Adapting a telescope to a fixed zoom lens HDV camcorder.

[Ed Marrs]

If a still camera lens can be focused on a spinning CD (Agus35), or on a wax screen (glasscreen.com), or on the M2/M3 HDV adapter (redrockmicro.com) so that a zoom lens camcorder with a close-up adapter (Century Optics, 58mm #2 diopter) could record the image on an HDV camcorder (JVC GR-HD1, or in the future probably a Sony HVR-A1U); then a telescope such as a 5 inch C5 Celestron spotting scope should be able to focus on the same focal plane just as well (without any eyepiece lens)?

Naturally if the camcorder’s lens were removable there wouldn’t be any optical adaptation issues. The Celestron telescope works very well with a C mount video camera, with no lens in place and no telescope eyepiece, with an optional set of C extension tubes in-between. However that’s not HDV; so I've tried the Scopetronics MaxView 40 eyepiece - camera lens optical adapter; which is designed to adapt fixed lens cameras to telescopes. However it didn't work at any focal length or alignment with the JVC camcorder’s optics.

Unless somebody has some practical experience with such long focal lengths on HDV; or perhaps something ready-made such as: the Redrock adapter or perhaps the Glasscreen element, or the Argus35 spinning CD; -- would any of these be better for HDV? I guess that the larger the image on the intermediate focal plane the sharper the image, to some degree?

Thanks for any thoughts, Ed 5-30-06

[Bob Hart]

Ed.


Here is a path which does work for the MTO f10.5 1080mm mirror lens into a PD150 and FX1. I use a design based on Agus Casse's original device.

However any relay device which uses a groundglass and close-up lens for a relay should work as an aerial image device with the groundglass removed.

The MTO lens is the closest I have to your Celestron.

It is a bit furry when an AGUS35 adaptor is used with the groundglass removed to work the adaptor as an aerial image device. It is also soft into 16mm motion picture cameras.

I also used the adaptor with a 300mm f3.5 Tamron, 500mm f8 Nikon mirror lens, Sigma 50mm - 500mm f4 -f6.3 zoom and the MTO.

Image path :-

MTO Lens or other lenses (Nikon Mount) >> 46.5mm to groundglass or aerial image plane >> approx 120m - 150mm to front of close-up/macro lens >> Century Optics +7 Acromatic Dioptre >> HDR-FX1.

My device also has two right-angle prisms in the path to flip the image.

The Century +7 close-up lens is only available for a 58mm 0.7mm thread pitch filter mount as used on the PD150. I made a 58mm to 72mm adaptor ring for the FX1 and the lens worked okay in this situation as the camcorder lens is operated fully zoomed in.

Zoom on the cameras has to be fully in to telephoto to eliminate the vignette of all the long lenses. There are some lenses which won't work regardless. You likely will still get a brightness fall-off in the corners, especially with the FX1 in 16:9 mode but it is useable.

There is no advantage to using the groundglass with such long lenses or refractive telescopes as the available aperture from these lenses is tighter than the f5.6 practical limit of groundglass relay systems. You suffer a brightness and resolution loss which you don't want for astro-videography.

I have used the Sigma 50mm - 500mm on its own or with a 2x for ground to air videography of aircraft in flight. The images don't seem to be as sharp as direct-to-camera but at those magnifications, environmental effects such as haze or mirage also soften the image.

You will discover a need to use the switchable ND filters on the camera a lot more. If you do not use the ND filters, the auto gain on the camera will reset the exposure okay but distortions from specks on the SLR lens/telescope will be introduced.

A typical situation where this is necessary is the transition from an aircraft from take-off to flight when the image chages from darker background to a brighter sky backlight.

Manual gain is also better over auto as the camera's automatic features seem to be a little confused and tend towards slight over-exposure.

As with the groundglass relay, the SLR lens has to be left with aperture wide-open otherwise the vignette creeps in tighter.

Compared with the same lens into a JVC KY-F50 "C"mount camera, the relay loses a little apparent contrast and sharpness as one would expect from sending the image through more pieces of glass.

The camcorder autofocus remains functional with some limitations. You need to initially set your SLR lens to infinity, then manually set the camcorder focus until the image is sharp.

From that point on, it is best to use the SLR lens focus to follow movements but if it proves difficult, such as when using the zoom on the 50mm - 500mm, the camcorder autofocus can be used. It is best used in short switch-ins rather than being turned loose to manage the focus entirely.

It doesn't take much of a brush with a close object to send the camcorder autofocus looking for the closest sharp object which ends up being specks of dust on the rear element of the SLR lens/telescope. The camcorder focus must be manually reset.

When looking towards the sun, all the extra pieces of glass increase flare so a sunshade or lens hood on the SLR/telescope is desirable.

Hope this helps.

[Ed Marrs]

It certainly sounds desirable to eliminate any intermediate glass or plastic between the camcorder and the telescope. I gather that I simply need to focus the telescope at some distance behind the scope (probably without any eye-piece lens), and then focus the HDV camcorder with a close-up diopter onto that same plane in space (aerial image), naturally all enclosed perhaps in a tube to keep out any extraneous light, to get a focusable image to the camera.

Since I have a Century Optics +2 diopter, I’ll try that; which probably means I have to keep the camera a bit further back; unless I can’t focus close enough, otherwise I’ll get the +7 that you used; and if that works I’ll align the camera and scope on an aluminum bar.

Thanks for the research and comments

[Bob Hart]

It may be helpful to have a temporary groundglass, (wax paper or plastic pringles chip cap) in the path to establish the backfocus of the telescope and for the camcorder via the +2 dioptre to find, then remove the groundglass for the aerial image thereafter. Otherwise it might be too hard to get everything to line up and be at the correct distances and then focussed.

In my prototyping, I found it helpful to use the pringles tube or a piece of plastic or cardboard tube, cut a slot in it for a removable groundglass, in my case a microscope slide dressed with 5 micron aluminium oxide on one surface to make the groundglass, set up the distances and focus from either end for the lens and camcorder, then pull the slide out of the slot.

The +2 dioptre may be a bit low power and you may only get a very small vignetted image.

I am surprised the Maxview didn't work. It's relay might have been a bit vignetted but I examined this albeit over the internet, before I went with another option for my project which was a night-vision adaptor for the PD150. For that I used a generic 50mm 2" eyepiece lens set (part number SW5042??) for Tasco scopes. I think the exit pupil diameter for that is about 44mm or thereabouts.

This placed the image plane at between about 20mm and 25mm from front of the eyepiece glass element which was right up inside the removable extension tube on the eyepiece body. In fact the image plane seen by the camcorder fell even closer within the non-removable body of the eyepiece.

I also found it necessary to reverse the glass optics as a complete group within a custom remake of the eyepiece barrel and place the optics as close to the camcorder lens as I could. But this was for viewing a 2/3" diameter plane object, the display tube, not aerial image. Of course this arrangement also contradicts the original design of the eypiece but happened to work right for the task I was putting it to.

It may be that the image plane from your telescope fell within the extension tube on front of the Maxview and that you could not get close enough to the front lens with a groundglass to be able to focus the camcorder on it.

Also, when you set up the Maxview, did you zoom right in? With the SW5042 optics I have to go in almost all the way for the 2/3" tube display.

I have used the same lens set in the same arrangement for a closer couple of 12mm - 18mm focus range from the camcorder focus system for a non-erecting AGUS35. It works in aerial image mode in this form.

[Bob Hart]

Deleted = double post.

[Bob Hart]

Ed.

I should have mentioned that there is an image of my adapted SW5042 lens set as the fifth image down the page in the .pdf article I uploaded to savefile.com. The .pdf file is a big one so it will take a long time to download.

It should be at www.savefile.com/projects/618011
or www.savefile.com/projects/338360. I am not sure which one.It will be the project which is not related to the independent feature "The Cage".

[Ed Marrs]

Bob,

No doubt the 1.25 inch Scopetronics Maxview 40 eyepiece to camera adapter in the telescope can’t set an appropriate focus for the JVC zoom lens all by itself, although zooming to full telephoto does help quite a bit. I did move the Maxview 40’s glass almost to touching the JVC’s front element; however the images were still largely unfocused with some of the losses at the sides being minimized by going to full tele on the zoom. Perhaps a close-up +2 or +7 diopter on the camcorder’s lens might help.

In early 2004 Bill Turner of Century Optics mentioned trying a negative lens element in front of the JVC zoom, from an optical company such as Roland Optics; at that time Century was considering introducing their 400% HDV telephoto adapter by mid 2004. However this June 2006, I called and they didn’t know about nor introduce the 4x unit; and no doubt resigned it to ancient history; I gather due to bulk and weight. This telephoto adapter required that the video camera be at full tele-zoom; nevertheless the few nature videographers that had the lens on loan for testing didn’t care to give it up; perhaps since it offered a sharp image with low light loss. In any case Turner mentioned that adapting optics to a complex and complete optical system is a complicated problem that often leaves many problems that can’t be resolved. So I gave up, until I recently noticed the still camera lens to video adapters being described here, and greatly appreciate your very helpful observations on the details.

When I have a day or so, perhaps next month I’ll give the Maxview 40, the wax paper and the aerial image approaches all a try, and mention the results. If a sharp image doesn’t look feasible, then perhaps a small HDV camera with a C or CS lens mount (preferably without any recording features) would be far easy to interface to a telescope. Although far from HDV, I do like the results I get on Supercircuits PC164C-EX .0003 lux, 600 line resolution black and white camera for night use. I also gather that Sanyo and a German division of company that I couldn’t locate on the web have small HDV cameras; which may be a better approach if it’s not more costly than an HDV camcorder, and it has a digital video recording interface available (Firewire or USB2). If it uses Rockwell’s 1920 by 1080 progressive imaging chip (Camarillo CA), I’ll definitely pay more, (which is not a particularly costly chip and has been available for well over 4 years). Perhaps such a small HD video camera or perhaps an HDV security camera with motion detection is more than a year off. Toshiba's web security cameras have such HDV resolution, however with low-frame rates at about $1000; and D-link’s (704 by 576 pixel) DCS-3220g camera, $226 US, has a C mount but without HDV; (which incidentally is very difficult even for D-link’s technical specialists to configure in order to send email or FTP images, even in their home offices, and incidentally is generally impossible for most anyone else).

I gather I don’t have the correct software modules installed or available to view: www.savefile.com/projects/338360, so on my PC, apparently Apple’s QuickTime decided that it was going to be my viewer of choice and played the clips as a blank screen; then it repeatedly asked that I upgrade its 35 megabytes; so I went along with that; then it wouldn’t load all its required modules and thankfully it’s now completely nonfunctional. So I guess I should be thankful that there aren’t 7 different JPEG formats for still pictures that have to kept-up with.

[Bob Hart]

Ed.

Sorry about being the trigger for the Quicktime crash. The blank screen may be related to the image being a full frame iimage.

I have the same problem with my PC. It only has the muscle to cope with a half-size frame H264. You may fin that if you select "play" then pause the image before it fails, you can step through still-frames.

I have some .jpg stills from some overhead aircraft so I will see if I can get them posted in the near future and let you know here when they are up.

With the Maxview, you could try reversing the eyepiece itself and letting the camera see into the eyepiece from the telescope end, ie., look through it backwards. This is how the SW5042 lens set works for me. Otherwise, the +7 achromatic dioptre from Centry might be the best bet. I know that it works with the Sony HDR-FX1/HVR-Z1P into Nikon mount SLR lenses.

[Bill Porter]

Originally Posted by Ed Marrs

It certainly sounds desirable to eliminate any intermediate glass or plastic between the camcorder and the telescope. I gather that I simply need to focus the telescope at some distance behind the scope (probably without any eye-piece lens), and then focus the HDV camcorder with a close-up diopter onto that same plane in space

Ed, I hate to say it but it doesn't and can't work that way. There is not such thing as an aerial image. A lens projecting its image onto one plane in the air, doesn't produce an image that can be filmed by a camcorder focused on that same plane.

For example, look at the link you provided, glasscreen.com. Notice how the projector is focusing an image onto the screen? Imagine if you took the screen away and focused your eyes or a camcorder in the air at that place where the screen was. You wouldn't see an image.

That's the problem with the term "aerial image" around these forums. There is no aerial image produced by a lens; it can't project an image into thin air a la R2D2 and his Princess Leia hologram.

"Aerial" means "of, in, or caused by the air." Unfortunately, around DVinfo people use it incorrectly, when they mean that their screen (ground glass, wax, etc) is so transmissive that one can see through it, rather than a piece of glass frosted/ground so thoroughly that it cannot be seen through.

[Bob Hart]

Bill.

You are more than likely correct on the definition "aerial image" which I may be using and abusing incorrectly. I am neither an industry professional, nor optical technician.

All I can say is that if I take the groundglass out, zoom in closer to eliminate the vignette, the videocamera can "see" through the telephoto lens in front to a subject and quite effectively at that.

My device is faulty in that the front (rear-facing) prism has to overhang the disk and there is about a 0.5mm gap in centre and about 1mm gap at frame edge.

When the camera zoom is retreated to widen the image off the groundglass, if one overshoots and picks up the edges of the path, the disk edge also comes into the frame. The projected image and the "aerial image"?? okay lets call it "direct image", is also visible.

The image is brighter by about a half-stop and is sharp and is largely unresponsive to the SLR lens setting. Where the image falls on the groundglass, this is responsive to the SLR lens setting.

Removing the groundglass after setting up a correct relationship between the camcorder and SLR lens yields a useful image into the camcorder.

The image from the SLR lens, via a Nikon-C mount adaptor into the JVC KY-F50 camera is sharper as it should be and has better contrast.

I tend to use the AGUS35 on the video-camcorder, without groundglass when I want the convenience of being able to record sound without all manner of leads and adaptors and power supplies having to be carried about.

When porting the JVC camera head signal into the camcorder in VTR mode, you lose the microphone inputs and a separate audio mixer or recorder has to be used.

In a weak moment of ego-centrism I emailed P+S Technik and suggested they might offer in the Mini35, the future option of being able to swing their orbiting groundglass out of the image path for direct-to-camera imaging as it is a really handy feature.

It is a feature which could be switchable with a CD-R sized disk device.

If some electronics wizard could devise a video driven sync device to run a highly controllable disk motor at an exact speed, then sectors on a single groundglass disk, finished to different grades of opacity with one clear panel could be selected.

This would also deal with the flicker problem associated with disks and the swirl problem sometimes referred to with the Mini35.