Well, after many hours in the shop, my "Guskers35" is pretty much done. It's a spinner with a DCX in front of the GG and achromat on the cam side. I cooked up a rail system to run the camera either inverted, or right side up to support everything nicely. I'll post up a clip or two once my wife starts talking to me again!

Props to Wayne Kinney and Dan Diaconu for some tips and ideas along the way. I spent way too much time reading through several years of posts here...but man, there's a treasure trove of ideas and experiments posted here.

Pic1 (http://www.pana3ccduser.com/attachment.php?attachmentid=1120&stc=1)
Pic2 (http://www.pana3ccduser.com/attachment.php?attachmentid=1121&stc=1)

Rail system mock up (http://www.pana3ccduser.com/attachment.php?attachmentid=1103&stc=1)

I have some aluminum and stainless stock left over so if anyone's interested in a rail system, let me know.

Well done Dennis,

Is it a spinning blank CD or some other material for the GG?

I'm trying a variety of materials for GG. The only glass I can find is 2mm and the cutting is a bit off, so I didn't use it. In my explorations of optical properties of different materials I found that acrylic has as good transmittance as glass, sometimes better. We also found a manufacturer of "optical grade" acrylic willing to cut and drill them to a fairly tight tolerance. Right now I have 2mm in there, and a bunch of 1.5 and 1mm rigid blanks on their way. The company also does optical coatings, but I haven't tried these yet.

Surface prep is either wet sanded (with an air powered "jitterbug") and/or I'm also experimenting with fine grade media blasting. I need to fine tune my procedure here as distance, media type (I'm using very fine glass beads) rotational speed and air pressure play a fairly large role in the coarseness of the finish. The current version is excellent in higher light as it's fairly coarse. I'm seeing about 2 F/stops total loss with the F1.4 lens and adapter in place. I've focussed so far on getting the adapter built and the vignetting in 16:9 mode (the GS400 does the "real" thing) under control. This was achieved basically by moving the achromat back a bit further and extending the cam portion of the tube about an inch. I can zoom in well beyond the vignette now. The F1.4 50mm is a lot easier to deal with than my F3.9 28-130 zoom.

Image tuning is next.

Good show Dennis. Clean and FAST flip! Way to go.

Thanks Dan. The image flip was bit of a surprise. The image is recorded correctly, but the cam senses it's upside down, and rotates the 3.5" LCD image upright in normal operation. That means that with a 35mm adapter attached, the LCD image is always inverted..regardless of the cam's orientation. bummer. My external monitor works fine though in that the image is upright with the cam inverted.

I know, I know, you're shaking your head and mumbling to yourself "Grasshopper, you should have used an oscillating fresnel. Tisk, tisk"

That's for the next version....although I'm not holding my breath on getting the motion right. The spinner is actually pretty small, and very quiet, so if can live with 2 f/stops, I may just stop here.

Looks really good Dennis..

Q: how do you power it? can see an LED etc there but no power socket.

Any chance of some demo footage with the inverted cam.

Thanks Quinn.

There's two lithium AA's in the case. The CD motor is run at 3 volts, the 2V LED is wired in with a 12K resisitor. That's it.

I'll be doing the image testing this week so I'll post up some footage shortly. So far, the image looks very good, but I need to do some carefull checking for distortion and/or chromatic aberration. I need to spend some time with a grid and EIA1956.

Looks great Dennis, nice and slick. Could be commerical!

Here's the first footage...a quick clip playing with focus: http://z10.zupload.com/download.php?file=getfile&filepath=5095

Do you have the rods 60 apart and the lens 85 to center?
The image is good! What's the frame size and what lens did you use?

Rods are 60 to center, but I didn't sweat the 85 to center with the cam inverted...should I be? Lens was a 50mm F1.4 at 1.4, the GS400 in wide mode interlaced at F4. This was a new GG that I "blasted" tonight. What frame size are you referring to?

but I didn't sweat the 85 to center with the cam inverted...should I be? Only if you plan to use a MB (camera inverted or no).
Frame size on the GG? (18/24, 24/36mm? The GG looks OK. Post a frame (full rez) from the footage and get an even lit BG. A rez chart if you have.

the cam senses it's upside down, and rotates the 3.5" LCD image upright in normal operation.

I tried turning my GS400 upside down as well a while ago with my adapter on it... The cam doesn't sense that it's upside down and flip the image back, it doesn't know which way is up! The problem is when you turn the cam upside down you also turn your monitor (i.e. the LCD) upside down as well, so the image is still upside down when you look at it.

frame grab here: http://z16.zupload.com/download.php?file=getfile&filepath=5473

Dan, I'm assuming the frame size is 24x36 as this is a standard 35mm lense. To be honest, I'm not sure what exactly it is as I just zoom in past the vignetting. I don't have a proper EIA1956, just a vectored laser print...but I'll post a grab anyway.

What's "an even lit BG"??

Dennis,
The links dont work for me, would love to see your footage....

Ack, free bandwidth is having problems....

looks like those links are up...

Link still not working. even light on the back ground (or frame), wall, sky, etc.

I'll repost these elsewhere tonight. Dan, what's the purpose for the evenly lit footage? I had a slight mishap with my latest GG (slight contact with case flange=swirl mark in middle of frame) last night, so I may have to wait a day or two for the new blanks. I can only imagine how careful you must be with your screens! For me, it costs about $2, for you ... a lot more.

Dan, what's the purpose for the evenly lit footage?

Im guessing he is asking you to film that in order to test for any vignetting.

Looking foward to seeing your footage

That's right! Vigneting. (That's why I asked for frame size earlier as well).

I get it. I just used a 500 watt on a white wall to check myself. I had to move the achromat back a bit and extend the tube about 1 inch to allow the GS400 to zoom beyond the vignette in 16:9 and still allow focus. I can now zoom in well beyond the vignette at the expense of a 1" increase in length. Shooting 16:9 (with about 15% larger field of view on this cam) requires a bit more zoom then shooting 4:3.

There's a difference between my F1.4 50mm and the 28-130 F4.0 zoom lens in terms of the rear lens aperture and effective GG area.

Other than masking a "test" GG with a 24x36 frame, is there any other way to determine the actual effective frame area?

Dennis,
I would put a 36x24mm frame in the adapter. You say your zooming past the vignetting, but try not to zoom into a smaller area then 36x24mm, as the more you crop this size, the less benefit you get from the adapter in terms of DOF.

One "other way" is to use a real camera (mounted with the same lens) and frame something to have a reference for the framing. Top/bottom, L/R.
That is what the 24/36 shows you. Then, (same distance, same lens on adapter), see the video framing. Zoom-in and out till you get the same frame as before. Zoom-in (most likely) to avoid vigneting and "guess" the frame size (by proportions)

Good ideas fellows. I'll just use my trusty Minolta SLR body on a tripod mounted next to the adapter. I printed out a couple of sheets of paper with a 1/2 inch grid. I should be able to just set up the two cams equidistant from each grid for comparison. Those links appear to be up again...

Dennis,
I dont think your ground glass is diffusing the light enough, as it looks like your getting ghosting issues.

Check this by holding up your glass to a light bulb and check if you can see through it. If you can read the writing on the bulb then its not diffused enough. This badly efects DOF.

Wayne, are you referring to the red fringing on Tigger? If so, the cam seems to do this with or without the adapter. Apparently it's a DV compression issue. The GG is a temp one at 2mm, but it is very diffuse. It was blasted with fine media, whilst spinning 1000rpm on my "jig", at about 60PSI, about 1 metre distance.

Is there something else you're seeing?

Dennis,
On the lunch box, I can see as it goes out of focus, you are getting ghosting. This is where some aerial image is passing through the glass and not being diffused. The lines on the box should be more blurred/diffused. This ghosting gives a very hazy/smeary effect in the 'out of focus' areas

You have a good eye Wayne. So you're thinking more diffusion is required...

I'll have the new GG's in a few days, so I'll experiment a bit. Other than a visual check, are there any other methods of determining whether diffusion is sufficient? I 've seen a few POC tests using lasers shone through the material onto a backing screen. Being that I don't have a laser, have you found anything else that would give an indication of diffusion levels?

Dennis,
Just to confirm if im correct, please try holding the ground glass up to a bright bulb as i said, you should not be able to see details coming through.

Another thing I do is hold the glass about 5cm away from the green LED on my PC speakers. It should be a complete blur. If you can see the hard outlines/edges of the LED through the glass then it is not diffusing enough. You dont want any aerial image getting through.

Is the method your using simple sandblasting? If so what grit size are you using? It maybe that it just needs blasting for longer to complete the 'etching' on the acrylic surface.

I really think your on to a winner with your adapter here!!!

I'll agree. Sometimes we get people who rant and rave about their new homebuilt adapter that "just needs a few tweaks." Then the sample footage has the most vignetting you've ever seen, aberration up the wazoo and a horribly wrong focusing screen distance.

On the other end of the spectrum, there's yours, which looks really nice and clean. Thumbs up.

Ben's right, the diffussion seems to be the only issue. The image is sharp, colours are good and i see no flickering from the spinning GG, the image looks solid.

How did you go about making your acrylic disks and centring them on the motor (what type of motor)?

Thanks gents...your feedback is very encouraging. I've been tossing ideas back and forth with François Poitras over at pana3ccduser on most of the adapter components so he deserves props on this too. It's great to collaborate on a project like this both from the buying/ordering side and the problem solving side.

The GG I have in there is not the optical acrylic and is 2mm thick which I'm thinking is not ideal. I picked up few samples from a local plastics specialist. The motor is from an audio CD player, and interfaced using a 5mm prop shaft adapter. In one of the vibrating GG threads one of the posters (can't remember who) mentioned checking runout by spinning up the GG and examining the reflection of light on the unground side. This GG shows some wobble in the reflection...so I know it's not perfect. It got pretty hot during the cutting process, so it's likely warped a bit.

I'm cutting the discs pretty simply by placing the acrylic stock on a flat surface, and drilling a 5mm hole through the material and into my plywood work surface. I take the drill bit out of the drill and insert the shank into the hole, creating a spin point. I'm lazy, so I've been using a dremel clamped to an overhanging board as the cutter. The acrylic is thus rotated in place, centered by the drill bit pin, past the dremel cutter, resulting in a perfect circle. I have a router that would work a lot better, but I'm being lazy. Wayne, your drill press rotating glass cutting method got me thinking in the right direction. The optical grade acrylic blanks we have coming were pre-cut by the vendor to a .1mm accuracy so I likely won't cut my own too often.

For blasting, I'm using a very fine (100 grid I think) glass media. The acrylic disc is affixed to a disc sander surface with two sided tape. I leave the protective sheet on the non-blasted side to make sure it is not damaged. By rotating the disc at about 1000rpm while blasting, my thinking is that a consistent surface is far easier to achieve. I've been playing a bit with gun pressure and distance from the media to achieve a good finish.

I believe Bob Hart posted a blurb on pitted surfaces being better than sanded surfaces for diffusion, but I can only find one reference to a journal abstract dealing with this topic. I'd be tickled if someone had data on this.

I've read a few posts on which side the GG surface should be facing. Mine is facing the 35mm lens. Is there any theory on which is technically better?

There are arguments for both approaches, but I don't think it makes much of a difference either way.

Dennis,
Thanks for sharing your cutting method there, very similar to the SG35 build process.

The SG35's ground side of the disk faces towards the camcorder, this is to minimise internal reflection.

Dennis, since we are testing GGs with different thicknesses, I would think it is best for now to put the ground side towards the camcorder and to set the back focus/infinity like that. This way, when you replace a 1mm GG with a 1.5mm one, the lens flange-GG distance should in theory stay the same (provided the collet holding the GG does not move).

Dennis.

Ben Gurvich has my original pressed disk but I went over it again with a finer finish, so there were effectivly two grades of finish on it. In his short film he made with his AGUS version, the outdoors scene where the father answered the door was very sharp.

2mm may be a bit thick because any internal reflection may be more evident as a fringing artifact, but that is speculation on my part. I understand a similar issue called halation happens with film which is why they have the anti-halation (anti-reflection) black surface on the back of it.

I have been trying to dress split DVD+R disks and spacer disks out of DVD+R packs, which are from actual disk production. They take some starting but can be split. Once they are split, they are very thin and run true if the RPM is high enough. They tend to drift off in violent camera moves if motor speed is high

When dressing them in abrasives for glass, they seem to have some surface hardening or some treatment which resists the abrasion and kills the abrasive brew off pretty quick. You have to carefully search for the side which has the guide tracks on it and dress this surface.

It took me as long to dress a DVD+R as it did an actual glass disk.

I got a very nice result from it but damaged the disk trying to re-mount it. I have not successfully made one since.

Your blasting method might be the answer. Have you considered mixing abrasive media. Big pieces of softer media to get inertia to press very much smaller particles against the plastic.

You would need to have this arrangement inside a cabinet and some protection for the motor which will last a very short time indeed once the grit starts to fly.

Thanks for ringing in on this Bob. I've read and appreciated a lot of your posts in this forum section! I'm curious if you have any research on pitted vs sanded diffusion surfaces? Fortunately we have 1mm and 1.5mm precut optical acrylic discs coming...

Wayne and Francois, I've been mulling over the GG issue, and having the GG towards the camera makes good sense. The extra air-polymer interface after diffusion cannot be a good thing. Logically, diffused rays will be exiting the diffusion surface at all kinds of angles..which is why the fresnel works so well to immediately redirect them. If you have an air-polymer interface after that with some highly incident rays, there's bound to be some diffraction occuring. Add in a slightly wobbling GG and you're likely guaranteed ghosting.

Wayne, my research for (and subsequent design) of a 16:9 hood for my .66 WA lens tells me that your suggestion to put a 24x36 frame immediately adjacent to the GG is a good thing. There's no optical use for the extraneous areas of light in the circle surrounding the frame. It hadn't occurred to me, but it seems pretty obvious now that you've pointed it out. A film SLR shutter pretty much does the same thing.

The scene I taped is easily reproducible, so rest assured I'll be reposting a clip with the rigid 1mm disc and fine tuned diffusion surface. Your feedback here gentlemen is a very positive catalyst for thought...

Dennis.

After I did the pressed version and did a quick finer dressing in my glass machine on that later, I did not perservere with plastic as I had by then started working on glass.

Using any paper backed abrasive sheet on plastic makes a pattern of scratches. Dressing using a grit in a fluid suspension tends to make pits as the cutting action of the grit is more of a rolling action across the two surfaces rather than a dragging action.

I found it very hard in fluid suspension to get a good pit finish on plastic. There was always a scratch there somewhere. Then they started making this newer plastic which seems to be harder to dress using suspended grit.

The pressed method was simply using a 600 grade silicon carbide wet and dry paper, placing the CD-R spacer disk on a smooth clean semi soft surface like paper, then placing the wet and dry face down on it.

Then I would move a round tipped stylus shaped stick or a small alternator bearing over the wet and dry with lots of pressure on it and making sure the wet and dry did no slide on the CD-R and scratch it. The wet and dry had to be lifted off and reset again and the process repeated.

You got out of this a texture which was finer than that the actual dispersion of pieces of grit on the paper.

I tried finer paper but was unable to get good penetration into the CD-R with the method I used. Other methods might be more successful. After starting with the glass I moved on in that direction.

I did some work today on a 1 mm GG. The good news is that I've refined the "blast" procedure and am very happy with the GG surface. I did the tests you suggested Wayne...the discs are diffusing very well. With the adapter about 2 inches from a red LED (sans 35mm lens), GG off, the entire 24x36 frame is red when viewed through the cam's side of the adapter. I also did some work on the prop shaft adapter so that it spins true with the disc on it.

So in subsequent tests, I was quite disappointed to see what you had described as ghosting..until I tried the test with the GS400 alone. The same artifacts were present! I've seen this before on this cam...but didn't think to try just the cam until tonight. I believe what we're seeing here is plain old DV colour blooming. I found that as I diffused the light (the shot used a 500 watt halo..no diffusion) the artifacts were reduced. In fact, the adapter footage had less blooming than the straight cam. Grabs here:

Bare GS400 (http://www.fortvir.net/gallery/album05/vlcsnap_376266?full=1)

Adapter attached GS400 (http://www.fortvir.net/gallery/album05/vlcsnap_376203?full=1)

GG (http://www.pana3ccduser.com/attachment.php?attachmentid=1150&stc=1)


You can right click and save the image to zoom in.

Dennis,
Cool. Could you please setup a shot with an object as close to the lens as possible and the background as far away from the object as possible, then pull focus from the close object to the background. It just does not look like your getting shallow DOF, but its hard to judge from these grabs. What SLR lens are you using? Keep it up man!

Ha, I know the DOF camp has been disappointed with my grabs. Once I get outside (still snowing!) I'll do some more DOF type shots. I was more concerned with image quality and critical focus with these little grabs. The few DOF shots I've done in my editing nook are typical of what folks are seeing with these adapters. I have a few Minolta MD lenses...an F/1.4 50mm Rokkor, F/2.0 45mm Rokkor, and an F/4.0 28-130 zoom. These grabs all used the F1.4 50mm. It has a min. focus distance of about 1.5 feet....so prepare for some seriously shallow DOF.

The upside to all my anal retentive fiddling is that I've got the GG plane very close to the 43.5mm flange to GG distance so that my lens focus marks are quite accurate now. Given the difficulty focussing in some situations, I know I'll be using them. I bet if I had $.05 for every time someone came back to the editing suite with out of focus adapter footage, I wouldn't be making my own adapter ....

To be truthfull, the first time I fired all this up, I was pretty disappointed! Designing the rod system to flip the cam and finding out the LCD will never display the image correctly, initial issues with the achromat too close, the blooming etc. Zooming down the barrel to that little light at the end of the tunnel, you would never guess these things can actually generate a decent image! Whenever I get discouraged, I pull out a few of my SLR DOF shots and then I feel better....

Dennis,

OK, would be great to see the pull focus shots and shallow DOF, i think your adapter is a winner!!!

Finally, here's the clip (http://www.fortvir.net/gallery/album05/adapter) demonstrating some DOF. You may need to click the "launch in external player link" if you're using firefox...and it's a wmv, not an avi. It's indoors, at night, F/1.4 50mm. The new 1mm GG needs a bit more work as I think it's a bit light on the diffusion side...but I think it's almost done.

Wow, that is a nice grainless image!

Thanks Leo. I've redone the GG and reposted the clip (same name). This one's much better.

Dennis,
The footage looks good, very sharp and you have satisfied my questioning of the diffussion. You got to be feeling pretty good about your efforts, eh?

I guess so... I haven't used any other 35mm adapters so I don't have anything to compare to. Your advice on the diffusion factor was very helpful as I think now I know what (and what not) to look for. I'm looking forward to a more comprehensive test soon. We have about 30cm of new snow outside so things are looking pretty scenic for a few shots.

Ok, here's the res charts. The EIA1956 is a laser print, but is fine for comparison. I did the GS400, and then the "G35" (Guskers35). For my first test, I set up the G35 (50mm F/1.4 lens) with correct framing as per the LCD display. F/4.0 was selected by the camera. I then removed the adapter, set the tripod to the correct height and adjusted framing with the GS400 zoom. The camera chose F/5.6! I was a bit surprised as I expected more light loss with the 50mm on.

My lighting was from behind and to the left and this shows. I should redo these really as the cam was not the same distance from the chart in both tests (it was in the first test where I saw the F4.0 to F5.6 change but I neglected to manually WB)....either way, they were framed the same.

Frame grabs are from VLC, nothing is touched except to add the labels. How do they look?

GS400 (http://www.fortvir.net/gallery/album05/GS400?full=1)
Guskers35 (http://www.fortvir.net/gallery/album05/guskers35?full=1)

Link (http://z13.zupload.com/download.php?file=getfile&filepath=4362)
Alternate (http://www.filefactory.com/get/v3/f2.php?f=2597ae8b86ca0c402ebdb8f3)

DOF sample footage.