Mini35 Type Adapter Discussion

For those of you who don't know, it is a proven that you can now make an adapter that will allow you to use 35mm lenses of any type, movie primes or still, of any mount. The specifics are not important, only the theory is, and the theory works.

I may or may not cover the history of this adapter in detail here, but that is not whats important, what is important is how it works.

PART I. Intorduction and theory

A. Backstory

You may have seen the PS Technik Mini35 adapter for sale or for rent and you may have realized that it is priced so out of line with typical budgets for DV productions that it really is cheaper to actually shoot in 16 or possibly even 35 for real. Nevertheless, their product opened the door for the internet community to brainstorm and determine how their system works and how to build their own for much much less.

The first I heard about this adapter was actually the fan film Marla marlathemovie.com . Their system I tried first, basically taking the viewfinder off of a SLR, and shooting the focusing screen with my video camera. Surprise, it works! But its impracticle. One, the camera needs to be mounted at a 90deg angle to the SLR, which makes it totally impractacle for anything but totally stationary shots. Two, there is too much variables to consider, its tough to clean, it wont work particularly well on anything but a Nikon F3.

So then it dawns on me to find out how it works. And here's the theory.

B. Theory

The prime works like this. The image comes through the front, through the lens, and is projected out the back, but the projection is not linear. The image only forms at approxamately 2" behind the rear element on the lens (which is just magically where the film is located behind the lens in an SLR, or the mirror that sends the image up to the viewfinder. ). It is crucial that whatever is to capture this image is in the EXACT position where this image is projected, because 1 mm either way means that the focusing will move too far towards the close up, or too far past infinity. So in other words, it must be perfectly spaced for the focusing to work correctly as it is indicated on the focus ring of the lens.

So now we just hook up the lens 2" away from the video camera lens, right? No. You will technically see through the 35mm optics but that is not going to take advantage of the DOF or FStop of the lens at all. Not to mention the image would be too small to do much with at all. So what then? Well remember how it works in an SLR. The projected image coming from the lens actually exposes the film, not the image in the lens itself, the film cannot "see" it only is exposed by the light that hits it. So we must STOP the projected image where the film would be. Seem hard? Don't worry, I did the work for you. What you need is the same principle as the Large and Medium format cameras, stop the image with some type of Ground Glass.

Ok so now we project onto SOMETHING at about 2 inches back from the lens (try it if you want, it will work no matter what you use). But how do we get the image into the camera? Here's where it gets a little tricky.

C. Why you can't just hook up the lens to the video camera

See the image needs to hit this plane of material to be stopped. But it would be impracticle to shoot the image from the side where the lens is. So we need to make this whole thing inline and use a ground glass as I said. In order to do this effectively we need two things out of the ground glass or whatever we use to stop the image.

1) High light transmittance. If the glass is too opaque, you wont hardly see the image because it will be too dark. But it cannot transmit too much light straight away otherwise we will just see the camer lens through the ground glass. So it needs to be diffused heavily on one side, but thinly.

2) Low Grain. If it is too visible with grain, you will see it in your final image every time you move your camera, and if its real heavy you will see it in static shots too. Its not like ever moving film grain, it looks like a thousand specs of dust that someone forgot to clean on the lens.

So it has to be a uniform, finely grained, diffused surface on one side of the ground glass, and the actual glass has to be thin enough to let a lot of light through without losing much.

Here is where the static and moving GG (Ground Glass) adapters diverge.

D. Static GG's and Moving GG's - the difference

Static adapter creators have tried everything from Beatie Intenscreens to acid etched glass to aluminum oxide scratced glass to just about anything you can think of. So far the BEST success I have seen has come from the Beattie Intenscreen, but I have only seen what was posted from Steev as mov's. I have been using a Nikon Type D Focusing screen. They are basically an all in one magnifier (plano convex lens which I will talk about in a moment) AND ground glass. It is very efficient, but semi costly @ $30 per one, if you can find it. It is the same size as a 35mm frame, which is not by coincidence, which again I will explain in a moment. HOWEVER, the ground glass is incredibly thin coating so it does not withstand much wiping of any type or it will start to lose its coating. I have also found a perfect piece of injection molded plastic that seems to do maybe even a better job than the Nikon with eliminating grain, but the light transmittance is probably 1 fstop worse.

Moving Adapters incorporate a portable cd player motor, a clear or frosted cd or GG spinning on the portable cd player instead of just the static GG. This radically eliminates any grain, but introduces many new problems such as vibration, size, and necessity to understand basic soldering etc. Personally, I don't see advantage to it only because I have had so much success without needing to do that. Believe me, when all is said and done I think either I or someone else will find the perfect GG or equivelent that will eliminate the need to get rid of grain because there won't be any to be seen.

Play around with different materials to understand what all of this means, particularly clear materials.

E. Lens to the Ground Glass, now what? Plano Convex.

Where are we now? Well we have the image going through the lens, hitting the ground glass at a precise distance that is approxamately 2" away from the rear element of the lens. Open up the aperature completely on the lens and check it out, its an image! How cool! If you can, try and focus the lens, and OMG it focuses! We're close to having the whole adapter now!

Well, there a slight problem. Remember the magic size of the 35mm frame and the focusing screen being the same size? Remember how I said we need a magnifier (actually a plano convex lens)? Well its not by accident. See, every 35 mm lens projects an image that is actually round and takes up a much larger size than the size of the 35mm frame. But only the size of the 35mm frame, in the center of the projected image, is usable. The surrounding area of this, all the way to the outside of the image, gradually loses its light intensity, and is more or less unusable. This is called Vignetting. This is not good for your image, it looks very poor. You can *spread* the light, with a plano convex lens, and slightly magnify the image itself, making the whole image bigger for your video camera to capture, but you will not be able to pull *all* of the frame. This is exactly why I use the Nikon focusing screen. The size is exactly the maximum size you can get from the projected image, and the plano convex lens on the front of it maximizes the light all the way around the image. Not to mention it is already 4x3, so the camera frames perfectly.

F. Hey its a nice image now, we're ready to shoot it right?

Pretty close. Now you have to determine the capability of your camera as fas as what the minimum distance it can focus on is, ie. macro. Can you put something right up to the lens and focus on it? If you can then you are ready to read part II and start building the adapter. If not, you will need a macro lens. At minumum you should get a +4. A great one to get to start is the +4 +2 +1 Hoya kit that is relatively inexpensive. Or you can get a +10 Hoya which I just bought. So now I have +17 but I think it may be overkill. But wait until you read Part II so you know exactly what you need.
__________________

Sidenote: Wideshot, what sort of 35mm adapter did you make?

Static, kind of custom, but nothing is really different from the traditional design.

+1+2+4+10 Macros > Plano Convex > GG > Lens
 
Last edited:
Right now I'm using a Nikon D Focusing Screen. I also have some injection molded plastic that is absolutely perfect with no grain, but the transmittance is lower. I also just got a 4x5 piece of GG from Satin Snow, the grain is slightly more noticable but I'll have to wait until I get it in the adapter to find out what it really looks like.

Shaw, I'd really like to use an anamorphic, as switching to a wide angle is just not really possible. At 2.8 the aperature is just too small and a wide angle that will do 1.x is just way out of my price range.
 
Interesting choie. Does the focusing screen have a lot of visible grain? I'd love to see some screen grabs if you have the time to spare.

Interesting comment about the wide angle lenses. Have you tried using a condenser lens of some sort between the focusing screen and the 35mm lens? It might help a little bit. The adapter I'm working with will actual sit between the 35mm lens and the lens mount. The benefits this provides over an adapter which goes in front of the lens are just too large to ignore. Wish I could use anamorphic primes but, dang, those things are expensive!
 
Last edited by a moderator:
Yes, I'm using the plano convex, which is what I think you are talking about. Basically, the focusing screen as I said above has the plano convex seperated a minute amount from the GG. I cannot see any grain. I have not 110% tested it yet but I've done some outside stuff and inside and I see no grain.

So basically the anamorphic takes the image coming through the lens and condenses it down to the usable area? Is there any distortion? Would it be possible to use it as a filter adapter like the wide angles that exist now (but they suck)?

Oh and depending on the interest of this thread I might do all of my posting of info and screen grabs as well as maybe some mov's here, diagrams and such, or I might just do it on my website.

PS are you the same guy that is developing the anamorphics on that other DV forum (we can't mention it here)?
 
Last edited:
Very nice! I'll have to look into such a focusing screen myself.

Yes I am the same guy as on the other site :D.

I actually posted this on the other site in an obscure thread but I'll quote myself here as I think it explains the process a little bit. After the exlanation is another quote about the benefits and drawbacks of the system.

The system I am building is known as a Bravais or rear anamorphic system. It is deisgned to allow standard 35mm film lenses to be used to create a widescreen image. These have been utilized in some 35mm films.

In it's most basic form it looks and works like this:

http://www.weet.us/anamorphic.GIF

The design effectively performs an optical equivalent of a digital in camera stretch. This is not the same as an anamorphic adapter which you would screw onto the front of your lens neither does it loose resolution like you would with a digital stretch. It in essence takes the middle 70% or so of the _projected_ image and optically stetches it vertically (yes, vertically).

Lets assume we start off with this image projected onto the ground glass (no recording with the camera is done yet):

http://www.weet.us/recon2_small.gif

Then optically we stretch the image so it appears like this on the ground glass:

http://www.weet.us/stretch.gif

What the camera actually records then is only the central portion of the image (though it uses the cameras full CCD block to do so - no res loss):

http://www.weet.us/stretch_crop.gif

Now ignore the resolution loss in the following image. This was done with a normally shot digital still so this process looses resolution in the demo pics (I didn't shoot it anamorphically to begin with). This is merely for demonstration purposes.

This then is what gets recorded to the full CCD block:

http://www.weet.us/cropped.gif

When you stretch this out in post you end up with an anamorphic picture that is the full resolution of your CCDs.

http://www.weet.us/stretched.gif

What you do loose is a little bit of light (the light that no longer falls on the actually imaged area. This is, however, a small downside given the great benefits. This system also has NONE of the limitations that an anamorphic mounted in front of the camera lens has.


Front Anamorphic Disadvantages:

- Causes the lens to have two focal planes; one for each axis. Because of this you have to play around with both focal planes so they are both in focus at the same time. This can severly limit the ability to choose a specifc aperture and/or focal length for creative purposes (which partially defeats the purpose of a 35mm adapter in the first place).

- Can cause strong horizonal and/or vertical flare due to the lens shape and proximity to direct sunlight. Some people like the effect but it can also be rather annoying to have a huge flare across th entire screen from a light in the corner of the room.

- Very few matte boxes are designed to work with square lens shapes. Also, the wider anamorphics can cause severe vignetting unless you have a very, very wide mattebox.

- Limit your close focusing abilities due to multiple focal planes.

- Filters are harder to come by

- Need an adapter which can fit many different lens diameters. This causes the need for a very large diameter lens (and consequently more expensive/slower).


Rear Anamorphic Disadvantages:

- Out of focus points can become squarish instead of round. Some people find this to be bad others don't even notice. The amount varies. This effect can be seen in films such as "Apocalypse Now"

- More light loss. At least a 1/3 of a stop for 16:9 and more for higher compressions such as 2.35:1 (at least 2/3 of a stop)

- Depending on the mechanical design, could possibly extend the focal length of your lens slightly.

- Actually looses vertical field of view to create tha anamorphic image. Much like cropping the image in post but done optically so you don't loose res. This isn't a problem for most people as they would have cropped in post anway (again no resolution is lost only vertical field of view).


Front Anamorphic Advantages:

- Out of focus points are always oval in shape (note: not round as you would get with a straight 35mm spherical lens). Some prefer the stretched oval shape.

- Extends your view by approx 33% for a 16:9 adapter. This can be both good and bad actually. It can cause some strong barrel distortion at times.


Rear Anamorphic Advantages:

- Easy to use when made correctly. Just a simple piece between your lens and the body.

- Does not suffer from the strong artifacts that can plague front adapters in terms of distortion and flare.

- No need to have an adapter that can fit many lens diameters as it mounts to the rear.

- You can keep using your current matte box and accessories.

- Smaller, lighter, less expensive
 
Yup I mentioned that in my post. The system they used obviously worked but is pretty impracticle. The inline systems that we are using are much easier to work with, and dont require an SLR. I remember watching that movie a year and a half ago and thinking "good god that's awesome. Its not film look, it really feels like film."
 
Last edited:
You know Shaw, I had read that post. In fact I read pretty every post there in that subforum. I am really very intrigued by your adapter, but I have to ask because from your pics its not making sense to me:

So I install this adapter of yours,

Where exactly is this placed in the path from the Lens to the GG?

All I do is then upres the horizontal pixels in my NLE? do I lose a lot of res in the NLE, in other words is it visible loss?

If I understand correctly I gain horizontal picture because the adapter would grab more of the horizontal projected image and squish it inward. So then that extra image area no longer has the vinetting properties that it had without the adapter?

Is this an off the shelf Double Convex + Double concave or does it have to be a certain width?

Again, very intrigued although you did mention it would be $300. Still thinking you can hit that? Its a stretch for my budget but if I picked up 200-400 pixels of horizontal res I guess Ill have to consider it.

Have you built a prototype yet?
 
Well what's wrong with you poke? :D

Sorry, to repeat myself then Wideshot! I just wasn't sure how many people saw it :)

Well there are actually several ways this adapter could be built (and thus several ways it could be installed). The way I'm looking at right now would require no real effort on the users part. It would have bayonet mounts on either end and would literally mount inbetween the 35mm lens and the 35mm lens mount. Just snap the piece in place and you're off.

There's no need to upres anthing really. Assuming the adapter provides a 16:9 compression, you would simply open it in an anamorphic project in your NLE. The end result would be exactly the same as if you had shot with an anamorphic adapter from century or panasonic.

The adapter works like this:

-image comes through the 35mm lens
-the image gets stretched vertically by 33%
-The CCD picks out a 4:3 area of that vertically stretched image. Because it cannot pick up the whole stretched image it crops off 33%. This results in a 33% loss in vertical field of view though no resolution loss. This gives you an anamorphic image. A normal anamorphic would give you a 33% wider field of view. This gives you 33% less vertical field of view.

It shouldn't have much effect on vingetting though it might help in the vertical direction.

Now, if the adapter were to compress the image to a 2.35 aspect ratio you would have to manually unsqueeze the footage (16:9 can be unsqueezed by the editor by using an anamorphic project) and then import this unsqueezed footage into a new project. It would certainly not be the easiest route to go and would take some time to setup properly in your NLE. In fact, you wouldn't really gain anything from using a 2.35 adapter if you are going straight to DVD. The image would have to be downconverted to fit into the DVD spec. Where a 2.35 would be useful is in a film transfer or digital projection (where you project from a laptop where the resolution limit of the DVD format doesn't matter).

It's a complex issue to be sure. I hope that made a little bit of sense. It's rather difficult to explain! Feel free to ask lots o' questions :D. It certainly took me a while to figure out! I'm going to draw up some detailed diagrams showing the differences between the two type of adapters side by side which should help.

About the construction:

The adapter uses cylindrical elements so that is different. Both lenses have to be perfectly balanced in terms of compression and focal length to get the right compression in the end. If it is off by a little bit you will be screwed. Because of this stock optics cannot be used. It's going to take some custom pieces. I do not have any real idea of price yet. I'm still looking into it. Needless to say I'm trying to find the best price. The real expense in this project is the custom optics.
 
You can keep the footage squeezed in it's anamorphic state for editing and whatnot.. and a lot of widescreen dvds are anamorphic, meaning they're stored in the squeezed state, which actually reduces the amount of compression used because it's not wasting space compressing black bars.
 
yep! A widescreen project will display the squeezed footage in its wide form but it won't actually affect the image. When played on a TV the DVD player does the unsqueezing.

For 2.35 though you would have to unsqueeze. Unfortunately you can't make an anamorphic 2.35 DVD.
 
Right, so do I but they are actually 16:9 masked to 2.35 (shot on super35 and then cropped the 2.35). To get a 2.35 image you have to either mask a 16:9 image or stretch it out so it will fit with the right proportions within a 16:9 area.
 
OK I think I understand now. This benefits really only something that will be projected from actual film, HDCam or similar. What you gain, in the end, instead of just cropping the image in post, would be ~200 pixels of resolution. This of course, would not be entirely noticeable on a DVD project.

Whereas a traditional anamorphic adapter would gain you horizontal field of view and consequently, horizontal resolution when unstretched.

This is very interesting, nevertheless I don't know that I actually need to just gain overall resolution, I'd rather have the horizontal field of view. Is it possible to switch something around to gain this?
 
Shaw, it finally makes sense to me lol, as of yesterday, after trying to make an anamorphic DVD from 4x3. It will never come out right. You need to do one of two things, 1) Shoot with true anamorphic, or two, optically trick the 4x3 by giving it anamorphic pixels which can be stretched in post. What this does is allows me to do, in the end, is deliver a true 16:9 flagged DVD, which of course is letterboxed on a 4x3 set, or displayed in 16x9 on a capable widescreen display.

What ends up happening to make a 4x3 DV image anamorphic without one of your adapters is the following:

First Method: Letterbox, at 720x480, then set the aspect ratio of the pixels to 16x9. When it unsqueezes in a DVD player it looks messed up, straight lines get an unusual jagginess to them, and overall it just looks bad.

Second Method: Similar to above, but after setting the pixel AR to 16:9, downressing the frame. Downressing DV is bad. In the end your unsqueezed image looses too much res and ends up looking like a VHS copy.

Now if I started with, say, a Hi Def image, or a film image scanned to hi def, this would not be an issue at all. Hi Def is 7x the res of a DV image, and therefore has plenty of res to downres after setting your pixel AR to 16:9. However, with DV, you need to pull as many tricks out of the bag to just make the DV image as good as possible so it stacks up on DVD, and if you start downressing 480 lines of res, it won't look good on DVD.

The end conclusion is that your adapter is actually a poor man's squeeze 16:9. It is not true anamorphic and therefor you do not gain the horizontal field of view (which is a huge plus on true anamorphic), but it does allow you (or should) to create a 16:9 AR successfully on a DVD.
 
Missed your last post somehow! Wow, wasn't just ignoring you mate!

About anamorphic shaped pixels: the same number of pixels are used just stretched by the display device to fit the widescreen TV (I think you know that from your post just clarifying for my thinking process). So if you play normal 4:3 footage on a widescreen TV is will look like it's stretched horizontally - because it is. The anamorphic footage looks normal because it is recorded by the camera squished (as you mention in your post).

So you have several options to get this squished image:

1) Use and in camera digital squeeze like the DVX100A has which takes the central 2/3 of the CCD and stretches the data vertically to fit the full frame.

2) Use an anamorphic adapter to squeeze in 33% more information horizontally (it will record squished).

3) Use a rear anamorphic to get do the same thing an in camera digitial stretch would do except optically. No resolution loss but like the digital stretch vertical field of view is lost.

The rear anamorphic is better than the digital squeeze because it create the same effect optically - by optically stretching the light source vertically before it is recorded. It does not, as you point out, have the advantage of a wider horizontal field of view that an anamorphic adapter/lens would have. It does have many benefits though which I personally feel outweigh this issue but that's a personal opinion. You will, of course, have to look at the list of + and - to determine which is better for your purposes.

So, now that I have repeated information that you knew:

The end conclusion is that your adapter is actually a poor man's squeeze 16:9. It is not true anamorphic and therefor you do not gain the horizontal field of view (which is a huge plus on true anamorphic), but it does allow you (or should) to create a 16:9 AR successfully on a DVD.

Yes. It's optical so it's not the same as a *digital* squeeze (just making sure to differentiate between the two). No resolution loss. Very same idea though. It isn't true anamorphic in the sense of squeezing 33% more field of view in horizontally, but it is anamorphic in that is stretches the image in only one direction.

Yes, it allows you to create a 16:9 DVD :). So... I think you got it (despite all my confusing rambling)!
 
I definitely will post some pics and clips, I was waiting to see what the response was, but it doesn't look like many at IT want to know about this. Maybe after they see the results they will...
 
Back
Top