TSE 24mm & D60

[Rainer SLP]

[font color=\'#000000\']Yes, if you want to have the tilting (extended depth of field) in the same direction as the shifting you would have to loose the screws and turn it 90 degrees.[/font]

[Mike Spinak]

[font color=\'#000000\']great Question. I'd like to know, also. I intend to use the 24mm TSE lens with a Canon 1Ds, as soon as I can get them.[/font]

[Alan Davey]

[font color=\'#000000\']Thanks for your replies, but my original question related to possible problems using a tilt/shift lens with a cmos/ccd chip. eg. Does the shifting/titling of the lens produce any digital artifacts or other problems?
Any info appreciated.[/font]

[Rainer SLP]

[font color=\'#000000\']Hi Alan,

I own a TS 24mm f3.5 EOS lens from Canon and if I understand you question right, this lens should behave in the same way with a digital camera as a film camera. Why should you get any artifacts or whatsoever on your digital shot?

If the usage of a 16mm wide angle is no problem, why should 24mm be a problem?

Also shifting the lens up-, down-, left- or rightwards is no problem because the image circle is larger than a normal lens.

The only thing is you will not have 24mm but 24 x your multiplier factor which is I think 1.6? So you get a ~38.4mm lens.

I tried one day to make a panorama with this lens, never finished the project. I used only a third of each image for overlapping. The center third and it looked quite well to where I got.[/font]

[Andy Jones]

[font color=\'#000000\']Ray, Thanks for the info and pointing out that the fall-off only needs to be corrected after the stitching - I've found the stitching program from Canon works OK, especially if you specify the overlap points (editing the seams).[/font]

[Andy Jones]

[font color=\'#000000\']Woops - hit some key accidentally that sent the message before I was done!

Dan, I don't think I'm "dead wrong" on the moving the camera issue.  The easiest way to think of it is how your eyes work - you get three-D information from the world because each eye is in a different position and each eye gets a different image.  So the image you get changes as you move the lens position with respect to the scene, so if you shift just the lens and don't move the camera , each exposure is of a (slightly) different image and parallax issues might make the overlap tricky.  If you keep the lens in the same place, the image itself doesn't change but only the part that is recorded changes.  The relationship of the focus plane to the lens and sensor planes (Scheimpflug stuff) also doesn't change if you only move the sensor while keeping it in the same plane for each exposure - nominally you're just making a bigger sensor by extending it sideways in the same plane between shots.

So if you had your camera on a rail that let you move the camera sideways in a direction parallel to the direction you shift the lens, you could shift the lens left by say 20 mm and shift the camera right by 20 mm, the lens is in the same place with respect to the scene you're photographing and the image coming through the lens is unaltered, both in focus and composition (parallax).  So you take three shots - one of the centred lens and centred camera, the second with the lens 20 mmL, and the camera 20 mmR, and the third with the lens 20 mmR and the camera 20mmL.   If everything's aligned properly, no stitching or distortion will be needed - simply overlap in PS and flatten.

Incidentally, such a rail would be perfect for taking stereoscopic pairs if you moved the camera your interocular distance between shots with a regular (non T/S) lens and this is based on the fact that the images captured are different (one for each eye)  - these are way cool - using two projectors one with V polarization and the other with horizontal polarization onto a metallized screen and view with polarized 3D glasses![/font]

[Larry Smith]

[font color=\'#000000\']

I just don't see how to get the perfect mechanical parallelism required to make it work. I feel much more comfortable leaving the camera physically in the same spot, and shifting the optics.
Dan

Dan/Andy (Dandy?;-),

Thanks for inspiring me with your exchange  !!

I have several Bausch & Lomb stereo-map making mechanisms, which include various rack/pinion vertical/horizontal/fore-aft shifters, ...smooth, precise movements with mm scales, etc.

These are knob-twirlers dreams, with PLENTY of range for what is being discussed here. (Not something anyone would ever want to pay the original price for, that's for sure, but at university-surplus auction prices, they weren't too bad!)

Think I'll give some thought to their application possibilities re. this shift/opposite-shift idea.

In the absence of some fancy mechanicals, though, is the T/S lens (s) sturdy enough to be mounted in a fixed position(by whatever means) with the camera being supported off the back of the lens, and therefore with the CAMERA doing the actual shifting, while the lens is held still?

And if so, isn't this, practically speaking, the same thing that Andy is suggesting?

I won't be shocked to learn I'm missing something here, ...so comments are welcome  

Larry[/font]

[Rainer SLP]

[font color=\'#000000\']Hi Ray,

If I understood correctly you want to put your camera vertical and want to have the shifting possibility in a horizontal way. Right?

You do not need to unscrew anyting. There is a chromed tiny lever near the base where the bajonett is. Push down (toward the camera body) this little lever and you can turn the whole lens in steps of 30°. It even has a stop on every 30° station.[/font]