Thanks Bob - I understand it much better now!

I just tested the P+S relaylens for the XL-1 and it allows the GG to be about 20mm x 15mm!!! Which isn't all that big. But this is how the mini35 works so that is good enough for me!

Which has the greatest lightloss (please put them in order with the least stealing lightoption first):

2 rightangled prisms (expensive!)

1 pentaprism and a mirror (expensive!)

4 mirrors

or

1 rightangled prism and 2 mirrors ???

Please do send me a picture of your mirror test.

Anders.

I don't know enough about the subject to make a qualified opinion.

Ranking would be a guess.

Least Loss.

Surface coated mirrors.
One prism with coated side faces + 2 surface coated mirrors.
One prism without coatings + 2 surface coated mirrors.
Two prisms with coated side faces.
Two uncoated prisms.

Given that P+S Technik are scanning a smaller gg image, they have probably arrived at a compromise between bulk and performance and may be using a much smaller prism and a mirror pair to compensate for the crosswise image in the third and fourth stages of the relay. They'd get away with a prism about 22mm thick across all faces and a hypotenuse length of 35mm.

I've wondered for a while why some of our Agus/Aldu images are sharper than the Mini35 and I think the smaller gg image may be the reason. The Mini35 would also avoid the hotspot problem by framing inside the dark corners.

I could not find an email to send the attachment so I might send it to Chris Hurd to post on www.dvinfo.net/media/hart

I don't know why my e-mail suddenly doesn't show but it is:

anders@floe-svenningsen.dk

In some ways I'm glad that I don't have to bother with a condenser. Still it would be nice to be able to use medium format GG but its too late now. Anyways I like the results the mini35 gives so...

I think I'm gonna go with a coated 25 mm right angle prism and 2 coated mirrors (just like the P+S). Do you think that I will loose significant light with 2 coated rightangle prisms compared to one and mirrors????

I am really not qualified to advise on this subject. If P+S Technik went that route and given the time and effort of their R & D, I'd say there was a good reason for it. Minimising light loss and maintaining colour fidelity may have been a consideration.

Two mirrors in the 3rd and 4th reflective stage before the relay lens would allow for building in of some adjustments for slight mismounting of the camcorder. there might even be a little skew adjustment available.

With the gg image being smaller, you will have to move that groundglass somehow but keep your eyes open regarding wax groundglasses. If I could have economically made two dead-flat disks I would have gone for the wax composite disk in a heartbeat. As a spinner it was just so much better, except for the flicker.

I'll send the .jpg to the email address. The filename will be agusmrcp.jpg. It is a composition of four frame grabs from a PD150.

I looks easy, cheap and effective :-)

Doesn't the image reduce in size for each mirror??? So perhaps I will see a much bigger part of the GG than i thought.

Perhaps I should just go with the mirrors instead! But will the image travel as far as with a prism (I need the image to travel 12 cm between the GG and my relaylens). I think it ought to be the same distance right?



Do you have a link to somewhere you can buy coated glass (perhaps a pair premounted in a 90 degrees angle?)

Also which kind of coating are you referring to?

Would these aluminium coated mirrors work?:

25mm x 25mm
http://www.thorlabs.com/ProductDetail.cfm?&DID=6&ObjectGroup_ID=890&Product_ID=26986

38mm x 38mm:
http://www.thorlabs.com/ProductDetail.cfm?&DID=6&ObjectGroup_ID=890&Product_ID=26985

Have a look here for a 4 mirror configuration diagram used in the new Olympus digital camera.
For testing, why not using some 90 degree aluminium profiles that can be easily mounted and adjusted.

Anders.

I can't make an opinion on the 4 mirror path in Valeriu's post. My browswer just chases its own tail forever as the reload speed is too slow on this landline.

Given a choice, I would prefer to stay with prisms and live with the light loss as most of the precision work has already been done for you. They will break but they don't deform or crack right across in a primitive mount like a surface coated mirror may if you are not gentle with it.

If you bond two right-angled prisms in 90 degree offset together at the half-hypotenuse faces and this is done precisely, the only alignment issues you have to worry about are that the SLR lens axis and camcorder lens axis are correctly centered and parallel.

With mirrors there is prospect of misalignment and distortion with each one before you even start lining your lenses up.

To get down to my 40mmm prisms I used larger mirrors as a test bed and incrementally masked off the outer edges of the included 90 degree angle between the pairs with black felt tip marker ink until the image became vignetted or crowded at the edges.

You will observe from the .jpg, the axes of the 90 degree bends in each mirror pair (prism corners) co-incide with the inner edges of the mirror pairs so by masking off the outer edges, the inner image edges don't change. You have to bring your camcorder and SLR lens axes in sideways and downwards slightly to recenter in the available frame.

I think your recoverable groundglass image frame will remain small as the relay lens itself is the limiting factor. All the prisms or mirrors do is bounce light which shortens the distance between groundglass plane and camcorder without shortening the distance the light actually travels.

I don't have any supplier info or links.

The mirrors I used for the prototyping were only household grade mirrors. If you look closely you will see the chips where I broke them down to the right size with glass cutter and a pair of pliers.

Optical grade surface coated mirrors.- Subjectively, there seems to be an increased whiteness or perhaps lower contrast with mirrors than with prisms. P+S Technik perhaps chose to blend the benefits and deficits of each reflective method to achieve the best overall outcome.

Bob,
I am sorry for the link. It seems to work from this end.
I got an intriguing piece of glass from one of my clients (a glass designer).
They call it "fused glass" and is imported from Germany. It looks like a VERY THIN piece of glass fused on a 3mm one. There is NO grain looks more like an opal glass.
I will cut it to fit an old medium format camera that I will probably hack and test it for digital capture.
I thought that you might be interested to see it and test it on one of your rigs. Email me outside the list at valeriuAToptushomeDOTcomDOTau to let me know if you want a piece.

I think that the best solution would be to go with a coated mirror and a coated pentaprism. But that solution would be expensive.

I think that each mirror colours the the image. (ie like when you look at 2 mirrors pointing at each other the end of the infinite tunnel becomes quite green - perhaps this is only with cheap uncoated mirrors?)

I think you would loose less light by a pentaprism rather than 2 right angles prisms. Also it would be really easy to allign - even easier than the 2 rightangled prisms. But then again it requires a damn good quality pentaprism.

Valeriu, have you bought the pentaprism you linked to earlier?? If you have then does it distort the colours in the image??? By the way great link it really helps to understand the pentaprism (at least for me!)

Do you have any idea why P+S doesn't use a pentaprism??

I have made two quick drawings of the two rotating image options I'm considering.

Did I understand it right - and which would work the best?

2 rightangled prism version:

http://www.backbonepictures.com/2%20right%20angle%20prism%20version.jpg

Pentaprism version:

http://www.backbonepictures.com/Pentaprism%20version.jpg

Anders.

As far as I can see the right-angle prism version should work. My final arrangement is the rear prism (faces gg) is upright and the front prism (faces camcorder) is horizontal. I did it this way to mount the camcorder high on the appliance rather than low to help with stability when handholding.

If your prisms are same width on the short sides as thickness across, you will need to make sure the corner of the common path of the two prisms is on the inside corner of the 90 degree angle of the prisms' crossover.

I can't verify the pentaprism arrangement as I have no experience with them.

Anders,
The pentaprism will offset the light only on vertical axis. The 2 x mirror / prism arrangement will producean offset both vertical and horizontal. Beware of stability issues arising from this combination.
AFAIK the prisms should have the reflective surfaces metal coated to deliver better light transmission. Have a look on the website I mentioned in my previous post. They say that will incur an extra cost as well.

Bob - how are you going to mount the porro prisms? Are you going to some how merge them or keep them with a little distance?

This confirms your theory:
http://en.wikipedia.org/wiki/Porro_prism

Anders.

As far as I know, porro-prisms are right-angle prisms but they are cut so as not to waste good glass in unused areas like my arrangement does.

I think they are cut so that the thickness across all faces is the same as the half hypotenuse distance but I am not sure of this.

My present mounting is a piece of finegrained particle board. After drawing my design in Turbocad in 1 : 1 scale, I printed it and used the paper to make an exact template. The Lexmark Optra S1250 and other members of that printer family print 1 : 1. So does the Canon BJC-7100. Most of the modern printers should too providing the document page setup is correct for the paper you print to and the printer setup is to the same paper.

The prisms sit in the channels which are exact width and are retained by thin metal strips like binocular prisms.

When playing with the prisms, make sure you do it over a soft carpetted floor as the chances of dropping one are high.

I then cut a channel across the front face of the board with a router, then cut another channel across the back of the board to cross over the first channel at a right angle (90 degrees).

Where they cross over there is only about 3mm of board. I cut a hole which is an "L" shape which provides the three ports for gg >> rear prism / rear prism << front prism / front prism >> camcorder.

Once I have this accurately established, I then cut and trim the board which in my case fits into a PVC tube which has a cap on either end. One cap has a mount for the SLR lens cut into it (or a camera lens mount screwed on, the other has the camcorder mount.

The rotary disk front enclosure is glued onto the tube and the rear enclosure fits to the front with screws and dome nuts.

There is a .jpg of both the non-erecting and erecting AGUS35 Australian Plumbers Versions as I call them on www.dvinfo.net/media/hart

Does anyone know where to get coated colorcorrected mirrors (40mm * 40mm)???