Part III: Diffraction, Anti-aliasing and Conclusion

Canon 5D Mark III   17mm TS-E   f/10   0.4s   ISO 100:   Click for larger image


There is a lot of conflicting information about diffraction. A high resolution 36MP sensor is NOT more diffraction prone. Diffraction is solely caused by the lens aperture and it will always be the same regardless of the resolution of the sensor. Of course, the more resolution you have, the more you will see the effects of diffraction evident. When using the same lens with the same aperture, the Nikon D800 will always have more resolution than the Canon 5D Mark III or any other current digital SLR on the market.

Canon 5D Mark III   24mm TS-E   f/10   1/15s   ISO 100:   Click for larger image


Shooting at the lake made me wonder, how would my images look without an ANTI-ALIASING (AA) filter. Nikon is set to release a D800E which cancels out the effects of an AA filter. However, there must be a compelling reason to include AA filters when designing a digital SLR. The filter blurs the "optical" image before it is sampled making sure there are no spatial frequencies present beyond the ability of the sensor to correctly render it. This is currently the most effective way to combat aliasing without noticeably impacting resolution. So basically, the filter function is there to block spatial frequencies that the sensor can not resolve while allowing what the sensor can resolve to pass through. AA filters are not 100% effective, there's always a small loss in resolution and some aliasing that's noticeable in your images even when the filter is present.


Canon 5D Mark III   24mm TS-E + 1.4x III   f/9   1.3s   ISO 100:   Click for larger image


The worse aliasing offender is Moiré but aliasing can also manifest itself in the form of jaggies, stair stepping, wavy lines, and sparkling, inaccurate, unreal detail. It really turns up the digital "fake" look, which is not for me. In my opinion, removing the filter is not beneficial for certain applications. For example, with architecture, fashion or any other subject filled with man-made patterns, aliasing is more apparent and may require post-processing correction which is never 100% effective. You can't effectively remove aliasing in post-processing because any higher spatial frequencies beyond Nyquist (sensor resolution) will be combined with the real detail, distorting and contaminating it. The best you can do is "band-aid" it as much as possible in software.


Canon 5D Mark III   17mm TS-E   f/5.6   8s   ISO 100:   Click for larger image


Moiré aliasing is not so obvious in landscape or nature images because we have "random" patterns. But, other types of aliasing artifacts can still mix in with real detail, especially when shooting with high resolution lenses at their optimal apertures (usually from f/4 to f/8). When shooting at smaller apertures diffraction will blur your image thus mimicking the effects of an AA filter, which is also not desirable.

Who knows, maybe one day we will have even higher resolution sensors making the necessity of having AA filters irrelevant. As for now, I'm glad Canon kept it on the Mark III and would like to see how the Nikon D800E performs without it.

On the last two days of my trip, there was so much snow falling that I had no choice but to hit the slopes. The cameras were forced to sit this one out. Ok, I did feel a little guilty neglecting them in the dark closet, but hey how often does a guy from Brazil get to go snowboarding?


Canon 5D Mark III   17mm TS-E   f/10   0.4s   ISO 100:   Click for larger image


The next step would be to get home and start printing. I mostly print my pictures in-house at 17x22 inches on my Epson 4900 and occasionally larger 24x36 inches. In my controlled tests, I shot with both cameras using the same Zeiss Distagon 21mm f/2.8. Thanks to a high quality Novoflex EOS adapter, I was able to mount the Zeiss 21mm (Nikon Mount) on the 5D Mark III. I printed in my preferred size formats and then even larger to test the capabilities of the 36MP sensor. I wanted to really see how the new sensors compared when using the same glass. There was no visible difference based on the smaller (17x22) print size. Small nuances detected in controlled environments were not perceptible on the prints. However when printing crops, interpolated to a larger scale print, Nikon revealed it's superiority in resolution and detail.


Canon 5D Mark III   17mm TS-E   f/10   0.5s   ISO 100:   Click for larger image


After an exhausting week, it was finally time to unpack my gear and look over everything I used on my trip. I had both the Canon 5D Mark II and 5D Mark III bodies with these lenses: 17mm f/4L TS-E, 24mm f/3.5L TS-E II, 90mm f/2.8 TS-E, 70-200mm f/4L IS and the Canon 1.4x III extender which works great with all TS-Es.
I'd also like to note, that in a majority of the shots portrayed in this review I used LEE Filters consisting of: Grad Neutral Density (ND) Soft Filter Set 4 x 6", 4x4" ND 0.6 Resin Filter and 4x4" Circular Polarizer Glass Filter. For the Canon 17mm f/4L TS-E, since it does not take filters, I was able to come up with a customized workaround.

I also took the Nikon D800 and the Zeiss Distagon T* 21mm f/2.8 ZF.2. Which I recommend for landscape shooters. I did try the Nikon 14-24mm f/2.8 lens and the 24mm PC-E, but both gave me blurry corners. Not sure if my was my copy or if the 36MP sensor was "seeing" more flaws. Aside from the corner softness, the 24mm PC-E revealed serious "CA" when tilting or shifting. Therefore, I deliberately did not use them that much and concentrated on the 21mm prime instead.

As for tripods, I brought both my Gitzo's GT3542LS and Gitzo's GT5531S Ball-heads were the RRS BH-55, Acratech Ultimate and Sunwayfoto XB-52.


Canon 5D Mark III   24mm TS-E   f/3.5   23s   ISO 1600:   Click for larger image


Overall, it was a smooth trip to Yosemite in the spring. The waterfalls were flowing and the weather had it's fair share of up's and down's. The first 2 days, the skies held nothing inspiring so I was able to concentrate on getting to know the new cameras better. My final days were covered in snow and with the storm, came the dramatic weather I wanted as I headed out towards Mono Lake.

As for the Canon Mark III, the 22MP sensor is not that different when compared to the Mark II. You might be able to detect a half stop DR improvement in the high ISO settings (above 6400) but this is not applicable to my needs and nothing to write about at the low ISO 100-400. But, the build quality on this new Canon is impressive and its auto-focusing system is a terrific upgrade. Personally, I don't rely on the camera auto-focusing system, but if you shoot events, weddings, photojournalism or even sports, you should seriously consider this upgraded camera.


Canon 5D Mark III   24mm TS-E   f/7.1   1/500s   ISO 100:   Click for larger image


As you can see from my samples, I mostly stuck with the Canon Mark III on this trip. Mainly because of the difficulty I faced with Nikon D800's poor LCD Live View performance in low light. For my photography needs, this was the Achilles' heel of an otherwise superb camera. But, if you don't rely on live view, then this will never be an issue for you. Aside from that, my only other desire would be for Nikon to release an ultra wide-angle tilt and shift lens and upgrade the current PC-E lenses so that the shift and tilt could be rotated independently.


Nikon D800   21mm   f/8   1/30s   ISO 100:   Click for larger image


There's no question that Nikon produced an exceptional camera. In fact, many photographers will gravitate towards it because of the high resolution. As far as I'm concerned, the big attraction is the incredible dynamic range. Especially in the 100-400 ISO range, which is critical for landscape photography. Nikon's ability to recover shadows and highlights without a noticeable penalty is a huge bonus.


Canon 5D Mark III   24mm TS-E   f/3.5   24s   ISO 800:   Click for larger image


Finally, I leave you with an image of my silhouette, captured without my consent by the illumination of a full moon that guided my path as I said goodbye to my favorite falls. The bottom line, is that these are both amazing tools for photography. There are good points and bad points to both. Nothing is ever perfect and the best advice I can give, is for you to evaluate your needs and make your decision based on what you primarily shoot. There are workarounds to every problem but ultimately a photographer needs to know the camera's strengths and weaknesses in order to get the most out of it.



Here is a list of my personal gear used on this photography journey. If you are interested in purchasing any of these items, please use the below links. It supports my site when you make your purchase from our trusted partner B&H Photo.