1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

PPI and DPI

Discussion in 'AP Magazine Feedback & Suggestions' started by PhotoEcosse, Jan 28, 2014.

  1. mark_jacobs

    mark_jacobs Retired

    A. It's a base figure. For the sake of brevity and submission assessment/storage capacity. Regards APOY and pagination/space it is unlikely that a file with a greater resolution (300ppi/150lpi) would be required. Should it be the photographer would be contacted

    B. AP prints @ 300ppi/150lpi. Beyond 'wiggle, world changing exclusive, room' etc. ;), there is no argument on this. 2,700 pixels (longest edge) @ 300ppi/150lpi allows a 'rule of thumb' halftone repro of 9"/23cm. The AP base 'resolution'.

    Halve the resolution and try to print 150ppi/150lpi and the 'preflight' software will 'squawk'
    'Push' it through and then pre-press human will point out the quality/cost implications should it go wrong*.

    C. It can be done, as can printing with potatoes :)

    m **

    *Explanations may also be requested by the photographer/agency

    **Regards halftone offset printing
     
  2. P_Stoddart

    P_Stoddart Well-Known Member

  3. PhotoEcosse

    PhotoEcosse Well-Known Member

    Right - I don't expect this to be the last word on the subject, but at least it should provide a definitive example.

    Below is a photograph of a framed photograph hanging on a wall. Ignore the fact that it is 800px on its long edge and has a very bad reflection of me in the glass and was taken in poor artificial light at ISO3200 with a compact camera.

    The point is that the photograph in the photograph was taken with a 12Mp dSLR. The original Raw file was 4256x2832px. It was heavily cropped to 2140x1571px. According to Chris Gatum, that should produce a print of around 7" x 5".

    What you are actually seeing in the photograph is an image measuring 16" x 10" (to get those dimensions it is actually cropped a wee bit more by the mount aperture) in a 20" x 16" mount.

    The little flag in the lower left corner of the mount shows that it was accepted for the RPS Nature Group Exhibition 2012. Believe me when I say that they do not accept poor quality prints. Yet it is printed to 4.5 times the size that Chris suggests should be satisfactory.

    [​IMG]

    Incidentally, the photo hanging to the left of the gannet was taken with a higher resolution camera. The full image measured 7360x4912px, for the print it was cropped to 3198x2361px and again printed to 16" x 12". So it has more than twice the "resolution" of the gannet photo but, quite honestly, there is no discernible difference in print quality.
     
    Last edited: Jan 31, 2014
  4. PhotoEcosse

    PhotoEcosse Well-Known Member

    Sorry - print dimensions in 4th paragraph of above post should be 16" x 12" (not 16" x 10")
     
  5. Benchista

    Benchista Which Tyler

    Are you being wilfully stupid, or can you just not read? It's now hard to believe you're not deliberately misrepresenting what Chris said for some unknown reason. I actually reproduced the key bit for you word for word in a previous post, so you haven't even got the excuse of being able to claim you missed it in the mag.

    There's an old saying - if you're in a hole, stop digging. It's rare to see someone dig a hole so deep and then deliberately bury themselves.

    What part of Chris saying that if the image quality ia good enough you don't have to print at 300 PPI to get a decent print do you not understand?
     
  6. Richard Sibley

    Richard Sibley AP Deputy Editor

    To my knowledge 300ppi is cited as the preferred resolution for printing due to the fact that the majority of humans cannot see much difference in greater print resolutions past this point (with the image at a reasonable distance).

    I saw a test somewhere that had some images printed at 350ppi and possibley even a 600ppi print, and in the sample on one person could actually tell the difference between the prints - i think they were viewed at around arms length.

    So, again - to my knowledge, this is why 300ppi is quoted as being the preffered resolution for photographic quality printing.
     
  7. Richard Sibley

    Richard Sibley AP Deputy Editor

    And again, just to back up what Chris said, I have printed a 6 million pixel image to a little smaller than A3 and it looks perfectly fine in my portfolio. However, I have no doubt it would have made a better print from 12 million pixel image without the interpolation.
     
  8. Roger Hicks

    Roger Hicks Well-Known Member

    Dear Richard,

    This has prompted another thread: "Good Enough".

    Cheers,

    R.
     
  9. P_Stoddart

    P_Stoddart Well-Known Member

    See what always blows holes in the "this shot a 3MP is fine print at A3" position is,

    why do camera makers produce high pixel count sensors then?

    Simple when you take a 3MP image and feed into a print system than can show 16MP of data then alot of guessing is going on by the printing system or the user has to upscale the image before hand again alot of guessing.

    Capture raw data a point of shooting is much better. Provided the lens is upto the job it can exploit the increase MPs.

    Prof Newman talked about this when the D800 came out. His conclusion was that even old glassware will gain from the extra MPs.

    So where are you likely to see this improvement? In printing. Because at this time in technology printing can show the extra MPs.

    Display technology is not upto the job yet. :(

    If you look at another thread Nick (Benchista) tried this very thing with 21MP and 12MP camera. He states that even at A4 he could see a improve on the 21MP downscaled to A4 which is around 9MP if using a 300ppi system.

    Many great images have technical flaws. When I was at the Landscape Photographer of the Year exhibition which has alot of large prints to display work (I think they are A2) I saw Kaushiklal Koria Tracks in Snow, interesting shot but taken handheld at 1/6 on a D200 where it showed camera shake. :( Another shot again camera shake handheld 1.8sec. The shake did not add to the shots in anyway IMHO. :(

    You look at it like this. Photography is a form of painting. You can paint a great image with cr*ppy paint but it will still be a great image. But it much better to paint that image with better paint or the best paint you can afford.

    I fairly sure if your (PhotoEcosse) 12MP image cropped to 3MP (4x crop) had started out on a 24MP camera and ended up at 6MP presuming the lens was upto the job. You would notice the different between the 6MP and the 3MP on your 16x12" print. So it could be a case of artist content overcoming the technical limits of the result.
     
    Last edited: Feb 1, 2014
  10. MickLL

    MickLL Well-Known Member

    Is that true Fen?

    I always thought that a monitor displayed at a particular, unchanging, dpi
    whatever. I have 72 (or maybe 96) in mind. It's why a 'small' file won't disply any worse than a 'big' file on a monitor (when displayed at the same size of course).

    So I thought that the number in the first post (displaying at 87.5) couldn't be right.

    I'm happy to be told that I'm talking rubbish - I'm no expert.

    MickLL
     
  11. daft_biker

    daft_biker Action Man!

    Aye. LCDs have a fixed resolution. (the same error is in the OP but I figured one thing at a time.....)
     
  12. mark_jacobs

    mark_jacobs Retired

    I believe that Fen was referring to the number of recorded pixels an image contains e.g. 7360x4912 and any chosen resolution (ppi) rather than the fixed dot pitch of the monitor/screen on which it is displayed.

    Such an
    image can have its resolution set to 72ppi, 7200ppi or whatever. It will still contain 7360x4912 pixels. The image resolution simply alters how those pixels are spaced/rendered at 100%, be it on paper or lcd.

    Forgive my clumsy attempt at the following explanation, Fen'll understand ;)

    You have a pint of beer. It fills a pint glass
    (bar person dependant). Consider this as a 7360x4912 pixel image.

    You pour said pint into a shallow vessel (a) with a volume of one pint. You still have one pint (7360x4912 pixel image). For the sake of argument at this point we will describe the pint as having a 'resolution' of 300ppi.


    You then pour it in to another wider shallow vessel (b) with a volume of two pints.
    You still have one pint (7360x4912 pixel image) however the pint is now dispersed over twice area. You have reduced the pints 'resolution' from 300ppi to 150ppi.

    Return to pint glass.

    Drink (crop) half of said pint and pour back in to vessel (b) You now have half a pint (3680x2456 pixel image) with a 'resolution' of 75ppi.

    Continue until pixelated ;)

    m
     
  13. PhotoEcosse

    PhotoEcosse Well-Known Member

    Wow.

    I'll reproduce something I said in another thread a few days ago, which may explain to others why I do not get offended by your inordinate rudeness:

    Let me quote, word for word, what Chris wrote (AP, 1 February, p57, "AP Glossary"):

    The first two sentences are correct. It is in the remainder of that quote that the misconceptions, that I started this thread with reference to, arise. Pixels per inch relate only to the display of an image on a screen; not to the printing of an image.

    To pick up on what Mark Jacobs said, when I send an image in for APOY I comply with the instructions to make it 3000 pixels on the longest side. So it might be, say, 3000x2000 pixels. It doesn't matter tuppence what ppi figure was in the Lightroom export dialogue box when I exported that file to send to AP (it was probably 96 because that is what it is currently and I never have any reason to change it). If my photo is judged good enough to get into the top 30 in that APOY round, Mark will have no difficulty printing it in the magazine. All he needs is that 3000x2000 pixel Jpeg - he is going to print it at whatever size suits his page layout. My image file has no "linear inch" (Chris's words) data in it that will affect how it is printed. That becomes relevant only when the print job is set up.
     
  14. mark_jacobs

    mark_jacobs Retired

    Because the resolution will be set to the 'rule of thumb/repro profile' for AP of 300ppi. At 100% the image (3000 pixels longest edge) will then print at 10"/25.4cm x ?.

    If left at your (probable) default of 96ppi at 100% it would print at 31.25"/79.38cm x ?

    As Chris Gatcum said

    With the best of intentions regards:

    See above.

    Regards
    m
     
  15. Roger Hicks

    Roger Hicks Well-Known Member

    You're right. It doesn't matter tuppence what ppi figure was in the Lightroom export dialogue box. But so what? As soon as you want it on the printed page, you need 300 ppi/dpi (equating ppi/dpi by a strictly sloppy, but in the real world sufficient, rule of thumb).

    In other words, your dimensionless, noumenal file is completely meaningless until it is displayed; and in order to be displayed, it must be in pixels or dots; and if it is displayed, that display must have dimensions; and if there are pixels or dots in a given dimension, there must be pixels or dots per inch (or other unit of dimension). The ppi/dpi dimension is inherent in the real world. Those of us who deal with the real world are having some difficulty in understanding the relevance of your theoretical files.

    Cheers,

    R.
     
  16. Benchista

    Benchista Which Tyler

    Rudeness, perhaps; inordinate, definitely not. I can assure you I'm only rude when it seems to be the only way to get through to somebody. It's perfectly acceptable not to understand something, as you clearly don't; but when it's been explained to you that it's you that's wrong, not the other person, to continue to insist you're correct goes beyond the literal definition of ignorance and into the more colloquial one; it's thus YOU that's being inordinately rude, because there's simply no excuse for you to continue to be so pigheaded.

    I'm not sure how to say this in words that you will understand, but you're wrong, wrong, wrong, wrong, wrong, and Chris has it right. Pixels per inch are not the same thing as dots per inch, nor is he saying that - quite the opposite, in fact; but of course pixels per inch are involved when printing! What he says is absolutely correct - PPI defines how big a print is, DPI does not. Read this week's installment, it should help with your understanding.

    The reason why AP asks for that file size is because they can print it to the biggest size necesary at 300 PPI! Surely that's not too difficult a concept for you to grasp?

    The only misconceptions are yours, I'm afraid - Chris has managed what is certainly the best summary of the issues I've seen.
     
  17. Richard Sibley

    Richard Sibley AP Deputy Editor

    And that is why it is relevant.

    If you have it set to 72ppi and make a 100% print it is going to be a lot larger, and more pixelated (when viewed at the same distance as) than a print at 300ppi at 100%

    If when you print all you do is select a print size then you may never need to look at the ppi value.

    However, when you come to making an A3, A2, A1 print (or a smaller print from a cropped image) then using the 300ppi value at very least allows you in the print dialog to see how much the image is being scaled.

    For example if you set it to 300ppi and you print at A3 and the dialog says <100% scale then you know you are ok.

    If it says 200%+ then you can quickly see that you are increasing the size of the image significantly that it will have a reduction on image quality.

    It is up to you what you consider acceptable (see Roger's thread)... but 300ppi is the standard that most in photography and print aim for.

    Not everyone has 36 million pixel cameras capable of easily producing an A3 or A2 image.
     
  18. Fen

    Fen Well-Known Member

    Really?

    [​IMG]

    I'm a bit surprised at that. If you look at your photograph it clearly shows that the picture frame isn't rectangular, there is quite a distortion there as you can see from the rectangle I've drawn over the image.

    There is also a strange dark area which I've highlighted in red. Looking at again, I can also see some dark areas on the right hand side as well.

    I'm wondering if the curve of the frame and this dark area have been caused to water damage... If so did it happen before or after the RPS panel saw it? I can only guess that it was afterwards.
     
  19. Catriona

    Catriona Well-Known Member

    I'm guessing here but I would think the distortion is barrel distortion of the photo of this image, and not the picture frame. I could be wrong though. Also the dark areas are reflections of the person taking the picture! Anyway, I am still puzzling over the request by two recent Salons for images of 300dpi quality. They got 300ppi so they can like it or lump it!

    Kate
     
  20. mark_jacobs

    mark_jacobs Retired

    I suspect/hope that via PM the OP and I may reaching a point of agreement/clarity :)

    Regards
    m
     

Share This Page