Discussion in 'News' started by NHSN, Jan 28, 2022.
Everything runs on iPhone or Facebook these days...
Hi @Bill C, many thanks for taking the time to explain things so clearly (again)! The concept of 'aerial images' was new to me but I now understand the principle of how 'digital phone backs' for film cameras photograph 'aerial images'. Fascinating!
I didn't find it easy to find more information on 'aerial images' via Google because most search of the top results refer to images taken by aerial photography. But I did find a few (by including optics and physics in the search). Interestingly, one of these, written by a teacher giving classes in optics, advocated starting optics courses by introducing students to 'aerial images' much as you explained!
One of the first references I came across via Google was to Wikipedia page on 'aerial images'. This was just a so-called 'stub' with just a couple of lines and no references. So I'll include the 'sources' I found as a result of your explanation on the Wikipedia page.
Do not let this make you believe that there is something special going on. 'Aerial image' means nothing else than that the image produced by a lens is not in the same location as something else, such as a screen, film or sensor. It just 'floats in air'.
Special things can happen when you can put a (tiny) object at the same position, making both it and the aerial image the subject, allowing you to get a sharp photo of both, say, an ant and the landscape. Something dof will never allow.
And that really is not that special too. It is what is also known as front- or rear projection, but then without a screen.
Thanks for this comment! I've never understood much about optics (other than convergent and divergent lenses) and I've never really thought much about these in relation to photography (or anything else). I've always just assumed that lenses with (a range of) different focal lengths do basically the same job of allowing photographers to manually or automatically focus an image on whatever medium (different film formats or digital sensor sizes) that they prefer.
Up until now, I've never questioned the principle of substituting camera exposures recorded on film by those recorded on like-sized 'traditional' digital sensors. So for example when a 35mm film camera is (temporarily) fitted with a 35mm (full-frame) digital sensor and memory 'back'. I have the impression that the main market for this kind of 'digital backs' has until recently been mostly professional large-format photographers.
But - ever curious - I was intrigued by 'the optics' of how a tiny iPhone camera could capture a 'reasonable image' that would normally be projected onto a film or digital light sensor many times the size. From browsing the internet on 'digital backs' based on digital phones, my impression is that there's much more information on 'what' the products do than 'how' they actually do it.
Following @Bill C's explanation, I did track down a 2013 reference to 'photographing a screen'. The 'screen' was presumably the equivalent size of a film frame or digital sensor. I have no idea whether this technique is still used today.
One of the main insights for me from @Bill C's explanation is that it's possible (with the right equipment) to photograph 'aerial images' before they reach any physical medium. I've also learned that 'aerial images' may - in the future\- play a role in 'virtual reality applications'. In the sense that images and videos remain 'invisible' until people walk/drive into certain locations at which point relevant images/videos suddenly become visible.
All in all, the concept of 'aerial images' - and their potential applications - is something that, even at my advanced age, interests me.
The thing is that lenses change the direction of rays of light passing though them. Behind the lens, that light is not invisible. But it only forms a sharp (and upside down) image at one exact distance behind the lens. The image gradually grows into sharpness and out of sharpness again, with growing distance from the the rear of the lens. The unsharpness you see through a viewfinder when focussing a lens is the unsharpness of the image seen from the incorrect distance.
When that light hits a diffusor, such as a focussing screen, it it scattered in all directions, the screen thus acting as a source of light that is easy to focus on for our eyes. Without such a diffusor, our eyes do not know where to focus, and it is rather difficult, but by no means impossible, to see and focus on the aerial, sharp image. There are also clear glass 'screens' that only bear marks (lines) to give the eye something to focus on, used to focus equimpment (cameras, microscopes) that only transmit very little light. It helps, there, that the image often is at high magnification, and the depth of it is very shallow. So when the eye sees the image, it probably is almost in focus as far as the focus setting of the equipment is concerned. But for critical focus, you must sure both the markings and the image are in the exact same plane. You use perspective, parallax, to achieve that (as long as the markings move relative to the image when you move your eye, they are not in the same plane).
Using and capturing an aerial image is not really special. But i do not know what they do to focus the iPhone on the aerial image. Nor how the iPhone lens matches the task of capturing the image at the distance required (poorly, judging by the amples. Bad vignetting).
Screen photography, by the way, was quite common.
Tons of Polaroid (expensive) and 35 mm film were used to capture, for instance, the images on CRTs on oscilloscopes and such. It is very cheap to replace that with digital capture (i know, because i have helped people convert their old style scanning electron microscopes to digital capture devices. Very cheap, off the shelf devices and some fancy anlaysing and programming was all it took. And a lot of time was wasted, then, finding out the experiential way, that cheap $15 items do their work better than dedicated components that cost many, many times as much).
But if you have a camera that allows seeing the focussing screen directly, you can see how extremely simple it is to take photos of an image on a screen. There is nothing to it.
And as an P.S.
There is nothing special, or new, about aerial images. There is no unexplored potential. There is no hidden magic. It's just what lenses produce. And they are known and exploited ever since they were first invented.
Lenses project images. If you put a screen where the image is, you'll see the image on the screen. If you substitute the screen for a film, it is exposed on film. If the substitute is a digital sensor, it may be captured by that. If you look at it, either directly or through another optical system, you will see it.
Again, nothing new, and no magic.
It's easy to take picture of an image on the screen but some how I am having problem taking picture of an aerial image.
Point a lens at the window. Point a camera at the rear of the lens, and as soon as you adjust distance such that you see sharp through the lens, press the button.
A random example on the internet:
Robot or human?
A pretty simplistic lens. But an excellent example of a photo of an aerial image.
There is nothing special, magical or difficult about it. Really!
To science it is not magic to an artist & viewer it can be magical. The first time I saw an exterior image focused on an interior back wall from a door keyhole in my darkroom it was special. Not a new idea but wonderful and mysterious for me even tho I knew how science explained it. It led to some interesting experimental images.
Let me then stress the bit that there is nothing difficult about it. It's not, erm... Rocket Science. It's more like tying-your-shoe-laces science. Not even that. Easy. Not hidden, but plainly visible. And accesssible to anyone who could make a photo such as the one i linked to as an example.
If you are seeing only a small circular image in the center then most likely your problem is field lens related. (As I mentioned, this is not something that photographers would generally understand.)
If you wanna answer a couple of questions, I'll make some suggestions. 1) what is your general setup, meaning main lens, desired image size, and the setup you are using to photograph the aerial image? And 2) do you have some junk optics laying around to play with? (What sort of stuff? A set of "close-up filters," ie, screw on diopter lenses for a 35mm camera would be useful.)
Parabolic Mirror Illusion
has a good explanation of aerial image.
But actually in optics, there is real and virtual image, and I am not sure that aerial
image is always the same one. Real image you can put the film (or sensor) there
and get an image, virtual you can't.
To get a good aerial image, you need big enough lens or mirror.
Hi, I more or less agree with most of what you've said, except... the simple application of trying to photograph an aerial image from behind a camera lens is not gonna be able to be any larger than the clear aperture of the camera lens. Yeah, that part is easy.
But if you wanna photograph the full-size image, such as the lens might project onto a piece of 35mm film, then it's not so easy.
Fwiw I designed the optical layout of such a system in the earlier days of "still video" systems, before digital came on the scene. The problem was to match the image size from a 35mm long-roll portrait camera, using a remote-controlled zoom lens, with a small-sensor video camera. Long story short, we piggybacked another zoom lens on top, mechanically linked to match the zoom, then picked some off-the-shelf optics to get the correct image size on the video. They worked great, couple thousand of em. The operator ("photographer") framed the subject via TV monitor. When they took the photo, the video camera was able to take a simultaneous "flash grab," identical to the film image.
A company we had worked with quite a lot, PhotoControl of Minneapolis, had another solution to the size-matching problem. They designed a so-called "ported lens," using a beam-splitter block within the lens. The optics straight through the lens gave one focal length, for whatever film size size was being used. But the image being ported out the side used a different set of back-half optics, designed to match whatever sensor size was being used. Much better image quality, I'm sure, but our system was plenty good for "proofing," which was a pretty big deal at the time. Up until then a portrait customer might come back a week or two later to find that their kid had blinked on every shot.
Yes, in my view an aerial image is always "real." But... it's possible to put in a field lens such that the image is within the field lens, making it "virtual." But most people would probably still say they are photographing an aerial image. If we don't allow this sort of loose usage of language it's difficult to function in the real world. Imagine trying to tell a word purist that you just developed a roll of film. ("What?!! Just developed it? You didn't fix it? Etc.). Everyone knows what you really mean, more or less.
I was not addressing that. Only pointing out that there is nothing new, esoteric, or magical or whatever about aerial images. Nor about seeing or taking pictures of them.
Wanting to capture all of it on a smaller format, and preferably without light fall off, is more difficult. In the case of the glass orb, its practically impossible.
But, as you know, that doesn't make aerial images hard to see or photograph, viz. the image 'through' that orb.
Glen, "But actually in optics, there is real and virtual image, ..." that piqued my interest and led me to do some research. It took me some time to wrap my head around the simple & clear explanations i found. Interesting google rabbit hole, thanks.
I will take you up on this at a later time perhaps thru private message. But reviewing the video the guy did have a screen at the film plane. So he didn't take a picture of the areal image.
Reminds me when I was young, stories about a Polaroid back for Nikon SLRs.
The image should be bigger than a 35mm frame, and also has to be farther back.
That would have been Nikon F days. I figured some type of lens system.
But now I found a picture of one for sale. It seems that it has a fiber optic block
to bring the image from the (previous) film plane to the Polaroid film. A solid block
of optical fibers, arranged to bring the image back without distortion, called
coherent, though unrelated to the common optics meaning of coherent.
Yeah, long time back we wanted to do something like that for our studio cameras - instead of waiting an hour or two to get film processed it seemed maybe useful for our repair shop to have quick acces to an enlarged Polaroid. I roughed out a design, but given all the intricacies involved, we figured to just have one made by people with experience. Which we did... I think maybe it was NPS. Now I know enough for a basic layout, but by no means could I design a high-quality relay lens, so this is for the specialist outfit. Anyway, as I recall, they just used an El-nikor inside. Ultimately the device turned out to not be that useful and spent most of its time on the shelf.
Regarding the fiber optic plate... this is something I actually considered with respect to our in-house digital camera interface to an existing film camera. Conceptually this would allow transfer of a hard-to-reach image plane to the bare face of a digital sensor. But there would always be the issue of imperfect alignment, etc. So in principle a relay lens arrangement is much better, quality wise. Provided that you have the physical space available.
Fwiw I was already familiar with the fiber optic relay means from some earlier equipment we had used. In the early days of digital printing an outfit called Sienna Imaging, as I recall, developed what they called a FOCRT, a linear CRT with a fiber optic faceplate. So essentially photo paper could be "contact-printed" by being advanced against the face of tri-color linear CRTs. As I recall the CRTs were maybe 11" wide, and could expose a nominal 8x10" print in a half minute or so. One of the first feasible digital printing systems for RA4 paper. We ran em in a couple of portrait studios as a trial. I think the most usable iteration was marketed by Gretag as the NetPrinter. (Maybe some info still exists on the internet? This stuff has a way of disappearing from internet history. )
Fwiw you used to be able to buy tapered fiber optic "magnifiers" maybe from Edmund scientific. So the same idea of image transfer through the fibers, but with a larger diameter at one end. No, I don't know of any practical use for those.
Separate names with a comma.