D300 vs D3 - which has the reach?


The DX crop reach


The Nikon D300 has earned a somewhat cult status among certain photographers. The professional grade autofocus, robust build and fast frame rate are the camera's most noteworthy features. And of course because it is a DX crop camera, it has superior 'reach' or 'focal length multiplier' compared to FX format cameras.

Here are a few samples that let you see with your own eyes the crop-factor-reach in action. I took a D300, its contemporary D3 and an old DX-lens-in-disguise Nikon Zoom-Nikkor AF-S VR 70-200mm f/2.8. I positioned the rig on a sidewalk and aimed the lens towards a demolition/building site. Below are some crops for evaluation and the full size images too.


How to do a fair comparison?


The first choice that comes to mind is to shoot the images at exactly the same settings. I did that and ended with an image pair shot at ISO 200, f/2.8, 1/800s at 135mm (because the lens has a reputation of being soft at 200mm).

At this point someone will cry out 'foul'. They will claim that the D300 has deeper DOF and the D3 image has less noise because it received more total light. To remedy this I stopped the lens one stop down on the D3, raised ISO one stop and re-shot the scene at ISO 400, f/8 and 1/800. (Note: the DOF is the same with the same settings, because the focal length, aperture, magnification and the distance all remain constant.)

Everybody is happy now, I guess. No, I hear a loud cry from the back ranks. The comparisons are not fair because the D300 should use 1.5 times higher shutter speed because of the crop factor. Ok, I hear you. I did that and then some. I raised the shutter speed one stop and compensated by raising the ISO by one stop and re-shot the scene with D300 using ISO 400, f/5.6 and 1/1600.



The image crops


The D3 images needed to be uprezzed to the same final size as the D300 images, so I uprezzed them with Photoshop bicubic to 152 %. I could have downrezzed the D300 images, but that might have destroyed any fine resolution benefits the images have, so I kept the D300 images at native size.

So, in the end we have D300 native vs. D3 uprezzed. Here are the crops from three different parts of the image and below them the four full-size non-uprezzed images. You might want to try uprezzing/downrezzing them with better tools than PS bicubic and see if the results differ. You can also try to sharpen and deconvolve them with all the fancy tools available. Please report back, if you can eke out more resolution out of the D300 images than the D3 images.

From these samples we can see that the cult status of D300 is somewhat unwarranted. At least when it comes to the magical focal-length-multiplier and reach superiority that the crop factor allegedly provides. The uprezzed D3 images look just as crisp and detail-rich as the D300 images.




The crops and full size images

From left to right:









D300 1/800 f/5.6 ISO 200 - Click for full size in new window



D300 1/1600 f/5.6 ISO 400 - Click for full size in new window



D3 1/800 f/5.6 ISO 200 - Click for full size in new window



D3 1/800 f/8.0 ISO 400 - Click for full size in new window


Why is this happening?


Why isn't there a more clear reach advantage by the D300? It has a lot higher pixel density. Well, there are a lot of factors in play here, but I have two candidates.

The lens on DX is not 1.5 times sharper

To be able to really extract the benefit of more pixels in a smaller sensor, the lens in use should be 1.5 times sharper. That is, on a DX camera the lens should be able to render the same resolution measured in line pairs per picture heigh (lp/ph) on a sensor that's 1.5 times smaller. Hence, to reap the benefits of DX 'reach', we need a lens 1.5 times sharper.

But we are using the same lens here, so it won't mystically transform into another lens. Perhaps my Nikkor AF-S VR 200mm f/2 would have been sharper on the D300? But hey, that could be mounted on the D3 too.

Noise

The D300 sensor is pretty noisy already at base ISO 200. I have never seen it render "clean" skies. At ISO 400 the noise has already permanently destroyed any and all resolution advantage there might have been.

Camera shake?

Not really at these shutter speeds while the camera is on tripod. Looking very closely at the images there seems to be no camera shake.

Occam's razor

Usually the simplest or most obvious answer is the right one, instead of an elaborate fabrication or conspiracy theory. Here it seems to be that despite the vocal cult following the D300 really does not have any more 'reach' than a D3. The D300 sensor is so noisy that noise eats away the resolution gained by higher pixel density.


What the test sites say?


DxOMark's rigorous test results seem to agree with my real world shooting. D300s mounted with a AF-S VR 200/2 II gets only a score of 20 and that is almost 'poor'. The Sharpness rating is 10 Mpix. Click here for D300s & 200/2. On the other hand a D3 with the same lens gets a score of 27 and that is at least average. The sharpness rating is 12 Mpix. Click for D3 & 200/2. The results were similar when DxO mounted a AF-S VR 70-200mm f/2.8 II to a D300s and D3. See for yourself: D300s & 70-200/2.8 and D3 & 70-200/2.8.


Conclusion?


The quick conclusion seems to be that D300 does not offer any 'reach' benefit over a D3 despite its higher pixel density (or crop factor). So many armchair theorists have calculated the 'reach' difference in so many ways: crop factor benefit, pixel density benefit and what not. But the proof of the pudding is in the eating. D300 offers no reach benefit over a D3 (or a D700). Probably due to the fact, that the pixels in a D300 just aren't good enough. Quantity does not conquer quality.




© 2014 Peter Forsell