Epson 1280: Printer head height setting

[andy_piper2]

I just made the following discovery - which has substantial

implications for printing quality, so I thought I'd pass it along.

 

Using an Epson 1280 printer - black-ink-only mode - 9-10 mil paper

weights: Setting the print head height to the "+" setting produces a

very significant decrease in the apparent size of the printed dots and

a corresponding improvement in tonal reproduction.

 

Here's how big the difference is - a print made at 360dpi and at the

higher print head height setting produces finer dots and smoother

tonality than using 1440 dpi at the "0" print head height setting.

 

The effect is consistent and repeatable, and occurs predictably across

a wide range of paper surfaces in roughly the same paper weight class:

Epson Heavyweight Matte, Epson Premium Glossy, Epson ColorLife (and the

equivalent Ilford Classic Glossy and Pearl papers) and Tetenal Portrait

Pearl Lustre papers.

 

I'm no expert in fluid dynamics, but it appears that as the ink crosses

the wider gap (roughly 1 mm) when the print head is higher, it

disperses into a jet of fine droplets instead of a single drop.

 

This is somewhat consistent with another 'print head height' effect I

noticed printing color on Colorlife-type polymer papers: a very faint

dusting or "5-o'clock shadow" of ink in the white border on the side of

the print nearest the right side of the printer (where the controls

are) when the print head was at the (+) setting.

 

At first I though this was mechanical 'rub-off' of ink due to the

swelling of the ColorLife polymer surface, but it actually disappeared

when the print head was lowered (which should have INCREASED any

mechanical rubbing). Which makes me think the ink cloud formed by the

wider gap at the "+" setting was really being "blown" off the edge of

the print area in a fine cloud by the 'wind' of the print head passing

on the back stroke. (Again, I'm no expert on venturi effects or fine-

particle dynamics).

 

At any rate - two significant effects based on where the printer head

is set, which may be of interest if you're using ColorLife papers or

experienting with black-only printing - at least with a dye-ink

printer. Someone might experiment with a pigment printer to see if

anything similar occurs.

 

It also implies that if you use papers significantly heavier than the

standard 9-10 mil Epson/Ilford papers, you may get a visible increase

in dot size and decrease in tonality even at the printer's "+" setting,

since the thicker paper will reduce the jet-to-paper distance.

 

And it implies that if you've tried black-ink-only printing of B&W

pictures and thought the dots and tones were much too gross - you might

check your print head height setting. At the "0" setting they are

indeed gross, but at the "+" setting they improve by an order of

magnitude.

[ronb]

Andy, how does one change the printer head setting on a 1280? Is it possible to do so on an Epson 1200?

[andy_piper2]

On the 1280: Right-hand side, underneath the flip-up main printer cover (the one you raise to change the ink), on the top above the various control buttons, is a lever that moves forward and backward in a slot - the back setting is marked "+" and raises the whole print-head/ink carriage away from the paper, the front setting is marked "0", which lowers the print head closer to the paper.

 

I don't know for sure about all other Epson models, but on my old Epson 600 it's in the same place, so I assume it stayed there for all the models in between.

[angel_o.]

From the fluid-mechanics point of view it doesn't make sense. It's a nozzle, so the further away it is from the paper, the bigger the dot.

 

Maybe the "pump" is not powerful enough to provide enough fluid (=ink), so that the available volume is torn (over the increased distance) away into a spray covering the supposed "bigger-dot area". This would explain your observation: "...fine droplets instead of a single drop". So it makes sense again... ;-)

 

There you go, a pump should never be under-motorized...

[peter nelson]

As tbe Car Talk guys say . . . . <B>"B-OH . . . <U>GUS!"</U></B><P>

 

I tried it with my 870 (see my comments elsewhere in the forum for why it makes no sense, so I don't know why I let Andy talk me into this . . . ) and there was no improvement.

 

Using Epson Heavyweight Matte paper, 1440 DPI, High Speed turned OFF, black only, I did one with the height adjustment on "0" and one with the adjustment on "+". I then scanned the results on a flatbed. The images, respectively, are at:<BR>

http://studio-nelson.com/tmpimage/bo0.jpg<BR>

and http://studio-nelson.com/tmpimage/boplus.jpg<BR>

The image is part of a test image I use - the numbers represent the RGB (R=G=B for grayscale) values of the patches.<P>

 

As I said below, this is just another GIMMICK to try to get around the fact that black and white printing with inkjet technology isn't quite there yet.<P>

 

Personally the best results I've had are to print with my 2200 black-only on artist's media like Canson cold-pressed watercolor paper, or Canson toothed paper where the ink spreads out and blurs together anyway, or to print grayscale with the full inkset and default driver settings on toned media (e.g., sepia-toned gesso-on-canvas) to swamp the 2200's metamerism problems. But at least I admit that those are gimmicks until decent B+W inkjet printing comes along.

[peter nelson]

On a quasi-related note, in another thread, when I complained about the 2200's metamerism problem Daniel Staver said <I> You need a RIP to get rid of metamerism. iProof PowerRIP 2000 and ImagePrint 5.0 will both give you prints free of any metamerism. Rumors have it that the new Epson StylusRIP will do that too, but it's not available yet.

They achieve this by not using the yellow ink to mix neutrals. The yellow ink is supposed to be the main cause of metamerism on the 2200.</I><P>

 

Anyone care to explain this? In color theory Cyan + Magenta = Blue. So if you take away the yellow ink why wouldn't that produce blue-toned prints?

[peter nelson]

BTW, the scale on the two sample images I posted is about 2.5 cm horizontally.

[daniel_staver1]

I just tested this on my 2100. I do get a slightly finer pattern if I set my printer to thick paper mode. The difference was visible at a normal viewing distance from the page.

 

As I mentioned in another tread it's possible to mix a neutral gray without yellow because the light black on the 2100 is very warm brown, not neutral. The matte black is also warm, but not nearly as much as the light black. My guess is that this wouldn't be possible if the light black was completely neutral. Of course, then we wouldn't have to use all colors to get neutral grays anyway...

[peter nelson]

<I>I just tested this on my 2100. I do get a slightly finer pattern if I set my printer to thick paper mode. The difference was visible at a normal viewing distance from the page.</I><P>

 

By "thick paper" mode do you mean the manual feed setting? Also what exactly do you mean by a finer pattern? Do you mean the dots for a given grayscale value are closer together, or the dots are smaller, or what? As you can see from my sample output of the 870, the dots were neither smaller nor distributed differently. Could you post a sample scan like I did?<P>

 

<I>As I mentioned in another tread it's possible to mix a neutral gray without yellow because the light black on the 2100 is very warm brown, not neutral. The matte black is also warm, but not nearly as much as the light black. My guess is that this wouldn't be possible if the light black was completely neutral. Of course, then we wouldn't have to use all colors to get neutral grays anyway...</I><P>

 

I called ColorByte to ask where I could see sample output of ImagePrint RIP before plunking down a pile of money on their product but they said they don't "do" that as a matter of company policy, nor do they have a dealer network where such prints can be seen unless you live in LA or Chicago.<P>

 

I would be very suspicious of any company that has a policy of selling a such a pig-in-the-poke. The fact that there are individuals who claim that it fixes the metamerism problem is not very convincing because there were perfectly respectable people here who claimed the 2200 didn't have such a problem in the first place! People's eyesight, viewing distances, ambient light, choice of subjects (a gray cat in the fog versus Aunt Martha on Aruba on a sunny day), and tolerance for color shifts, vary so much from individual to individual that personal testimonials mean nothing.

[andy_piper2]

Peter: Thanks for posting examples. I can't post any yet myself since I don't have a flatbed scanner or SLR mit macro lens (I shoot rangefinders) to make a high-res image of my printer dots - but I'm working on it.

 

I agree with you that inkjet photo printing (with any number of inks) still has its weak points - but I'm not sure 'gimmick' is the right word for creative workarounds.

 

One might just as well complain that: "Graded photo paper is just another GIMMICK to try to get around the fact that film can't automatically adjust develop for each frame to the proper contrast range. Even after 165 years silver-based photography really isn't quite there yet."

 

I can't think of anything more non-intuitive (or perhaps 'gimmicky') than using 4 or 6 or 7 COLOR or gray-shaded inks to try and produce a simple black-and-white image - but is it really that much more strange than making B&W prints with COLOR filters in the enlarger (i.e. on multi-grade paper)?

 

You do what you need to do to bend the materials to your will.

[daniel_staver1]

By thick paper mode I mean that I switched the lever on the right hand side of the paper feed to the envelope symbol. I assume this is for thicker paper. I tried to make a sample scan, but the scans aren't sharp enough to illustrate the difference, so all I can do is encourage anyone with a 2100/2200 to try for themselves. I printed the picture at 2880 dpi with matte-black on Epson Archival Matte paper. For BO prints 2880dpi shows much finer detal than 1440dpi.

[peter nelson]

<I>You do what you need to do to bend the materials to your will.</I><P>

 

I call things "gimmicks" if they don't really work or only serve to distract us from the real problems. That's why I mentioned people who print on toned paper or who digitally "sepia-tone" their prints to obscure the color casts that they would get if they tried to make neutral black and white prints on their inkjet. Or people say they LIKE the dots they get with black-only printing because it makes it look "crisper" or reminds them of old, heavily-pushed Tri-X. In other words, it's a desperate attempt to make a virtue of a necessity because black and white inkjet printing is still smoke and mirrors.<P>

 

You're not bending the material to your will if it doesn't work. So far we have zero evidence that your technique actually works, and at least some evidence (mine) that it doesn't. Even if it works for you it may be due to some peculiar problem with your printer that's not reproducible by anyone else.

[marc_genevrier]

For what it's worth...

I made similar experiences on my 1290 printer when trying to get better prints on certain non-Epson papers, particularly with Tetenal glossy papers. With the lever in '0' position, even with the naked eye, I had very clearly "lines" of dots which were very distracting. I went to the local Epson service shop and they told me that one HAS to play with the lever for all papers thicker than 164 g/m². However, they admitted that the lines on my Tetenal prints were among the worst they had ever seen. So I played with the lever, always at 1440 dpi setting and I found that I can obtain the best results when positioning the lever at about midway. This setting varies however with different papers.

Coming from an Epson serviceman, I wouldn't certainly speak of 'gimmick'.

 

Also: my printer works wonderfully when printing with ImagePrint, that places the dots in a different way on the paper. So I'm not sure that my particular printer is defect.

[daniel_staver1]

Peter: You seem very eager to prove people wrong, and highly suspicious of any new technique presented in your postings. Just because you can't get it to work on a _different_ printer model than the original poster doesn't mean it doesn't work for anyone, on any type of printer.

 

It works just great on my 2100, and from the post I learned that this setting can affect the quality of my prints, so I'm going to try that more on different papers from now on. I personally found this to be very useful information.

[peter nelson]

<I>Peter: You seem very eager to prove people wrong, and highly suspicious of any new technique presented in your postings. Just because you can't get it to work on a _different_ printer model than the original poster doesn't mean it doesn't work for anyone, on any type of printer.</I><P>

 

But it doesn't mean it DOES work, either. Skepticism is the heart and soul of virtually every advance in our civilization since the Enlightenment. That's why science demands double-blind study protocols, peer-reviewed journals, repeatibility, etc. That's why we have democracies, free presses, different political parties, etc, so the competing candidate or party can hold the other side's feet to the fire or point out their inconsistencies. That's also why courts have rules of evidence. That's why the investment community demands independent auditors and why they get so pissed off then auditors fail to be independent, e.g., with Enron. I could go on and on, but you get the idea: skepticism is a vital tool for getting at the truth of something.<P>

 

Anyone who has been involved in inkjet printing as long as I have (I write image processing software for a living, and I used to write inkjet printer drivers for an industrial computer graphics company as far back as 1985) has heard an <B>AMAZINGLY</B> wide range of wild claims on matters of inkjet printing. So I treat everything that's said with a HUGE grain of salt, as should any thinking person. There's more snake-oil being flung around in this topic than at an 1880 medicine tent show!<P>

 

Andy made the apparent claim that large, fuzzy dispersed dots made a better photo than small, sharp dots. This flies in the face of every effort by inkjet makers to have smaller dots. Why do you think we prefer 2880 DPI to 1440, or 1440 to 720? So from a theoretical stndpoint we have good reason to be skeptical about his claim. Just the same I also took an empirical approach and posted images and invited him, you, or anyone else to do the same.<P>

 

<I>It works just great on my 2100,</I><P>That's funny - yesterday you said "I just tested this on my 2100. I do get a <B>slightly</B> finer pattern"</I>. And you still haven't told us what you mean by "finer".<P>

 

My undergraduate major was neurophysiology/physiological-psychology. I specialized in visual processing (single cell recording in the occipital lobe) so I'm well aware of how psychological factors can play a role in our perception of how sharp or what color an image is. Did you see that post in the Unmoderated section last Friday by that woman who was looking at some 400x400 pixel JPEGs and was totaly blown away by how "sharp" they were? That's why I (and everyone here should) demand objective evidence.<P>

 

<I> and from the post I learned that this setting can affect the quality of my prints, so I'm going to try that more on different papers from now on. I personally found this to be very useful information.</I><P>

 

I thought you're using ImagePrint, so why are you even bothering with black-only printing?

[peter nelson]

<I>This setting varies however with different papers. Coming from an Epson serviceman, I wouldn't certainly speak of 'gimmick'. </I><P>

 

Adjusting the height, per se, isn't the gimmick. Adjusting it in a manner to deliberately disperse the dots into a fine spray, as Andy was apprently advocating, is the gimmick. It's like when I print black-only on Canson cold-pressed watercolor paper - the ink spreads out and blends together so you don't see the individual dots. That's MY gimmick and I'm only doing it because the printer can't make a decent black and white print in the conventional way. Neither one of us would be using these hacks if our printers could make decent black and white prints in the standard manner.<P>

 

You weren't using a gimmick - you were simply adjusting the printhead height to get the best performance from your printer for your chosen paper. Andy was using a gimmick - he was deliberately degrading the performance of his printer to create dispersed dots because he was forced to print black-only because these printers can't do a decent grayscale.

[daniel_staver1]

I am all for being skeptic, and I always test things out for myself to see whether any given technique will work for me or not. However, being skeptic is no excuse for being rude towards people. You may think this technique is bogus and useless all you want, but keep in mind that it actually produces a visible difference at least on some printers, and that some people actually might like the effect, even if you don't.

 

By finer pattern I mean that I got the reported effect that the dots are spread out more and looks like it's composed of several smaller dots, and solid color areas with little texture looked smoother to my eyes. I didn't see a difference in sharpness. And no, I can't post any proof of this. Like I said, my scanner simply isn't able to pick up enough fine details to show the difference.

 

I use PowerRIP, not ImagePrint. I still bother with this because I'm always curious about any technique that can affect the quality of my prints, and even though I never make BO prints myself I figure if it works for BO maybe it can make a difference for PowerRIP prints as well. That's on my list of things to try.

[andy_piper2]

Evidence attached. It's a largish file to preserve detail.<div></div>

[loweskid]

Altering the head height or the thickness of the paper will alter the head alignment setting. Head alignment should always be done with paper of the same thickness as the finished print. My guess is if you re-align the heads when you alter the head setting you will end up with identical results. I could try it myself but I can't be bothered!

[peter nelson]

<I>(How print-head height affects dot size in 1280 -- 401 x 897 photo)

Evidence attached. It's a largish file to preserve detail.</I><P>

 

It looks like the original theory - that you are blurring or fuzzing the dots - is correct. The problem is that your test image is soft to start with. Notice how there's no real loss of detail in the 360 DPI image compared to the other ones. In other words the spatial frequency of the image <B>content</B> (the picture - the hands) is so much lower than the spatial frequency of the dots that you can afford to fuzz the dots without any loss of image detail on this shot. Try this with a sharper picture and you'll probably notice that when you fuzz the dots you'll lose actual information content in the photo.<P>

 

To use an audiophile analogy, you're getting rid of hiss (high frequency noise) by turning down the treble control. Normally that would also eliminate the high frequencies (cymbals, sibilants, etc) in the music, but the music you've selected to demonstrate this on is a standup bass - it doesn't have any high frequencies.<P>

 

This is why I was saying before that this flies in the face of efforts by inkjet makers <U>for years</U> to increase DPI and decrease drop size. Smaller dots allow you to record more detail. By fuzzing the dots you're doing exactly the same thing I do when I print on artist's watercolor paper - we're using a low-pass filter.

[andy_piper2]

Peter: I guess my question would be - if I'm 'fuzzing' or 'blurring' or 'degrading' the dots, how come there is clearly a little more (and certainly not less!) detail visible using the "+" setting?

 

In the image where the print head was set on "+" (the left-most one), I can see: that each fringe on the woman's net vest has two strings; that the kid's fingers have fingernails, that the woman's topmost bracelet has a spiral twist to it; and that she's wearing a false nail on her thumb.

 

All of these details are obscured by the dot size in the middle image, where the setting was on "0".

 

I could buy the audio analogy with regard to tonality and dot size alone, but I can't imagine that having a higher resolution image would punch more detail through those great clots of ink in the center example.

 

For all I know, the "+" setting is, and always has been, the "correct" setting for 9-10 mil photo paper - it's certainly thicker than typing paper. I may just be demonstrating that the "0" setting is the wrong one, not that the "+" setting is mystically better somehow.

 

At any rate - I know where MY printer head will be set from now on - at least for B&W work.

 

Everyone else is free to do otherwise, of course.

[peter nelson]

<I>In the image where the print head was set on "+" (the left-most one), I can see: that each fringe on the woman's net vest has two strings; that the kid's fingers have fingernails, that the woman's topmost bracelet has a spiral twist to it; and that she's wearing a false nail on her thumb.</I><P>

 

I can see the fingernails just fine in all three images. I can also see the two strings in the fringe in the 360 DPI image, which means the spatial freqncy of that detail is very low. The spiral twist is also clearer on the 360 DPI image than on the 1440. Also the dots are smaller and more evenly placed on the 360 DPI image than on the middle image, which suggests you have some variable you haven't accounted for, or perhaps something wrong with your printer in the middle setting.<P>

 

As I mentioned in another thread in this section, I made a test image I use for printer testing. One feature of it is that it has a series of lines of specified width and specified distances between them, e.g., 1 pixel wide, 1 pixel apart, 2 pixels wide, etc, both vertical and horizontal orientations. So I tried Daniel's suggestion with the 2200 - I printed one with the regular paper setting and one with the Envelope setting which he claims produces an effect similar to yours. The images are at:<BR>

http://studio-nelson.com/tmpimage/pap2200.jpg<BR>

and<BR>

http://studio-nelson.com/tmpimage/env2200.jpg<P>

 

As you can see, there is no practical difference between the two. Also, note that the individual dots, especially in the light areas, are easy to resolve, unlike the weird unresolved clumps in your middle image.<P>

 

So what is <B>your</B> theory about what's going on with your printer and settings?