Large Prints and EOS 5D Mark II

Discussion in 'Canon EOS' started by romain_j., Mar 16, 2011.

  1. Hi all,
    I wanted to know about your experiences to produce large format prints with the Canon EOS 5D Mark II white USM lenses. By working in RAW format images. Allows it print very large sizes approx. 1m x 1.80m? What type of extension support it from a good file? Any advices ?
    Thank you,
  2. Romain,
    This is a very argued over question. Basically it depends on how far the print is realistically going to be viewed from and just as important, what the subject matter is. Any camera image can be enlarged to a billboard, but it might only look good if you view it at 55mph from 200yards away on the highway!
    I regularly print 20"x30" and sometimes go to 24"x36", these I expect to be viewed quite close. I'd print a 70" print if the viewing distance was further. Fine detailed images suffer more, so trees and foliage are more difficult and don't enlarge as well as sky, for instance.
    The best thing for you to do is print a small section of your image at the full size, say a 12"x12". Tack that onto a wall and see if the quality will work for you.
  3. Thank you for your reply.

    I noted that many photographers who work with this camera sell their edition of prints of their large size, approx. 36x48 inches (for larger).
    To make prints suitable (which will look at 3 / 4 meters away), from images rigorously taken (tripod, iso minimum, RAW), the 5D may be issued prints are acceptable?

    Scott, have already tested for prints larger than 24x36in. ?
  4. What Scott said. 3-4 meters viewing distance is pretty far. I imagine you could do 70" at that distance easily.
  5. Romain,
    No I haven't printed above 36" on a single capture. But I agree with Daniel and think that for a 3-4 meter viewing distance then 70" should be fine. But please do a small area test print first, this is all very personal, some people won't print a 5D MkII file over 14"x21", others are very happy at 64"x96".
  6. Hello....
    You can see what your output will look like at 1x1.8m for the cost of paper and ink at 8 1/2" by 11".
    1. Open file in photo editing program.
    2. Select print.
    3. Use the same premium paper that you will use for the full sized print. Set the paper type inquiry for the proper type, i.e. super glossy, set print type to highest.
    4. Back at main printer window, select center on page. Then type 1m into either height or width boxes.
    5. Print.
    You'll get the center 8 1/2 by 11" of a 1 Meter+ print. One can also do it with 4X6 paper, which is obviously about the fourth of materials cost for 8 1/2 by 11. However, you may not get enough square area to make a good decision on whether you've sharpened enough or too much, etc.
  7. I just did the above with a Canon 5D II "sample" file photo off of here:
    The image is 3744 Pixels by 5616 pixels. I set the short side to scale as if the print would be 40" (close to 1M) and printed the center section on a 8 1/2" x 11" piece of Epson Glossy Photo Paper with an Epson 1400.
    It looked pretty good. Why not try it with your printer and premium paper.
  8. Art,
    That is not a good image to test the enlargement potential with, it is a 200mm image shot at 1/200 and shows definite signs of camera shake. A sharp image will look much better.
  9. Thank you for your advice.

    Unfortunately I have no printer, so I inquire here. I usually make my prints in alaboratory. So no paper tests.

    Art your link does not work ... : (
  10. Romain,
    What kind of stuff do you shoot? If you tell me I'll email you a section of a big print, get it printed by your lab and then you will know for sure.
  11. I use a 5DmkII with good lenses,tripod,RAW, good trusted lab,lambda print, Fuji paper and never go above 36x24 and only then at a stretch. Any bigger and the print stops looking like a photograph and starts looking like just a big image. The 'flat' appearance that all DSLRs produce starts becoming really apparent above that size and is a problem for me. If its just the resolution that really concerns you, you could probably go a little bigger before the detail degrades but not much.
    On the idea of an ideal viewing distance that you hear people talking about- like a big print has to be viewed at a certain distance...How does that work, I don't get it. What you going to do put a rope around it to stop people looking close? If you watch anyone look any print(and especially big ones), they stand back for a bit to look at the whole thing and then they go close when they want to examine something they've seen, then they go back a bit and so on, constantly changing their distance as they respond to the image and examine it. If when they go close to look and all they can see is a degraded,smudgy detail thats generally an unsatisfactory experience and people notice that. This only applies of course if what you want is detail ( which it sounds like you do ).
    My point is whatever your 'ideal' viewing distance is, your viewer is probably going to have another idea and that will be the real experience of looking at your photograph.
  12. The image was the one that was handy and had some detail in the center where it would be put in an 8 12 x 11 center snip of the 40" x 60". It still looks pretty good. Try Canon 5D II sample 32 and see the good line definition.
  13. Link does not work...
    Truth, It did work for about 5 minutes and then I got a timed out message also. Go to Dpreview, Reviews, Canon 5D Mark II and check out the sample page (about the last page in the review). You can tap on a thumbnail, get a larger view and from that page select the full size. Save and open in your photo editing program and try the printer routine at whatever size you want to. The printer will only print out the paper size of the requested print size.
  14. All the above being said, for myself I try not to print at less than either 300PPI or 360PPI depending on the printer input specs (Lightjet, HP and Canon @ 300PPI, Epson @ 360PPI and even my Fuji Pictograph @ 400PPI and Sony dye sub @ 403PPI). That is why I usually use 6x7, 6x9 and 4x5 FILM cameras for large prints.
    3.8" @ 3150PPI scan gives me 11970 pixels on the short side. 11970/300=39.9"
    4.8" @ 3150PPI scan gives me 15120 pixels on the long side. 15120/300=50.4"
    and produces a tack sharp 40"x50" lightjet print.
  15. I have made 40x50 inch prints from the original 5D. My assistant had the camera and didn't really prep them as well as I think they could be. They were incredible even up with your nose in them. There were a couple of places where they could have been better, but you had to look hard to find those places.
    I have printed 40x60's with the 1dsmkIII, same size files as the 5dII, and they have no flaws at all. Depending on the image, I think you could even go larger, but in most cases I do think that is close to where some issues might start to show in most images.
    Although I haven't seen them, I do understand that Gursky's very large prints=over 8 feet aren't all that wonderful up close, but sell in the millions. What is there to be concerned about?
  16. From your experiencer can print to the 40x60inch approx. without too much damage.

    Art Thomas: that you use MF and LF cameras for better image quality. Review foryour test with the 5D should be fairly accurate (as you know the quality produced by a4x5).

    By cons for such expansions with the 5D, the resolution must be reduced to 72DPI tobe large as 40 "x60" size image? no?
  17. also, if you're worried about large prints and high image quality, I would use prime lenses.
  18. How big can also depend on the DPI setting you have to print at. 200, 250, 300 can make a big difference.
    I've been using OnOne Software's Genuine Fractals (now renamed as Perfect Resize). It takes much of the guess work out of equation. The results are much better than anything you get out of the native Photoshop tools.
  19. Also, forgot to mention, if what you need to blow up is a static type of scene (nothing moving), then don't forget the option to break the scene into multiple smaller photos and then stitch the photos together.
    I've had great success at that. Took 24 pics in a 4x6 grid with 30% overlap and as able to blow it up native to 40x60
  20. My Photoshop reports that a panorama I did had a native size of 30" x 195". see here:
    The example is a reduced size, but zooming in will show a huge amount of detail, based on stitching 10 or more images.
    Trouble is, I have no printer that can do it justice.
  21. My first digital camera had a 0.97Mp sensor. In my (then) blissfull ignorance of all things digital, I printed a nice shot of my daughter on my consumer color printer at 8.5x11. It looked great. I calculate that this picture has barely a hundred pixels per inch stretched over those 11 inches. This experience gave me a considerable skepticism about the holy 300 dpi rule. It's the shot that counts. Use the best lenses you can afford, stop down, use a tripod -- all these things count for more than how many dpi you print at.
    Oh, and get yourself a decent letter size printer. They're cheap now. What everbody above says about using a test print at the same enlargement you plan for your big print is on target. Really big prints are really expensive, and even the best monitor isn't good enough to show you what you'll get in the print.
  22. @John A
    Gursky, like most of his fellow members of the Dusseldorf School is famous for shooting on Large Format (8x10) negative film( with the possible exception of Candida Hofer who occasionally goes down to MF). Also Gurskys images are intensively manipulated, possibly involving the stitching of multiple 8x10 negatives together in a single image for extra post- modern grandeur. The detail is of course, immaculate. I doubt if you'd get any of them giving more than a passing glance at my lowly 5D mkII.
  23. I have an Epson 7900 and do my own printing. This is a 24" wide printer. I shoot a 5D2 and also work with earlier images form a 5D.
    As mentioned earlier in the thread - and no, I have not read the entire thing - the definition of "good" is a very subjective thing. It depends a lot of factors including the subject, how the images will be displayed, and your individual ideas about where the boundaries for excellent image quality lie.
    Most of my work is shot from a tripod and shot fairly carefully with excellent lenses. I know what I'm doing in post in regards to the factors that affect the objective and subjective levels of sharpness in the prints. My feeling is that - for me - a 24" x 26" print of quite high quality is a reasonable expectation when all of the factors that affect sharpness are well handled. I also know that - for me - not every image will necessarily support my ideal print quality standard at this size.
    That said, if the image will not be inspected too closely or if the subject and display are such that totally optimal print resolution isn't necessary or with originals of exceptional quality, it should be possible to print a bit larger. However, if your target print output is regularly going to by in the 1m x 1.8m range (about 50% longer in both dimensions than what I mentioned above) then I think you might want to consider MF digital as long as your subject is amenable to that format.
    Of course, you can always try making a print at this size from your best original and see what it looks like. Alternatively, you can go through the process of generating the print file for such a print and then print a small crop out of it.
  24. Unfortunately I have no printer, so I inquire here. I usually make my prints in alaboratory. So no paper tests.
    You can still do the test as described - just have the lab make the letter-sized test image for you.
    On the other hand, I recommend that anyone who is going to use a lab also try to have a small high-quality printer in house, at least if optimal print quality is important to you. It is not possible to fully know what a print will look like if you only preview on the screen.
    BTW, as is often the case when this question comes up a) the "answer" is highly subjective, b) you'll have to try it to know how you feel about it, and c) you'll undoubtedly read a lot of "common wisdom" that is just plain wrong or at least very much open to argument. For example, I now see a mention of "300 dpi" and "360 dpi" standards for good prints. That is just plain untrue. In fact, some of the commercial print shops are using processes that natively use lower print resolution than that. With very small prints - likely to be held in hand and viewed very closely - higher resolutions may be useful. With very large prints - even those that must retain sufficient detail to stand up to closer inspection - most who print a lot will agree that falling below a native resolution of 180 is risky, but that excellent work can be done with resolutions between that value and commonly quoted higher values.
  25. For those who feel mentions of 300 dpi (if you're referring to me, sonny, I carefully used ppi because many inkjet pixel areas are made up of several dots of ink), and 360 dpi as being just plain untrue, I obviously disagree. If you had said it isn't necessary for you, I'd say fine. If you had said a lot of people would not appreciate the finer pitch provided by those numbers (if properly given in ppi), you'd be right. If you had said that many people's visual acuity does not measure up to those standards and therefore it is useless to them, you would be right. But that's not what you said.

    Yes, I know the world, and especially America, is dumbing down and rapidly. Some years ago, when either Nikon or Canon first came out with a pro-bodied extremely expensive SLR of approximately one megapixel, one of the self-proclaimed professional photographers in town bought one. I considered him to have more money than brains. He often used it to put out 20X30" supposedly professional photos that looked pretty sharp, but only from across the street. I remember one he brought in and proudly showed off at one of the local photos store that was either 3X4' or 4X5'. They looked terrible and yet he was so proud of his work. He bragged about how he could palm these off on people who didn't know any better and it would save him so much darkroom time and expense. He sneered at the ignorance and poor taste of his customers and proclaimed that this type of work was essentially the wave of the future. He was right and people who make improperly definitive statements about lack of quality are his supporters.

    Back when many large corporations and even normal everyday American values felt they had to adhere to a certain high level of standard (yes, I'm going wa-a-a-ay, wa-a-a-ay back), a Leica had a goal or standard of providing print acuity of 7 lp/mm. Yes, it took pretty good and young eyes to fully take advantage of that value, but it was still their value. I think Kodak even rated "acceptable sharpness" for enlargements up to 4 lp/mm. Four lp/mm is close to 200 PPI; obviously 300 PPI is pretty close to 6 lp/mm, which is pretty close to the Leica standard.

    Just for curiosity, I took a scanned medium format slide of around 8500 pixels/side and printed it at 6'X6'. I was surprised at the detail, even from a foot away. If it was in a gallery with a rail of rope 10-15' from the print, I suspect that only some of the youngest military pilots with their keen eyesight would be able to see any unsharpness or lack of detail.

    Yes, everything is relative. You have your standards and you're entitled to them. But it doesn't make somebody who adheres to a traditional higher standard as saying something that is just plain untrue. You would have more credibility, or maybe even some credibility, if you said that you, a product of the dumbed down society, choose to lower your standards and accept work of lesser quality. You could have also used a few words from your prior sentence that said it was open to argument because an argument can take into account such things as personal standards, distance of viewing, and pride in workmanship.

    When you say that a standard is just plain untrue, I'm afraid I beg to differ with you. But on the other hand, why beg? Your statement doesn't deserve any begging, so I'll just say I differ with you and have pointed to evidence, rather than conjecture.

    P.S. Why 300 or 360 PPI? Take Hewlett-Packard, for instance. Some of its better printers, at 300PPI, put down 4 ink droplets in a square box per pixel. The ink droplets are so small that the eye picks up that combo of droplets as a particular color by averaging out the color of the various ink droplets. If you input to an HP printer at 75 PPI, 150 PPI, 300 PPI, or 600 PPI, the printer software can evenly divide out or multiply the input material and come up with a reasonable solution to keep from lessening resolution more than necessary. On the other hand, what if you input at 360 PPI? The printer software is going to have to interpolate as it reduces the pixel information by an uneven proportion. Straight lines are going to find themselves a little squiggly. Curves are going to be jagged. It is the same when inputting 300 PPI material into a 360 PPI machine. Again, all the pixels are going to have to be reformed in an uneven basis. You can't just double length and width of the pixel, ending up with 16 ink droplets, 4 of each of the original 4 colors. The printer driver, printer and computer have to make some decisions, which won't necessarily make the output look exactly the same as the input. This is especially true where there are vertical or horizontal lines very slightly off axis, and all curves. It can also do such things as show a window with some panes wider or narrower than others, as it tries its best to reproduce a decent picture, considering the intelligent programming that went into the driver and software.

    All of the above being said, other people are still entitled to their opinion, as I think I am, but calling it an absolute, through the superiority of ignorance, is a bit much.

    Opinions are like rectal orifices, most everybody has one. I obviously think my rectal orifice is just as worthy as anyone else's.
  26. I'll attempt to avoid references to anatomy and the level of intelligence of "the world, especially America."
    The 300 ppi resolution notions come from commercial printing - (e.g. magazines and newspapers; not photographic printing in the sense we discuss here) which uses an entirely different process of creating photographic images. It is not a number that evolved from or was discovered through photographic processes.
    A number of very critical photographers and photographic writers have tested the performance of inkjet printers extensively. While you can find still a few people who, like you, hold to the view that inferior photographs will be the result of using resolution lower than 300 ppi, you will find many, many more who have discovered by comparing output at different ppi that this seems to be more myth than reality.
    Of course, anyone who thinks that you or I or both of us are wrong can easily produce test images at the various ppi values and check the results themselves. The best way to judge the resulting images is, by the way, to make sure that you don't know which you are looking at. I recommend perhaps making three images, possibly two at 240 ppi or whatever value you want to try and one at 300 ppi or your choice of higher value. Then hand a reasonable test subject the three prints and ask him/her to line them up from right to left in order of increasing quality, or perhaps to determine which two are the same and which is different along with evaluating which is better.
    I've done a bit of testing myself. I originally was very concerned about maintaining the 300 or 360 ppi value in images that I sent to my printer or to my service. I jumped through the usual hoops and then some to get there. Then I heard Jeff Schewe (look him up if you don't know his background) make the seemingly controversial claim that it is better to print at whatever resolution naturally results from resizing rather than interpolating to get one of those famous 240, 300, or 360 ppi values. For example an image that might have a resolution of 410 at small dimensions might end up a 210.56 at some larger dimension. His notion was based on the observation that a print interpolated in software before going to the printer gets interpolated twice, the observation that 180 or higher produces a very fine print, and the recognition that the highest ppi is most important in the smallest prints rather than in the largest.
    I thought he was nuts.
    I decide to try this approach - basically to prove that his idea wasn't right. I tried it. It worked very well. I learned something that changed my approach to printing and which caused me to think a lot more critically about some of the "witchcraft" that surrounded the inkjet printing process.
    Back to the OP's question, based on my own printing I'll stick to my opinions about relying on a full frame DSLR to regularly aim for a target print size of 1m x 1.8m. My view seems to be echoed by others who posted in the thread.
    For what its worth, most who see my prints remark on their high technical quality.
    (Apologies for my accidental transposition of the "dpi" term. After years of use I'm still trying to train myself to not make that error when I write. I understand the difference between the two terms when I write, for example, that I might print a 200 ppi image at 1440 dpi.)
  27. 300-360 ppi required for good quality prints? Here 's a link to a blind test conducted to determine if it is really required in practice.
  28. Hmmmm. Certainly a lengthy and well-written response with a lot of thought put into it. I also see it as a challenge to either subtly or blatantly put in another reference to anatomy. After all, it's all good fun.
    Yes, 300 PPI is thought to equal approximately 150 screen lines, which is considered a higher standard printing resolution usually associated with high-quality glossy-paged books and magazines. As long as I'm answering a pseudo-logical common knowledge reply, I'll do the same and say, "Gosh, why is 150 screen lines any type of a gold standard for printing? Could it be that printers long ago realized their reader/viewers would recognize that as sufficiently sharp and, in fact, be willing to pay additional monies for the additional effort?" Sure, I was around to see World War II era newspaper photos with extremely low quality and low numerical line screens. I remember the first color funnies, whose dot pattern could be duplicated by a bold-pointed ballpoint pen. I enjoyed seeing the pictures and reading the funnies, but I also know things have gotten better, clearer, crisper, and, to me, nicer.
    Even at my age, I have unusually acute sight. Perhaps in World War II I could do better, but today, I still exceed 10 lp/mm. I enjoy and appreciate the difference. I have a friend about my age who could not see detail, even with glasses, on many pictures we took while together. In fact, he would claim it just wasn't there ("If I can't see it, it can't possibly be so."). I was frustrated, but thought I would do a little research before I tagged him as being either obstreperous or ignorant. I bought a chrome-deposited-on-glass test target from Edmund Scientific and laid it on a light box. Glasses and all, from nose close to about a foot, he could sometimes discern 2 lp/mm but usually 1. I may be dumb, but at least I figured out anything that required more visual acuity to him than 1 lp/mm would be lost to and of no value whatsoever. Still, that does not mean that a 300 PPI standard for good sharp prints is invalid simply because some people simply do not have that visual acuity.
    Likewise, age and large bore gunfire has not been kind to my hearing. In my youth, I could discern most of the individual instruments in a live symphony. Today, I'm lucky to differentiate instruments where the band consists of a set of bongos and a Jew's harp (all right, all you pendejos who would call me out for political incorrectness, please also stop using terms such as Irish stew, Mexican food, and Italian dressing.). That does not mean that a major symphony can't use a third and fourth violin. I would also go out on a limb and say that a major city's symphony that only employed one violin might be understaffed or even substandard. On the other hand, they might be able to play music that was very pleasing to my beleaguered ear.
    I believe I did mention above that I had test-printed a file of 8400 pixels on a side at a size of 6' by 6'. Oops, it was actually 7' by 7' and there's a lot of detail available. A foot away, it looks a little soft, but sitting in my computer chair, 7' away, especially without my glasses, it looks pretty sharp. I'll bet it is sharper than the best color print I could make, back in the MSDOS days when my color printer was a Star 10 with a nine-pin print head and a four-color ribbon. I thought that Star color printer was pretty neat stuff.
    Yes, I know that today, all America is asking, "Why have standards?" Most people I see are more concerned with getting by, doing the minimum, asking the minimum, and being the minimum. With the sincerity of a weak, but trembling voice, they say, "It's hard."
    Me? I understand your points and they are not invalid, especially in today's society. However, I reserve the right (realizing that it may be taken from me) to apply reasonably high standards and to speak of them in the various discussions. Yet, I can still enjoy reading a 60 year old funny paper with their large and crude dots. I just like to continue thinking I know the difference.
    P.S. I confess...I did put in an anatomical reference, just like you knew I would. Hey, smile and laugh. Photonet is meant to be informative and enjoyable. We can all have different opinions, and even see different truths, and still have the camaraderie of our mutual photographic interests.
  29. Yes, 300 PPI is thought to equal approximately 150 screen lines, which is considered a higher standard printing resolution usually associated with high-quality glossy-paged books and magazines. As long as I'm answering a pseudo-logical common knowledge reply, I'll do the same and say, "Gosh, why is 150 screen lines any type of a gold standard for printing? Could it be that printers long ago realized their reader/viewers would recognize that as sufficiently sharp and, in fact, be willing to pay additional monies for the additional effort?"
    300 ppi does not equal 150 lpi. PPI is a statement of the density of continuous tone imaging elements. LPI is a statement of dot spacing in a commercial CMYK printing press. The 300 ppi meme comes from the rule of thumb that artwork resolution in ppi should be 2x the line screen frequency for a press run, and from the fact that 150 lpi is common in magazine work. This is related to the mechanics of producing a halftone screen and does not indicate equivalence. A modern photo laser or ink jet print of a 300 ppi photograph will yield resolution, sharpness, and tonality far superior to a 150 lpi magazine print of the same image.
    That rule of thumb is not hard and fast even in commercial printing. 1.5x the line screen is typically just fine. The number is certainly not a photographic standard and not of any importance when printing on photo laser or photo ink jet printers. Photo lab laser printers are continuous tone devices, and photo ink jets do all kinds of tricks with dot placement which shatter the rigid halftone screen found in commercial presses.
    So what is the ideal number for photographic printing? You can't tell from looking at high contrast resolution charts or line art because visual acuity is highly dependent on detail contrast. With B&W line art people can generally tell the difference past 1,000 ppi. But with photographs many people fail to notice differences past 200 ppi. 180 ppi is good for photographic prints, and 250 ppi excellent. There's not much to gain in photographic prints past 250 ppi unless those prints contain a great deal of high contrast, monochromatic detail.
    And there is no need to interpolate images for photo laser and ink jet printers unless you have an exceptionally low resolution image, in which case you can reduce pixelation and jagged edges through interpolation. For images of sufficient resolution the printer drivers are more than capable of mapping the pixels to the print engine's pixels or dots. It's better to leave any possible artifacts from your own interpolation out.
  30. "Gursky, like most of his fellow members of the Dusseldorf School is famous for shooting on Large Format (8x10) negative film"
    Andrew, I am not sure of the point, however, I watched a documentary on Gursky making his Miners image (it is on YouTube). He debated between using his 4x5 or his Hasselblad H system with a digital back. He ended up shooting the main scene with the 4x5 film and then some of the "add-ons" with the Hasselblad. My point isn't, or wasn't, that he doesn't ever use an 8x10, but that the large prints, very large ones, suffer in quality from what I understand and that perfect quality isn't always necessary.
    The funny thing is that today many of those we associated with the 8x10 are shooting digital in some form. I have seen Alec Soth using a 35dslr and Misrach with the Hasselblad system--apparently he is using this for his latest work and he also used a 4mp digital camera to make his Katrina series--which I have seen in person.
    The most important thing is to upsample to an appropriate dpi for the printer, generally that means at least 200dpi for a Lambda and 300dpi for certain rips used with inkjet prints (mine).
    As to Genuine Fractals, I haven't tried the latest version, but found it to be better, but neglibibly, than using Photoshop's native algorithm for upsampling--tested on 60mb(8bit equiv) file to 850mb. I could only see a difference when the upsample was viewed at 100% (equiv of 18' wide print) and it was very minor at that size.
  31. Art dissembled:
    Yes, I know that today, all America is asking, "Why have standards?" Most people I see are more concerned with getting by, doing the minimum, asking the minimum, and being the minimum. With the sincerity of a weak, but trembling voice, they say, "It's hard."​
    Right. When others hold a different opinion and you have no other meaningful response, what remains is an ad hominem.
    I'm done.
  32. Exactly Dan -- no doubt Art is seemingly coming from far Outer Space or something... it reads that way and with his fixation on orifices like an alien prober?
    No doubt Art would tell me that my seemingly high quality (my words) 19x13 inch prints from my lowly 10D are only hi Q in my mind because of my "low" standards (not because they are better than nearly any large print I've ever seen anyone else attempt with such a high noise, 6MP sensor).
    LOL. I'd like to see Art's online portfolio somehow too.
  33. The 300 ppi meme comes from the rule of thumb that artwork resolution in ppi should be 2x the line screen frequency for a press run, and from the fact that 150 lpi is common in magazine work​
    I believe the 300-360 ppi number comes from the fact that a person with 20/20 vision can resolve between 6-7 lp/mm at a viewing distance of 10 inches. 6 lp/mm is about 300 ppi, and 7 lp/mm is about 355 ppi, rounded to 360 ppi.
    Kodak mentions in its film datasheets that "the standardized inspection (print-to-viewer) distance for all print sizes is 14 inches, the typical viewing distance for a 4 x 6-inch print". At 14 inches one can resolve about 4.83 lp/mm which is about 245 ppi. So a print resolution of 240 ppi, is great for most prints.
  34. I believe the 300-360 ppi number comes from the fact that a person with 20/20 vision can resolve between 6-7 lp/mm at a viewing distance of 10 inches. 6 lp/mm is about 300 ppi, and 7 lp/mm is about 355 ppi, rounded to 360 ppi.
    The actual numbers are 335 ppi for 6 lpmm and 391 ppi for 7 lpmm. (Your Nyquist multiplier should be at least 2.2x and 1 inch = 25.4 mm. Though I do understand that most people use 2x and round off.) 6-8 lpmm is the traditional range given for a high quality analog print, with the caveat once again that this is printing test charts and human visual acuity will be lower with a real photograph but higher with monochromatic text or line art. Most people will notice the difference in text or line art printed to 300 ppi, 600 ppi, and 1200 ppi. Repeating B&W lines simply do not test or mimic the full range of conditions which influence a human's ability to perceive detail.
    Still, the commonly quoted 6-8 lpmm figure is not the source of these numbers. LPI is related to the difficulty of press registration over a run and ink bleed for a particular paper. 150 lpi was found to be a good compromise between viewer acuity and paper, ink, and press costs for the magazine industry. And once again, 150 lpi != 300 ppi even though that's a commonly given value for input into the halftone screen. I have photo books printed at 200 lpi and a print from my Epson 3880 at 300 ppi, or even 180 ppi, still looks much better on close inspection.
    That these numbers fall within a certain ballpark range is governed by the fact that all this technology is for human eyes. But they're still not equivalent or derived from each other in the way most people assume.
    360 ppi comes purely from Epson btw. It became a meme that you needed to feed Epson photo ink jets images at a multiple of the print engine resolution of 1440 or 2880. If that was ever true with early print drivers when PS might have done a better job scaling, it most certainly is not true today.
  35. The actual numbers are 335 ppi for 6 lpmm and 391 ppi for 7 lpmm. (Your Nyquist multiplier should be at least 2.2x and 1 inch = 25.4 mm. Though I do understand that most people use 2x and round off.)​
    I don't see how the Nyquist sampling frequency applies here as we are not sampling anything. 6 lp/mm=6*2*25.4=304.8. Similarly 355 ppi holds for 7 lp/mm. The 6-7 lp/mm range is derived from the fact that 20/20 vision can resolve 1 arc minute at 10 inches.
    Still, the commonly quoted 6-8 lpmm figure is not the source of these numbers. LPI is related to the difficulty of press registration over a run and ink bleed for a particular paper. 150 lpi was found to be a good compromise between viewer acuity and paper, ink, and press costs for the magazine industry.​
    I am aware of the 150 lpi number used in the printing industry. However the 6-7 lp/mm resolution limit for the normal human eye provides the theoretical justification for the 300-360 ppi image resolution required for a sharp print. But as I mentioned in my previous it is not really necessary.
    In fact a better measure for print resolution is to use the SQF (subjective quality factor) by determining the resolution at which the image retains good contrast, and printing at an enlargement that can provide 3 lp/mm for that resolution. For example if you can get strong detail in an image at 2400 lw/ph, then you can get a very sharp print at 16x24. The latest dslrs can do it.
  36. "Ad hominum" ....Hmmm. Just to make sure I was answering to what the term really means rather than perhaps its slang or common usage, I looked it up. Under the word, Wiki had ad hominum abuse, where a statement would be applied to an individual specifically to demean him. Yes, that would make someone mad. However...
    I was not using the lack of standards and dumbed down at the person with whom I was debating that which is necessary for a clear photo. I was making a statement that my fellow Americans, with their changing mores, have really dropped standards in general. The public is influenced by the media non-informational talking heads and that which is written in articles concerning something of interest to the reader. Besides the overall social attitude, work attitude, etc., these people influence what is considered acceptable, good, and so on. My statement ending in "It's hard" is what I have heard from several people whom I have asked to do such things as:
    1. Reprint of my 5X7s. A camera shop that used to turn out clear, well and properly colored prints now has a new operator after they cleaned house of their tenured people. The operator, rather than select grain suppression electronically on the machine, appropriate for the input, simply maxes out grain suppression on every order that goes through. This makes all of their output look soft and out of focus and like it was made on the disposable camera with shaky hands. Since I buy good lenses, try to use appropriate shutter speeds and most often use a tripod, I don't want my results looking out of focus unless I goof. I asked for grain suppression=0 on each order I hand the counter person and write it in felt tip pen or bold pencil on the front of each submission envelope. Invariably it comes out soft and fuzzy and I either have to ask them to do it over again, or take it to Costco, which seems to be able to provide reasonably crisp prints. When talking with either the store owner or the operator, trying to get a decent print because I want to support the last camera store in town, I get a version of how really tough it is to remember to set their machine according to envelope instructions. It's said in a put-upon whiny voice.
    2. Auto service. No, you don't want to read that also, but it is the same thing.
    3. A nice upper middle class/price restaurant. Shortened version is that we customers (ourselves twice to three times a week) had to listen, almost every trip about how they were being put upon by their customers when good help was hard to find. Having made it a "regular place" for over five years we finally quit going there three weeks before they closed their doors. We were loyal to the business in their hard times and we would see fewer and fewer of the other regulars. Life was just too hard for the owners.
    Dissembled...? No, I'm being straight forward. I do hate to see society, especially the America that I grew up in, participated in, and tried to leave behind with more than I took/used.
    Now, as to the person I thought I was debating/discussing the PPI/dpi issue with...This was not an attack on YOU. If you find the equivalent of 240, 200, 180 or even 50 PPI acceptable to you, that is your right. Even I conceded that my 100PPI 7'x7' print looked pretty good a few feet away. You get to enjoy your pictures as you see fit. But... I still think that 300PPI is a good and worthwhile standard that when printed with good work and equipment will look properly sharp and capture good detail for the majority of viewers (even if I prefer Sony's 403 standard for myself).
    I have an Epson inkjet that produces 13"x19" prints. I sometimes make 13x19s from my 8mp, small sensored Panasonic FZ30. They look pretty good. Most people are surprised such a small camera can make such a clear large print. But if I scan some 35mm film made in a good camera at 4000PPI of the same scene, there is a lot of small detail on that print that is not on the FZ30 print. When viewed side by side by the "average Joe", many would not have known that the detail should have been there on the lesser print, some marvel at the difference, and others can see no difference.
    I love the fully populated 400PPI 8"x12" prints from my Fuji Pictography 4500, sharp and crisp. I try to maintain a lower limit of 300PPI for my inkjet prints (but sometimes cannot reach that), and if I need a large print from my FZ30 I'll go ahead do it and be glad I have that print rather than nothing at all.
    Most of all, G Dan Mitchell, I mean you no harm, no disrespect, no insult. I'm glad to have folks like you participate in a meaningful way on Photonet. I value intelligent discussion much more than agreement.
  37. I don't see how the Nyquist sampling frequency applies here as we are not sampling anything.
    When you convert from a line pair to the resolution necessary to represent that line pair in a 2D fixed grid of imaging elements you most certainly are sampling.
    6 lp/mm=6*2*25.4=304.8.
    2x is not sufficient to represent a line pair as we are discussing photographs, not computer generated vector art. If you want to know what resolution in ppi is truly equal to a photographic resolution given in lpmm you must take Nyquist into account.
    The 6-7 lp/mm range is derived from the fact that 20/20 vision can resolve 1 arc minute at 10 inches.
    Once again, this is true for a particular test target. (The photographic community puts too much into that one particular form of testing.) It varies widely given targets with different detail shapes, sizes, and contrasts. If you doubt this, try printing some small text and intricate, monochrome vector clip art at 300 dpi, 600 dpi, and 1200 dpi from a photo ink jet or a laser printer onto paper which can handle these resolutions. According to a line pair test the text should look the same to you. If your vision is 20/20 or better I guarantee you will observe differences.
    Now try scaling a digital photograph to a print size that gives you 600 ppi, then print three down sampled versions at 180, 250 and 300 ppi. You'll see some difference between 180 and 250. You might not see any at 300, depends on subject matter. And good luck detecting any real difference jumping to 600.
    BTW, one of my most important clients is a printing company. All of this is old hat to me. I've done these types of tests and comparisons multiple times.
  38. C. Sharon wrote: "In fact a better measure for print resolution is to use the SQF (subjective quality factor) ...."

    When making prints, as opposed to theorizing about what it might be like to make prints, subjective judgment becomes
    the bottom line. I know from my own experience that it can be easier to rely on "rules" about things like image resolution
    than to look at images, understand what you see, and evaluate how well it works.

    Photographs are very different from one another and prints can be produced with a variety of intended uses. There is
    no absolute number that defines what is universally "best" for image resolution. Most all calculations that attempt to
    arrive at a particular definitive vale make a ton of assumptions and end up being less useful than a pair of good eyes.

  39. @John A
    Hi John, I had a look on you tube and think I found the doc you must have seen - the one where he's making 'Hamm,Bergwerk Ost'. I was very interested to see this because I never could work out exactly what this picture was of - thanks for directing me to it.
    I must have missed the bit where he was discussing which camera to use but what I did see was Gursky fastidiously making a composite photograph from several 4x5 negatives ( this must be the case because the polaroids he was seen to take represented a fraction of the final image) and then like you say using a digital Hassleblad for some 'add ons' on the bottom. I can't say this differs enormously from my knowledge of his process. He wasn't using an 8x10 in this case and he's moved with the times with his use of the Hassleblad ( though it only took a minor role in the image).
    My point as you put it,was I suppose that you remarked on poor image quality in his photographs based on something someone had told you and that differed from my experience of them.
    Moreover of course a 5D mkII isn't going to come anywhere close in terms of quality or ultimately, control over the image ( one thing you can be sure of is that every square centimetre of his photographs has been fretted over and decided upon).I maybe sound like a big fan of Gurskys here which I'm not. I do think he's a great artist and find his images very interesting but theres a kind of grandiose element to them which does put me off.
    Regards A
  40. Describing Gursky prints as being poor is bizarre and, in a way, embarrassing.
  41. "Describing Gursky prints as being poor is bizarre and, in a way, embarrassing."
    Wow, talk about "my way or the highway!" When I made my anatomical reference concerning opinions some posts above, I was doing it with a twinkle in my eye, tongue-in-cheek, thinking it might evoke a smile or perhaps a laugh. Was I to make it here again I guess the twinkle would be absent.

    Since any opinion of disagreement seems to be worthless and invalid , may I add my worthless and invalid opinion? I would not exactly rate the prints or work of Gursky as poor. Mediocre, uninteresting, not worth looking at or just "there" comes to mind. Maybe he has some other stuff I haven't seen that is a little better? And, given enough time and reflection, I too, may even feel embarrassed. Hmmm, No, probably not.

    However, if someone else likes it or thinks it is swell, maybe they see something in it that does not occur to me. Maybe they have different tastes and or interests. That would be good. The world would be pretty dull if we were all alike.
  42. Do you recommend the software onOne Software Perfect Resize to gain resolution?

    The best definition files supplied by the 5D (RAW) is 10x16in. for a resolution of 350pixels / inch. To make a 36inch high print, the resolution drops to approx. 100pixels/pouce,which is very little.

    Any experiments?

Share This Page