DSLRUsers.com

First of all I should say that this was done quickly (30 minutes from shoot to download to convert while I was watching Insight on SBS...) and was all done at the lenses' closest focusing distance, and I have noticed that the lenses perform better focusing at infinity since most of the fast primes loses bite at the closest focusing distance. I tried to put these lenses through their worst possible (that I can think of anyway) scenario so this is about corner sharpness, centre sharpness is definitely better on both of these high speed legacy primes.

I only went to f4 because personally I would rarely go to or beyond that aperture and I do think for these lenses f4 is the start of the peak aperture range, so at f5.6 and (maybe) f8, these lenses should be even sharper than what they are at f4. Also the focusing point was on the model's left eye.

I will list several parameters for this test
- All metered through centre weighed metering and manual exposure (not an important variable)
- Both lenses were tested at their closest focusing distance, or at least I tried to get closest as possible
- All were shot handheld on the D300 (might affect sharpness a bit)
- Manual focusing (of course this is the biggest variable and I did my best)
- All were from 200 ISO except for the 50mm at f4 which was at 400 ISO
- All shots were straight out of the camera from RAW which was processed through Capture NX, no other processing have been done on them
- All crops at the largest size on FLICKR are displayed at 100%

The 35mm f1.4 have coma (think this is what Bjorn described as internal flare) issues but it is still useable. The 50mm f1.2 is just an amazing performer, starts to be quite sharp at f2 and just have such magnificent optics. Although for most of the practical situations the AFD 50mm f1.4 is fine even for my uses (it performs very very well), and even though I swapped it for the 35mm f1.4, I have ordered another one of the AFD 50mm f1.4s along with another 50mm f1.2, I think I'm going crazy (I have since ordered an AiS 28mm f2 as well...).

Here are the results for you guys to decipher.

The original scene


In case of hogging bandwidth, I'll only post the links to the images
AiS 35mm f1.4
-At f1.4: http://www.flickr.com/photos/grev/2490087503/
-At f2: http://www.flickr.com/photos/grev/2490088493/
-At f2.8: http://www.flickr.com/photos/grev/2490907130/
-At f4: http://www.flickr.com/photos/grev/2490916568/

AiS 50mm f1.2
-At f1.2: http://www.flickr.com/photos/grev/2490916940/
-At f2: http://www.flickr.com/photos/grev/2490917358/
-At f2.8: http://www.flickr.com/photos/grev/2490917832/
-At f4: http://www.flickr.com/photos/grev/2490918380/

Of course they will perform worse on the edges of a D3.

Lastly, draw your own conclusions.

I think you will be the AIS king of the board, thanks for posting your tests

Hi Grev,
nice to see you playing with the fast stuff

Coma is not the same as flare. Flare is caused by stray light inside the lens, due to insufficient coating and baffling, whereas coma is an aberration that is inherent to the optical formula of the lens. It's basically spherical aberration affecting far off-center light rays (which you only find in lenses this fast). Coma will cause loss of resolution and comet like tails protruding from bright spots towards the edges of the frame, whereas flare reduces contrast and colour reproduction anywhere it appears (not just edges).

Most lenses decline at close focus, that's correct. The 35/1.4 also shows pretty strong field curvature there and needs to be stopped down to obtain even sharpness. It gets better from 1 or 2m on. The 50/1.2 is different. Even though it's not a macro lens it performs best from close to about in-door distances, but gets softer towards infinity.

If you find the time, maybe you can repeat the test? It appears that all 35mm shots are overexposed and show not much detail to compare. Perhaps you can include a medium distance subject too, like the classic brick wall This would also show geometric distortions of which both lenses should have very little.

I wish I had thought of doing a direct comparison while I still owned the 50mm...

Cheers
Steffen.

Well it was quick and on manual exposure in centre weighed metering, it said it was fine with exposure. I think you can still see details on her right chin and face.

The 35mm is too good for general purpose though.

And for AiS king, I don't know! I just wasn't convinced of many of the build quality of today's lenses, I much prefer the older lenses. My 28mm f2 should be here later on today.