting in the end of the nozzle will dry out and plug the jet, necessitating difficult maintenance. I have no doubt printer makers are working more on this. However, the motivation for such research is reduced by the fact that right now you can print at resolutions finer than the unaided eye can see. What you are proposing, which is deliberately to print detail that can't be seen unless you use powerful enhancements, is not a common market need.

But going back to your original query, why would one want to print more detail so that it can be examined with magnification? If the aim is to preserve a lot of information in a smaller space, maybe another technology is the better route. For example, rather than printing a high-resolution photo, why not store the data electronically as a computer file? I realize this gets us into the huge discussion of how to archive data and ensure it is preserved accurately and "permanently" AND can be read later when needed. But that is no different from how to archive printed materials. In terms of resolution of detail in the data, you always will be limited by the resolution of the original data at the time of acquisition - you can't make more later, but you must preserve faithfully all that was available from the start.
 

PaperclipGod

Banned
Apr 7, 2003
2,021
0
0
Originally posted by: Paperdoc

I almost never say we have reached the limit and can't make further progress. Anyone who says that is guaranteed to be proven wrong. As far as developing papers for high-resolution printing goes, the potential improvements are most likely to come from the surface modifications through coatings, and not from the paper itself. But that's not new. In the coated fine papers business, it is well recognized that papermakers are selling surface, not paper. We have passed the point where surface roughness irregularities interfere with ink laydown. Use a strong magnifying glass on a magazine photo and you'll see easily the regular arrays of colored ink dots that create the image. You also will see that the underlying glossy surface looks pretty flat compared to the ink dot size. Now, if you go to a photo from a good ink jet printer with that same glass, you will find it hard to see individual ink dots clearly, except where there are only a few dots in an otherwise white area. But using more powerful magnification like a stereo microscope you will get to seeing each ink dot clearly, plus some really tiny surrounding ink bits caused by "splatter" of the wet ink as it hit the paper. On that photo paper you will see very little surface irregularity compared to the ink dot size.

So, the limiting factors on resolution currently are dot gain and ink droplet volume. That is, once the ink drop hits the surface, how does it spread out as it soaks into that surface? Now, it certainly is possible to prevent that soaking and spreading, to keep the ink as just a little spherical drop sitting on the surface. With enough time its solvent (water) will evaporate, leaving a film of ink dyes on the surface with a diameter close to the original droplet size. The problem here, though, is that this situation normally means that the ink components are so completely different from the surface that they refuse to interact chemically, with the result that the dried dyes have no bond to the surface and will simply slide off the first time they are touched by anything. Someone may find a way to solve that.

Then we have the droplet size. Printer makers constantly work to make even smaller droplets so they can produce finer-resolution prints. That involves making even smaller ink jet nozzles, but as you do that you increase then probability that, when not being used, the ink sitting in the end of the nozzle will dry out and plug the jet, necessitating difficult maintenance. I have no doubt printer makers are working more on this.

Ahh... that's really interesting! So, using current technology, the only way to print at higher resolutions is to reduce ink droplet volume... but those ink droplets actually spread (capillary action, I guess?) when they hit the page. So if you've got an ink droplet with a size of "10", it might appear on a finished photo at size "30", right? Is there a general rule for calculating ink droplet size to ink surface diameter size? e.g. 30% larger?

If the problem with smaller ink nozzles is clogging... I can think of a few things (all surely wrong or inapplicable, but I'd be curious to hear why!):

How about using inks which have an extremely low viscosity when hot, but extremely high when cold? So then you could heat the inks while inside their cartridges/nozzles so they wouldn't be prone to clogging, and use some sort of refrigeration method to cool the paper just before the ink is applied. The fast cooling of the ink and it's increase in viscosity should mitigate the spread of the ink, right?

Or instead of all that, what if there was a way to keep the ink in the nozzle/cartridge constantly moving via a pump or something like that? If the ink isn't just sitting in the nozzle, then it wouldn't have a chance to dry/clog, right?

Or how about using a stamp-type device? So that the ink isn't applied in a giant droplet and allowed to soak into the page, but instead a very small amount of ink is physically pressed onto the paper. Less ink = less spreading, right?

What about using a laserjet type of technology... but instead of heating the entire page surface in one pass, what if you had heat "nozzles" which would thermally bond the ink particle to the paper in the exact spot you want it? So you wouldn't really be using a laser at all, just "dry" ink, but applied to the page in a manner more precise than laserjets....
 

PaperclipGod

Banned
Apr 7, 2003
2,021
0
0

Originally posted by: Paperdoc

(...) However, the motivation for such research is reduced by the fact that right now you can print at resolutions finer than the unaided eye can see. What you are proposing, which is deliberately to print detail that can't be seen unless you use powerful enhancements, is not a common market need.

But going back to your original query, why would one want to print more detail so that it can be examined with magnification? If the aim is to preserve a lot of information in a smaller space, maybe another technology is the better route. For example, rather than printing a high-resolution photo, why not store the data electronically as a computer file? I realize this gets us into the huge discussion of how to archive data and ensure it is preserved accurately and "permanently" AND can be read later when needed. But that is no different from how to archive printed materials. In terms of resolution of detail in the data, you always will be limited by the resolution of the original data at the time of acquisition - you can't make more later, but you must preserve faithfully all that was available from the start.

I was actually just thinking that this would be a really awesome way of looking at photos. Imagine if you've got a wallet-sized photo of a beach from your last vacation. It's looks decent with the naked eye... but if you put a magnifying glass over it, you can see high-rez detail of the location as if it were right in front of you on a 100 sqft mural.

Or, speaking of murals... imagine being able to put the classic "solar system" or "space shuttle" mural on a kids bedroom wall. But, instead of being limited to that one view... you could walk up to any point on the mural with a magnifying glass and view stars or constellations that would otherwise be too small for the naked eye to see. If the moon is represented as an image 2 inches in diameter on the mural, with features hard to resolve, you'd be able to zoom in with your magnifying glass and see the undulations of each crater on the surface!
 

Paperdoc

Platinum Member
Aug 17, 2006
2,499
374
126
There is no rule relating final dried ink spot to volume of the wet ink droplet. And that's because the increase in droplet diameter, once it is on the surface, depends heavily on the chemical interactions between ink components and surface components. It is in the details of this interaction that most printer makers work to improve their print results. And that is why they tell you that, for best performance, you MUST use BOTH their inks and their papers.

Consider these design considerations for the first generation of papers optimized for ink jet printers. They use a pigment called silica because it is tiny particles of irregular shape that are full of little pores. The ink is mostly water that has to go someplace. With plain paper the water flows over the surfaces of the paper fibers and some sticks where it is, while more keeps on flowing to find a dry fiber surface. That spreads the ink a long way and even distorts the image because the ink tends to flow in little straight lines along fibers. It also means the fibers that get wet try to swell, causing the paper to wrinkle. But with the microporous silica pigment on the surface, all the water flows into the little micropores quickly and then, slowly over many minutes, evaporates back out again leaving the ink dyes behind bonded to the silica particles. However, even that process means the ink dyes are carried into the pores and some of it is "hidden" from light because it is inside the silica pigment particle pores. So the next refinement is to use electrical charge effects. The ink dyes tend to be anionic - that is, their molecules have several negatively-charged parts exposed. So the silica pigments are prepared with cationic additives on them that give their surfaces a bit of an opposite charge. Now when the ink drop lands the ink dye molecules are rapidly stuck to the oppositely-charged silica particle surfaces and held there, instead of flowing with the water down into the pores. So the water absorption and evaporation mechanism (to handle the large water volume involved) is still there, but the dyes carried in the water are selectively bonded to the surface instead of being carried down into the pores. With dyes moving around even less than the water, the effect is to reduce even more the spread of the colored dot, improving dot sharpness by keeping its diameter smaller. The additional benefit is that the dye is right where the light hits, so its impact on light reflection back to your eye is improved.

Even more sophisticated techniques are used in the newer less-expensive coatings for ink jet papers and inks, plus even fancier designs especially for the high-performance photo inkjet systems.

Regarding your idea of temperature and viscosity, oddly there was a system that used something similar a while ago, but I have not seen it much lately. Tektronix marketed a line of high-resolution color printers in which the ink really was a specially formulated stick of wax with dye in it. (They used different shapes of sticks so you could not put the wrong color ink into a slot.) They produced very good color because the ink did not really soak into the paper, and they did not require special papers. The did not use cooling. To print, the head heated the waxy ink and shot out a liquid droplet that solidified rapidly by giving up heat to the paper when it landed. I do not remember just how fine the droplet size and resulting ink color dot was, but I don't think it was as small as current inkjet printers can achieve. The prints had one aspect that some found disconcerting - the resulting image felt slightly rough because it was composed of an array of tiny waxy ink mounds sitting on the paper surface. Some of their printers actually ran the final printed sheet through a pair of heated rollers to "iron" the surface flat before discharging the printed sheet, but this had the disadvantage of spreading the fine dots out. OK for smoothness and for photos where smooth gradations among colors is desired, but not OK for fine detail in line drawings.