Contax 645 Lens Evaluation

Greetings from Anchorage,

 

I recently purchased a Contax 645 with a few lenses. Yesterday I used

a local brick wall (not easy to come by in Alaska) to evaluate the

performance of the 80mm F2 Planar. I was working in a windless area

with a level sturdy tripod and ballhead and was using mirror lock-up,

but was surprised by the poor performance from wide open through

f5.6. With an 8x loop corners were fuzzy until f8; center performance

increased significantly from f2 through f5.6. Barrel distortion was

also quite evident. Focus distance for this trial was ~1 meter. My

questions:

 

Should I expect poor performance from this lens at close focusing

distances? Is barrel distortion more evident at close distances? Can

film flatness contribute to the pattern I've described?

 

I also tested (on the same roll of 220) the 120mm F4 Macro at a close-

focus distance (~1:2) and found it to be (as reported by many)

amazing - even wide open corners were sharp. This suggested to me

that a problem with film flatness is not contributing significantly

to the problems observe with the 80mm lens.

 

Thank you for you input,

 

Eric James

Eric:

 

I have the 35 mm, 55mm, 80mm, 120 mm, and the 210 mm, and I love them all. The 80 mm lense I think is only second to the 120 mm in terms of resolution. The one I have is very sharp. MF has very shallow DOF so make sure you focus correctly when using big apperture. I shoot landscape more so I use mostly f11 or smaller. It should be a sharp lense. Use a cable release as well. I did get some bad pictures when I used MLU but without cable release.

I would not say that 80mm lens isn`t sharp.

 

Nevertheless I use it wide open for portraits and very rarely open it wider that 4 for landscape. For portraiture it works just great

 

All Contax lenses have a center sweet spot and lpm at corners may be quite significantly less than at center area - check internet, I think for 80mm lens it may be about of 60 lpm in center area and 45 lpm in the corners. 140mm and 55mm lenses are better than 80mm, 120mm is the best of them all, 210mm is quite soft.

Nevertheless I think that 40lpm is a great resolution for MF lens.

 

Look here for more data - http://medfmt.8k.com/mf/contax645.html

 

Speaking about your test if you had your wall 1 meter away from the lens at aperture 2 edges of frame were slightly out of focus.

Place camera 3 meters away from the wall at least and repeat the test.

I've used all the Contax 645 lenses. As far as the wide angles were concerned I've never found anything better. And like you and many others I regarded the 120mm macro as an optical gold standard. The problem, if you can call it that, was with the longer lenses, the 140mm, 210mm, and 350mm. They all gave a beautifully harmonious transitions from the in-focus to the out of focus areas, with a very mellow character that was perfect for portraits. Compared to the 120mm however (or to say the 180mm Hasselblad lens from Zeiss) they never really delivered the same biting sharpness.

I think I've seen similar results on colonial stone buildings in the Philadelphia area with my Mamiya 80mm. My theory was that depth of field at wide apertures was not enough to cover the difference in distance from the lens to the center of the wall

(focusing point), and the lens to the points on the wall farther to the sides corresponding to the corners of your frame. If you focus on the center of the wall, the distance to points towards the corners of the wall is greater. Can you try another brand 80mm lens for comparison?

Lenses were OK, test was wrong.

 

At 1 meter, aperture 2 DOF is 2cm. Let`s draw a triangle - 1 meter to the wall, 20degrees to the edge of the frame - long side is about 1.11 meters. Hope I rememeber my school classes quite right.

Naturally it will be blurred - it`s out of focus.

Speaking about cable release I prefer to use 2 second delay mode - it works much better - raises mirror first then takes a shot. Nice.

 

Honestly speaking this is a N1 reason why I love this camera so much - it`s the most convinient tool I ever had for landscapes.

"At 1 meter, aperture 2 DOF is 2cm. Let`s draw a triangle - 1 meter to the wall, 20degrees to the edge of the frame - long side is about 1.11 meters."

 

Modern lenses (except Holgas) are well corrected for this phenomenon and their plane of focus ist a plane not spherical like this explanation asumes. If only points equidistant to the film plane will be in focus the focus "plane" will be spherical. This is not the case. You probaly thought about focussing on the corner vs. the center where you will find the calculated difference.

 

The test may nevertheless be faulty. If the frame covers 1x1m on the wall and DOF is about 2cm tilting the camera as little as about 1mm is sufficient to end up with unsharp corners.

 

Stefan

Oups. Looks like I should read some basic level books again. :)

It`s unbelievable how fast such details evaporate from the memory.

 

Me wrong. Sorry.

I appreciate all of your answers!

 

(By the way, I had used the very convenient 2 second shutter delay for these shots.)

 

I will repeat the test this weekend but increase the focus distance to ~3 meters. I had used a Bogen flash shoe bubble level to level the camera but I may not have leveled the camera as precisely as my close focusing would have required. I hadn't considered John and Paul's theory and perhaps this geometric consideration contributed to some degree to my results. The information gleaned from Paul's URL (In a nutshell: that most of the Contax 645 lenses have a very sweet spot in the center and image quality (in lpmm) significantly toward the corners) may explain most of my observations with the 80mm Planar.

 

One of my main objectives in running a few test rolls was to investigate the issue of film flatness. My results with the 120mm Macro at 1:2 wide open lay my concerns to rest.

 

Can anyone offer insight into the barrel distortion observed at my close focus distance?

 

Regards to all,

Eric

Barrel distortion (or pincushion)will not be affected by focus distance. Distortion will also not be affected by diaphragm position.

 

As far as DOF being planar; my Pentax 67 lenses have spherical DOF and I sure don't see the name Holga on them.

My feeling is that Paul retreated too quickly.

 

My understanding is that the correction that Stefan described varies depending on the intended use of the lens.

 

Since depth of field is non-linear, the out of focus effect at the corners decreases as focal distance increases.

 

Lenses that are not designed for macro work do not need, and do not have, a strong level of correction. Hence, the permitted close focus distance (designed in by the manufacturer) is greater. The permitted close focus distance of macro lenses is much shorted because they are corrected, as Stefan described, for that distance.

 

Some of the Mamiya wide angle lenses have a floating element that allows for the curvature of the focal place to be adjusted according to the distance.

 

I suppose what I am saying is that I agree with Stefan and Paul, but in the latter case before his retraction!

Risking being stoned shortly after this posting... Eric may just be stating facts. What he finds matches the infamous PopPhoto Mamiya Contax/Zeiss lens comparison. The Contax Planar performs only at a mere "acceptable" between F2 and F5.6 - in contrast to the Mamiya...

 

Yes, I know about PopPhoto, but on the other hand, if even PopPhoto finds that significant of a difference, why shouldn't Eric? The test is posted on Mamiya's web site, for those who don't know about it :).

 

BTW, I am not saying that PopPhoto is right or that the Mamiya lenses are better overall. Anyway, check it out!

... adding to my previous post: "acceptable" in the corners for the Contax Planar 2.0/80mm compared to "excellent" in the corners for the Mamiya. Both the Planar and the Mamiya were "excellent" in the center. Which is exactly what Eric was finding. The values in the corner for the Planar were unfortunately quite low :(.

May I offer a hint from inside Zeiss:

If you want to assess lens performance in a meaningful way, you need to put the lens to test under conditions it was designed for. Standard lenses are designed to deliver optimal performance at "infinity". For lens testing purposes "infinity" is sufficiantly well approximated if you shoot from a distance of 100x focal length of the lens. This means for the Planar 80 mm: Shoot your house wall from a distance of 8 meters, and you can expect to obtain valid results.

 

When I shot test targets with a Planar 2,8/80 for Hasselblad from 8 m distance on Agfa APX 25 film, the lens resolved 160 linepairs per millimeter at f/8. I would expect nothing less from the Contax Planar 80 at f/8.

 

Wide open, film flatness errors will likely reduce the resolving power, and so will lens aberrations. However, I would expect to find some 60 lp/mm resolved in the center of the image at f/2.

 

Last week I tested one of our 50 mm lenses, which resolved 250 lp/mm (on Agfaortho 25) in the center at f/5.6, and 160 wide open at f/2, just to give you an example of what can be achieved even at f/2.

 

From what I found out so far, the "comparison" at the Mamiya website is based on two different PopPhoto tests which were performed half a year apart and under quite different conditions. PopPhoto once stated this, however, I never found that disclaimer on the Mamiya website. Wonder why...