Thursday, January 22, 2009

Nikon D300 vs D700


Nikon's D300 and D700 digital SLR's are very similar state-of-the-art cameras that differ mostly in their sensor format. The D300 uses a cropped DX sensor (23.6 mm x 15.8 mm) whereas the D700 uses a full frame FX sensor (36.0 mm x 23.9 mm). Both sensors contain about the same number of photosites, roughly 12 megapixels. Due to sensor size differences, the D700's photosites are about 2.31 times larger than the D300's photosites. This allows the D700 to perform at significantly lower noise levels at higher ISOs. The D700 is also about US$ 1000 more expensive than the D300. It is a bit larger and heavier, and requires the use of FX lenses (the D300 can use DX lenses which are typically smaller and lighter).



What is the difference between these two cameras, in practical terms? What do I get in return for the extra cost, size and weight of the D700 vs the D300? It boils down to the different focal lengths, apertures and ISO levels that are required to achieve equivalent images in the two cameras, and the quality of the resulting images. Equivalent images are those with the same Field of View (FoV), Depth of Field (DoF), perspective, and exposure.


Crop Factor




The crop factor of the D300 is the ratio of the D700's diagonal to the D300's diagonal. This turns out to be 1.5215. A value of 1.5 is commonly used.










Field of View (FoV)




As shown above, the view captured in the DX crop is smaller than that captured by the FX sensor when the same lens and focal length are used. In order for the DX sensor to capture the same view, the focal length has to be reduced by the crop factor as illustrated below. Thus, in order to obtain the same FoV, the focal length used in the D700 should be approximately 1.5 times the focal length used in the D300. For example, the D700 at 300mm and the D300 at 200mm offer the same field of view .






Depth of Field (DoF)




The depth of field is the distance between the point nearest to the camera that is considered to be in focus, and the point furthest away from the camera that is considered to be in focus. This concept is excellently described in Wikipedia, including the formulas that can be used to calculate it.



An analysis of the formulas, under the reasonable assumption that the focal length is very small when compared to the distance between camera and subject (the point theoretically in perfect focus), leads to the conclusion that in order to maintain the DoF constant between the two images with equal FoV (e.g., the D300 at 200 mm and the D700 at 300 mm), the f-number used for the non-cropped image should be increased by a factor equal to the crop factor. So, extending the prior example, the D700 at 300mm, f/3, and the D300 at 200mm, f/2, offer the same Field of View and Depth of Field.



Exposure



Increasing the f-number by the crop factor is equivalent to stopping down by a number of stops equal to the base 2 logarithm of the crop factor squared (1.21). So to achieve the same Depth of Field, the D700 must use 1.5 times the focal length and the aperture must be stopped down by 1.21 stops relative to the D300. In order to obtain the same exposure, the D700 image must decrease the shutter speed by 1.21 stops, or increase the ISO by 1.21 stops. Assuming hand held shooting, the shutter speeds should remain constant for equal shake-induced blurriness, so the only recourse is to increase the ISO by 1.21 stops (multiply by the crop factor squared, or 2.31). Thus, equivalent images can be obtained with the D300 at 200mm, f/2, ISO 800, and the D700 at 300mm, f/3, ISO 1880.



Thinking in terms of lenses



I like to look at this in terms of lenses. What lens do I need to use in the D700 to get images equivalent to those produced by the Nikkor 18-200mm f/3.5-5.6 in the D300? I would need a 27-300mm f/5.3-8.4, and would need to use an ISO increased by a factor of 2.31 (1.21 stops).


Looking at it in the other direction may make the relationship clearer. What lens do I need to use in the D300 to get images equivalent to those the Tamron 28-300mm f/3.5-6.3 provides on a D700? I would need a 19-200mm f/2.3-4.2, but would be able to use a smaller ISO, decreased by a factor of 2.31 (1.21 stops).



Assuming you are a D300 user, what enjoyment, size, weight, and cost would you associate with an optically perfect Nikkor 16-46mm f/1.8? Well, you can get equivalent images from a D700 and a Nikkor 24-70 f/2.8.


On the other hand, in order to get images equivalent to those from a D300 with a Nikkor 200-400m f/4, you would need to attach a Nikkor 300-600mm f/6 to the D700.




Image quality



Equivalent images do not imply equal quality even assuming lenses with equivalent optics. The noise levels and vignetting may differ, and for either of these what is better may not be obvious. How does the noise level of a D300 at ISO 800 compare to the noise level of a D700 at ISO 1880? While vignetting may be higher when the lens is attached to a D700 than when attached to a D300 using the same focal length and aperture, does stopping down 1.21 stops and increasing the focal length by a factor of 1.5, when attached to a D700, compensate for this? Further, most lenses perform better when stopped down, so insofar this is true, that will improve the quality of a D700 image relative to the equivalent D300 image.



In order to test these ideas, I set up a Nikkor 70-200mm f/2.8 on a tripod. I attached the D300 and took pictures with focal lengths set to 70, 80, 90, 102, and 120 mm, and with apertures from f2.8 to f7.1 in 1/3 stop increments. Then, without moving the tripod or 70-200, I attached the D700 and took pictures at focal lengths set to 105, 120, 135, 150, and 180 mm; and apertures from f4.5 to f11 in 1/3 stop increments. I took a total of 45 pictures with each camera.


The apertures I used in the D700 to match the D300 were 1 and 1/3 stops smaller. For each D700 image, the focal length, aperture, and ISO ideally required to produce an equivalent image are listed below the actual values used for the image. The calculations used to determine the ideal parameters use the true crop ratio, and explicitly consider the ratio of focal length to subject distance, rather than assuming it is negligible. The focus in all images is set to the number "6" in the green "six ball", which was 9.5 ft away from the camera. All pictures where taken in manual exposure mode at 1/8 second, letting the camera adjust the ISO appropriately with auto-ISO. Auto white balance was used in all cases. Below are 4 pairs of equivalent pictures along with comments. If you click a picture you will be taken to Picasa Web Albums, where you can download 1600 pixel images.



Comparison 1



D300, 80mm, f3.5, 1/8 sec, ISO 400




D700, 120mm, f/5.6, 1/8 sec, ISO 900
(required for Equivalency: 121.7mm, f/5.41, 1/8 sec, ISO 954)



A comparison of the two images above, clearly shows that the D700 version suffers from significant vignetting compared to the D300 image. This is a known deficiency of the Nikkor 70-200 when used with FX sensors. Peeping at the two pictures side by side, will show that the even at ISO 900, the D700 image has noticeably less noise than the D300 image at ISO 400. In the center, the D700 image is sharper, but this is probably due to the fact that the lens performs better at 120mm f/5.6 than at 80mm f/3.5. It also appears that the DoF of the D700 image is a bit larger, although the "10" in the "10 ball" looks crisper in the D300 image. Overall, with the exception of the vignetting, very similar images.


Comparison 2


D300, 90mm, f/4.5, 1/8 sec, ISO 640





D700, 135mm, f/7.1, 1/8 sec, ISO 1400
(required for Equivalency: 136.9 mm, f/6.96, 1/8 sec, ISO 1532)

The differences between these two images are similar, but not as pronounced, as between the first pair. The D700 still has noticeably more vignetting. One change is that the DoF of the D700 now appears to be a bit narrower.

Comparison 3


D300, 102mm, f/7.1, 1/8 sec, ISO 1600





D700, 150mm, f/11, 1/8 sec, ISO 4000
(required for Equivalency: 155.2 mm, f/11.01, 1/8 sec, ISO 3849)

These two pairs of images are very similar. The D700 still shows a tad more vignetting. The noise levels are almost identical. The D300 is a little sharper at the center, but again probably due to better lens performance at 102mm than at 150mm


Comparison 4


D300, 120mm, f/2.8, 1/8 sec, ISO 280




D700, 180mm, f/4.5, 1/8 sec, ISO 640
(required for Equivalency: 182.6 mm, f/4.36, 1/8 sec, ISO 678)

Again the images are very similar.

Conclusion

The D300 and the D700 perform close when outfitted with lenses capable of delivering equivalent images. To obtain equivalent images, the D700 must use a lens that has 50% larger focal length, and about one and a third stop slower. You may not be able to find a lens for the D700 that can provide equivalent images to the D300 with a super telephoto attached. Similarly, you may not be able to find a lens fast enough and wide enough to attach to the D300, that will provide images equivalent to those attainable with the D700 and a fast normal or wide lens. Both are great cameras. If you want to have it all, then you need both.

9 comments:

  1. Thanks for the explanation, i learned a lot. I am an amateur and own a D300.

    ReplyDelete
  2. Nice music selection btw.

    ReplyDelete
  3. You can just buy a Nikon D3x and you will have both in one camera.

    ReplyDelete
  4. So the FX needs a smaller DOF to have a low-light advantage over DX at the same Field of View?

    ReplyDelete
    Replies
    1. It turns out that light advantage, either way, depends on weather you have the appropriate lens for the situation. With current typical lenses you are likely to be able to shoot with less light on FX, but as you said, using a larger equivalent aperture. The 16-46/1.8 that you would need on DX to match a 24-70/2.8 on FX does not exist.

      Delete
  5. Thanks Julio.
    I'm thinking about going from DX to FX and I'm trying to understand how FX lenses (on an FX camera) would compare to my current DX set. Especially for low-light. So I thought I had figured out that, simply put, a 24-120 f/4.0 FX lens would "compare" to a 16-80 DX f/2.7 lens.
    As I understood from your explanation, this is true but will not give the same picture because of a smaller DOF. So with that in mind, a 24-120 f/4.0 FX lens WILL give the same FOV and better low-light results as my current 16-80 f/3.5-f/5.6 (okay it's a 16-85, but this is just for comparison) Thanks again!

    ReplyDelete
    Replies
    1. It will not give you the same picture wide open, but you can always stop it down and get closer to ideal aperture in the process.

      For example, you can stop the 24-120 down to f5 at 24mm on FX to get the same picture you would get on DX wide open at f3.5 with the 16-85 at 16mm.

      Delete