How big is the resolution on a normal 35 mm negative?

Discussion in '35mm Cameras' started by Sebastian Zamorski, Dec 7, 2003.

  1. I'm trying to find the benefits when it comes to a normal 35mm film and
    compare it to a digital picture.
    Could anybody tell me how big a resolution a 35mm negative has - I've heard
    something about 20 mill. pixels (?)
     
    Sebastian Zamorski, Dec 7, 2003
    #1
    1. Advertisements

  2. Sebastian Zamorski

    Mxsmanic Guest

    It depends on the film, the camera, the lens, the shooting conditions
    (with tripod or handheld, etc.), and the scan.

    For fine-grained film, 22 megapixels is easy enough to achieve. But
    unlike digital, there is no sharp dividing line for film. You don't
    have x megapixels of high quality, and nothing beyond. Instead you have
    x megapixels of high quality, or a slightly larger number of lower
    quality, or a still larger number of even lower quality, and so on.

    Still, when done right, 35mm film easily matches current high-end
    digital SLRs. It takes a lot more effort to get the best results from
    film, however.
     
    Mxsmanic, Dec 7, 2003
    #2
    1. Advertisements

  3. Sebastian Zamorski

    Rafe B. Guest

    Although what you describe could of course be
    achieved after the fact with either a film scan or a
    digicam capture.

    Goodness. An entire post by msxmaniac with
    which I fully agree.

    Sebastian, please have a look at sample film
    scans and digicam outputs at:

    http://www.terrapinphoto.com/jmdavis

    mr. manaic, long ago you posted a very
    crisp LS-8000 film scan, of a chrome, taken
    with you 'blad. Would you mind terribly
    emailing me a small 0.25" x 0.25" raw
    scan snippet of a sharp scan of a chrome?
    (eg 1000 x 1000 pixels @ 4000 dpi raw scan.)

    thanks.


    rafe b.
    http://www.terrapinphoto.com
     
    Rafe B., Dec 7, 2003
    #3
  4. Sebastian Zamorski

    Mxsmanic Guest

    I was already including the scan.
    No problem, if you can tell me which image you had in mind. Do you
    recall the content of the image? I have several MF photos out there,
    and a few 35mm photos as well.
     
    Mxsmanic, Dec 7, 2003
    #4
  5. Depends on lens, film and how it is developed. With the Nikon macro
    and Tech Pan, you can probably get up to 150 lp/mm if everything is
    downhill and in your favour.

    That is in the 20Mp ballpark.

    Please note that you can not equate film and digital resolutions
    unless you specify almost all the conditions, they are not a 1:1
    relationship.

    (Also note that the *slightest* misfocus, vibration, haze or anything
    else will slaughter this number)
    --
    Paul Repacholi 1 Crescent Rd.,
    +61 (08) 9257-1001 Kalamunda.
    West Australia 6076
    comp.os.vms,- The Older, Grumpier Slashdot
    Raw, Cooked or Well-done, it's all half baked.
    EPIC, The Architecture of the future, always has been, always will be.
     
    Paul Repacholi, Dec 7, 2003
    #5
  6. Sebastian Zamorski

    Rafe B. Guest


    You pick. The aim is to see how much detail can
    be pulled out of 0.25" x 0.25" of film. Have a
    look at what's on the test site already.

    That area of film is a bit arbitrary, but convenient
    for most typical scan resolutions and monitor
    resolutions. It gives a snippet of 1000 x 1000
    pixels at 4000 dpi, which is viewble at 100%
    within a browser on a 1600 x 1200 screen.

    The image I'm recalling had some kind of tower
    or balcony or terrace, very fine detail in the railing.

    What I'm hoping to collect are raw scans, without
    sharpening or any other post processing, and at
    the native scan resolution.

    Having that, folks can download the raw scans
    as they please and try different post-processing
    tricks as they see fit.

    Saved as high quality jpg, the snippet should easily
    be under 2 Mbytes or so.


    rafe b.
    http://www.terrapinphoto.com
    scan snippet collection:
    http://www.terrapinphoto.com/jmdavis
     
    Rafe B., Dec 7, 2003
    #6
  7. Sebastian Zamorski

    Mxsmanic Guest

    Okay.

    Keep in mind, though, that my scanner doesn't necessarily extract
    everything, although it's a very good scanner. I think it gets about
    95% of most fine-grain films. It would not get everything from Tech
    Pan, though.

    Try

    http://www.mxsmanic.com/scantests.html

    This is a slideshow of seven scans, with both full-size 1000x1000 scan
    excerpts and a thumbnail of the original image. I've included examples
    of MF and 35mm, slide, color negative, chromogenic B&W, and conventional
    B&W, handheld and on a tripod, day and night, exterior and interior.

    The scans have no sharpening applied, although I have optimized the
    curves for each of them (because that's how they are archived). The
    thumbnails are sharpened.
    See if the above helps. Note that each image is about a megabyte in
    size.
     
    Mxsmanic, Dec 7, 2003
    #7
  8. Sebastian Zamorski

    Rafe B. Guest


    The image I was thinking of isn't there, but no
    matter. I don't see anything in this collection that
    pushes the envelope beyond the samples I've
    already gathered. The LS-8000 is at this point
    well represented (or over-represented.)

    The image I was recalling (through the fog of
    memory probably 3 or 4 months back) was
    intensely sharp, and I remember being impressed
    at the time.

    I'd sure like to get some more drum scan snippets.

    Overall, my conclusions so far:

    1. Either film just doesn't have more, or 4000 dpi
    is more than adequate to catch what's there.
    2. The 10D more or less matches the best film
    scan snippets I've got, either from chromes or
    Reala.
    3. The 1Ds exceeds 35 mm scans and (in my wife's
    opinion at least) beats my LS-8000 scans of 645 film.
    Me, I'd call it a draw.

    I've got a couple of 1Ds snippets on the site
    at this point.


    rafe b.
    http://www.terrapinphoto.com
    scan comparisons at:
    http://www.terrapinphoto.com/jmdavis
     
    Rafe B., Dec 7, 2003
    #8
  9. Sebastian Zamorski

    Mxsmanic Guest

    I've added one shot that may be similar to the one you recall. See the
    eighth shot in the list at

    http://www.mxsmanic.com/scantests.html
    I don't have any myself. Drum scanners are expensive, and drum scans
    are expensive, too.
    My examination of Velvia and Provia transparencies under a microscope
    indicates that the scanner is not the limiting factor at 4000 dpi. My
    guess is that lens resolution is starting to be the weak link at that
    point (78 lp/mm). These two transparency films can record more than
    4000 dpi, but on my transparencies there is virtually nothing beyond
    that resolution, and I suspect it is the lens, because the dye clouds
    under the microscope appear to be capable of holding considerably more
    detail.

    One of these days I'll mount my 90mm Summicron on a tripod and test with
    that--it is phenomenally sharp. I don't like the way Leicas mount on
    tripods, though.
    That's hard to believe, since the 10D doesn't come anywhere close to the
    resolution of scanned film.
    Obviously resolution does not enter into your evaluation. I don't see
    how 60 or 80 megapixels is going to be worse than a 1Ds.
    Too many of the film scans are of color negative film, which is always a
    waste of time. Only transparency films produce really nice images.

    The 1Ds looks very much like film. The differences I see are less noise
    in the 1Ds image (although the 10D _does_ show noise similar to film),
    smearing of colors in the 1Ds image, and artifacts in the 1Ds image.
    The pixels are of better quality, though.

    It is possible that both the film and 1Ds images are limited by lens
    quality. If any of these shots were taken handheld, they can be
    excluded from the comparisons as well.

    I'm not really sure what you are trying to prove or determine, either.
    If you are trying to prove that digital is better than film or vice
    versa, it can't be done, as there are too many variables.
     
    Mxsmanic, Dec 7, 2003
    #9
  10. Film doesn't have pixels. Depending on the type, some of the sharper,
    finer-grained films (Kodacrome, slow/medium speed B&W) absolutely
    trounce anything digital can do.
     
    Michael Scarpitti, Dec 7, 2003
    #10
  11. Sebastian Zamorski

    bmoag Guest

    This is a pointless question as film has to be converted to pixels and so
    the number of pixels is determined by the scanner and not the film. The real
    issue is that the amount of information that can be captured by a digital
    sensor is fixed as is the resolving power of that sensor. The color gamut is
    also fixed and ultimately determined by hardware limitations and the
    processing algorithms chosen by the camera's designers (and to a limited
    extent by the user if working with RAW files). Digital users are hung up on
    comparing their equipment to 35mm film/camera combos: digital can produce
    pleasing, great images but technically cannot compare to the information
    capturing cacpacities of photograpic film. Step from 35mm to medium format
    and larger films sizes and the comparison of film to digital information
    capturing abilities is even more absurd. It is also utterly beside the
    point. Is color better than black and white because is captures more
    information about the light source than black and white film?
     
    bmoag, Dec 7, 2003
    #11
  12. Sebastian Zamorski

    Deathwalker Guest

    So the big question is what res film scanner is required to acquire all the
    image info that there is. Some say 4000 some say 8000. Also film to scan
    or dslr route. How many megapixels before the sensor carries as much info
    as 35mm film would with the same lense and the same magnification factor i.e
    1 not 1.5 or 1.6 or 1.3 So in otherwords all things being equal, ccd sensor
    resolving power, lense resolving power and focusing on a scanner what are we
    talking about. I beleive it would be 4000 for scanner and 20megapixel
    camera sensor. 8000 scanners simply have better focusing etc to retreive
    the 4000 worth of info that is there. Lets face it, are there 4000 halides
    per inch of film? Or for colour are their 4000 r g b c colour couplers per
    square inch of film.

    Once we answer the above questions we then get a "ball park figure". Just
    what do we have to wait for before the film/digital playing field is level
    and its down to resolving optics microlenses algorythyms whatever to fight
    it out.

    As someone who gets asked this a lot i would like to pass on the best advice
    possible. Currently i advise anyone wanting value for money on scanning
    large quantities of slides to purchase at least a coolscan iv or better
    still the v or the minolta 5400. individual prints are 99 pence for 6x4.
    Cibachrome is even more and takes longer. However kodak cd scanning is only
    4x5.33 at 300dpi and is therefore inadequate.

    on the digital front i advocate quality built compacts against dslrs due to
    the lense design being superior on the custom designed schneider and carl
    zeis fitted compacts as opposed to a very expensive dslr and then a £100 kit
    lens.

    My final advice to customers is megapixels aint everything. If you plan to
    come into the shop and get your chip downloaded and printed to 6x4 and 5x7
    then get a 2 or 3 megapixel with quality lens rather than spending the money
    on 4 megapixel but with only a digital zoom.

    4 and 5 megapixel cameras are only good for a3 size prints or a lot of
    selective enlargement. Usually if it takes too long to explain selective
    enlargement then they probably don't need it.

    on 6x4 prints a 3meg camera will be just as good as a 5 meg. actually alot
    of 4 meg cameras have 5 megs shoehorned into them. according to dpreview
    the canon ones don't benefit cos the resolving power or whatever is shown up
    as inadequate when you put the 5meg sensor in last years 4meg camera.
     
    Deathwalker, Dec 7, 2003
    #12
  13. Sebastian Zamorski

    Leicaddict Guest

    After film has gone through a digital film scanner, it does.
     
    Leicaddict, Dec 7, 2003
    #13
  14. Sebastian Zamorski

    Leicaddict Guest

    A 4000 dpi film scanner will capture 24 MP under ideal conditions. A 5400
    dpi film scanner will capture @ 44 MP under ideal conditions.
    Ideal conditions will be if the information is there to be captured. 12 MP
    is more than enough to make an incrediable 8.5x11 photo inkjet print @ 300
    dpi
     
    Leicaddict, Dec 7, 2003
    #14
  15. Sebastian Zamorski

    Rafe B. Guest

    Because a mere count of megapixels turns out to be
    meaningless. Scanned film has a second optical
    generation that digicams avoid.


    That would be yours and David's opinion, not at all universally
    shared and not at all born out by the scan samples. And there
    are several scans of chromes, both mine and Dave's and from
    Jim Davis.

    Are you claiming that (for example) fern1000.jpg has less
    useable detail than 0.25" x 0.25" of Velvia? If so, I beg to differ.

    At reasonable ISO values and light levels, the 10D is
    nearly noiseless. At low light levels or high ISOs, the
    10D starts looking more like film.
    Really? Call me blind, I sure don't see this.
    Indeed. On the 10D, about 3:1 better than a film scan.
    On the 1Ds, about 6:1.
    That could be the case, but then I would ask, what the
    hell kind of lens is it going to take to get film to show its
    "true potential"? You're using Zeiss, I'm using Nikkor
    and Pentax...

    It's not a given that a handheld shot must be soft.

    This is merely a collection of sample scans and digicam
    captures with an attempt to standardize the film area or
    equivalent digicam image area. It's certainly not a
    scientific study.

    The effort began as a comparison of CCD film scanners to
    drum scanners. But I've been rather underwhelmed by the
    drum scans that Jim Davis provided. It also didn't help
    that Jim chose to "showcase" an image from Gold 400,
    or that the drum scans from Jim are mostly at a lower
    resolution than the LS-8000.

    What is your objection to the film vs. digicam comparison?

    The only arbitrary notion is that the digicams are "scaled"
    to a 35 mm frame. Which means that film still benefits from
    larger imaging area, but digicams don't.



    rafe b.
    http://www.terrapinphoto.com
     
    Rafe B., Dec 7, 2003
    #15
  16. Sebastian Zamorski

    Mxsmanic Guest

    The "second optical generation" is specious. That link in the chain
    easily surpasses the other weak points. And so-called digital
    photography involves a conversion from analog as well.
    The chromes look much nicer.
    Both are limited by lenses, I suspect. But color negative film has more
    grain and poorer color resolution, as a general rule.
    Then the example on that page must have been at an unreasonable ISO or
    light level, because the noise is pretty obvious.
    I see it, and that's all that matters to me.
    I'm not sure where you are getting these numbers.
    Some films aren't served by any current lenses. It's difficult to push
    Technical Pan to its 300 lp/mm upper end with any existing production
    lens.
    It can generally be taken for granted. I certainly see the difference
    regularly for most types of photos. Extremely high shutter speeds can
    help, but then you have to worry about image noise.
    Gold 400 is a very poor choice.
    Drum scans usually shine in their ability to handle a broad density
    range, but resolution can be an advantage as well.
    I don't really object to it; I just don't see that it really proves
    anything, either.
     
    Mxsmanic, Dec 8, 2003
    #16
  17. Sebastian Zamorski

    Bob Monaghan Guest

    yes, Msxmanic makes a very good point, namely, that limiting elements in
    the scanner system chain limit the resolution below what film actually
    achieves, so we are really testing our scanner losses and techniques ;-)

    to get a better idea of what film can do, not limited by the scanner lens
    resolution issue (e.g., 78 lpmm in above post), one could check a laser or
    drum scanner (photomultiplier tube..) produced image. In medium format,
    one such scan can occupy the entire CDROM with data (540 MB), though it
    may be difficult to find an output device that can handle or use so much
    data ;-)

    [78 lpmm * 2 d/line * 25.4 mm/inch ~=4,000 dpi ]

    the answer to these questions is easily determined, simply because the
    film mfgers have rated their films for resolution at standard contrast
    ratios, see http://medfmt.8k.com/mf/lenslpm.html These are observed and
    tested film resolution results at given contrast ratios. At the high end,
    zeiss has achieved over 200 lpmm with slow films (ektar..) and their
    lenses in real world photographs demo'd several years ago.

    finally, kodak and others have rated film rather higher than 20 MP, for
    example pop photo tests rated 100 iso color print film over 40 MP (March
    2001 p55) equiv. for 35mm film format, and fine grain film at 25-50 ISO
    would likely push that over to nearly 125 MP equiv (per AFIPS researcher's
    study). However, for fast color print films at 400-800 ISO+, most would be
    reasonably close to 20-25 MP, so it depends on the film you specify ;-)

    regards bobm
     
    Bob Monaghan, Dec 8, 2003
    #17
  18. Sebastian Zamorski

    Rafe B. Guest

    Do you have a better way to extract a 2D digital image from film?

    No thanks. I'm interested in what it actually does, in practice,
    not only with the gear I own (not inexpensive, at that) but also
    gear I might some day be able to acquire.

    I'm still hoping someone can show me a film scan that
    seriously challenges the ones I've collected so far.

    Perhaps, as you say, it can't be done with current scanners.

    Which, to my mind, is another good reason to go with
    digital capture.


    rafe b.
    http://www.terrapinphoto.com
    film scan & digicam comparisons
    http://www.terrapinphoto.com/jmdavis
     
    Rafe B., Dec 8, 2003
    #18
  19. Sebastian Zamorski

    Rafe B. Guest


    The evidence I'm collecting doesn't seem to bear
    out your claim. Rather the contrary, in fact.


    rafe b.
    http://www.terrapinphoto.com
    film scan & digicam comparisons:
    http://www.terrapinphoto.com/jmdavis
     
    Rafe B., Dec 8, 2003
    #19
  20. Sebastian Zamorski

    Rafe B. Guest

    <big snip>

    All kinds of bogus info there -- like suggesting a compact
    digicam over a dSLR, but I'll let it pass.

    If you're truly interested in these numbers, please have
    a look at the scanner and digicam comparisons I'm
    collecting at


    http://www.terrapinphoto.com/jmdavis


    By all means, if you or anyone else have scan or
    digicam snippets that can outclass those I've gathered,
    I sure want to see them.

    Eg., I really would love to see some first-rate drum
    scan snippets of some sharp chromes.


    rafe b.
    http://www.terrapinphoto.com
     
    Rafe B., Dec 8, 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.