JPG vs TIFF Resolution Test

Discussion in 'Digital Photography' started by Jerry McG, Dec 4, 2003.

  1. Jerry McG

    Jerry McG Guest

    Ok, now how can this be:

    I shot two identical images using a Nikon 5400 on a tripod and the identical
    ISO (sensitivity) & exposure, same WB. One resulted in a 14.5MB TIFF file,
    the other a 600.2KB "basic" JPG file. I brought them both into PS7, and
    printed them at 300DPI on an Epson 2200 using 8x10 Epson Archival Matte
    paper ....both prints are identical, even when observed under magnification.

    Now, why the %$#@ would I bother to shoot 14MB Tiffs, when I could shoot
    600k JPGs and get the same output?

    Please solve the riddle, is this simply because of the capabilities of PS7s
    interpolation? Whatever it is, the little file rocks!
     
    Jerry McG, Dec 4, 2003
    #1
    1. Advertisements

  2. Jerry McG

    John Guest

    magnification.

    I don't know, but I'll try:

    Each pixel in the TIF file has 24 bits, or 3 bytes, of color information.

    14,500,000 / 3 = 4,833,333 pixels.

    Your 8 x 10 print has 80 square inches.

    You TIF image, printed on 8 x 10, has 60,416 pixels per square inch.

    You print at 300 dpi, or 90,000 dots per square inch. The print of the TIF
    image is nearly perfect (or as good as the TIF imkage).

    The JPG file has been compressed using interpolation techniques. The
    complexity of the original image is a factor in whether the interpolation
    results in a satisfactory image. Compare several shots - say, perhaps an
    artificially lighted portrait versus a group picture, a scenic view, and a
    vase of roses.

    Hoe do they look?
     
    John, Dec 4, 2003
    #2
    1. Advertisements

  3. Jerry McG

    mark herring Guest

    Jpeg compression does not degrade resolution--at least not directly. The
    effect is most noticeable in light areas near edges. I'm surprised at the
    compression ratio. In my 2Mp Canon A40, an uncompressed file would be just
    under 6MByte. I ALWAYS use the highest quality jpeg setting, which gives me
    files of about 1Mbyte (6:1 compression).

    Your example is about 24:1 compression ratio---there has GOT to be some kind
    of artifact.

    What kind of scene is it? For example, a picture of a white sheet of paper
    might have very high compresssion

    --
    ******************
    Mark Herring
    Pasadena, CA, USA
    private e-mail: just say no to "No"

    *
     
    mark herring, Dec 4, 2003
    #3
  4. Jerry McG

    Jerry McG Guest


    So far I've done several shots using ambient or room lighting, macro and
    infinity focus and I'll be damned if I can see any difference between the
    little jpg files and the big tifs. I have usually shot Nikon NEF (raw) files
    with my D1x, but now I'm wondering if this isn't a complete waste of space &
    time. Hmmm .....
     
    Jerry McG, Dec 4, 2003
    #4
  5. Jerry McG

    Matt Guest

    Your observation is correct.

    But the main reason for using TIFF images is when editing and
    re-editing are involved. Successive editing and saving of JPG images
    eventually results in image deterioration that becomes more noticeable
    with each generation. From my experience, there appears to be little
    gained by shooting an original TIFF as opposed to a hi-q JPG, but once
    these are brought into an image editing program they should be worked
    on as bitmap (TIFF etc.) images to prevent quality loss.
     
    Matt, Dec 4, 2003
    #5
  6. Jerry McG

    Ken Weitzel Guest

    Hi...

    I can suggest a little experiment for you to try that may
    be helpful if you're game...

    Load the tif into psp, save as a copy in jpg format
    with very very little compression.

    Save another with a different name with say mid compression
    levels, and yet one more with real high compression, up into
    the 90 or even 95% range.

    Then load and look at each of the jpg's. Guaranteed that
    you'll see the artifacts getting worse and worse.

    Hope this is helpful.

    Ken
     
    Ken Weitzel, Dec 4, 2003
    #6
  7. Jerry McG

    Jerry McG Guest

    FWIW, the 5400 has a 5.1MP chip. I took several close-up pictures of items
    with text, like a can of dust off under tungsten light, and a picture of my
    house, which is Victorian and has a lot of gingerbread detail & color. When
    viewed on screen or printed as an 8x10 there's not no discernable difference
    in artifacts at all between the two. I've gone back and triple checked all
    the camera settings and all were shot at the identical exposure, white
    balance, contrast settings, etc. In short, how can this be???
     
    Jerry McG, Dec 4, 2003
    #7
  8. Jerry McG

    Jerry McG Guest

    Yeah, I think what I'm getting at is the capture end, that is, I'm taking a
    compressed, basic jpg 600k file, expanding it to 14.5MB (8x10), then taking
    a 14.5MB (8x10) tif, viewing & printing them and seeing no qualitative
    difference. From the capture side, why would I bother with big tifs, when I
    could just shoot little jpgs and get the same results? I generally shoot RAW
    files with my D1x, as a matter of fact I don;t think I've ever shot JPGs as
    serious images with it. Maybe I've been wasting space & time?
     
    Jerry McG, Dec 4, 2003
    #8
  9. Jerry McG

    Jerry McG Guest

    Thanks Matt, yes, I always work files as Tifs, I was just surprised that
    uncompressing the little JPGs caused no image degradation. I'm sure if I
    kept opening them up and editing I'd see a sharp fall off in quality.

    I have been shooting and saving Nikon NEFs (RAW) with my D1x, but converting
    them to Tifs before bringing them into Photoshop for editing. (I use the
    Nikon Capture software to alter exposure or WB settings .) Now I'm thinking
    it may be a lot easier just to shoot standard size JPGs and going from
    there, sure would be a LOT faster workflow.
     
    Jerry McG, Dec 4, 2003
    #9
  10. Jerry McG

    Ray Murphy Guest

    ----------
    RM: I'd put it this way: The main reason for using a TIFF instead of a
    JPEG is because the TIFF holds the best possible image, and the JPEG
    ~always~ holds an inferior image because of compression.
    This can easily be proven by copy/pasting a section of the JPEG
    version onto the TIFF it was made from, and then comparing the colours
    by using the eyedopper tool.

    I'm pretty sure it's an urban myth about JPEGS losing quality when
    saved because I've done experiments recently where I saved a JPEG over
    and over again, and nothing was lost IF the highest JPEG quality was
    selected.
    To do this in Photoshop you have to "Save As" instead of "Save" --
    and then use either the same name or a different one.
    If the "Save As" method is not used, you do not reach the panel where
    you CAN set it for highest quality.

    To do this in Paintshop Pro, there is no need to use the "Save As"
    method, but you MUST use the "Options" which are available prior to
    saving. ie. Click "Save" -- then click on the Options button and set
    the quality to highest.

    If anyone gets different results after experimenting and measuring the
    precise pixel colours, let me know.

    The danger involved in re-saving JPEGS is that one can forget to set
    it for the highest quality and the program will use the low quality
    default.


    Ray
     
    Ray Murphy, Dec 4, 2003
    #10
  11. Jerry McG

    MikeWhy Guest

    You could be. Wasting time and space, that is. If you don't typically edit
    the pictures much, JPG could suffice. The advantages of TIF are 16-bit color
    and lack of compression artifacts. It holds fine gradations better after
    color and gamma adjustments, and has fewer artifacts to worry about when
    sharpening. The advantage of RAW is you can play with WB, contrast,
    saturation, and sharpening as much as you like up to the point of exporting
    or saving. Even afterwards, by simply reconverting. On a TIF or JPG that the
    camera got wrong, or just less than perfect, you have no choice but to
    damage the original image. My printer probably won't see much difference
    between a 16-bit TIF versus RAW, but 8-bit JPGs start out marginal and only
    get worse. Just the same, small curve or level shifts won't hurt them much
    or noticeably, if that's all you're doing.
     
    MikeWhy, Dec 4, 2003
    #11
  12. Jerry McG

    Ray Murphy Guest

    ----------
    RM: Ok, so you are taking almost identical photos in 2 modes - JPEG
    and TIFF
    RM: It's only a waste of space and time if you are looking for
    second-rate results.
    If however you want the best possible result there is only one option
    - a TIFF.

    All digital camera JPEG photos are second rate, but are still very
    good looking pictures a lot of the time if they are made on a decent
    camera.

    Ray
     
    Ray Murphy, Dec 4, 2003
    #12
  13. Jerry McG

    Todd Walker Guest

    There is a big difference between shooting TIFF in-camera and shooting
    RAW. IMHO, using TIFF in-camera has no value whatsoever and only wastes
    storage space. The only thing that a TIFF has over a JPG is that it
    hasn't been compressed and today's digitals use such excellent
    compression algorithms that you won't be able to tell the difference
    unless you make a very large print.

    RAW however is a much different story. If you shoot in RAW, you are
    capturing much more information, usually 12 bit vs. 8 bit for TIFF or
    JPG. This increases dynamic range, which increases shadow and highlight
    detail. Second, RAW files are completely unprocessed by the camera,
    leaving all the processing to you and your conversion software. This
    means that you can control the white balance, exposure compensation, and
    all other settings AFTER the fact, which is an indispensable capability.

    To sum up -- TIFF is a waste of space, RAW is always best.

    --
    __________________________________
    Todd Walker
    http://www.toddwalker.net
    Canon 10D page:
    http://www.toddwalker.net/canon10d
    __________________________________
     
    Todd Walker, Dec 4, 2003
    #13
  14. In short, how can this be???It is because JPEG is well designed so that the errors it does introduce
    are not easily visible to the human eye under normal viewing conditions.

    We have been using Basic on our Nikon 990 and 5700 cameras and whilst it
    is indeed fine for about 98 - 99% of shots. The worst errors on the
    remaining 1 - 2% are in clear blue sky areas where some blockiness or
    quantisation does show. If you reduce the resolution by viewing on the
    screen (i.e. showing a 2048 x 1536 image on a 1024 x 768 display), the
    errors will be even less visible.

    It's a matter of knowing how to get the best from the overall system as
    _you_ use it!

    Cheers,
    David
     
    David J Taylor, Dec 4, 2003
    #14
  15. Jerry McG

    Ray Murphy Guest

    ----------
    RM: I knew nothing about the RAW format until a week ago when posters
    on this newsgroup made it very plain that RAW is the best to use for
    quality; however it raises the question - What advantages are there in
    getting cameras to produce TIFFS when they can be made in Photoshop or
    some other image editing program?
    Could it be simply a convenience thing?

    Ray
     
    Ray Murphy, Dec 4, 2003
    #15
  16. Jerry McG

    Steve Guest

    It's definitely not an urban myth. You can see the degragation after only a
    couple of saves.

    You were doing the following when executing your test?

    * Save the document
    * Close the window
    * Open the saved document
    * Repeat

    Simply repeatedly saving the same image won't degrade it. Also, if you don't
    close the window, some software will just use the copy of the image that is
    in memory and not actually load it from disk again.

    Regards,

    Steve
     
    Steve, Dec 4, 2003
    #16
  17. Jerry McG

    Ron Hunter Guest

    No, degradation of .jpg images is NOT an urban myth. If you will set
    the compression to about 75% on the save and load the image, and do some
    kind of edit, say crop a 1/4 inch off one side, and save, then repeat
    this several times, and then compare a section of the picture containing
    solid or slowly changing color, you WILL see the difference. Just
    loading and saving with no changes will not cause recompression on most
    good software.
    nn
     
    Ron Hunter, Dec 4, 2003
    #17


  18. Hmmm, is right, and I've been saying this all along. I see other wedding
    guys shooting raw, then doing all that post processing, and I guess they are
    more anal than I am, got lots of more time on their hands than I have, but
    I find that I do fine at 1:8 compression, getting 300 shots on a 256
    megabyte card. Like you, I just can't see the difference between a TIFF and
    a 1:8 jpeg, even up to 11x14.

    Patrick
     
    Rudy Von Tschudi, Dec 4, 2003
    #18

  19. On the computer is one thing, but at print is all that matters, in my view,
    since I'm selling prints.

    I don't see the difference. I swear. My 1:8 compressed jpeg prints are
    silky smooth, to the naked eye.


    Patrick L.
     
    Rudy Von Tschudi, Dec 4, 2003
    #19

  20. I shoot maybe 500 images on a wedding. I just don't have the time to dick
    around with that many shots. I get them as close to a proper exposure as I
    can, maybe do a simple curves adjustment, if they're underexposed a bit,
    and about 25% of them are, and that is it, out I'm outa there. 1:8
    compression, jpeg, gets me 300 shots on a 256 card, and it works just fine.

    Patrick
     
    Rudy Von Tschudi, Dec 4, 2003
    #20
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.