Okay, it's late and so this may not be the brightest idea I've ever had, but I was just sitting here thinking and came up with something.

From my understanding of optics, the virtual image becomes smaller and smaller until it reaches the focal point, correct? And after passing the focal point, it begins expanding again, but now inverted (because the individual rays continue travelling in a straight line). Couldn't we take advantage of this? In one of his recent posts about his adapter, Jimmy said that (as would be expected) placing a lens between the 35mm lens and the GG changed the focal length of the 35mm. So theoretically, a properly-placed lens could shorten the focal length of the 35mm sufficently that the image could completely flip over before reaching the GG, giving us a properly-oriented image on the GG.

Anyone with a more thorough knowledge of optics care to comment on this?

You do have a point. However, it is how feasible and WHERE and how FAR this "invertor" lens would be placed? Would it make the whole, shebang too long? - At the moment I can't understand why this hasn't previously been employed. I'm sure we will find out though. Loss of lumens can't be a factor as a prism must have at least a similar lumen reduction factor?

I'm listening.

Grazie

Yes, I considered that. I think it would probably have to be a fairly powerful lens to shorten the focal length enough that we wouldn't end up with a 12" long adapter. It might have to be an achromat to avoid distortions too, I'm not sure. Nevertheless, it seems doable...

Edit: I took the lens off my SLR and tried it out with a mirror, and sure enough, the image flips over after passing the focal point. Although that was a bit of a no-brainer I suppose...

Sounds like an interesting idea, but I wanted to say something slightly off the topic... Why don't more people use first surface mirrors for flipping (not prisms)? It was my understanding that there's less light loss with a mirror, or is it the opposite?... I'm using mirrors, myself.

Go to this link and play around. http://webphysics.davidson.edu/applets/Optics/intro.html
It looks to me that the second lens would have to have a much shorter FL than the first in order to keep the focal point somewhat manageable (with equal lenses, the focal point seemed to be as far away as the object was from the first lens). Considering the cost of mirrors or prisms, it may just be the economics of keeping CA down, let alone any distortion problems that may crop up, and having a second piece of expensive glass trying to have short focal length and pristine optical characteristics. The shorter the FL, the harder it is for a set diameter lens to keep distortion and CA to a minimum, and that may be a factor too.
Anyway, if you haven't seen applets like this before, they're a lot of fun.

Glen

Is the whole obsession with flipping the image just to avoid the trouble of shooting upside down? If so, why not avoid the prism, mirror or whatever to gain some f-stops and bring the distortion level a notch or two and just use a extra LCD screen upside down attached to the camera? For post, Cineform flips the image during capture so it doesn't even matter.

Or just run the camera upside down (which is what I do). There have been several solutions using surface mirrors including ?? the Letus flip. It's not a big deal, you just need space to do it....which means a larger box. Two porro prisms (like older binocs) , as well as a mirror/roof pentaprism are other options.

Why not a telescope erecting prism . they usually are at a 45 degree angle but if used with a small enough camera ?

The projected image is 24x36mm...

they have 2.25" d. elliptical erecting prisms commercially available. I think that would be big enough

Commercially available 2.25 " erecting prisms are common - I think that would be big enough- Kurth

Michael: For me the "obsession with flipping" is two-fold: First, to avoid having to have an unwieldy LCD screen (the more stuff you stick on the camera the harder it gets to move around. Plus you have to power the thing) and rotate in post and secondly just because I can. ;) Technology was never advanced by those who were content with the current ways of doing things, was it? ;)

Or just run the camera upside down (which is what I do). There have been several solutions using surface mirrors including ?? the Letus flip. It's not a big deal, you just need space to do it....which means a larger box. Two porro prisms (like older binocs) , as well as a mirror/roof pentaprism are other options.

Although using mirrors may save some weight here are some possible drawbacks:
1) You need surface coated mirrors in order to reflect as much light as possible. These mirrors are extremely prone to scratches.
2) These mirrors have to be fixed inside some carefully designed cavity and may appear to become "the temple of dust and dirt", having not only the four mirror surfaces as a dust collector but also the wall of the housing, so that the effective dust&dirt gathering surface may amount to 20x the surface of the ground glass focusing area.

My conclusion is the following: any flip design should minimize the amount of dust gathering surfaces.

I own a Letus35A flip. the picture and colour rendering is outstanding. However dust and dirt poses problems. So flipping has a price...

Michael: For me the "obsession with flipping" is two-fold: First, to avoid having to have an unwieldy LCD screen (the more stuff you stick on the camera the harder it gets to move around. Plus you have to power the thing) and rotate in post and secondly just because I can. ;) Technology was never advanced by those who were content with the current ways of doing things, was it? ;)

You don't need to use an extra LCD if you don't want. The magnet trick on the stock LCD seems to work for everybody. Flip in post? What’s the problem with that? As I said, there are even capturing applications that will do that for you automatically during the capturing process, so what’s the problem?
Because you can? Well I guess that's up to you if you want to lose 1 or 2 extras F-stops and add distortion to your image just to have a cooler gadget. For me, I really don't see the reason to go through all the trouble when you are actually downgrading your performance, but to each his own.

if you want to lose 1 or 2 extras F-stops and add distortion to your image just to have a cooler gadget. For me, I really don't see the reason to go through all the trouble when you are actually downgrading your performance, but to each his own.

Actually, there is potentially less distortion from the flip version. Most if not all barrell distortion from the letus or SG35 for that matter comes from the PCX lens inside. On the flip version, you are increasing the length of the light path and as a result, you dont need such a strong PCX, a PCX with a longer focal length can be used to give the same level if corrrection to vignetting. This gives you almost unnoticable distortion and much less chromatic abberation.

Your correct about the extra light loss, but im sure that is an exceptable sacrifice for better image quality.

But then again, with a flip version using 4 mirrors, there are more things that can potentially go wrong, dust being 1 of them.

Actually, there is potentially less distortion from the flip version. Most if not all barrell distortion from the letus or SG35 for that matter comes from the PCX lens inside. On the flip version, you are increasing the length of the light path and as a result, you dont need such a strong PCX, a PCX with a longer focal length can be used to give the same level if corrrection to vignetting. This gives you almost unnoticable distortion and much less chromatic abberation.

Your correct about the extra light loss, but im sure that is an exceptable sacrifice for better image quality.

But then again, with a flip version using 4 mirrors, there are more things that can potentially go wrong, dust being 1 of them.


Does the Letus use 4 mirrors?
Considering what you are saying about the PCX and distances, a shorter adapter like the MPIC should have a lot of distortion, but it doesn't. So I’m not sure I understand it.

Yes the letus flip uses 4 mirrors.

Regarding Dan's adapter, he is probably using something different to a PCX lens, hense his high price.

So I would think an adapter without a PCX would be the best. Ben Winter mentioned something about not using a PCX in his Letus with an interscreen didn't he?

Yes, the beattie uses a fresnel lens, apparently.

So the beattie seems to be the bet set up. I wonder why adapters are not using it. If it needs no PCX and probably no achromat (didn’t Ben say he could use it without one as well?) it seems the way to go to me.

Only experiementation can tell.

Lots of opinions about the use of fresnel's. Some say they loose resolution and the 'rings' appear in the image, especially with HD footage. I know Steev used a beattie in his adapter, weather the fresnel was attached or not I have no idea.

Dan is able to use the MPIC on a GS200 without an achromat...but it has a min focus distance of 1.6 inches...so a 24x36mm will fill a 4:3 frame from this cam at that distance.

I believe Ben is using the achromat to address slight vignette on his GL2 with his Beattie mod'd Letus.

Here's my 2c on the adapters based an a whack of reading...and a lot of experimentation (although none personally with a Beattie)

The only downside I can see of a vibrating Beattie is that you have 0 control over the diffusion surface, squirmy grain (Letus has it too), and have a need to tightly control motion. A medium format lense system with a larger Beattie, vibrating, would be the bomb IMO. The spinning GG systems have .5 to 1 f/stop greater light loss, are larger, but will accomodate higher shutter speeds (depending on GG), and will allow "tunable" bokeh.

Pick yer poison.

A fresnel lens just acts as a PCX lens. The Beattie has a fresnel with a focal length around 120-150mm (depending on the screen).

So the Beattie is using a PCX, but in the form of a fresnel (I would assume this is to increase the light transmission since fresnel lenses are very thin).

Regular focusing screens use normal PCX lenses and show no distortion.

Alain,
Any experience with fresnel type lenses? I was just wondering how it effects resolution. I may try it out for fun.

Alain,
Any experience with fresnel type lenses? I was just wondering how it effects resolution. I may try it out for fun.


I heard the MPIC uses a fresnel screen and it looks nice in the resolution department. Dan even has some rez chart shots.

The only experience with fresnel lenses I have comes from larger projection setups where the fresnel is just used to condense light from a bulb, so I do not know how it would affect an actual image.

Theoretically, there should be no loss in resolution provided that the fresnel has very thin grooves. But fresnel lenses can have much more distortion than a regular lens.

I really think an aspherical lens should be the best solution with regards to coma and spherical aberration (or perhaps an achromatic condenser?), but I am waiting on some lenses. (plus this isn't the topic of this thread :) )

Classical fresnel lenses, as described by Alain have a regular pattern. It is possible to imagine fresnel structures with local random orientation, thus bringing them closer to ground glasses.