50mm f/1.4: LEICA vs Canon vs Nikon Systems
Let's compare how three 50mm f/1.4 lenses compare on three of the best digital cameras on Earth that you'd want to use in the field.
Let's see what the best each from Nikon, Canon and Leica can do.
The examples shown below are crops from what would be 58" (1.5 meter) wide prints, presuming you're looking on a standard monitor, on which these crops are 7-1/4" (18cm) wide. Each is shown at the same size, as you can see by the trees staying the same size.
We have three contenders:
I had no current LEICA SUMMILUX-M ASPH 50mm f/1.4 with which to make this a fair comparison, so I used a 45 year old lens instead.
I shot the newest 50mm f/1.4 AF-S on a 27MP-equivalent FX body.
I had no access to a D3X (24MP), so since I'm only looking at the center of the frame, I gave Nikon an advantage and used a D300s, which has a higher pixel density than any of the other full-frame cameras.
When you consider the higher pixel density of the D300s, it would fill the FX frame out to 27MP.
Therefore, I'm showing something of which Nikon is not capable. The D3X has only 24, not 27MP, and the D700 has only half that. The D700 and D3 look like mush compared to the D3X, and the D3X doesn't look as good as the fake 27MP FX camera I've rigged up here.
I shot everything on a tripod at a special optical test range under controlled conditions. There was no wind and no heat shimmer.
Each camera was shot at ISO 200.
I shot raw and opened all the images in Adobe Camera Raw in Adobe Photoshop CS4. Each was opened at default settings, with "as shot" white balance. Each camera was set to DAYLIGHT WB.
(I tried opening these in Photoshop CS2, but ACR in CS2 can't be upgraded to read the Nikon or Canon files; only the Leica used the standard DNG format and would open. The Leica files looked better in CS2.)
The Nikon file, as usual, was overexposed, so I let Nikon cheat and I dropped the exposure back 1/2 of a stop as I opened it in ACR.
These are from the very center of each image.
At f/1.4, the ancient Leica lens has more resolution than Nikon's newest 50/1.4 AFS.
Canon isn't in focus, typical with the EF 50/1.4 and the 5D Mk II. Even when the AF system says AOK, every once in a while it's not.
At f/2, the new Nikon lens has caught up to Leica's 1964 lens.
Canon is still out of focus, and I refocused for each shot.
At f/8, lenses perform about the same. At f/8, we're more comparing cameras than lenses. These again are from raw files opened in CS4 ACR.
At f/8, the Leica lens is limited by diffraction; it was sharper at f/5.6. With 25MP, the Nikon looks a tad better than the 18MP Leica.
The Canon looks fine, since f/8 eliminates the focus problem before, or the Canon may have gotten it right for this shot. (I refocused for each shot.)
Since the Leica DNG files would open in CS2, in which the M9 files look sharper than in CS4, I also opened the Leica in CS2, at which point it also looks about the same as the Nikon and Canon.
Don't go blind splitting hairs. At f/8 these are all too similar to worry about, especially since your choice of software makes as much difference as the cameras.
If I showed a 12MP camera like the D300s or the D3 for comparison, it would look like mush as I showed here.
Against all odds with only 18MP, the lowest-resolution LEICA M9 with the oldest lens from 1964 have as much— or more — fine detail than the best from China (Nikon) and Japan (Canon) today.
Tricky devil that I am, what I've really shown is that the 18MP LEICA M9, with a 45-year-old lens, performs as well or better than the newest 21-27MP full-frame DSLRs from the Far East.
The M9 sensor is made in Rochester, New York, USA. GO USA!
What If? top
I should have used a modern lens on the Leica. Using a 45-year-old sample puts the Leica at quite a disadvantage, but tough, it still comes out on top.
Nikon looks better than it should at f/8 because I used a mock-up of a 27MP camera. THere is no way to get this level of detail from Nikon, even in a D3X. Yes, I used a D300s, but the D300S can't take a full picture: it only sees the middle half of the image (DX). See Crop Factor.
Yes, I could have used better Canon lenses, like Canon's 50mm f/1.2 L, but then again, Leica makes the LEICA NOCTILUX 50mm f/0.95 ASPH. I stuck to 50mm f/.4 lenses here. Nikon's 50mm f/1.2 isn't that great, and Nikon hasn't made an aspherical 50mm lens since the last decade.
Shoot a Leica. It weighs less, looks better and costs less than a D3X, even with ancient lenses. Imagine how well it will work with modern ASPH lenses! See A Canon vs. Leica vs. Nikon Sharpness comparison with modern lenses.
If shooting the 5D Mark II, make sure whatever lenses you use give good focus. With Canon, some lens and camera combos result in misfocusing a small percentage of the time, even when AF OK lights up.
Any of these cameras is worlds away from 12MP cameras, either in FX or half-frame, when printed this big, but then again, do you really care how sharp something is when printed 5-feet (1.5m) wide and looked at this close?
Even if you get a D3X, remember that here I'm showing something of which the D3X isn't capable: the D3X has only 24, not 27MP, and the D700 has only half that. The D700 and D3 look like mush compared to the D3X, and the D3X doesn't look as good as the fake 27MP FX camera I've rigged up here. With Nikon, anything other than the D3X looks a lot worse than this comparison.
Ultimately, none of this matters. All that matters is getting the right picture with whatever camera you have with you. I have no problem printing this big from 6MP cameras.
In-Camera JPGs top
While I've got your attention, lets see how each camera makes JPGs. These three images are taken from each cameras' JPG files. I shot the Leica and Nikon at f/4, and let the Canon shoot at f/8 since it was having a focus problem.
Clearly my clever 27MP Nikon camera looks the best. In this comparison, the Nikon has the most pixels, and in my experience and as shown again here, Nikon makes the best-looking JPGs in-camera.
The Leica is a little soft, but hey, while the Nikon was shrunk down from 27MP, I had to blow up the Leica from 18MP.
The Canon is sharp, and as typical with Canon, looks artificial when seen this big. See how all the edges are harsh and obvious, while subtler textures aren't there? Canon makes the worst-looking JPGs at high resolution.
I'm showing each image at the same size.
If you have a Ph.D. in mathematics with a specialization in image processing, read the rest of this section if you care, but if you don't, just skip this since all its going to do is confuse laypeople.
Since each camera makes images with different numbers of pixels I had to get all three to the same pixel size so we could make an intelligent comparison at what would be exactly the same print size.
I got lucky: Adobe Camera Raw has an upscale option that resized the Leica and the Canon each to exactly the same horizontal pixel size: 6,144 pixels. I opened each of these cameras at that setting. (The 5D Mark II opened at 25.2MP and M9 opened at 25.1 MP because there are tiny differences in vertical size.)
I opened the much higher effective resolution D300s at its native image size. I increased the canvas size to 6,475 pixels to replace what was missing from FX in DX (see crop factor) and scaled it back down to the same 6,144 pixels wide.
In these cases, the math and pixel-pushing isn't doing much to these images, but if you're a mathematician, you'll note that the Leica is given the worst opportunity here, having to be stretched the most from 18MP to 25MP, the Canon is in the middle, being stretched slightly from 21MP to 25MP, and the Nikon has it easy, since I'm actually squeezing the pixels a bit from 27MP (FX equivalent) back down to 25MP, which would make the result a tiny bit sharper.
Help me help you top
I support my growing family through this website, as crazy as it might seem.
If you find this as helpful as a book you might have had to buy or a workshop you may have had to take, feel free to help me continue helping everyone.
The biggest help is to use these links to Adorama, Amazon, Calumet, Ritz, J&R and when you get your goodies. It costs you nothing and is a huge help to me. eBay is always a gamble, but all the other places have the best prices and service, which is why I've used them since before this website existed. I recommend them all personally.
Thanks for reading!