How are reach and sensor size related and what are the merits of different options?

WestEndBoyWestEndBoy Posts: 1,456Member
edited January 2014 in Nikon DSLR cameras
We have went off topic in the D3300 discussion, so I am starting a new thread. I will quote the last few relevant posts for continuity.
Post edited by WestEndBoy on
Tagged:
«13456

Comments

  • WestEndBoyWestEndBoy Posts: 1,456Member
    @donaldejose: "that is a lot of performance for the buck. It even makes me think about picking up one of those bodies to use on my 300mm f4 giving me the reach of 450mm at 24mp (can even crop from that perhaps making the 300mm into a 500-600mm) for birds."

    Are you sure it increases your "reach" or is that just an illusion? To increase your "reach"--as if you had increased the focal length of your lens--requires an increase in magnification, i.e., measured in pixels, objects in the FOV should be larger, not just seem larger. As long as the sensor pixel pitch isn't too different, cropping changes the angular size of the FOV without changing the magnification of the image. For example, an image of the Moon taken with the D800 in 1.5X crop (DX) mode is exactly (or nearly exactly) the same size in pixels as an image of the Moon taken in FX mode (i.e., without cropping). In other words, if you want more "reach", you will have to buy a longer lens, not a lower grade camera with a smaller sensor.
  • WestEndBoyWestEndBoy Posts: 1,456Member
    @donaldejose: "that is a lot of performance for the buck. It even makes me think about picking up one of those bodies to use on my 300mm f4 giving me the reach of 450mm at 24mp (can even crop from that perhaps making the 300mm into a 500-600mm) for birds."

    Are you sure it increases your "reach" or is that just an illusion? To increase your "reach"--as if you had increased the focal length of your lens--requires an increase in magnification, i.e., measured in pixels, objects in the FOV should be larger, not just seem larger. As long as the sensor pixel pitch isn't too different, cropping changes the angular size of the FOV without changing the magnification of the image. For example, an image of the Moon taken with the D800 in 1.5X crop (DX) mode is exactly (or nearly exactly) the same size in pixels as an image of the Moon taken in FX mode (i.e., without cropping). In other words, if you want more "reach", you will have to buy a longer lens, not a lower grade camera with a smaller sensor.
    All you have to do is shoot an FX camera in DX mode and you have your "reach". However, I certainly agree with Donaldejose, the D3300 is a lot of bang for the buck. If I was in the market for an introductory camera, I would have to think seriously about what I was getting before spending more.
  • WestEndBoyWestEndBoy Posts: 1,456Member
    @jshickele: No, sorry, that is incorrect. The proper definition of "reach" implies that it increases only if you increase the magnification of your image. Shooting an FX camera like the D800 in DX mode reduces the angular size of the image without increasing the magnification, which is why, in the example I cited, the diameter of the Moon IN PIXELS is the same whether you shoot it in FX or DX or 1.2X mode. The same would be true of birds or other animals. If you doubt me, do the measurements yourself, as I have. In fact, try it with an MFT camera. I have compared pictures of the Moon taken with my Panasonic FZ200 superzoom, my Panasonic G6 MFT camera, my D7000 (which I no longer own), and my D800. For the same equivalent FL of the lens, the diameter of the Moon is nearly identical IN PIXELS for each of those cameras...which is why I chose to use my FZ200 to photograph a total solar eclipse in 2012 and left my D800 and its long lenses at home.
  • WestEndBoyWestEndBoy Posts: 1,456Member
    @babaganoush .. I dont really understand your logic. I have a Nikon1 and my 200mm lens definitely provides me with more "reach" when I use the lense on it..

    PS : this reminds me .. I was planning to shoot the moon with my nikon 1 tonight if the sky is clear .. ! got to get my big tripod out and stick the 400mm on it ! with my nikon1 I should get 1080mm !
  • WestEndBoyWestEndBoy Posts: 1,456Member
    @BabaGanoush, read these two web pages, as they both explain what is meant by reach.
    http://en.wikipedia.org/wiki/Nikon_DX_format
    http://en.wikipedia.org/wiki/Nikon_FX_format
    Also, consider this. A 36mp FX sensor when put in crop mode yields 16mp on subject. If a 24mp DX camera were scaled up to FX it would be 54mp. In other words you would need a 54mp FX sensor in order to be able to take a 24mp DX sized crop out of it. In the end what matters is mp on subject or field of view which is increased by 1.5 on a DX sensor. If this doesn't answer your question, we should probably take this off topic to a different thread.
  • WestEndBoyWestEndBoy Posts: 1,456Member
    Those were interesting to read, thanks @Ironheart.
    Explaines the differences so clearly. Very defined.
  • WestEndBoyWestEndBoy Posts: 1,456Member
    The relevant comments have now been transferred over.
  • WestEndBoyWestEndBoy Posts: 1,456Member
    @jshickele: No, sorry, that is incorrect. The proper definition of "reach" implies that it increases only if you increase the magnification of your image. Shooting an FX camera like the D800 in DX mode reduces the angular size of the image without increasing the magnification, which is why, in the example I cited, the diameter of the Moon IN PIXELS is the same whether you shoot it in FX or DX or 1.2X mode. The same would be true of birds or other animals. If you doubt me, do the measurements yourself, as I have. In fact, try it with an MFT camera. I have compared pictures of the Moon taken with my Panasonic FZ200 superzoom, my Panasonic G6 MFT camera, my D7000 (which I no longer own), and my D800. For the same equivalent FL of the lens, the diameter of the Moon is nearly identical IN PIXELS for each of those cameras...which is why I chose to use my FZ200 to photograph a total solar eclipse in 2012 and left my D800 and its long lenses at home.
    BabaGanoush, I mostly agree with you. My comments were directed at Donaldejose's statement.

    Heartyfisher, I agree with you too. In fact, your argument is consistent with BabaGanoush.

    When shooting the moon, you probably should have taken your D800. However, this is not because of the camera, but because of the lenses. For example, I doubt that there are any DX lenses, with the possible exception of the Macro Primes, which resolve close to the ability of the 24 megapixel DX sensor. I suspect that this is even more of a case with Micro Four Thirds. And a camera phone with 40 plus megapixels is just a sick joke. Even with FX, there are some lenses that do not benefit by upgrading from the 24 megapixel D610 to the 36 megapixel D800. Spend a few hours on DXO mark and you will see what I mean. However, there are some lenses, particularly the better primes, that still resolve better than the D800's sensor and would benefit from further megapixels.

    What I am saying is that the while you may get the same pixels with the smaller sensors, you will get more pixels of fuzz because of the inherent limits in the lens.
  • IronheartIronheart Posts: 3,017Moderator
    edited January 2014
    If you use a FX lens on a 24mp DX body I don't think you are approaching limits of the lens, in fact you are getting the "sweetest" part of the lens, namely the center. Look at any MTF chart, and you can see that the sharpest part is in the center. I don't think we are at the point yet of "out resolving" most lenses. I'd love to see scientific evidence if we indeed are. I'd say this is even more true on CX bodies, you are using an even smaller part of that FX lens. Heartyfisher may be able to give some examples
    Post edited by Ironheart on
  • WestEndBoyWestEndBoy Posts: 1,456Member
    If you use a FX lens on a 24mp DX body I don't think you are approaching limits of the lens, in fact you are getting the "sweetest" part of the lens, namely the center. Look at any MTF chart, and you can see that the sharpest part is in the center. I don't think we are at the point yet of "out resolving" most lenses. I'd love to see scientific evidence if we indeed are.
    Agreed. I was referring to DX lenses on DX bodies. It is pretty easy to make that statement. Most are zooms and zooms almost never have the resolution of primes (though there are exceptions, the 14-24 2.8 is better at 14 than the 14 2.8).
  • IronheartIronheart Posts: 3,017Moderator
    Gotcha, I should have read your post more carefully. There are a few good DX zooms that will do the 24mp sensor justice, some from other than Nikon. The DX primes seem pretty good too. I think we are in agreement.
  • WestEndBoyWestEndBoy Posts: 1,456Member
    The perceptual megapixel on DXOMark.com is thought provoking. I am not sure how much stock to put into it.
  • donaldejosedonaldejose Posts: 3,865Member
    It should be obvious that a 24mp DX sensor puts more pixels on the same size subject than any FX sensor cropped to DX size. Whether or not the lens can make good use of those additional pixels would depend upon the lens.
  • PB_PMPB_PM Posts: 4,494Member
    We are getting away from the point of those thread. Yes the DX bodies have more pixel density than FX bodies. That does not equal more magnification (reach). Only a physically longer lens, with high magnification can give you more reach.

    Think of it this way, from a purely optical standpoint, because that is what we are really talking about here. Example, if you use a pair of binoculars you are putting lenses in front of your eyes (sensors). Now, does squinting make what you view through the binoculars more magnified? No, of course not! All that happens it that you narrow your field of view. The magnification of the binoculars does not increase just because you squint, nor does using a DX sensor camera with a 300mm lens give you more magnification. You just have a squinted view at the same magnification, nothing more, nothing less.
    If I take a good photo it's not my camera's fault.
  • AdeAde Posts: 1,071Member
    edited January 2014
    Different analogy:

    Two people taking turns sharing one binocular.

    The first person has poor vision, can only make out coarse details.
    The second person has perfect vision, can resolve small details.

    The first person needs binoculars with stronger magnification, to match what the second person can see with lower magnification.

    Simplifying from a pixel density perspective, a D600 can resolve details as small as 6 microns, while a D7100 can resolve details as small as 3.9 microns.
    Post edited by Ade on
  • donaldejosedonaldejose Posts: 3,865Member
    When I used the term "reach" I simply meant a DX sensor can turn a 50mm FX lens into the equivalent of a 75mm DX lens. When putting an FX lens on a DX body 1.5x marked lens mm = FX equivalent lens mm. If you had a 300mm f4 lens on a 24 mp D600, took it off and put it on a 24mp D7100 it would be as if you now had a 450mm lens with the same 24 megapixels covering the area of the viewfinder. If you were photographing a bird that bird would now look 1.5 times closer (hence you have "more reach") and you would have more megapixels on the bird. To obtain "more reach" you can get a bigger lens or you can put a smaller sensor on the lens you have. As a side benefit you also have more resolving detail as Ade notes.

    Why even think about this? For birders if you have a D600 and a 300mm f4 lens and want to get more pixels on birds so you are lusting after a 500mm f4 lens for more "reach" but don't have the $8,000 it would cost why not consider getting a D7100 body for $1,200 to put your existing 300mm f4 lens on that body to obtain a similar 450mm "reach"? I don't know how well this would work but you would be shooting out of the center part of the lens so IQ should be good and you have a DX sensor with good high ISO IQ. Will downsizing the sensor work just as well as upsizing the lens as long as you have a very good sensor (and all the current generation Nikon DX sensors do seem to be very high quality)? Even a 24mp D3300 sensor may produce very good "450 mm" images when used with a moderately priced Nikon 300 f4 lens.

    My point is rather simple. When we want higher quality images of small distant subjects we tend to think we need expensive super telephotos. Maybe a very high quality DX sensor on a high quality moderately priced FX prime will work just as well as long as we are not going to print the image beyond 16 x 24 inches.

    Another comparison to think about. We have all seen very nice birds shot from a D3 at ISO 200 with a 500mm f2.8 lens. That was then, this is now. Today could we take a D7100 at ISO 400 with a 300mm f4 lens and produce equal images? Maybe. I have not tried it but it is worth thinking about for those of us who are never going to be able to afford Nikon's super telephotos. Think of it this way. Can "super DX sensors" reduce the need for super telephotos? I would like to think so.
  • PB_PMPB_PM Posts: 4,494Member
    Different analogy:

    Two people taking turns sharing one binocular.

    The first person has poor vision, can only make out coarse details.
    The second person has perfect vision, can resolve small details.

    The first person needs binoculars with stronger magnification, to match what the second person can see with lower magnification.

    Simplifying from a pixel density perspective, a D600 can resolve details as small as 6 microns, while a D7100 can resolve details as small as 3.9 microns.
    Bad analogy, since focus (just like the adjustor on an optical viewfinder) can be corrected to your eye sight on binoculars... just saying.
    If I take a good photo it's not my camera's fault.
  • AdeAde Posts: 1,071Member
    edited January 2014
    Poor vision != focus. E.g., cataract, macular degeneration, etc.
    Post edited by Ade on
  • sevencrossingsevencrossing Posts: 2,800Member
    edited January 2014
    until I joined NRF I had never heard of "reach"

    no one ever suggested to me that 35mm cameras had more "reach" than a medium format

    It is term is still do not understand

    it is not true that all DX bodies have more pixel density than FX bodies ( the D90 is still listed by Nikon UK and had less pixel density than my D800 ) may be the latest cameras do , but that can change

    To me, reach seems to infer the ability to photograph a distant object

    My D800 fitted with a FX 28 - 300 will have a similar reach as a D7100 fitted with the equivalent DX 16 -200

    The very big advantage DX has, is the cameras are cheaper lighter and smaller

    FX produces a bigger "negative" and throughout photographic history, a good big'ne has always beaten a good little'ne; the gap may be very narrow and many people may not notice or care about the difference, but there is and always will be, a gap
    IMHO

    Post edited by sevencrossing on
  • framerframer Posts: 491Member
    You own an 8x10 view camera with a big lens, doesn't matter the size. You take a 8 x 10 picture. Now you put on a 4 x 5 back and you take a 4 x 5 inch picture. You now go to a paper cutter and cut the center out of the 8 x 10 so it's sized to a 4 x 5. Compare them. Are they not the same. That all that's happening between FX and DX. You lose data anytime you use DX.

    framer
  • WestEndBoyWestEndBoy Posts: 1,456Member
    We are getting away from the point of those thread.
    Kudos and point to you for getting us on track.
  • AdeAde Posts: 1,071Member
    You own an 8x10 view camera with a big lens, doesn't matter the size. You take a 8 x 10 picture. Now you put on a 4 x 5 back and you take a 4 x 5 inch picture. You now go to a paper cutter and cut the center out of the 8 x 10 so it's sized to a 4 x 5. Compare them. Are they not the same. That all that's happening between FX and DX. You lose data anytime you use DX.
    That would ONLY be correct if we didn't change film type between the two cameras.

    But suppose we load the 4x5 with film having very fine grain (e.g, Ilford FP-4), while loading the 8x10 with a different, more coarse grain film (e.g., Ilford HP-5). Now when we compare the 4x5 film with the crop of the 8x10, they will NOT be the same.

    Similarly, the sensor density between DX and FX are not the same. A 24mp DX sensor is more "fine-grained" than a 24mp FX sensor.
  • tc88tc88 Posts: 537Member
    edited January 2014
    (1) If there is perfect film with infinte resolution, then magnification doesn't really matter. One can take a picture with a perfect 10mm lens, and enlarge and crop it in print to have any magnification needed. Now film was never perfect, and digital is discretized. So in theory, only the pixel density matters, regardless the magnification of the camera/lens.

    (2) It may not be possible to resolve all pixels. That's what Ade's analogy is about. My 70-300 at 300 can resolve no where close to my 300 f/4. So it really has less "reach" since I can use 300 f/4 to take a picture of the same object at a further distance and digitally zoom in to have the same quality.

    That's why while Nikon 1 should provide some additional "reach", I don't feel it has the "reach" at the ratio of the formats. Both the pixel count (which translates to pixel density) and the resolving power of the lens/cameras count. Note that I also included the camera since the ISO noise will also affect the resolving power.

    Really the word "reach" as referenced by format and focus length is a misnomer in digital age.
    Post edited by tc88 on
  • donaldejosedonaldejose Posts: 3,865Member
    Nomenclature issues such as "reach" set aside; has anyone tried a Nikon 300 f4 with a D7100 or D5300?

    Anyone compared the image from a 300mm f4 on a 24mp DX sensor with the image from a 500mm f4 shot with an FX sensor?

    I am thinking the most recent 24mp high ISO IQ DX sensors may give a new lease on life to the old and inexpensive 300mm f4 lens.
  • CoastalconnCoastalconn Posts: 527Member
    I will share my perspective being a birder with this simple Sparrow shot from my D7100.. It was cropped down to 1735x2690 from 6000x4000 resulting in about 4.6 MP. On a D800 it would have been 1388x2152 about 3 MP.. on a D600 it would have been about 1136x1764. My point being that a D7100 offers the advantage of putting more pixels on your subject that allows for more resolution. Which could make a big difference in prints. That is how I quantify the "crop factor" or "reach" of course just my humble opinon..
    House Sparrow in the sun
Sign In or Register to comment.