The Next 35mm Full Frame Digital Camera?

Discussion in 'Mirrorless Digital Cameras' started by neil_poulsen|1, Aug 7, 2003.

  1. Canon has scored quite a coup introducing the first full frame 35mm
    digital camera. In spite of the high price, this throws down the
    challenge to all other manufacturers.

    Has there been any information as to who will be next and when? I'm
    looking forward to the substantial price drop that will occur.
  2. Kodak has the 14n that's already out and about half the price of the Canon... though the reviews i have read say that the image quality isn't as good as canons.
  3. Actually Contax had the first "full frame" (as in 24mm x 36mm) digital image sensor
    built into a 25mm SLR like body -- they were also the first to stop production as well.
    I nthe studio where you can work at ISO 100 to 200 the Kodak 14n produces as good
    if not better iamge than the EOS 1Ds accordingto the tests I've seen run. I doubt tha
    you will see a substantial price drop occur any time soon. Sensor technology and lens
    technology are geting better on an almost monthly basis (for example the the new
    extremely low noise technology and fast on / fast write technology incorporated into
    the recently announced Nikon D2H. My suspicion neal isthat the market is going to
    drive which way the camera makers go: if someone can produce a smaller (lower
    cost), high enough resolution sensor that satisfies the vast majority of the high end
    amateur and professional market ,and there is a bigger profit margin in making such
    a camera a 24 x 36mm sensor might be an orphan technology. <P><P>
    On the other hand Bill Gates allegedly once said something to the effect of that an 8
    bit operating
    system was plenty of power for anyone who used a desktop computer.<P>
    24x36mm imaging chips in digital backs for meduium format systems way out
    perform what an EOS 1Ds can do, more room for better cooling and other factors are
    part of the reason, as is better software.<P>
  4. The few comparisons I've seen between the 1Ds and the Kodak 14n, some shot by people here show that the 1Ds is the same or better at low ISOs and produces better image quality than the 14n over ISO 200. In ultimate spatial resolution, the 14n is superior and it does show in some types of images.

    If you're doing production shots or studio setups where you want high quality and won't be using high ISOs the 14n is a good camera. I don't find it quite acceptable for running about with because the noise is still high and i value that over spatial resolution in many instances.
  5. I am not sure that I understand the argument about small sensors being good enough. I really want a full frame sensor (or even bigger) for underwater stuff.
  6. Ellis, could you please use a spell checker! Anyway, I agree, "sub-frame" sensors might be the future, sicne they're lower cost while offering sufficient quality. Certainly, I will at some point buy a full-frame system for myself (the largest format I shoot with is 4x5" anyway), but looking at the specs and lenses for the new Olympus 4/3 system, I can see the appeal of these smaller formats: light, very fast lenses, good image quality. It really boils down to choosing the right tool for the job.

    I don't expect any large price drop: a full frame sensor requires and area of 24x36 mm minimum and given that wafers are 8" or 12" (in newer foundries) in diameter, it's easy to calculate how many chips a wafer gives. This and the yield of the chips will limit the economies in such a way that fast and dramatic price reductions aren't possible.
  7. I don't own the Canon EOS 1D but I have owned a Kodak Pro Back Plus for over a year and have been shooting more than 80% of my commercial work digitally with that back.

    Just last week I purchased a Kodak 14N to use on location and in the studio when I don't need view camera movements or do need a wide angle lens capability.

    All of my pre-digital commercial work has been shot on 100 iso film, 2 1/4, 4x5 and until recently 8x10. So for me a limitation of the 80-100-200 iso on the 14N is not a problem. The Pro Back has a nominal iso of 100 also but, using Kodak's Capture studio software, you can adjust raw files from 50 - 200 iso in 1/10 stop increments.

    The ProBack produces a 13.5" square image at 300 dpi. The 14N produces a 10"x15" 35mm ratio image at 300 dpi. The 14N has a slightly higher resolution that the ProBack in that regard. If you crop the 35mm ratioed 14N image to 8x10 or 8.5x11 then you do lose some image and the Pro Back cropped to those sizes is better.

    Although I have only had the 14N a week, I haven't noticed any noise in the shadow areas when shooting at 80 or 100 iso so far. In fact I actually like the color, saturation and extra sharpness (no anti-aliasing filter) of the 14N slightly better than the Pro Back.

    The downside to the 14N is the N80 based body which is not as robust as it could be though I am careful with my equipment and don't abuse cameras very much. And the N80 body only allows 1/2 stop brackets which is a huge bracket for digital capture.

    The other dislike I have for the 14N compared to the Pro Back is the bundled capture software. Capture Studio for the Pro Back is much better, more configurable and more elegant in use than either Camera Manager or Photo Desk which come with the 14N.

    Shooting tethered to the computer with the 14N, I have to run both Camera Manager and Photo Desk at the same time to enable both hard drive capture and a viewable contact sheet. Capture Studio handles both these functions by itself.

    Still, at $4800 for the 14 N I'd rather have two 14Ns than one Eos 1D.
  8. Anyway, I agree, "sub-frame" sensors might be the future
    That's good, because Bill Van Antwerp said,
    I really want a full frame sensor (or even bigger) for underwater stuff.
  9. Even the sensors on the digital MF backs are small compared to MF film, although I recently saw a press release for one about the size of 645 film. In the medium format community 645 is often called "Medium Format APS".<P>

    Recently I was doing a model shoot where I was using 35mm for some of the shots and my old Mamiya RB67 for the others. And I was reminded once again what a pleasure Medium Format is. On a camera where a portrait lens might be a 180, you have exquisite control over depth of field. And all that light-sensitive real estate means you can capture images that can be blown up to huge prints, or heavily cropped, way beyond what 35mm or digital can handle. Even if we allw for MF lenses being a little softer than 35mm lenses, so we only scan at 3200 PPI instead of 4000 PPI, that still produces a 7040x8640 pixel image - i.e., a 60 megapixel image!
  10. I have an basic suggestion and request for those D-SLR manufacturers :

    Using the right lens system for the corresponding sensor size. i.e.

    35mm format lens for 35mm size sensor (e.g. Canon 1Ds) WITHOUT focal length multiplier
    APS lens system for APS size sensor (e.g. Pentax *ist D-SLR) WITHOUT focal length multiplier

    and I don't want to see 35mm lens on APS size sensor (Nikon D2H) with 1.5x focal length multiplier which confuse and doesn't match the focal lengths of my existing Nikon AF lens.

    I really don't want to buy the transition products (lens) in order to matching the right focal length as used in my existing 35mm film system.
  11. Mr. Ho,<P>As of today (August 8, 2003) your choice is clear: you want the Kodak
    14n. what will tomorrow bring? Send me a cashiers check for US $50,000 and once it
    clears the bank I will tell you all that I know.
  12. Ellis,

    Check's in the mail !

    What's the latest scoop ? #8^)
  13. I didn't buy my F5 because I wanted "sufficent quality" and I won't buy anything less than a full frame Digital SLR even if it means selling all of my Nikon Gear!
  14. Rick,

    You can have a full-size 35mm sensor and not have to sell your Nikon lenses if you buy a Kodak 14N. #8^)

    No offense, but in my line of work as a commercial advertising shooter there have been plenty of times, almost every time actually, when 35mm film (even when shot with my F3HP) didn't have "sufficient quality" for the jobs I do.

    That doesn't seem to be the case with the 14N.
  15. Brooks, when you refer to the EOS 1d, are you meaning the 1Ds since its more of a competitor to the 14n? Even though IMO there's still no contest between the two.
  16. Carl,

    Yes, I did mean the 11mp chip 1 DS. I'm not a "strong swimmer" with the names and number designations of the Canon line of cameras. I'm sure the 1 DS is a fine camera. My point was that having 2 14Ns (I only have one) with one as a back-up would be better than having 1 1DS which costs about the same as 2 14Ns and has nominally less resolution but does have a more robust camera body than the 14N.
  17. Peter,

    As you will note by re-reading my response, I'm referring to what Ellis Vener wrote in my first paragraph. I'm actually mentioning Ellis' name in the first sentence. There is no typo: the sub-frame reference is to Ellis' answer and in conflict with Bill van Antwerp's opinion.
  18. While it's by no means a general-purpose camera, Horseman's Digiflex cameras ( or ) provide 11-16 megapixels using Kodak's DCS Pro Back, Imacon's IXpress, or PhaseOne digital backs. It's a bit of a hack, but it uses Nikon's lenses and most MF-type digital backs, and gives a 24x36 frame size.<p>
    Of course, the backs alone cost more than the 14N or 1Ds, and the camera's another $2-3K...<p>
    CL Ho: if they list the physical focal length and angle of view (not that stupid "equivalent", I'd be delighted.
  19. True, true. But if I'm investing in high end digital I would either go all out or not bother. Which is part of why I'm not doing that right now. From my use there's no doubt the 14n isn't a bad camera, but in almost every way except spatial resolution and cost the 1Ds is an incredible machine and a much better one than the 14n. I won't be buying either, the technology isn't mature enough for me to want to sink any money in to it at the DSLR level.

    I still use them of course, but I'm not buying in to anything for a few years yet at least.
  20. Something I've been wondering and can probably answer for myself, but if you have, say, a 200mm lens at f4 on a 1.5x DSLR you effectively get a 300mm f4 lens - but does it have the same (narrow) depth of field as a 300mm at f4 or is it the same as a 200mm?

    Similarly, if a 50mm lens on a DSLR has the same field of view as an 85 does on film, what aperture would I need to get the same DOF of an 85mm at f2.8?

    I guess what I'm really asking is if the difference in DOF at equivalent angles of view, between full frame/film and a small sensor chip, significantly affects your results or shooting style.
  21. The lens doesn't same when you use a smaller sensor. You just have a different field of view. Its the same as using an 8x10 lens on a 4x5 camera, or vice versa (not good idea generally :p). The lenses optical properties do not change, the DOF remains the same, the same aperture gives the same DOF. What does change is the same area is "magnified" more than with a 35mm full frame sensor/film.

    You get the same DOF at the same aperture. It would be the same as just using the central 21by17mm of 35mm film... aka cropping.

    No magic here.
  22. Something I've been wondering and can probably answer for myself, but if you have, say, a 200mm lens at f4 on a 1.5x DSLR you effectively get a 300mm f4 lens - but does it have the same (narrow) depth of field as a 300mm at f4 or is it the same as a 200mm?
    Look for Bill Atkins' articles on d-photog and DOF. If I understand correctly, if the difference between the sensor and 35mm film is 1.5x, you'll find that you'll have to stop down by a factor of 1.5 to get equiv. DOF. So if you want the DOF of f5.6, you'll need to stop down to 5.6/1.5 or 3.7. Think of how you have to translate between MF and 35mm focal lengths 50mm on 6x6 is wide angle; 80mm is normal, and the DOF is a lot narrower in the larger format.
    Narrow DOF is important to me, which is why I shy away from the smaller sensor d- cams.
  23. Actually it's "Bob Atkins", but I appreciate the plug anyway!
    I didn't buy my F5 because I wanted "sufficent quality"
    Well actually you did. Otherwise you'd have bought a Pentax 645 or a Hasselbald or a Pentax67 or a 4x5 or 8x10 field camera. Obviously 35mm was "sufficient quality" for your application or you would have had to go for a larger format.
  24. Of course- Bob- what was I thinking?! Many apologies! (I think I was thinking of an
    early Apple software developer Bill Atkinson who created MacDraw...)
  25. I believe the marketing departments are missing the point. I want a sensor the same size as the film it replaces, so all my precious wide-angle lenses don't turn into short telephotos!

    I really don't care how many zillion pixels -- this is a case where SIZE MATTERS!

    I would prefer a 36x24mm low-density sensor to a 24x16mm high-density one which makes my $1800 14mm rectilinear ultra-wide do the same job that my $400 20mm does with film.

    Then again, maybe *I* don't get it... The marketing department will be happy to sell me a new 9mm rectilinear ultrawide for $3600 next year...

    And I won't even look in my MF and LF camera bags...

    Bah! Phooey!
  26. If digital SLR technology were driven by marketing departments the very first one would've been full-frame ...

    A 25mm x 36mm chip is HUGE by IC manufacturing standards and getting enough yield to sell them at reasonable price is going to take time. Whatever the resolution.

    At the same time shrinking sensors while maintaining adequate sensitivity is a challenge.

    A smaller sensor's always going to be cheaper, though, just for reasons of yield, the number you can cut out of a wafer, etc.

    Someday they'll be cheap enough so it doesn't matter ... I think too many people own too many expensive wide angle lenses for the manufacturer's to orphan us by switching to APS-size sensors and equivalent lenses wholesale in the near future. Though the temptation's got to be there.
  27. What Don says...

    You will get something less than 29 24x36 mm sensors on an 8" silicon wafer. The
    sensor area is about 860 mm2. To put that in perspective, the semi company I work
    for (about $10B sales) considers die areas > 100 mm2 large. That's greater than 8
    times larger.

    Of the 29 sensors I bet the yield is *much* less than 50% cuz defects are exponential
    with die area. Maybe a lot worse. And you will probably still have individual defective
    pixels - hopefully clumps are rejected. As a side note, I wonder if 300D sensors have
    more allowable defects than the same 10D sensor? Silicon wafer costs are pretty much
    fixed. $2K to $3K, depending on a lot of things, and likely to stay that way. 12"
    wafers will help, *if* the economics make sense. Which means that full-frame sensors
    (and cameras) will always be expensive. Now, how about that 60 x 60 mm MF sensor
    people want? Four of those will fit on a wafer. The yield will be MUCH worse than a
    24x36 sensor. Those, and the underlying camera infrastructure will be real expensive.

Share This Page