Photo Stuff

Here are  four Modern Photography tests of Rolleiflex cameras and lenses. The tests originally appeared in Modern Photography from 1957 through 1963. These are scans of photocopies from a library xerox machine. They are 150 dpi, and should print fine.

• Are the New Rolleis Really Better? - a review of the then-new 3.5e, penned in ‘57 or ‘58 by the late great Herbert Keppler. This review does not contain lens test charts. It does contain the clearest and most concise explanation of the old lvs uncoupled meter I have ever seen, done with a series of 8 illustrations on 1/2 of one page. Worth printing out if you have an E-series Rollei.
• Great Cameras? Fact or Fiction Rolleis have always had a great reputation. Do they still deserve it? Another teaser question by H.K., this time in a review of the entire 1957-8 Rollei line - the 2.8e, 3.5e, baby and ‘cord Va.
• 4 Different Rolleis, 4 Different Lenses Lens tests of the 2.8 and 3.5 Planar and Xenotar. This review includes a bonus: ‘Mutar means more lenses for Rollei’,
• Rolleicord Vb can Doff its New Hood Includes a test of the four element Xenar.

At the time Modern did not print lpmm in their tests, instead using ‘excellent, very good, good, acceptable’. They did provide lookup tables giving lpmm corresponding to their ratings for the different classes of lenses. I penciled in some number from a medium format normal lens table dating to the same year as the test.

Back in the 60’s and 70’s, very few photojournalists used in camera meters, even though excellent ones were available from Nikon, Canon and Pentax. Both Herbert Keppler (the noted photography writer, recently deceased) and Galen Rowell thought this was due to sheer cussedness.

That can’t be all … there must have been a reason PJs didn’t go for ttl and kept their old Gossen Luna Pro and Weston Master hand held meters. I think that the way pros used hand-held meters was just plain faster and less fussy than using a match-needle ttl meter.

Here is the key:

“It is not necessary to take a new reading with the meter every time you are ready to make an exposure.”

- ‘Fritz Henle’s Guide to Rollei Photography’.

Instead, the pro took one basic reading for each scene, ideally from the primary subject. That was it. The pro then adjusted his exposure depending on the light falling on the subject, as observed through the viewfinder, using rules of thumb:

Open up 1 stop for sidelight, 1-2 stops for backlight, 2-3 stops for open shade.

If clouds were passing over the sun at intervals, the pro would take one sunny reading and one reading when the clouds came over the sun - and use those for the entire shoot.

Taking one careful meter reading was part of setting up for the shoot - not part of the shoot itself. That is why an in camera meter seemed a useless and expensive toy to the old time PJs.

Imho, the old pro technique is a good way to use a meterless camera and hand held meter, provided you stick with print film. Slide film has a narrow margin of error, and rewards a more elaborate metering technique … but that is a subject for another post.

Just using the cameras, the only differences are

• The D300s 1/2″ larger screen - which is nice, but imho greatly over-rated.
• The D300 has a great wealth of useless features, such as Live View. At least they don’t get in the way.
• The D300 feels even more ‘put together’ and solid than the D200.

In terms of performance, the D300 has a slight to significant edge in most categories:

• Autofocus is slightly faster and easier to use.
• The D300 has a noticeable edge in image quality
• The D300 has 2/3 stop more usable iso (iso 1000 vs 640).

Buffer depth and card write speeds are about the same for both cameras. I measured the d300 buffer depth at 19 frames, raw, and card write speed after that at about 1.3 frames/sec. This with a udma card.

The Kodak gives about 7 1/2 stops of usable dynamic range. Thats pretty good. Amateur photographer, using the same technique I did, gave the D80 7 stops of dynamic range. While the latest generation of Nikons (d3/d700/d300) give about a half stop more than the Kodak, their effective dynamic range is pretty close - noise in the deep shadows robs the Kodak of that extra bit of range the d3/d700 have.

Here is the curve I got testing my own slr/n:

There is a nice ‘toe’ to the curve at the high end, with the Slr/n going to 255 at +3 1/2 stops.

Whats happening at the low end? Noise. I set white balence off the gray card used for these tests and that balance held to +-1 until -5 stops - at which point the rgb values went all over the place (17-10-12 at -5 stops) - and stayed all over the place. That range is reflected in the chart.

I’d recommend setting your black point in photoshop levels to 17 or 18 when using a Slr/n - that should eliminate a good portion of the shadow noise, and you won’t be losing any information.

Having been using the Slr/n a few months, I’ve found its really at its best at its base iso of 160. However, even at iso 320, great looking images can be produced, especially with the Noise Ninja photoshop plug-in available to rescue problem files. At iso 400 and above contrast drops dramatically and noise increases more and more rapidly.

My neighbor’s roof at ISO 320 (left) and ISO 400 (right) showing loss of contrast at 400.

The quality of the Kodak noise is a problem: its present almost entirely in shadows and is bigger and blotchier than what other digital cameras produce. This makes the noise look even worse than it is.

Noise Ninja is perfectly suited for the Kodak’s noise signiture - if ‘filter course noise’ is checked. According to Noise Ninja, this setting filters very low frequency noise. At a high setting (10), Noise Ninja knocks the Kodak noise right down, at some cost to texture sometimes. At a lower setting (4-6) texture is unaffected, and the noise is broken up into finer dots more like film grain.

Noise ninja also has a quick mask built in, although I find it easier to run it on a duplicate layer and then erase the parts where detail was lost, revealing the original image beneath.

My noise ninja plug in generates a noise profile for an image with an overall noise ‘number’. Lower is better. I found a good test image with a sunlit foreground and trees behind in deep shadow (see my neighbor’s roof, above). Using both a Kodak slr/n and a Nikon d200, I took a series of shots at increasing isos. Here are the results:

I also evaluated some scanned Velvia 100 slides - they showed values of 24-30 in Noise Ninja.

While the D200 goes to iso 1600, 800 and up isn’t really usable, while 400 and up isn’t really usable on the Kodak.

Its interesting to note that the D300’s numbers aren’t really better than the Kodak. The D300 noise has  a fine ‘film grain’ like quality (left), while the Kodak noise is heavy and blotchy (right). Both at iso 800.

Also notable is that there is no difference between the D200 and D300 below iso 800.

This superbly constructed lens has a serious weakness: its tripod collar. When the D200 is hung off the end, the tripod mounted 200 ‘rings like a bell’. My tests showed camera shake as high as 1/180^th of a second. How much? I got 62 lpmm at middle apertures using the camera’s tripod mount, but that dropped to 49mm on the lense’s tripod collar.

I tried bolting on a much heavier collar I had left over from a 300 f4.5 ai. It showed the same issues. Holding the lens down at the tripod coller to dampen vibrations helped some of the time, but not others, depending on exactly where my hand rested. The lens must set up sympathetic  vibrations from the D200 mirror and vertical run shutter

Simply taking  the tripod collar off cured the problem.

The 200 alone without its collar is quite lightweight for a macro: 22.6 oz. That’s only 3 oz heavier than the 105 2.8 AF – light enough so that its not really a noticeable weight when hung off the front of a 2 lb tripod mounted D200.

It’s a long lens, though – over 7”. I find it considerably more awkward to hand hold this lens than my 15.6 oz , 4” long Sigma 180 f5.6 or 10.2 oz 55 f2.8.

What the 200 f4 does have is wonderful quality and feel. . The 200s silky smooth and unbelievably light internal focus ring wraps around 330 degrees from infinity to ½ life size. That’s much better than the modern 105 f2.8 af, which only goes 120 degrees to get to ½ life size. It makes a huge difference in manual focusing ease.

Working distance at 1:2 is 19 ¼” – excellent for tripod work. It helps to have a micro focusing rail … or a slider – an extra long arca-swiss mount with its own arca-swiss clamp. The slider works well, but is not anywhere near as smooth as a real focusing rail. Nor is it anywhere near as heavy.

————-resolution tests follow————-

Here are some numbers (all from USAF 1951 chart, using a d200, which is sensor resolution limited at 62 lpmm)

(200 f4 micro, tripod to camera body mount, steadied)

f4 - c55 e49
f5.6 - c55 e55
f8 - c62 e55
f11 - c62 e55
f16 - c55 e55
f22 - c55 e55
f32 - c49 e43

200 micro ai (lens tripod mount, not steadied)

4 55 49
5.6 55 49
8 49 49
11 49 43
16 49 49
22 49 49
32 43 43

(lens tripod mount, steadied)

4 55 49
5.6 55 49
8 49 49
11 55 49
16 55 49
22 55 49
32 49 43

Its that tripod mount that has given this lens a bad reputation for sharpness.

This little Sigma is only about 4” long and weighs only 15.6 oz. That’s a bit less than Nikon’s 60 micro. For me, this under 1 lb lens is perfectly balanced on the D200 for hand-held work out to 1:2 magnification. Anything heavier feels awkward and I have to use a higher shutter speed to avoid camera shake. With the Sigma, 1/250 of a second is good, and I can go a bit less in a pinch.

In fact, I find the Sigma 180 as easy to use hand held as my little 10 oz 55 f 2.8. In the macro range, camera shake increases with magnification, not focal length. This light 180 feels the same as a 55 or 60 Nikkor.

The Sigma has a huge advantage in working distance: 15 3/4 inches at 1:2, compared with 4 3/8 inches for the 55.  That’s better working distance than just about all 105 macros. The Sigma is an internal focusing design, meaning it does not get longer as you focus closer. The f-stop remains at 5.6 when you go to 1:2. This makes that slow maximum aperture less of a limitation than it seems at first glance. I’m almost always stopped down to f11 or f16 shooting close-ups. And f5.6 gives me enough light to focus and compose.

The other limitation of this lens is its 1:2 maximum magnification. That’s less of a limitation with a DX sensor. 1:2 on DX covers 32 x 48 mm – that’s the same coverage as ¾ lifesize at full frame, quite a bit more effective magnification than the 48 x 72 mm full frame gives you at 1:2. Any more magnification isn’t really workable unless you are on a tripod.

And this little lens is very sharp. Sharpness extends to the close-up range.

Somehow this little 180 ends up in my ‘go’ bag a lot. Its still a Sigma, though, with that cheap Sigma feel. I’ve never had any trouble with mine, but I’m sure its possible to break it.  Don’t try to focus manually with the camera on af, and don’t drop it.

There is a lot to like about this little lens. If only Nikon had built it…

The 70-180 micro is two things: a first rate macro lens, and the only pro-quality Nikon autofocus telephoto zoom that isn’t  really big and heavy.

70-180 micro at 135

f5.3 c-55 e-55

f8 c-55 e-55

f11 c-62 e-55

f16 c-62 e-55

f22 c-55 e-55

f32 c-43 e-43

70-180 micro at 180

f5.6 c-55 e-55

f8 c-62 e-55

f11 c-62 e-55

f16 c-62 e-55

f22 c-55 e-49

f32 c-39 e-39

As you can see, there is absolutely nothing wrong with this lens. It gives better results than Nikon’s older f4 pro-quality telephoto zooms in the normal range. Pop photo tested this lens with their sqf equipment – it rated just a hair behind the 80-200 f2.8.

The 70-180 is a mixed bag used handheld, however. On the one hand, manual focusing is a real joy. The smooth, light focusing ring is the best I’ve ever encountered on an autofocus Nikkor. In fact, its right up there with the best manual focus lenses from the classic era. The zoom ring is also light and positive. The two rings work very smoothly together. Someone obviously put a lot of thought into the usablity of this lens.

On the other hand, the tripod mount is always a little in the way, even if you swing it up over the lens. I cut off the end of my arca-swiss mounting plate so that the tripod mount and plate fits in the palm of my hand, with my fingers grasping the manual focusing ring. That’s fine, until I switch to a vertical shot. Then the tripod mount is in the way again.

Put this lens on a tripod for macro work, though, and its in its element. You set up, and zoom in or out to get the precise magnification and framing you need. This zoom becomes just as versatile a 105 macro with a focusing rail. Of course, it also weighs as much as a 105 macro with a focusing rail…and costs as much too.

The working distances of the 70-180 are in about the same range as a 105 macro:

70-180 at 180 1 to 1 1/3 – 4 ¾”  (30 mm subject width on D2x)

                              6 1/8” (35 mm subject width)

                              8 ¼” (42 mm subject width)

                              10 5/8” (27) (48 mm subject width)

48 mm is 1:2 – where working distance is a respectable 10 5/8”. Go for more magnification, and your working distance rapidly deteriorates – this is due to the complex internal design of  the 70-180, which shortens focal length as it focuses closer.

In use, this all translates into the equivalent of a rather heavy 105 macro.

But a 105 macro doesn’t transmogrify into a high-quality telephoto zoom when you take it off the tripod. So you have to carry a 70-200 f2.8, too, as well as the 105 and focusing rail - suddenly you have 5 ½ pounds of kit to lug around instead of 2 1/4 lbs of 70-180 zoom.

This beautifully constructed, reasonably sized lens was just replaced by a much larger and more expensive vr version. There isn’t a reason in the world I can see for putting vr on a 105  - you don’t need it for distance shots and it won’t work for close ups. Save your money - buy the older version used. Its built like a tank, lots were manufactured, so its easy to find a good one.

The 105 focuses to 1:1 without attachments. This means a lot to most people, I guess. Personally, I found 1:2 adequate in 35 mm full frame and more than adequate in DX, where it gives the same subject area (32 x 48 mm) as ¾ life size in full frame. That’s why I prefer the 180 f5.6 sigma – its lighter with almost twice the working distance.

The 105 has a reasonable working distance of 5 ½” at 1:1, going to 8 5/8” at 1:2. By comparison, the 180 sigma has 15 ½” working distance at 1:2, but that’s as close as it goes.

If you really need 1:1, the older 105 micro af is the most versatile choice. The 60 has under 3” of working distance at 1:1, making it virtually unusable at that magnification. And the 200 is a great heavy beast that weighs over twice as much as the 105.