Luminous Landscape Forum
Raw & Post Processing, Printing => Digital Image Processing => Topic started by: JimAscher on September 13, 2010, 12:35:54 pm
-
I don't know how better to formulate my query other than the somewhat inadequate phrasing above. Please excuse.
I use an Epson Stylus Photo 1400 which prints with 1.5 picoliters for my black-and-white work. (I do no color.) I print in size about half a sheet of 8.5 by 11 inches, which averages usually 8.5 by 5.5 inches. I am trying to emulate the size of a 5 by 7 inch film contact print, for mounting on 8 by 10 matt board for hand-held portfolio viewing.
With film contact printing one usually gets extremely fine definition and gradation. However, within the limits of a picoliter printer, does one lose out in definition and gradation through printing the smaller size? In fact, are definition and gradation possibly perceived as INCREASING with printing in larger sizes?
-
In fact, are definition and gradation possibly perceived as INCREASING with printing in larger sizes?
They definitely do up to a certain point... Especially taking into account that your 1400 doesn't have grey inks.
You'll have to test for yourself the optimal adequation between print resolution in ppi and output quality, but may end into the classical 200-400ppi territory.
More generally, B&W contact printing may still be challenging some inkjet technologies (and why wouldn't you convert your 1400 to pure B&W with grey inks?), see eg http://www.custom-digital.com/2008/09/bw-print-quality/ .
-
For quality monochrome printing you need a printer with several shades of grey.
On an Epson printer print at 360 or 240 original camera pixels per print inch
-
This is a very interesting topic, and one which I have pondered a great deal in recent years. The whole issue has been clouded by the fact that most people seem to wish to print at HUGE sizes these days, and the small print is very much out of fashion. Which is a great shame, because I have always loved the jewel-like quality of a fine B/W contact print from 5x4 or 10x8 film. And small prints often suit the intimate nature of photography far better than trying to compete with painters and their huge canvases. Many of my own pictures would gain nothing worthwhile by printing them larger than my usual 9x7 ins plus border on A4 paper.
Unfortunately, small prints are precisely what ink-jet printers are worst at. The smaller you print the more detail and resolution you lose, because the resolution of the printer stays the same, and pixels simply get thrown away. Whereas in the darkroom, the detail gets smaller, but nothing is lost. Sadly, there is no simple answer to this, and the more megapixels your camera or DB has the worse the problem will be.
So for small prints, for those of us who love them, the silver-halide process still rules.
John
-
And small prints often suit the intimate nature of photography far better than trying to compete with painters and their huge canvases.
John
It is because I want to "try to compete with painters and their huge canvases" that I have ordered a 60Mpx H4D-60, that I can use shift-and-stitch on my Sinar.
I hate over-enlarged prints, but hope to get that "Jewell-like" quality even in single shot @ 360 original camera pixels per print inch @ 18*24"!
-
I really appreciate the thoughtful responses I have so far received to my initial query. I suppose I should have mentioned initially that my Epson 1400 is loaded entirely with different dilutions of black ink, but I have not yet achieved sufficiently satisfying results for myself in using all six dilutions in the same print. I have so far found satisfaction in a curve combination of only two dilutions -- 100% and 9%. This may very well be a result of my preference for small-size prints. And while I have recently upgraded to cameras with larger sensor sizes, I am finding that many of my most favored prints are a result of using a point-and-shoot camera with a smaller sensor size. Maybe for my type of work, trying for prints with finer resolutions and definitions beyond a certain point is a fruitless endeavor.
-
Jim
Unfortunately, you find yourself in the position of wishing to pursue a perfectly valid branch of photography for which there is almost no popular support. I have the same feelings of frustration myself, most of the time. Many times I have posted about a problem I am experiencing on this forum, and the response has been mostly one of polite puzzlement as to why on earth I would wish to do "x" in the first place. But as you have found, ink-jet printers are only absolutely wonderful and the complete bee's knees if you wish to make very large and mightily impressive colour prints. Which is what almost everyone seems to want to do.
On the other hand, from my personal point of view, why on earth would I want to do that? I can put 10x8 or smaller prints, sleeved, in albums which I can pick up and browse. I can frame them to a maximum size over the frame of 11x14 ins, which are quite practical to hang in my tiny cottage with very little wall space. If I was printing 20x16 or larger, what could I do with the damn things? I certainly have no space to hang them, so they would just sit in a portfolio where I and no-one else would ever see them.
The other serious problem with even the very best ink-jets, which you have not so far mentioned, is that if you examine the prints with a loupe you will see that they cannot deal with very subtle highlight gradations. This is because, even with a light-light black ink eventually as the highlight approaches the uppermost zones, the printer can no longer lay down continuous tone but has to resort to spacing the dots wider apart. Unlike a silver-gelatine print, which is truly continuous tone even in the lightest highlights.
John
-
John: Inadvertently, you have hit upon one of my key problems in trying to extend gradation through use of all six of my ink dilutions. I quote you as follows:
The other serious problem with even the very best ink-jets, which you have not so far mentioned, is that if you examine the prints with a loupe you will see that they cannot deal with very subtle highlight gradations. This is because, even with a light-light black ink eventually as the highlight approaches the uppermost zones, the printer can no longer lay down continuous tone but has to resort to spacing the dots wider apart. Unlike a silver-gelatine print, which is truly continuous tone even in the lightest highlights.
This is precisely the area of my key difficulty. In those maximum highlight areas, all I get is a smooth light-gray tone (smudge?). When I don't try to get (or expect) definition in those areas, through use of my two-dilution approach, I can achieve an otherwise satisfying, but less defined, print. Does this make sense? I don't know how else to put it. In other words, what I have been satisfied with in the past are small prints with pronounced compositions which can be nicely achieved on the basis of significant ink contrast. And this is probably the approach I'd best stick to.
-
Jim
I did quite a lot more thinking, calculating and examining my own darkroom contact prints last night. I can give you the accurate figures if you like, but roughly it goes like this -
For a 5X4 ins digital print (representing a 5x4 film contact) there is no point in using a sensor with more than 3MP if you are using an Epson printer whose native resolution is limited to 360 dpi. Any additional information in your digital file will simply be discarded. And to get the same quality as film for a 645 contact print from my 39MP DB you would need about 3,500 dpi from the printer!
There is an absolutely shocking difference between film contact prints and the same scanned negative printed to the same output size even at my R2400's best output resolution. Viewed through an 8x loupe the inkjet print is just rubbish by comparison. The reason most people never notice this is that they never print small.
Sadly, we seem to be in a tiny minority of those who actually love small B/W prints. Notice how little interest this thread has generated? Which I don't really understand. Some years ago, my wife made a wood engraving of the lighthouse on St Agnes, Scilly. It is just two inches square, and framed with a very generous mat it is one of my favourite pictures.
So, with the current state of technology, if you wish to print small very high-quality B/W photographs, the darkroom is still the undisputed king. Shoot 5x4, 5x7 or 10x8 film in a field camera and contact-print the resultant negative. The results will be streets ahead of anything you could manage using digital cameras and ink-jet printing.
John
-
John: The results you report of your latest investigative effort are discouraging for me indeed. Have you formed yet any thoughts on the efficacy of making digital negatives from higher pixel camera output for darkroom contact printing?
-
John: The results you report of your latest investigative effort are discouraging for me indeed. Have you formed yet any thoughts on the efficacy of making digital negatives from higher pixel camera output for darkroom contact printing?
Jim
As far as I am aware, the only labs who do this laser-scan digital images to 35mm film. Now 35mm is just too small for the work we would like to do. So what we need is a lab that could laser-scan to 5x4 sheet film (or larger). There may be someone here on the Forum who has more knowledge of this process and could enlighten us further?
John
-
John: Somewhere recently in one of the other forums I monitor -- and I can't recall which one -- there was an informative discussion of making digital negatives ourselves through printing on transparent film with our printers. Does this method ring any bells for you? Jim
-
John: Somewhere recently in one of the other forums I monitor -- and I can't recall which one -- there was an informative discussion of making digital negatives ourselves through printing on transparent film with our printers. Does this method ring any bells for you? Jim
Well, you can, but then you are limited by the same dpi issues we have already discussed. All you gain is the ability to print on silver-gelatine paper, but you get the same inkjet quality of output.
You see, essentially the problem goes like this -
I have a Haselblad CFV-39 digital back which I use on my lovely old 500 C/Ms. This gives me stunning 39MP files from a sensor which my Zeiss lenses might just out-resolve (on a tripod, with every care taken) but only just. It's pretty much state-of-the-art for old lenses like these. But here is the catch - at 360 dpi (the native resolution of my Epson) the optimum print quality is achieved at 16x20 ins (or A2). If I print any smaller than this, I am throwing image quality away. Whereas in the darkroom, it is effectively the other way around. I lose image quality (through diffraction etc) the bigger I print.
John
-
John: Your cogent explanation should of course have been obvious to me. In either case, we are limited by the resolution native to our printers. But on that subject -- printer limitations -- and here I must display my ignorance, when all six (or fewer) ink jets in the printer are pumping away with different ink dilutions, isn't there some ink over-lap to fill in the pixel gaps, in order to provide smoother gradation? Jim
-
John: Your cogent explanation should of course have been obvious to me. In either case, we are limited by the resolution native to our printers. But on that subject -- printer limitations -- and here I must display my ignorance, when all six (or fewer) ink jets in the printer are pumping away with different ink dilutions, isn't there some ink over-lap to fill in the pixel gaps, in order to provide smoother gradation? Jim
Jim
Yes indeed, and this is what is called dithering. Otherwise you would need 255 ink cartridges in your printer to render the 256 gray scale from an 8-bit B/W file. And keeping track of the ink levels in 255 cartridges would be a bit of a pain. But I am no expert on these technicalities. To get into the nitty-gritty of this stuff, you need to be talking to someone like Ernst Dinkla over in the printing section.
Some of this article is now rather dated, but it forms a good starting point -
http://www.scantips.com/basics3b.html (http://www.scantips.com/basics3b.html)
John
-
John: Thanks for referring me to the very informative article. I am the wiser for it, but not much encouraged. I've been asking myself whether a dye sublimation printer could not provide me more of what I seek. They are of course primarily color printers (but some users report good black-and-white results) and are more costly per photo than from ink jet printers. Also, while the resulting photos are reportedly less sharp at the edges, the tones are smoother. Anyway, as you can see, I'm sort of thrashing around seeking a better solution to my needs in the small print realm. Jim
-
In my experience the problem is matte papers, not inkjet dithering. If you print with a current-generation inkjets and high-quality photo-black papers such as the newer fiber-gloss and baryta papers, you won't see dithering.
I have 24mp images printed at 6x9", which comes out to just over 600ppi. Using the 16-bit Canon plug-in with my ipf6300 and printing on Hahnemuhle Photo Rag Baryta, the resulting print is continuous-tone as far as the eye can tell, even for monochrome. Put a 10x loupe on it, and it starts too look a little grainy, but you don't see actual dither pattern; and wet prints are going to look grainy under a loupe too, unless they're 8x10 contact prints. Personally I don't much care what my prints look like under a loupe anyways.
-
As a (relatively) inexpensive experiment I've found and ordered from eBay a highly regarded, but discontinued, Kodak 1400 dye-sub printer for $85 plus shipping (from Texas), which is little more than the cost of the paper-package I'll need to also purchase to use it. I expect I'll initially be printing in glossy (although matte finish paper is also available, which is my preferred finish for my ink-jet printing, but few users seem to like in dye-sub). I think that if it doesn't work usefully for me, my wife will love it for her color snapshot work. So it's possibly a win-win situation for me.
-
It was mentioned here that there is almost no technical support/info available for printing with black and various dilutions of gray. I subscribe to a great Yahoo Group in their Digital Photography section. It is called DigitalBlackandWhiteThePrint. It averages 200 to 500 messages a month, and has a massive amount of expertise and experience available.
If you're not aware of it, it will be worth your while taking a look.
Brad
-
Brad: Many thanks. I am in fact not only aware of it (and its QTR companion) but have also posted on it and had private email contact with several of its leading contributors. It is indeed a wonderful resource for those like myself interested in black-and-white photo printing. But it nevertheless has to date not provided much of a significant solution to the problem I've elicited here regarding inherent ink-jet limitations for small photo printing. With the current technology of otherwise excellent ink-jet printers it appears just not to be possible.
-
Divide the horizontal pixel count (the number of pixels across) by the desired resolution. The result gives you the maximum print width (in inches).
So 9,000 pixels at 360 ppi gives you 25 inches... very useful if you use 24" roll paper!
-
Jim’s original point about the problems of making small inkjet prints was a perfectly valid one, and I for one will be very interested to know how he gets on with dye-sub printing as an alternative. Yesterday, though, I had a kind of Road to Damascus moment which will not entirely solve the problem but does make me re-think the whole issue.
I realised that I, like everybody else, have been trotting out this mantra that native resolution is 300 dpi for HP printers and 360 dpi for Epson. But then I remembered reading on Eric Chan’s site that resolution for the Epson 3800 can be either 360 or 720 dpi, depending on the “Finest Detail” switch. And somewhere else, I can’t think where, I seem to remember a discussion which came to the conclusion that all the modern Epson photo printers are 720 dpi. So did that mean that even my good old R2400 actually had a potential output quality which I had never previously tapped into?
Now that Lightroom 3x has the ability to re-sample and output to print at 720 dpi, I could easily set up a quick trial. Remember that all my work is B/W, printed via the Epson ABW mode using Eric Chan’s paper profiles. Whether that affects the results, I don’t know. It shouldn’t. For the test, I needed to create a situation where LR and the printer would be discarding information from the file, at least at 360 dpi, so I set up a small 5x4 ins print of a very detailed subject which was technically perfect (lighting, exposure, focus, no camera-shake). I then printed two versions at the highest-quality settings on the 2400, one from LR at 360 dpi and the other at 720.
Superficially, they look identical. But under an 8x loupe there is a clear difference, with more fine detail, smoother transitions and sharper edges on the 720 dpi version. In fact, I can even see the difference with the naked eye, because I have very short-sight and I can focus without my spectacles at around 6 inches. So the R2400 definitely does print at 720 dpi, otherwise it would simply have re-sampled the file to 360 and the prints would have been identical. Which presumably means that the 3880, 2880, 4880, etc can also do so.
Obviously, for most people printing large colour or B/W prints this is all pretty academic. At any sensible viewing distance, 360 dpi (or even 300 dpi for that matter) is quite enough. And as others have noted, you don’t normally view your prints under a loupe. But for printing small prints it does make a difference. As I mentioned before, to include all the detail from my 39MP files at 360 dpi I would have to print at (more or less) 20x16 ins, or A2. But at 720 dpi I can print at 10x8 ins without losing any of the detail in my file. And below that, at 7x5 or 5x4 there is still a gain compared to using 360 dpi. Small prints are viewed at much closer viewing distances, and that little bit of extra smoothness is noticeable to me.
John
-
On this kind of experiment, there is a (very good as usual) Ctein post on TOP :
http://theonlinephotographer.typepad.com/the_online_photographer/2010/01/how-sharp-is-your-printer-how-sharp-are-your-eyes.html
with interesting developments in
http://theonlinephotographer.typepad.com/the_online_photographer/2010/05/more-on-printer-sharpness.html
-
On this kind of experiment, there is a (very good as usual) Ctein post on TOP :
http://theonlinephotographer.typepad.com/the_online_photographer/2010/01/how-sharp-is-your-printer-how-sharp-are-your-eyes.html
with interesting developments in
http://theonlinephotographer.typepad.com/the_online_photographer/2010/05/more-on-printer-sharpness.html
I ran Ctein's experiment, printing contact sheets in both 360 dpi and 720 dpi and examining them closely with both an 8x and 10x loupe. And yes, the 720 is slightly moire finely defined than the 360. But how AWFUL they both looked under a loupe! I hadn't realized what the naked eye doesn't see (and how fortunate it doesn't) to preserve an illusion of detail. My dye-sub printer doesn't arrive until this weekend, although I'm entertaining few expectations that it'll prove to be significantly better. Tones possibly smoother, but edge detail less. But for me anyway it should prove interesting. Will report back.
-
Nicolas
Thanks for those useful links. I had read the articles, but had forgotten where they were. I do take issue with Ctein in one respect, though. There surely must be only one "sweet spot" for the best possible print resolution, which is where we send the printer a file at its native ppi so that the printer driver is not forced to resample the file again. If the Epsons use 720 ppi, then that must be it. Adding more ppi in our editing software should not produce greater quality of output, because the printer simply cannot make use of it.
For any branch of photography, we must be aware of the inherent limitations of the technology we are using. Once one moves out of the wet (chemical) darkroom, we are of necessity embracing a new world of photo-mechanical reproduction, which is essentially based on using a screen, just as it always was. And to date, there is no way any screen-based process can equal the smoothness of tone and quality of detail which (unenlarged) fine-grain film and silver printing paper can deliver. It just can't be done. This has nothing to do with the sensor in your camera or MF DB. Even if you had 1,000 MP in your sensor, you are still limited by the same old output quality of your printer at 7x5 ins or 10x8, or at any size below a 1 to 1 print at printer output resolution.
Mostly, this simply doesn't matter. The convenience and control which digital printing offers more than make up for the losses. And my own tests show that at normal viewing distance, a 10x8 B/W print from my Epson holds its own against the best I could do in the darkroom fom MF film. As far as really big prints go, there is no contest. The inkjet printers can roll out fabulous 20x16 and larger prints consistently time after time while I would still have been desperately trying to focus the damn enlarger on a piece of paper pinned on the wall six feet away.
But for small prints, amusingly enough, you will still get a better result by loading up your camera with Kodak Gold and taking the film down to your local Jessops for D & P.
John
-
There surely must be only one "sweet spot" for the best possible print resolution, which is where we send the printer a file at its native ppi
This is more or less addressed in the second link I gave... Seems that driver interpolation is done well enough for this not to be very significant.
But for small prints, amusingly enough, you will still get a better result by loading up your camera with Kodak Gold and taking the film down to your local Jessops for D & P.
I thought these people used now some Frontier or Noritsu digital minilabs? And therefore limited to 300 or 400ppi...
For me, it's only with contact print that film could have that (http://www.custom-digital.com/2008/09/bw-print-quality/) an edge.
-
I thought these people used now some Frontier or Noritsu digital minilabs? And therefore limited to 300 or 400ppi...
For me, it's only with contact print that film could have that (http://www.custom-digital.com/2008/09/bw-print-quality/) an edge.
Nicolas
I think you are right. I am somewhat behind the times. And another useful link, thank you.
So for any real progress on this issue, perhaps we should expect (or hope for) -
* Printers with a dot size of less than 1 pico-litre
* Variable dot size over a wider range
* Papers with minimal or zero dot-gain
* And a native printer resolution of 1440 ppi or better
* With 16-bit capability for colour and B/W
Don't hold your breath.
John
-
The lesson I am (with regret, if not actual dismay) learning from all the enlightenment I am receiving from others in this thread is that it apparently has benefited me naught (or very little, anyway) by "upgrading" from my wonderful little Panasonic Lumix LX1 to such larger-sensor cameras as a Sony DSC-R1 and Lumix DMC-GF1. (Although I am taking advantage of the ability with the GF1 of being able to utilize my Nikkor lenses from my old discarded 35mm Nikon F2.) My aesthetic preference for intimate, small-sized prints in this digital age to date is an apparently quixotic quest. (Hope I'm not becoming too literary.) Oh, well......
-
Jim
Without wishing to labour the point, but as a further thought –
If you would really wish to revise your notions of photographic quality, take the opportunity to examine historic prints made with a fine lens on a large field camera (10x8 or bigger), and contact-printed from glass plates (heavy and perfectly flat) on old-fashioned printing-out paper (POP, which holds detail in the shadows and allows extended development in the highlight areas). For B/W photography they define the ultimate in fine detail and perfect smooth tonality, and make our efforts now (100 years later) seem rather cheap.
John
-
I thought I'd contribute a few observations to this very interesting thread.
I think the 1.5 picoliter 1400 is about as sharp as any digital output device we have available to us for home use. It should be able to make a cleaner looking edge than the 3.5 pl printers. The printer's native resolution for the input image is 720 ppi, which is the same as all Epson desktop printers, I believe. Nonetheless, the drop size can make a slightly sharper looking image.
However, I have seen no inkjet that can match a contact silver print. The highest print resolution I've been able to get on the paper, using standard resolution charts, is about 15 lp/mm with the inkjet printers, versus about 25 lp/mm for silver prints.
Incidently, the best matte papers are often sharper than the glossy papers.
With respect to grayscale steps, etc., using the lightest 2% ink in the Eboni/carbon-6 inkset will help. A standard LLK is about the same as the 6% Eb6 dilution.
With the 2%, using the Epson driver with an ICC that contains a curve that uses the light inks to the max, you'll get lots of separation. With this workflow you have a high bit pipeline all the way to the pirnt, not the 8 bit pipeline ofthte ABW approach.
Paul
www.PaulRoark.com
-
I suspect that my comments now regarding my Kodak 1400 Dye Sub printer really belong at this point under another forum category, that dealing with Printers, Papers and Inks. Anyway, the printer arrived today and after an easy built-in calibration of it to achieve the most neutral gray tone, I printed off one of my .tif photos with pronounced gradation aspects. As a first run-through I was greatly relieved. No color cast to the gray tones at all. I printed on the same (fairly dear, i.e., costly) page one of my wife's color jpeg photos, which came out in a very close match with the color tones she had intended when initially processing it in Photoshop. I examined my black-and-white with a loupe and saw, as I had only somewhat expected (and hoped for), no ink jet pixel dots, but photo textures emanating principally from the textures of the objects I had photographed, but also with a surface texture mildly representative of the crystal grain one gets from wet photography. I believe my description here is likely rather vague, imprecise -- and inadequate -- but with my limited technical background, it's only what I'm really capable of conveying. So, there you have it, for what's worth. Regards to all.