Static 35mm Adapter Solution

[Don Berube]

More info about Prisms:
http://www.edmundoptics.com/IOD/Brow...8&FromCatID=10

Image Inversion:
http://www.edmundoptics.com/IOD/Disp...Productid=2057

Mounted Penta Prism:
http://www.edmundoptics.com/IOD/Disp...productid=2086

Ground Glass:
http://www.edmundoptics.com/IOD/Disp...Productid=1935

Sandblasted Glass:
http://www.edmundoptics.com/IOD/Disp...Productid=2356

Opal Glass:
http://www.edmundoptics.com/IOD/Disp...productid=1671

- don

[Nicholi Brossia]

Vendible,
I know I'm a little late in the game here, but I've been researching this stuff since Agus started the first thread. Like yourself, I have been experimenting with taking it a step further by producing an upright image by reverse-engineering an SLR camera and using its parts. Its actually strange how synchronized our ideas seem to be. Anyway, I was just wondering about how finely ground glass you've used. I planned on using 3 micron aluminum-oxide grinding dust to achieve the equivalent of 3000 grit. I've read that you tried 1000 grit and didn't like the results, and I agree 100%. I was just wondering if you'd gone all the way down to 3 micron (it sounds like you have). If you weren't satisfied with that product, then I know I won't be.
Also, I'm very curious about your non-ground glass idea. I've been reading about large format cameras (same technology) and see that many photographers use brightscreens (Bosscreen) with a fresnel. Is this the approach you're taking? or is it more of a coating like flashed opal diffusers?

Helen,
I'd also like to thank you for pointing out the difference between pentaprisms and roof-pentaprisms. I knew that SLR prism looked a little goofy, I just didn't know why.
I know this is age old technology and I'm basically just discovering for myself what people have known for years, but I am excited about making this adaptor and hope I can contribute to the final idea.

[Boyd Ostroff]

Don: Edmund's is a fascinating place. As a kid growing up in the Midwest I used to save my allowance to order cool stuff from the Edmund Scientific catalog. If you're interested in this type of thing and are ever in the Philadelphia area, their factory store is just across the river in New Jersey. Very cool, especially the surplus room with bins full of wierd stuff at low prices. A real must see for any of you tinkerers! You'll find directions on their website.

[Mike Tesh]

Hey everyone I got hooked as well. Trying to make an adapter. So far I've torn apart an old 35mm Minolta and a twin lens yashica medium format camera searching for a solution. I think I have one which may work but only half way.

http://www.visionengine.com/projects...5concepts5.gif

Once I had the roof pentaprism off my minolta and was able to look down at the 45 degree slanted mirror I noticed the image was upright. Although reversed horizontally. Even so an improvement on the original Agus35 design.

The above (badly hand drawn) diagram above shows the camera mounted up top pointing down through a spinning ground CD or ground glass. Below that is the mirror and then the lens at a 90 degree angle to it all.

If anybody sees any error in it please specify. Thanks

[Zac Stein]

I personally feel the BosScreen option would most likely work, as there is no grain structure in parraffin, which would also have no colourisation problems either.


Zac

[Zac Stein]

is it possible to use parts from a medium format camera, as the oversized prisms and stuff would allow the fumm 35mm frame though?

Zac

[Nicholi Brossia]

Hi Mike,
That idea might be easier to shoot visually than the Agus35, not having to mentally rotate the image while taping, but you'd still have to reflect the footage in post (which you already know). It will be difficult at first getting used to the right-is-left, left-is-right property of your current design. But you'll get used to it eventually, just like the weatherman does (that's how they see themselves in the monitor while pointing around at the green screen).
It is definately a possible approach and could be usable depending on your preference. I actually read about and saw pictures of an adaptor similar to yours quite some time ago (it may have even been before Agus posted on the spinning glass experiment... its been a while) but can't remember where. It had the same layout with the camera at a 90 degree angle in relation to the 35mm lens. Everything seemed to work okay if I remember correctly, but it sure was awkward and goofy looking. Of course that could be due to all the duct tape ;).
One thing to think about is your audio. The on-board omni-directional mic might work just fine, but you'd probably have to go handheld for a shotgun mic. Its worth some experimenting.
Also, make sure you take into account the size of the spinning ground glass/cd. You're trying to project a rectangle image (35mm x 25mm) onto a round object, so you'll have to make sure the corners are inside the curve. To get a visual idea of that, take the pentaprism and viewfinder "cradle" off of the SLR to expose the mirror. Now put a cd above the hole until it covers the entire rectangle. On my Minolta, its a very tight squeeze and I don't think the cd would be able to spin. Normally, you could move the mirror further from the lens and compensate by moving the ground glass closer to the mirror, but it looks like the SLR ground glass is sitting right on the top of the mirror. If you tried just moving the mirror back, you could still accomplish a good picure, but you'd lose a lot of your focus range. Try this: take the pentaprism off the SLR body again, leaving the ground glass in place. Point the camera at an object about 5 feet away, then set the focus to infinity. The picture should be fuzzy. Now slowly lift up the ground glass. After about 1cm, the image will come into focus. That means that the lens will only focus on objects as far away as 5 feet at that setting... or provide a really nice close-up lens. By moving the glass further, you can focus on objects much closer than the original setting. So maybe you could move the mirror back and just use a longer lens? That might be worth a try too.
Fortunately, there are numerous designs that can accomplish the same "film-look" end result. Basically its a personal preference. If you're satisfied with the result of your design, then by all means go for it. Yours won't cost much at all because you don't have to deal with the prisms, which is a definate advantage. I, on the other hand, plan to go all the way to achieve an upright image mainly for the challenge... and its going to cost me :).

[Louis Feng]

Nicholi,
It's great to know you are working on a similar solution. Yes, I have made GG from very coarse GG to the 3 micron aluminum oxide dust.

The coarse ones are good for diffusing the light to all the corners, but they leave those obvious noises on top of the video. The very fine ones like 3 and 5 microns, they have much less noticeable noise, but there appears to be a "hot spot" in the middle. This hot spot may become less obvious when you have them in a black box. But something else must be done to remove that. Maybe a fresnel lens would help.

Bosscreen could be another solution and probably won't make any visible noises. But I don't know where I can get hold of a small piece. Plus it probably will have the same "hot spot" problem.

Roof pentaprisms are hard to find by itself. Large roof prisms are expensive too. Pentaprisms are a lot easier to get. I have asked some optics companies, they can make custmized ones if ordered with large numbers (50+).

[Nicholi Brossia]

As you know, fresnel lenses are common in large format camera photography. That would cause me to view that as a possible solution. However, we are dealing with a much different scale in the sense that we'll be blowing up a tiny object to the size of a television. Just like you said in an earlier post, this will greatly exaggerate any imperfections in the glass, prisms, or any other part. My question is... would the grooves be visible after being blown up? It seems as though many people have different views on how the fresnel should be located relative to the ground glass. I have a feeling that this may come into play, even though I have seen a couple Agus35s that incorporated the fresnel and looked good... at least at the 320 pixel resolution on the computer.

Zac,
I also like the sound of Bosscreen's lack of grain, but has anyone blown it up on a television? I'm curious to see if imperfections are still non-existant. I'm also concerned with temperatures. During the summer, car interiors will reach 130+ degrees F. I don't plan on leaving any equipment in the car at that temperature, but if cars get that hot... will the inside of my adaptor? And if so, I'm sure the wax will melt/run. Also, just like Vendible said, Bosscreen is expensive and you can only get huge pieces... and you'd have to use a fresnel to diffuse that hot spot. Hopefully someone will run accross something worthy of a test.

As far as custom prisms, maybe we can get some kind of group buy going. I'm sure there are 50 people that will jump in on purchasing custom made prisms once the plans are tested and readily available. Also, I've found a company that could end up helping a lot. www.surplusshed.com has some great lenses and prisms that could be perfect for this project, and would most certainly provide a less expensive prototype. I actually just purchased items from them, and they've treated me very well. Currently, they have a 35mm x 25mm pentaprism, coated and aluminized (which just happens to be the perfect size for this project) www.surplusshed.com/pages/item/l2159.html
and a scope assembly that contains a 45mm long amici "roof" prism that could work as well http://www.surplusshed.com/pages/item/l3082.html
although it looks like that prism is not aluminized.

I'm curious to hear the results on your static diffusion screen tests, Vendible. Any variable that can be eliminated from this ultra-precise adaptor would help very much, especially that crazy spinning glass.

[Mike Tesh]

Nicholi yeah I've tried that. Moving the elements around. As you say in the minolta camera the ground glass (plastic) sits right above the reflex mirror. It's all such a pain. :)

Perhaps we are all going about this wrong. In the old world dealing with prisms mirrors were needed becasue film was a fixed format. However in the digital world things are different and we are dealing with digital cameras in teh end here. Maybe we can all petition some electronics manufacturer to build a battery powered box that has a firewire input and output along with an analog output. Where we can throw a switch and it will digitally flip and reverse the image to be recorded by a second DV deck.

That's a long shot considering the cost of R&D on such a device and it's limited market potential. But they could always market it to the masses as a firewire to analog convertor with that extra added bonus to it for us.

Even so it's along shot. But it would result in the best image as we wouldn't have bounce the light around losing stops and possibly degrading the image more.

Another idea (for people who already use 1 chip cameras) is to get a small Apple iSight camera and mount it upside down in the casing. Firewire it out to the camera to be used as the record deck. You would still later have to flip the image horizontally. But once again you'd get an image that is right side up and wouldn't have to bounce the light around.

Just some ideas.

[Nicholi Brossia]

Mike, I see you're imagination has been running :).

I agree that it would be best to avoid the prisms entirely. However, it seems that, in the adaptor's infancy, prisms provide the most effective means of accomplishing an upright image. You make a very good point saying that we're in the digital age and anything can be easily manipulated.
I think your firewire method may be a bit overkill, but you're on the right track. If this adaptor catches on, eventually the camcorder manufacturers will produce a camera that electronically rotates the image before recording onto the tape. That way, the 35mm lens could project directly onto a static "ground glass" and the camera would cover rotating the image 180 degrees.
I'm certainly not an expert on the subject, but it seems like that could basically be an in-camera menu setting, much like frame mode or any of the effects. I'm sure it wouldn't be too tasking, because they've already accomplished that exact idea with the on-board lcd screen (rotate 180 degrees and the image flips, then you can even mirror the image with a menu setting on some cameras).

As far as the iSight goes, there are a couple big hang ups with that idea. According to the stats, iSights can handle up to 640x480 resolution. However, it looks substantially lower in resolution. Many users claim that it just records a 320x240 resolution image and blows it up to 640x480. Even at that, NTSC is the equivalent to 720x480 resolution on a VGA monitor, so it would still be a degredation even at its highest setting (not as wide of a picture). Also, the iSight recieves its power from the firewire connection. I can't say for sure, but I'm guessing digital camcorders only send/recieve video and audio signals over firewire, not power.
Just as a side note, if you plan to make small web videos, the iSight is a neat little camera. Great low light abilities and close focus range makes it a great camera for smaller scale productions. Here are a couple sites to check out regarding "aftermarket" iSight software and accessories http://www.macdevcenter.com/pub/a/ma...re_isight.html and www.kaidan.com/pdf/iSight_Kit.pdf.

I doubt 35mm lens adaptors will ever become prominant in the consumer community (too much to fiddle with), but filmmaking is a different story. Hopefully, the camera manufacturers will notice filmmakers interest in such a device and even see how simple it is to produce, then supply proper equipment to fill that demand. If they would just hurry up and come out with 2 inch, 16x9 ratio, ultra-high definition ccd's, then we wouldn't have to deal with all this ground glass nonsense ;).

[Louis Feng]

Nicholi, thanks for posting the links. I already ordered both prism from them. Although I'm not sure about their quality (in a very picky sense), they are great for experiment.

I'm pretty confident that with the two prisms, a correct image will be produced. All my calculations and simulations confirm that. Although I would prefer a roof pentaprism/mirror combo over a roof prism/pantaprism combo (in terms of effeciency and cost), but for experiment, the minor loss of light in these reflection/refraction can be ignored.

In SLR camera, the Fresnel lens are used in between the ground glass and the condenser lens. My understanding is it will reduce the diffusion of light from the ground glass before it's passed onto the prisms. I still have to work out how the Fresnel lens' magnifying effect applies to the image, especially when there isn't a condenser lens in my design.

The way other people have used it is quite interesting, that is between the 35mm lens and the ground glass. I plan to get a cheap one to experiment first and try simulation to confirm the theory.

All this is so much fun to work out.

[Nicholi Brossia]

Yes, its a lot of fun... that will hopefully produce something :).

I've taken apart an SLR camera and plan to make my first prototype using salvaged parts. Of course, I will need to buy a close-up lens/diopter in order to view through the prism though.

I have a question about the 3 micron ground glass that you experimented with. You noted that there was much less noticeable noise with the downside being the hotspot. It seems as though, when enclosed in a box, the hotspot would become even more pronounced due to the lack of light "pollution" surrounding the area. My question is, was it the grain or the hotspot you didn't like about the fine grind? If it was the grain, then another solution should be found. If it was the hotspot, then hopefully the fresnel will turn out helpful. I'm sure that's exactly what you're in the process of experimenting with... and I plan to as well.

You may have already read this article, but I'll point it out anyway. http://www.wisner.com/viewing.htm has a good couple paragraphs describing the upside, downside, and proper placement of a fresnel lens. According to this site, the fresnel will create a full frame without hotspot (which we already knew), but will also reduce detail. This is what has always concerned me about using a fresnel. I guess I just need to experiment and see it for myself :).

Anyway... good luck with the experiments and have fun.

[Louis Feng]

The article you posted were very helpful. I bought one of these pocket fresnel magnifiers from officemax. It does reduce the hot spot, but didn't completely remove it. Maybe a bigger fresnel lens with higher magnifying power is needed. The circles were barely noticeable. I think the one I have probably has 100 lines/inch, I saw Edmund sell ones with 200 lines/inch.

Regarding the 3 micron ground glass, all I can say it's at the border line. If it's very close to the video camera, then yes, the grain is noticeable, but barely. When it's about 2 inch away from the camera, then the grain seems to disappear. I can't tell you for sure because I haven't tried it in a fixure, my hand shakes. If this is true, then it may work out because the light is going to tavel in the prisms for at least a few inches. I think it's farely promissing and certainly worth a try. I'm still testing other materials to give better light transmission. It seems the hot spot problem can be fixed with a fresnel lens.

I will be testing a new material in a few days, it's coming in the mail. It's not cheap. If it doesn't make any grain, then I think we'll have a winner.

[Filip Kovcin]

i'm not good in math, but maybe someone can find the answer...

can anyone tell the proper "size" of GG grain?

i mean - if we shoot with 1/3" or 1/4" CCD cameras, that means - we have specific measures of CCD - right?

1/3" is 4.8mm x 3.6mm

1/4" is 3.62 x 2.7mm

so that means that we can calculate the "size" of the grain and know which size will be visible and which NOT.

if the resolution is (in PAL countries) 720x576, we can calculate the SIZE of one pixel.
if the grain is SMALLER that pixel - it's invisible....

... or i am wrong?

any thoughts?

filip