Tilt Shift Lens Review: Hartblei 35 Super Rotator

Started Mar 15, 2006 | Discussions
Mike K Veteran Member • Posts: 5,525
Tilt Shift Lens Review: Hartblei 35 Super Rotator
1

Hartblei 35mm Super Rotator f2.8

The Hartblei 35mm Super Rotator lens has recently been introduced to the US through Kiev Camera and sells for $500. By comparison the Canon TS-E lenses are about $1100. The main attraction and unusual feature to this tilt/shift lens is that the tilt axis and shift axis are not coupled, and can be freely rotated in relationship to one another. This is a very qualitative review covering the description of the lens, how it works, and examples of the image quality relative to other Canon lenses. I have borrowed a friend’s Canon 24 TS-E for the purposes of comparison to another, more expensive tilt/shift lens. The Hartblei is available in Nikon or Canon mounting rings, is 10 cm tall and 8.5 cm in diameter. The glass elements are actually a bit smaller in diameter than the Canon 24 TS-E, but the aperture is a wide f2.8. The Hartblei is just a little bit heavier than the 24 TS-E, but not very much, the Canon have a pretty solid tilt/shift mechanism made out of sizeable chunk of solid aluminum.

For these tests I mounted it on a 1DmkII, a 1.3X crop sensor with RRS L bracket and Angle Finder C at 2.5X to aid in manual focus with a split screen focusing screen (Canon Ec-B). These focusing aids are a tremendous help in focusing as otherwise it is very difficult to discern optimum focus. Here is a picture of it with the camera on a Gitzo Explorer tripod, mounted to get a low perspective view of the deck: 6 degree tilt at 10-11 inches off the deck surface.

Here is the shot taken by this set up on a very rainy day, at reduced but larger size, f/11
http://www.fototime.com/FE33C0CFF29BF71/orig.jpg

and the same shot with a Canon 24-70L f2.8 lens at 35mm to demonstrate the added dof of the tilt feature on the Hartblei
http://www.fototime.com/073C7402E19C1E8/orig.jpg

with the Canon lens at f/11 the flowers of the middle orchid are sharp, but the pine needles on the near deck are not.

The Movements of the Hartblei Super Rotator design

Here are some close ups of the lens tilted 8 degrees downwards and shifted a maximum 10mm so you can easily see the movements.

The 4 adjustment rings on the lens all are different size and different textures so it is easy to tell them apart by feel when you are looking through the viewfinder. The focus ring is the big one with circular scallops on it near the outside of the lens. In these images the blue numbers are the feet scale, while on the other side of the lens are yellow numbers for the meters scale. The action of focus ring is really nice a smooth and very easy to fine adjust. The focus travel is quite far, but most of it covers a very small range near the closest focusing distance and the useful range from 2 meters to infinity is a very small travel. Overall I very much liked the action of the focus ring. On the second lens photo above you can see the DOF ring printed in yellow and the aperture ring on the inside of it in white. There are slight one stop indents from f2.8 to 22; it is very easy to move this ring by accident while you are messing with the tilt and shift adjustments.

Next come two rings with white lettering that say SHIFT is numbered 1-10 mm, and the ring closest to the camera that says TILT numbered in degrees 1-8 with small indents. These two rings obviously control the amount of Shift and Tilt. On mine the shift is quite stiff the last 2 mm of travel from 8 to 10mm, but the scalloped ring gives something to hold on to and twist. You can put screw on handles on to the Tilt and Shift rings to make it easer to turn them. I have put one on the tilt ring as it has no textured surface to grab hold of. You can see this knurled, thin handle near the bottom of the lens on the first shot with the tripod and at the bottom of the middle image. They stick out so much it makes the lens bulky to store so I only use one.

Here is the confusing part. See the silver tab on the top of the lens (at the 7 degree mark on the tilt ring) and the black tab between the shift and tilt rings? To change the tilt angle one pushes in the silver tab and freely rotates the entire lens. This disengages a spring loaded tab and allows the entire lens assembly to freely rotate in either direction 360 degrees. There are 18 stop tabs, in other words 20 degrees each. Thus one can move the tilt angle 0, 20, 40, 60, 80, 100, 120, 160, 180, etc. but not 90 degrees. Kind of odd, but usually the tilt axis is not very sensitive, I doubt if one could tell 80 from 100 degrees with the modest 8 degrees tilt max. When you rotate the tilt axis, the shift axis (the front of the lens) will rotate with it. The red dot in front of the silver tab in the last image is the index mark for the tilt, 8 degrees in this image. It is not normally seen as it is covered up by the lens at 0 tilt (the lens is tilted downwards here).
Please continue to part 2

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
OP Mike K Veteran Member • Posts: 5,525
Re: Tilt Shift Lens Review: Hartblei 35 Super Rotator

Part 2

Now in the pics you can see the black tab just behind the shift ring, press that and hold it in and you can freely rotate the shift axis either direction 360 degrees (and beyond). There are 16 tabbed stops or every 22.5 degrees. Thus relative to the tilt one can rotate the shift axis 22.5, 45, 67.5, 90, 112.5, 135, 157.5, 180, etc degrees. Now turn the shift knob to dial in the amount of shift you want, 1-10 mm. It only shifts one direction; outwards, or wherever you have aligned outwards to be. If you want to shift the same amount (say 10mm) in both directions, you have to rotate the shift assembly 180 degrees for the next shot. The shift mechanism only shifts one way, so a 180 degree rotation of the shift axis is needed to shift both left and right (or up and down). The shift angle rotation is relative to the tilt axis, while the tilt angle axis is relative to the camera axis in landscape mode. There are red dots to line up the tilt and shift axis relative to the camera. Here is photo of the bottom of the lens showing the tilt rotation stops, and the lens shifted 10mm, and the knurled handle near the tilt locking tab.

Notice how far over the lens is at 10mm shift! What if you really want to set the tilt angle at exactly 90 degrees relative to the camera body in landscape orientation and not 80 or 100? In the image above you can see the EF lens mount ring is attached to the lens with 4 tiny screws. You can undo the 4 screws, and underneath you'll find there's a second set of holes, so you can put the mount back on at 10 degrees rotation to the original position. Thus you can set it up either to get 20 degree tilt detents including exactly horizontal, or get 20 degree detents including exactly vertical, but not both horizontal and vertical at the same time. Why isn't the tilt axis tabbed for 16 stops of 22.5 degrees like the shift?

Because the shift and tilt portions of the lens freely rotate with respect to each other and the camera body, there is no electrical connection for aperture like in the Canon TS-E lenses. In fact there is no electrical connection from the lens to the camera at all. With the Hartblei the camera doesn’t even know it has a lens on it, so everything is totally manual. Wires would tether both the tilt and shift rotations as it does in the Canon TS-E, limiting total lens rotation (both axes together) to 180 degrees. Even the EXIF is manual, as I used the voice annotation feature on the 1DmkII after each shot to document what lens, tilt setting, shift setting and f-stop were used.

Shift Performance, 3 shot stitch

One of the big advantages of a shift lens is that it is possible to take 3 shot panos without parallax error and without pano stitching software. Using rotation about a nodal point (or more correctly exit pupil) allows stitching software to construct very large panoramas. Usually some amount of debarreling, small rotations, and alignments are used by the pano software to correct for perspective distortion and nodal alignment. However using the technique of moving the camera in the equal and opposite direction of the lens shift, has the effect of keeping the lens in the same exact place and shifting the sensor around to capture the larger image circle projected by the shift lens. One simply aligns the images without any corrections (flat stitching). Look at the last image in this tutorial, this camera L bracket movement is what was done in this test:
http://www.outbackphoto.com/workflow/wf_58/essay.html

Here is a shot off of my deck with variable cloud cover. I focused on the bare tree in the center using the Angle finder C with split prism focusing screen. The 1DmkII was on ISO 200 at 1/250 sec in manual mode with the Hartblei at f/11. This image was captured in RAW, but the others in this review were shot in JPEG Large. I took the first centered shot, shifted the lens 10mm to the left and moved the camera 10mm to the right. Took the second shot at the same camera settings, then shifted the lens 10mm to the right by rotating the shift ring180 degrees trying not to jiggle the camera or touch the focusing ring. To compensate, the camera was shifted 10 mm to the left of center and took the 3rd shot. For these camera movements I used a RRS L bracket where I made a centerline mark. I centered on the bubble spirit level of the RRS lever clamp and used that to measure the 10mm camera shifts. This RRS lever clamp is a big help because when half open, the camera plate can slide back and forth for this 10mm positional adjustment to either side. Here is the flat stitch; I was not at all picky about the overlap, so it’s not perfect and there was a bit of wind which may cause some additional blur in the seams.

I used the outer portions of the left and right images along with most of the center image. This is because the left side (center) of the right shift image is very soft, and vice versa. (See below under vignetting behavior.) No parallax problems at all, and previous stitch examples posted with this lens/method have also shown this to be the case. This stitch image will give you and idea of the field of view on a 1.3X body. Obviously FF will be wider and 1.5-1.6X crop will be narrower. Here is a larger (not full size) version for pixel peepers.
http://www.fototime.com/6EB24B073B4F06F/orig.jpg

See part 3

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
OP Mike K Veteran Member • Posts: 5,525
Re: Part 3 Hartblei 35 Super Rotator

Part 3
Image Sharpness: Comparison to Other Lenses

My initial test was done at night, with soft incandescent lighting. The scene:
http://www.fototime.com/F7AD28FE94140F0/standard.jpg

The initial comparison of the Hartblei 35mm Super Rotator, Canon 24L TS-E, and Canon 24-70L f/2.8 @ 35mm. I did a series of shots with similar exposure from f/3.5 to f/16; I show the comparison at f/8 and f/11 because I felt it was most representative of how I would use this lens. I shot the 24L TS-E first then moved the tripod back for the Hartblei 35 and the Canon zoom at 35. I ended up a bit too close for the 35mm lenses and those images are a bit larger as a result. Both Tilt/Shift lenses were manually focused using angle finder C at 2.5x using a split image focusing screen. All the images were post processed very similarly with considerable lightening and some USM. Not great exposures, minimally edited, and I hope a fair, but not exacting comparison. Smaller versions:

and 100% links of the same comparison:
http://www.fototime.com/9AB47E911248910/orig.jpg
http://www.fototime.com/2172A685217E160/orig.jpg

Well, so far the all three lens appear reasonably similar. The next test was the scene off my back deck, shown for the flat stitching example. Here I added a 17-40L f4 at 35mm to the test. Focus was on the bare tree in the center, all of the lenses at f/4 and the two T/S lenses are centered. The 24L TS-E of course has a wider FOV and the trees are smaller. A 100% crop of the center of each of these images:

There are some differences in exposure, contrast and cropping, but I tried to keep the processing consistent so the exposures were not exactly comparable. At f/4 the Hartblei is a bit softer than the other 3 Canons, which look more similar to each other. Here is the same scene with the Hartblei when changing the f stop from 4 to 11, again at 100% crops:

The Hartblei sharpens up a bit from f/4 to f/8 and is constant after that. Overall it seems to be close to the Canon lenses at f/8 and above in the center of the image.
Continue to part 4

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
OP Mike K Veteran Member • Posts: 5,525
Re: Part 4 Hartblei 35 Super Rotator

Part 4
Vignetting and Image Circle Considerations

The same scene was used to evaluate vignetting at f4 where some would be expected. The frame was centered and focused, a shot taken, shifted right 5mm another shot taken, then shifted right to 10mm, each time rotating the ballhead to refocus on the bare tree to the center. The entire right half of each image is shown and the image size reduced to get three of them in a row. Unfortunately the weather wasn’t cooperating and there were some exposure changes due to the moving cloud cover.

The vignetting was noticeable at f4 and about the same in the 5cm and 10cm right shift images. The vignetting was entirely gone at f/8 (not shown, take my word for it because it’s a boring result). One interesting observation in the f/4 composite above, is that for the 10cm right shifted image, the center tree is much less sharp, while the pine tree at the far right is quite a bit sharper! The 10mm right shifted image has the lens center shifted to the far right, while the image center is now using what used to be the edge of the lens when it was centered. Not shown, is that the far left side of the right shifted image is even less sharp than the center tree. All of this is consistent with the proposal that the center of the lens is working much more effectively than the edges of the lens. Here is the bad news, the center of the image at f/8 after shifting 0, 5 and 10mm to the right. Each shot was focused on the bare tree in these 100% crops.

While the 5mm shift shot shows a mild decrease in definition the 10mm shift shot goes to mush. But before getting too excited lets try a smaller aperture, which will increase the effective image circle. On the 1DmkII I have empirically found that the maximum aperture value was f/14-16 without seeing any diffraction effects. So in between the constant rain I got 3 shots at f/16. Processing was a bit different here since these shots were in RAW.

Redemption for the Hartblei 35! The center of the 5 and 10mm shifted images look close to that of the unshifted image; some loss of contrast, but not much loss of detail. Also when inspecting the 10mm right shifted image, (not shown) the sharpness was very similar on the extreme left, center and right edges. Here is a 100% crop of the center of the f/16 images shifted 10mm to the right, with some small amount of local contrast added (USM at 15, 20, 0) compared to the unshifted, centered image.

In the shifted image, we can see some change in colors from the local contrast enhancement, and the very slightest amount of chromatic aberration (when blown up to 300%). But in general, the detail is very close. From the comparisons at f/16 I would presume that the edge softness seen in the 10mm shift images at f8 is the result hitting the edge of the image circle, where image quality dramatically degrades. F/16 simply increases the diameter of the image circle, maintaining the good image quality seen at the center. The ability to utilize the extremes of movement on this lens is going to be in part, dependent upon how far you can stop down without being diffraction limited. The smaller the pixels, the effects of diffraction onset will be a stop or two sooner that those of us with larger pixels. For example the smaller pixels of a Nikon D2X will be diffraction limited at f/11 to the same degree as my 1DmkII at f/16. Check out this reference for an explanation:
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm#

The Canon 24L TS-E was tested with the same vignetting test at f/4 and f/8. Not shown, but the vignetting seen as darkening of the sky at f/4 was actually stronger than with the Hartblei. However, the FOV is significantly wider for the 24mm focal length too, so this is not unexpected. This is also observed as a decrease in exposure as there is less light coming in when shifted. Similar to the Hartblei, the vignetting completely disappears by f/8 and is a non-issue for practical f stops. However, the sharpness at f/8 when shifted was dramatically better than the Hartblei at the same aperture:

The sharpness is almost as good at 10mm left shift as it is centered at f/8.

I did not come up with any clean experiments on testing lens tilt performance, but presume that the image sharpness upon extreme shift may be diagnostic of the image quality when the edge of the image circle is used in extreme tilt as well. I would assume that since the movements of tilt and shift are independent that the amount of potential vignetting will also be additive. Extreme amounts of both tilt and shift will likely challenge the image quality of the Hartblei sooner than that of the Canon T/S, even at the smallest apertures.

In summary, when stopped down to f/8 the Hartblei 35 Super Rotator is comparable to Canon lenses in the center, but when the edge of the image circle is used the sharpness drops dramatically. When the Hartblei is stopped down to f/16 the image quality is close to that of the Canon lenses, even with more extreme shift movements. The Hartblei still has the edge on flexibility as the decoupling of shift and tilt axes is a nice design feature. Mechanically the Hartblei is sound, particularly given the much lower price, but comes with some optical compromises. Stopping way down to the limits of diffraction may be needed to fully utilize the tilt and shift capabilities of this interesting lens.

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
OP Mike K Veteran Member • Posts: 5,525
Re: Posted in wrong forum

Not that folks wouldn't mind finding out about this interesting lens, but I had planned to post this review in the Canon SLR Lens forum where it belongs. I apologize for the mispost.
Mike K

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
Roger Pfister Regular Member • Posts: 150
Many thanks for this Mike

Many thanks for this Mike

It's very definaltly on my most wanted list.

-- hide signature --
Habemus_Canon Senior Member • Posts: 2,437
Excellent review.

Thanks for that review! Excellent work.

 Habemus_Canon's gear list:Habemus_Canon's gear list
Canon PowerShot SD600 Canon PowerShot S90 Canon PowerShot S100 (2000) Canon EOS 350D Canon EOS 7D +14 more
OP Mike K Veteran Member • Posts: 5,525
Re: Excellent review.

thanks for the appreciations, Roger and Habemus.
Mike K

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
Robert Scultz Senior Member • Posts: 1,296
what about the 80mm super rotator

what about the 80mm super rotator ?

was looking at the 65mm, and asked how they sell,

and the guy at kiev camera said, he sells more than 2:1 the 80mm super rotator,.... unfortunatly I have the 20d, and 80mm whould be quiet to long, and the 35mm to wide, flare prone, mainly for portrait, and for full length shots, would need lots of space, but the question would be, at the long length how much the tilt/shift, will be usable ? that's why a lots of fashion shots, ar shot at above knee height, so it's not a problem, but I'd like to explore that avenue, for a litle bit more controll on depth, that reminds my of the other reason, pesky dust, if I can shoot under f12, would'nt be as bad, and
would not need to think of a bigger powerpak for the 7' octodome,...

thanks for the review,
Robert

denisb Contributing Member • Posts: 993
Thank you for the review, but

I will love to see this lens compared with the Arax sell by rugift.
http://rugift.com/photocameras/mc-35mm-tilt-shift-lens-canon-eos.htm

Thank you again for the review !
--
If you find my english wrong... you are right !
http://www.photo.net/photos/denisbergeron

Space Cowboy Contributing Member • Posts: 510
Thank you very much, Mike.

Really thank you! Very interesting, please keep it going like this!

Jon.

-- hide signature --

'You are not Ansel Adams.'

OP Mike K Veteran Member • Posts: 5,525
Re: Thank you for the review, but

denisb wrote:

I will love to see this lens compared with the Arax sell by rugift.

http://www.photo.net/photos/denisbergeron

The Arax uses the same (or similar) lens elements from the same factory. From another poster on a different forum:

"The Hartblei super-rotator mount is the same as what they make for the 45mm and 65mm medium-format lenses.

The optical elements are selected from those made for the Mir-20 35mm/2.8 Arsat lens, made by Zavod Arsenal in the Ukraine. Arax uses the same optical elements. Arsat markets a shift-only lens with the same glass. You wondered on the FM forum if they were the same and I'm answering here since I'm not an FM registrant. Hartblei usually recoats the glass with fancier coatings, and they claim to match the optical elements."

The main difference is that the Super Rotator mount allows free rotation of the tilt and shift axes with respect to each other. The Arax design is fixed at perpendicular oreintation of these axes. In the Canon TSE design the 2 tilt/shift axes can be perpendicular or parallel, but not at any other angle.

I was hoping the discussion of this lens could take place in the Canon SLR lens forum where I meant to place this post!
http://forums.dpreview.com/forums/read.asp?forum=1029&message=17612442
Mike K

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
OP Mike K Veteran Member • Posts: 5,525
Re: what about the 80mm super rotator

Here is a pretty good review
http://www.photozone.de/8Reviews/lenses/hartblei_80_28/index.htm

but no measurements were taken with tilt or shift.
Mike K

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
OP Mike K Veteran Member • Posts: 5,525
Arax examples

denisb wrote:

I will love to see this lens compared with the Arax sell by rugift.

Guy Manusco on the FM forums has some tilt examples of the Arax. He is using it for product shots on a Leica R9/DMR camera.
http://www.fredmiranda.com/forum/topic/367350

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
OP Mike K Veteran Member • Posts: 5,525
Re: what about the 80mm super rotator

Robert Scultz wrote:

what about the 80mm super rotator ?

I found out some more, from another post:

"Of course, all of these, when adapted to small format, will have an abundance of coverage. The 80mm tilt-shift lens is just a standard Arsat 80mm normal lens with Hartblei coatings mounted in a Hartblei barrel."
Mike K

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
ioannis Veteran Member • Posts: 6,618
How about flare ?

The photozone.de review that you posted show some serious flare problem for the 80mm Hartblei lens.

How is your 35mm Hartblei compared to the 24mm TS-E ?
--
Yiannis

When I was a very small boy,
Very small boys talked to me
Now that we’ve grown up together
They’re afraid of what they see
That’s the price that we all pay
Our valued destiny comes to nothing
I can’t tell you where we’re going
I guess there was just no way of knowing

OP Mike K Veteran Member • Posts: 5,525
Re: How about flare ?

It has been raining so much here there is very little sun! I did not expect to see any since I was shooting from the shade of a deck roof in order not to stand in the water. Sorry, I cannot comment upon this issue.

Go to the other thread in the Canon SLR Lens forum, Daniella may have some other experience from mine. Sorry to have parallel threads, my error in posting here and the review was so long I didn't want to repost.
Mike K

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
killersnowman
killersnowman Regular Member • Posts: 450
great comprehensive review

every one needs to read this
--
Tyler Brigham
take a look at my gallery http://www.csc.calpoly.edu/~tbrigham

 killersnowman's gear list:killersnowman's gear list
Canon EOS 5D Mark II Canon EOS 20D Canon EF 17-40mm f/4.0L USM Canon EF 35mm F1.4L USM Canon EF 135mm F2L USM
warp_foo New Member • Posts: 23
Re: Tilt Shift Lens Review: Hartblei 35 Super Rotator

Thank you for the review. I have one question, however. You state the lens is about $500 at kievcamera.com, but there does not seem to be a 35 2.8 SR which is made in EOS mount.

The 65 and 80 are listed f/EOS, with the 80 being $500.

There is a 35SR available for the medium format folks, but that lens is listed at $900 and does not mount to EOS.

Am I missing a link to the lens you have in the review?

Thanks

OP Mike K Veteran Member • Posts: 5,525
Re: Tilt Shift Lens Review: Hartblei 35 Super Rotator

Its not listed, you have to call them.

There is very limited availability of this lens, the owner of Kiev claims only 3 or so a month in the Canon mount. I'm kind of burned out on testing, and am willing to sell mine if you want one.
email me

 Mike K's gear list:Mike K's gear list
Canon EOS D60 Canon EOS 5D Mark II Canon EOS-1D Mark II Canon EOS 6D Canon EF 100mm f/2.8 Macro USM +9 more
Keyboard shortcuts:
FForum MMy threads