35mm Adapter Static Aldu35

[Filip Kovcin]

Wayne,

exellent work!

it looks that this solution may work!
wonderful idea!

filip

p.s.

i will go now thru all your mentioned links very preciselly. i think it deserves our attention.

[Wayne Morellini]

I meant holographic screens, it was noted the low angle holo screens got a darker image than the high angle ones (when it should have been the opposite). I summise that because the light is so undifused that much of it can go straight ahead and miss the sweet spot on the video cameras lens, which is often smaller than the 35mm frame projecting the light, so the outer portions of the frame will only give the little off axis light that hits the lens. So a conensor is needed to bring the off axis light into tthe lense.Another thing is the iris in the camera could possibly interfere with the low angle light, where as the iris in the still lense (before the light is straightened out) has the normal effect. There is also the possibility that there could be some dependence in the iris/shutter/gain mechanism on normal diffused light sources.

Pricing, quality, transmission:
Next or pricing, surplushed has some funky round 30mm something Fibreoptic plates for around $5-7US. But from the picture they seem dark and not of the quality of some of the specs I've seen. So it is a matter of hunting around to find a place3 thaqt sells good ones cheap (preferably surplus). I have an old Edmund catalogue here, and the ones listed don't seem anywhere near 40Million fibres, and I think max out at 100lines per mm. But apart from resolution there are a couple of other important factors. L:ight transmission is one, with values below 60% still acceptable compared to a GG screen, but better ones have much higher transmission. But also the type of fiber used will also effect this transmission, and also the spectral frequencies transmitted, like Matt said with diiferent colorisation effects so it has to be studied. More below.

There are two values from memory (name will be in those links). One is the transmission of light going through the fibre, another is the amount of light that comes into the fibre, remember there is still clading (some 1:16) around the fibers, and surface reflection etc, that reduces it. I would think that the clading plays a big part in the problem, with poorer face plates using thicker cladding etc.

Angle of light Acceptance:
Another value is what angle of light is accepted, to great of an angle the light starts bouncing off, or hitting the side walls of the fibre and being eliminated (because it can't be internally reflected). From the nice pictures on the princeton link and the income ones. You can see that parts of he face go black, this would be because that paret of it is at too great an angle that light is refected off, or he camera was outside of he transmission angle of the taper.

To get over the angle of light problem certain companies use etchings and patternes on the surface of the plate to increase light acceptance. This is somethign I was wondering why they haven't done until I found recent research into the area.

Image Inversion:
Another problem is that if you use an image inversion, apart from being a luxury feature that might be big money, the fibers will have to cross over somewhere, so unless there is a big buldge in it where the fibres cross over, or some apprartus on the front of the faceplate, I would expect poorer light transmission.


Now, Fibre optic plates have been used for decades, and in cameras for years. In matter of fact I looked at an old videogame system I had the other night, at it's monitor, to see the depth of glass before the phospors, from what I could tell it must have been really thin glass, which I hought was impractically unussual, now it occurs to me it must have been a CRT Fibre Optic Face Plate ;)

[Filip Kovcin]

<<<-- Originally posted by Wayne Morellini : ...In matter of fact I looked at an old videogame system I had the other night, at it's monitor, to see the depth of glass before the phospors, from what I could tell it must have been really thin glass, which I hought was impractically unussual, now it occurs to me it must have been a CRT Fibre Optic Face Plate ;) -->>>

wayne,

can you explain me quoted part of your previous text in a different way. i do not understand it properly.
if it's not too dificult for you.

thank you,

filip

[Wayne Morellini]

BTW don't give up on finding cheap solutions, optically Micro Cystaline wax could do simular things, but maybe not anwhere near as good in resolution and light transmission. I should have posted it over at the HD static thread too.

Just thought of a few more tips, sorry I hope this is the last lot:

Thickness required:
Another thing is that we only need a thin face plate that might further reduce prices, as I think that cutting might be less cost than the material itself (unless it is too thin). Also the thinner you go (and thinner fibers) the plate apparently becomes flexible. But there is a downside to becoming thin, the thickness will cause lightloss from light going out the side of the fibers, that will be greater or smaller depending on type of optic used. The thickness of the plate and the fibers should also much greatly effect the angle of exit of the light becuae of internal reflection, to a certain degree. Go to thin the light will come out at high angles (undesirable), go thinner again and much light should go through it like plain glass, because there will be too little internal reflection to mix the light.

Resolution versus Incidence of error:

It just occured to me again to mention this. If you have one fibre to one pixel you should have sparkly grain and over image stuff, A bit more and you will have softness. The reason why this is, is that the missalignment of pixels to to the entrace of the fibers, cause part of what should go into one pixel to be lost on the clading, or spread inbetween multiple pixels. To reduce this problem I have devised the following scheeme I use for most of my projects, that may or may not be accurate you decide:

Reuce image pixel error by increasing screen resolution:
- You need to reduce the amount of error in the image pixel to less than that of the lowest value possible per pixel. Probably by half. This is 8-bit+1 for most of us, though 8 bit cameras may use upto 12-bits (+1), which is more desirable as the camera does processing on the image in this region, but maybe not really needed (i.e. I don't know). But if you have a high niose level (say 6 or 8 bits worth, around 36db-48db) than that might be acceptable too you. After dertermining how much accuracy you have you have to increase the resolution past the desired image resolution enough to achieve this accuracy. With perfect accuracy, you should get a sharp image as smooth as silk (if your camera can do such a thing).

Problem is I can't remember the formular I use, as I am a bit under the weather at the moment. But I remember, it is actually something like this, you have to calculate the (maybe radom) arrangements of focus screen pixels so that the percentage of area that the outer most pixels (fibres in this case) consume is the same, or less, than the desired accuracy you want. You see, as things come into these outer fibers then to inner fibers, and vice versa, they will cause sudden jumps and falls in brightness as they cross these fibers. Even though the camera pixel will average these out, it will still produce niose (inaccuracy) in the pixel, that sort of flashy grain seen in the spinning and staic GG convertors. The bigger the fibre the worse it will be, the smaller the less nocitable it will be until you reach your signal to niose ratio/max accuracy, then it will contribute a little to the niose and virtually dissapear. For our eyes that is ussually the 8 bit image, but for pro camera, and post, colourisation, effects and editing, this probably should be as much as you can get, or at least 10 bits+1.
There are other ways to do this also, without having to increase the screen res so much, but it gets more complicated.

Has anybody considered doing a static on the new 3chip PV-GS65, I would be interested to see how it goes?

Have fun, tell me how it goes, I spent many hours going through much searching, so keep looking, it takes much time.

Wayne.

[Bob Hart]

Wayne.

I made enquiry some time back with the Electrophysics Corp who make the Astroscope Night-Vision systems.

The NV tubes which erect the image use what was described as a coherent bundle of optical fibres to rotate the image 180 degrees however there was doubt whether these coherent bundles could be had in diameters large enough to erect a 35mm still camera or motion picture camera frame.

[Wayne Morellini]

I too, am not a fan of going the full hog and turning the image via great expense, I also am dubiouse that it may cost quality/light (see above) , however why not just put a prism in the front of the back of the faceplate/taper, as you may have gotten rid of the need for a condensor/field lense you have more room. For me if I use a digital cinema camera type camera head (no native lense) the image canbe turn electronically for display and editing.

Of interest, One of the uses for faceplates are as interconnects in image intensifiers, to pass the image from one tube to the next.

I have been so involved in the cinema camera project I haven't read six molnths of all the adaptor threads, what advancements have been made? I will have to read it sometime, eventually.

Something has occured to me recently that maybe of some help. You also mentioned a relay lense system for the image intensifier, which I was advocated back in my first thread on dvinfo year before last (for some silly reason stuck in the panasonic group from general, when it was allways about alternative imaging, which led lots of Pana people to descend on it, maybe the moderator would like to move it here or shift it back to general group before I get back to it). It has occured to me recently that when you relay an image down to a smaller sensor through an optical lense image intensifier that you increase the effective aperaure (sending down the f rating) therefore dramatically reduce the DOF. This is simular to what the Drake digital cinema camera project did with using super fast (F0.75) 16mm lense, they were able to obtain simular DOF to 35mm film. So maybe we can get 35MM still camera like DOF by using this sort of shceeme (but now needing a good set of ND filters for mid-day shooting). This however, will not suit all cameras as some cameras may require straigher light for there microlense or sensor cell design, other wise the extra angle of the light will cause it to miss the pixel, or bounce off. Bob, so the question is if we can do the same DOF, and we can do the same FOV, as an 35MM SLR lens through this system, are there any other 35mm image qualities that could be compromised in doing this?


Thanks

Wayne.

[David Parker]

Just thought I would drop in to see what is going on as of late.

I am Dave Parker and own Ground Glass Specialties, there have been a few on here that have purchased my ground glass for their projects, both just flat stock as well as a couple that have purchase in filter rings which I have the ability to product now.

I was wondering if anyone that is using our ground glass couldpost some feedback as to how it has worked them or if I need to do a finer grind than I currently am doing, we are currently rated at around 1500, as I specialize in ground glass screens for view cameras, I have not heard if our current glass works correctly for the application you folks are using it for.

Thanks for any feedback that you can provide.

Dave Parker
Ground Glass Specialties
Satin Snow Ground Glass
www.satinsnowglass.com