Yahoo! I got to make post number 1,000 to the lengthiest, most intense thread we've ever had here at DV Info Net.

Bob's images are up! See http://www.dvinfo.net/media/hart.

he, he...

and i'm just first in new "millenium" :))))

Brainpick time if anyone can offer advice.

I'm attempting to invert an AGUS35 image in Premiere. The image was aquired on a PD150P in DVCAM in electronic 16:9. Whilst inverting I am letterboxing the image. I am also de-interlacing the image before applying effects in an attempt to reduce the jitters in the final product.

All efforts so far with two music videos have resulted in a severe loss of resolution by comparison with the image played straight off the camcorder into a TV set.

As well as a loss of resolution I am ending up with a severe stepladder effect on straight lines which run obliquely in the image, ie., guitar frets and strings. Te image appears to be segmented with vertical divisions which distort objects as they pass across these divisions.

Vertical objects such as poles will thicken and go thin again as the frame pans across them when small adjustments have been made in the framing.

Is this a product of letterboxing from electronic 16:9. Previous letterboxing has not introduced this problem but it was not AGUS35 footage and was shot with a anamorphic 16:9 lens.

Whilst this would hint that electronic 16:9 might introduce some faults but previous work I have done with electronic 16:9 has not introduced this problem.

Any clues anyone. Advice will be greatly appreciated.

Are you deinterlacing it in Premiere or in AE ? what i do is to first deinternalce, then pre compose, and rotate de image, and then precompose again, so i can apply the filters that i want.

Also be sure to turn off the steadyshot function. That will make some images a staircase problems in some parts.

For those experimenting and/or prototyping, you may want to try (if I hadn't missed this elsewhere in the related threads) making the focus screen slightly adjustable to tilt from the top and bottom to give it the tilt/shift adapter look.

Jim,

very nice idea, but this will be very difficult with agus35, because of rotating GG, but maybe in static solution....

it looks that this is much easier to do there.

just a thought

filip

Jim.

Moving the screen with a rotating CD is a dead-easy option if you use three fixed screws as pillar bolts, springs and adjustment nuts to set backfocus and alignment by moving the motor mount plate. Just add a push button with it's own return spring (maybe an old switch with the back removed and an extension added to the shaft) so that when you press it in, it butts against the motor mount plate and pushes it in momentarily, moving the focal plane in the required direction on one side.

Agus.

I'm using Premiere 6. I have selected "always de-interlace, but have not been ssaving a pre-view or rendering out before adding the other effects. Maybe this is it. Thanks for the clues.

The difference between doing it in Premiere then extracting a sample frame and extracting an un-adjusted sample frame and then doing it in Ulead Photo Plus 4 is profound. Any problems I have had with Premiere 6 have always turned out to be my mismanagement of the software. I'm sure I am mismanaging it again but have not worked it out yet.

I'll make up a comparison .jpg and send it to Chris because that best shows what is happening.

Jim why did you have to give that away. That is part of my design. I was hoping to surprise everyone. Oh well. Anyways yes it can be done but not too well with standard 35mm lenses. You'll need to step up to a medium format or larger. Although they do make a few tilt shift lens in 35mm. The problem with using a normal 35mm lens is that when you start getting into things like rise movements you need a lens that projects a image a fair size bigger than around 36mm by 24mm. Otherwise your going to run into major fall of problems. There are also other issues as well since our GG effected by the focal point of two lens (one on each side of the GG). Great thinking though Jim. Its funny more we work on this adapters more we come full circle to how a bellows camera looks and works.

RE: Bellows camera. I think I recall the term "camera obscura" used in one of the early posts. The bellows camera was a quantum leap forward from the pinhole camera.

I have emailed a comparison image set to Chris relating to the AP6 resolution problem I have experienced. This file will be "AGUSPROB2.JPG"

Would anyone be willing to accept a snailmailed .avi file or 42 minute DVCAM tape of raw non-inverted footage of the music video to experiment upon? I have the consent of the singer- songwriter-copyright holder, to post the footage or cause it to be posted.


The spec of the footage is :-

Running time 5 mins.
Head slate and tail slate.


Vision :-

Inverted, aquired via AGUS35
Electronic 16:9.
DVCAM PAL. (Sony DSR PD150P).


Defects in vision :-

Overexposed two f'stops due to changing light.
Some density variation and flicker apparent.


Sound.

48k 16bit Stereo.


Errors in sound :-

Location sound was amplified voice, distortion in beginning, guitar is clean. Camera manual record level was kept down to about -20db due to occasional loud spots in the performances.


If somebody is willing and able to exact the cure and perhaps make it web downloadable at a respectable resolution for the benefit of constructors, this would be appreciated. I will need to check again with Chris. My understanding is he would be willing to host downloadable motion footage at www.dvinfo.net

I am beginning to accept I might be a little undercapitalised and insufficiently skilled for this exercise. By the time I get my act together and have some AGUS35 motion origination people can make valid judgements upon, excessive time may have passed.

Hey bob, i never have the chance to thank you for the constant progress of the proyect... thank you very much..

btw are you still using a rotating ground glass ? i want to make a static adapter and test which one is better.

Hello Agus.

Re: Rotating groundglass.

Yes I am sticking with what I have for the present. I don't see the rotating groundglass and fixed groundglass as competing principles rather as tools for special purposes.

For less weight and small size, the fixed glass is better for small camcorders. Otherwise you defeat the very purpose they are made for, the convenience of light weight, simple operation and mobility.

A rotating glass versus a fixed glass for the same groundglass texture will always yield a superior image so any work done to improve fixed groundglasses is going to benefit the development of rotating versions. Getting the glass to run true requires a little care but is not impossible.

There are creative things you can do with a rotating CD groundglasses like varying the polish-back to create less dense patches in the surface. If the spindle centre is set off to the left rather than above or below the projected image frame, you can come close to the effect of the vertical varying density bands one sometimes sees in projected motion film if you adjust the spindle motor speed to almost strobe with the camcorder frame rate.

Testing which version is better is as simple as turning the spindle motor off. There is little if any difference to see but if you are not chasing an aggressive depth of field effect, you can close the objective lens aperture furthur before you see any defects from the groundglass surface.

Best practice seems to be to keep the objective lens open and use neutral density filters or gels to control the light instead of the aperture or the camcorder controls. Too much light onto the groundglass seems to spoil the contrast.

Slightly off-topic info follows. DVD-Video players which have to be a little more precise have an extra little trick you won't find on a CD player spindle motor though in most respects you will find little difference in the basic setup (still three Mabuchi style miniature electric plain-bearing motors).

To stop the spindle shalt jumping inwards and outwards in a cheap plain-bearing motor (end-float) they have come up with a cheap little low friction solution.

It is a small nylon vee-pulley which has a tiny stiff thin wire spring sliding in it. It has the added benefit of imposing a small side load on the shaft as well to take up chatter of the shaft in the bearing clearance which might be enough to throw off the tracking. One of the subjects I think I mentioned in earlier posts related to this being a cause of noise.

I was looking at this thread and let me say I'm very excited. However, I JUST started looking at it an there's well over 1000 posts (a bit overwhelming). I looked through the stuff for about an hour to see if Agus ever created that web page with a tutorial, or does any1 else have one. The idea is great, but I think I need a central place with a set of instructions in order to get this thing under way, and then i can sift through this thread and get ideas on how to improve it. So, if there is one, could you please direct me to it. If it's posted somehwre in this thread I appologize, but please understand at this point this thread is so big that it's a bit overwhelming and time consuming to search for a thing like that.

Matthew: Yes, following these threads can be overwhelming. There's a lot of info to digest, as you know. I would almost reccommend reading in a back-to-front sort of manner, because there have been several improvements on the device, and you don't want to start building based on the first posts.

Here are some links to get you started:

http://ideaspora.net/agus35

http://www.metafilms.net/upload_images/agus35.gif

http://infrastructuredv.com/g35d/

Hope this helps!

RE: the music video with the Agus35, I have found that my version of Premiere will create a Netscape page which puts up a little screen which plays the video, one of my failed experiments in the steep learning curve.

The resolution is dreadful but apparently this is the case for all downloadable material. The file size is about 10 mb for an original 1 gb clip so something has to give I guess.

I will try to email this to Chris for posting here at dvinfo. In combination with stills it may give some indication of the Agus35 in practical use. The video is not complex, just a simple one camera take of the whole song outdoors.

If Chris is around, maybe he will see this message and advise me of a method of uploading an approx 10Mb Quicktime video file. It currently consists of a folder titled "export00" which contains another folder titled "upload to HTTP address" and four files which will each play at various resolutions.

They all look fairly crappy.

Is the distance from the rear lens element to the GG super cirital? I made the first part of the Agus35 today and tommorow I'm going to make the 35mm lens mount. I'll be using Canon's EF lenses. I know there is a certain distance it should be, but how ciritcal is it, cause I feel as though this could be the most challenging part (lining up the center of the 35mm lens to the ceneter of the DV lens and making sure the distance from the 35mm to the gg is the correct distance). Thnx

With a simple prime lens, you'll get away with the backfocus being a bit off but there will be problems, maybe with not being able to get infinity focus or confusing operating conditions if your lens is too close and you focus past infinity and the image softens again. If you use a zoom lens it will be pretty much useless unless you get the backfocus right.

If you just want to rough up a simple demonstrator of the principle then nothing is critical. What can be frustrating is becoming encouraged to take it to a higher useful level, then having to do some work over again when a little more initial care or time taken would have avoided it.

Establishing the back focus distance or flange to focal plane distance (same thing) is as easy as having the camera the lens came off and measuring the distance from the lens mount rear face to the focal plane. On the case of your camera somewhere there will be a little circle with a line through it. That's the focal plane.

Bob -- I will try to get a temporary upload account established for you today. If you don't hear from me, please keep bugging me about it until it's done.

Note to all -- an off-topic side discussion involving email virus issues has been deleted. Thanks,

Bob,
I got a similar email from our system administrator to my personal email. The RED FLAG, I'm the system administrator and I knew I didn't send any such email out. Our system was clean as well as my email account but the email sent to me was a virus. This is the reason I said what I did about the email possibility being a virus spoofing the network ID.

I know this is a little off topic, however I feel it is important.
FYI: I got some new information on a revised threat. It does NOT require you to open an attachment to spread.

"Alert to control the spread of PE_BAGLE.Q.
TrendLabs has received numerous infection reports of this malware spreading in Korea and Japan.

This new BAGLE variant is capable of infecting files. It propagates via email in two ways. The first is by sending emails,
which do not have an attachment. Instead it contains a link, which upon opening the email, starts a series of events that
eventually downloads this file infector into the system. The second is that the email may contain varying subjects,
message bodies, and attachment file names, just like its earlier
variants."

John

John.

Thanks for the reply. My habit with anything strange and different is never to open it but delete it. Can this thing propogate from within the in-box on being received or does it still have to be opened?

Because this machine is a W95B, no virus software seems to support it and the on-line virus screens now require Explorer 5 or better to come out and play. Mcafee stays home unless Explorer knocks on the door.

According to a nearby vendor, it seems the only way will be to extract the hard drive, make a temporary slave of it in a W98 machine and give it a scrub there.

That's my video machine. Will it go something like "TIIIIIIMMMMMBERRRRRRR" and that pretty sky blue screen with white Courier typeface show its ugliest face forever after?

That is a tough one. I like and use online scans as well as what we have on the systems. The interesting thing, with all these new attacks some pick up on them quicker than others. In fact the new viruses and worms even attach themselves to antivirus programs on the system making them ineffective without your knowledge. I have found a few this way, (scanning online with different companies).

As for your W95 drive, the easiest way I know would be to install it into an external drive bay, USB or IEEE. You could then hot plug it into a system and scan it or even scan it online. When done, pop it out of the external bay and back in your system.

Not knowing if it is infected or even what variant of a virus or worm to look for, manual removal of a virus or worm is not an option. If you know you have a particular virus or worm, you can go to one of the antivirus software companies and look up the removal procedures.

To keep things on topic, I gutted a store display which had a cd player in a kiosk. Even though my first mini35 adapter is a static, I think I will also make a rotating disc type.

John. a furthur thanks for the info.

Today I took the music video DVD to a retailer to test it on their big projection screen before I submit to a local exhibition. I have the contrast and colours a bit too crashy for a television but on the big screen, the thing looks heaps better than I thought I had any right to expect, so AGUS35 footage with careful attention to focus will stand up in that exhibition environment.

new FAQ of mini35 in FRENCH wellcome to all "french" speaker

http://mini35.lickthetoad.org/

Okay... i know for a fact that this issue has been adressed, but it's so hard to look through the oodles of pages on this thread. Plus... the posts that I did see didn't really give the direct answer I need. Okay.. I have a consumer DV camcorder (Canon ZR-45 to be exact) which has excellent macro mode (it can focus on the dust particles on my UV filter... so it's min focusing distance must be less than a cm). When I built the Agus35 i still get some pretty major vintiging (sp?). I'm zoomed almost all the way (as far i can go without the focus going out) and by the time I get there I'm so zoomed in that free hand is impossible and my FOV is so small that it makes it extremely hard to shoot in tight spaces. My question is: is there anyway i can increase my FOV (I'm using a 50mm prime) and reduce the ventiging? I'm excited that I got the image, but I can't do anything with it because the FOV is so small. Thnx.

You can use a condenser lens. If you look through the last 10 pages on this topic, you'll see it disscussed. I'm pretty sure. I'm not sure which ones have been the most successful, that's why it's best to read these posts. I know the condenser lens, goes between the gg and your camera

Ideally, get a PCX lens (flat on one side, curved on the other) with a 50mm diameter and a 50mm focal length. This will make a *huge* difference in vignetting. Also, longer lenses suffer less from vignetting, but I'd think your 50mm should be fine with a condensor.

Other advantage of a 50mm lens is it can reportedly be mounted into a 52mm filter ring, which makes for a possible mounting solution... (I've only done it with a Static adapter, not sure how it would work in the rotating GG version).

Do you know where I can buy a condenser lens? (sorry if I sound like an idiot)

Surplus Shed is pretty good: http://www.surplusshed.com/ and of course, there's edmunds - great but expensive http://www.edmundoptics.com/

These are just off the top of my head - google for optics and PCX and you ought to find some other hits.

Matthew. Here follows what may be a silly question. If it is ignore at will.

What size is the image you are picking up off the groundglass?

Some of us are electing to go for the 4:3 35mm motion picture image frame which is 24mm x 18mm according to Brett Erskine's test chart. This image frame requires care with the groundglass texture when making it, to ensure it is fine and scratch-free.

Others of us are going for the 35mm still-camera image frame which is larger. This confers a furthur depth of field benefit at possible expense of a dark corners in the image and is a little more tolerant of coarsness and blemishes on the groundglass.

Whichever of the two or any size in between you choose, I suggest you draw your preferred target frame on a piece of paper, draw some dead straight lines across and up and down then shoot some test images. (ADDED CORRECTION TO THIS MESSAGE - THE TARGET FRAME I REFER TO REPLACES THE GROUNDGLASS DISK OR SCREEN. YOU DON'T TAKE A PICTURE OF THIS THROUGH THE WHOLE AGUS35, ONLY THE CAMCORDER PLUS CLOSE-UP LENS COMBINATION. SORRY TO BE CONFUSING).

Alternatively you can print Brett's test frame at its correct size and shoot that. This will provide a benchmark to work to. If you can extract an image frame and ask Chris to post it here, we can look at it and come up with something useful.

I have this condenser lens http://www.geocities.com/buhce5/cond.html, but if I was to fit it in an aldu, I would need another one to compensate, right?


Like this )( ?

Compensate for...?

Whats the focal length of that lens? It looks like it has a pretty sharp curve. I mention it because if the focal length is shorter than it's diameter your going to have spherical distortion problems. As I understand it the focal length should be at least as long as the diameter to give you a flat field thats distortion free.

-Brett

To compensate for any distortion that this lens may produce, I thought you had to have another one identical to it facing it. Perhaps I just read wrong. It's been a while between when I was reading about them on here, and actually gettig the items together. So I must apologise. This lens is 50dia and FL is 39mm.I guess this means problems

To be perfectly honest, I don't know anything about cameras.

So please don't try and hunt me down and slap me if this sounds stupid.

What if you ran the 35mm lens image through a two way mirror?

The image would be able to pass through the fixed GG like device hit another mirror and then reflect back on the two way glass fixing itself in a clear mirrored image, that the Camcorders CCD's could pick up.

Just a random idea.

Now it's time for you guys to tell me why that wouldn't be a good idea?

Oh wait, I think I know, something about the frosted glass/plastic that only allows enough light in to create an image, or something.

Oh well.

Trevor

I'm not quite sure what you are getting at. However, aren't two way mirrors tinted?

Brett Erskine- I have found a place that'll make me a condenser to spec. So you're saying if I had a 50mmdia condenser, it's focal length would also have to be 50mm? does the back focal legnth, have to be 50mm also?

Just disregard whatever I say.

I should stick to script writing, my lack of knowledge on the subject of cameras could make you guys dummer from having heard my lame brained ideas.

I'll leave this up to the professionals.

Trevor

Louis please post or send me a link to your resource.

BErskine@CinematographerReels.com

To hopefully answer your question about back focus. Take your condenser lens (PCX) outside and point the curved side down. Raise and lower the lens untill you see the image of the sun sharp in focus. Measure the distance between the lens and the ground to find your focal length. But I'm not sure if would call this your back focal length in this situation or not.

As far as your the lens you have right now....not going to work...well. It will do the job of removing the hot spot but will have serious barrel distortion problems no matter what 35mm lens you put in front of it. What you have there is a aspherical lens which, if you read the definition of a aspherical lens, sounds like it would work great at correcting barrel distortion but apparently in simple lenses it does not. Its strange because film lens manufacturers often call their wide angle lens a aspherical lens if it can correct for barrel distortion.

Hope I didnt give you more questions than answers.

-Brett

Trevor. Don't sell yourself short. Oft times it is the straight-to-the-point offering of a solution to a problem from left field which cuts through the complicated thinking of others who might be caused to think it can't be done for this or that reason. There is apparently a lens around called the Frazier which like the bumblebee perhaps should not work in theory but does in practice. The designer marched to the beat of his own drum and this lens became a stalwart of nature documentary making.

As a scriptwriter, you will have run into the one about a classroom full of chimps and typewriters and the possibilities or otherwise of a cogent screenplay being writ - eventually. Truly, no matter at how basic a technical level of understanding any idea might be, collectivisation of all ideas in forums such as this will get us somewhere.

You are not wrong on the mirror principle and in the form of prisms is a commonly used optical solution. I've proven it by building one. The problem is the quality of the mirrors which cannot be simple wardrobe mirror glass but precision surface coated mirrors which don't have internal reflections which cause multiple images into the camcorder.

The AGUS35 evolution presents an interesting conundrum which ahs had an airing here once or twice. For about US$30 you can build the simplest one. If care is taken in the construction you can get near to professional quality results from it and it thus liberates no-budget film-makers with handicams into a whole new area of very real creative visual options.

Taking it up through incremental levels to easier usability and reliability is achievable but at what point does the whole principle become more costly than the professional level products it performs against.

An AGUS35 can be thrown together from bits and pieces from a pawnshop and computer store in about two afternoons (or even less if one becomes practiced at it), about the same time it takes to get the professional gear from the other side of the country where it can be hired at. For the low/no-budge player who needs a good now result with maybe some in-field repairs with gaffer tape, that's all it needs to be.

Then there are those of us who enjoy a technical challenge who like to do things thoroughly and properly. To this end, pushing the evolution to its highest development when it can be grabbed anytime taken out and used with confidence, is just as valid.

Brett- My resource is well known to this forum I think, it's :

http://www.knightoptical.co.uk/acatalog/LensesGlassasphericcondenser-Fire-polishedgrade.htm

At the bottom it says,"Specific requirements
We can source Glass Aspheric Condenser lenses to most sizes or focal length at short notice, so if your requirements are not listed please contact us by telephone/ fax or email: stockrequirement@knightoptical.co.uk"

Also
http://www.knightoptical.co.uk/acatalog/LensesLargecondenserlenses-Qualitygrade.htm

It also says,"Specific requirements
We can manufacture Planoconvex Lens to any size or focal length at short notice, so if your requirements are not listed please contact us by telephone / fax or email: stockrequirement@knightoptical.co.uk"

This site is handy for me as it is in the uk.

I have made another attempt to get an AGUS35 demo clip posted. If Chris receives it intact and undamaged it may appear here in the address www.dvinfo.net/media/hart

It will also be accompanied by a .pdf file which contains some info relating to the clip and appropriate copyright cautions as to the musical performance.

The clip is a severely cut-down version and is a quicktime file of about 9Mb.

The files are titled "AGUSDEMO.MOV" and "AGUSDEMO README.PDF" (DOS filenames = "AGUSDEMO.MOV" and "AGUSDE~1.PDF").

It is not a piece of inspired film-making. The performer was in shadow, the lighting was alternating light overcast and sunlight.

Through error, exposure was about three stops over and aperture variation in the zoom objective lens itself was nearly two stops. The quicktime version has ended up being subjectively about 1.5 stops darker and of considerably poorer resolution than the .avi and DVD-Video versions, but as a practical example of the AGUS35 in action it may suffice.

In future, I shall favour using two or three pieces of neutral density gel across the back of the objective lens to try to limit flare from highlights in the disk image. Managing two levels of exposure control, one manual and the other an electronic pretender to manual, plus two layers of focussing and zoom control, again one purely manual and the other electronic, gets a bit tricky and requires vigilence to make sure something has not got bumped and shifted.

Furthur to the above, the video file and .pdf read me file made it and Chris has put them up. The video file was sent in a form which has to be copied in one piece.

Hi

Good job, and it is good to see that you guys intend to address the issue of optical problems/distortions.

I've spent over 20 hours (and around 300 web pages) on the 35mm adapter topic, though I still haven't gotten to read this thread fully (a bit reading impaired), but I have some questions.

I thought I saw something, in one of the posts, about a very bright filter/screen/diffuser that worked by having a surface of microscopic lenses/beads that increase light transmission. But I can't find it again and have spent many hours looking and searching through threads. Can anybody help me?

This diffuser reminds reminded me of a coating for TV and computer screens developed over 10 years ago to. The coating would blend in the shadowmask between pixels to hide it and give a pseudo High Definition look, it would be excellent for the projection surface in an adaptor, as it is a controlled substance that would let through the maximun amount of light (compared to the spread in a ground glass).

I have been thinking about optical stuff for a while and I have these ideas. That the courser and higher the diffusing surface the wider the angle of light spread you will get, includng back into the surrounding grain, producing bleed and washout. To reduce this it needs to be finer and thinner. Instead of condensers there should be some matarial out there that will take light from a wide arc and transmit it in a smaller arc. I think the holo diffusers (yet to read up on it) does this, but the micro lense system I mentioned above could also be designed to do this (if not just projecting through the back of it instead). The lenticular lenses could be designed to do this, but being plastic the optical qualities might not be beast.

So what is the best solutions so far for the projection surface? I hear that Louis and Dino are doing adapters with holographic diffusers. How is brightness, contrast and low light ability with these compared to spinning ground glass. I have seen demo footage from different systems on the web and there is a big difference, some are much better than others, but I am uncertian where ever it is the adaptor type, or personal or camera model style? The Marlar movie is great. I can see the light dropping off to black quickly because of the spinning disk, the glimmer of the spinning grain as their arch of movement sweeps reflections outwards, the circular streaking (presumably from the disk), or the encoder), the amazing film like softness and contrast levels. But the Dino one seems more like normal video in contrast and everything, , presumable from a lack of GG grain and softening. I saw another, with a racing cart that was just bright and glary like normal video. There was a number with the soft film look. Of course they all have good DOF, but which systems are best for visual performance?

Is it possible to have a website with a breakdown list of all the solutions and parts, with the science and their plusses and minuses, and links to parts, peoples adapters and example footage?

Thanks

Wayne.

Don't forget about post - Marla was pretty heavily worked in Post. Most of the things you mention as liking about the 'film look' are probably more apt to come from post-processing rather than the adapter's qualities. Color balance and saturation and blooming of highlights being two biggies. In the collection of examples, it would be good to see unaltered footage for better comparison.

I have seen footage before (I remember ceramic tea pots) that seems unaltered but simular. I think the difference is post, but also lighting, they have sequences with center lighting and leaving the rest go to black, and there is a lot of black, a style, I think is, suited to miniDV adapters. Even the new film look in cinema, I think, is also computerised in post production, so I am expecting post production. The problem is that different cameras and lense performances are also going to effect the amount of processing needed. I take your meaning, so we really need to decide which has the best raw footage. The raw performance is what we have to work with in post. So each adapter technique has to be tested to see what it will look like on a range of cameras, could I suggest that Sony PD170 and an XL1s have significantly different picture and probably would be the best reference cameras (for miniDV). The problem is that different adapter techniques coould simulate the XLS1 softness and range, and enhance the PD170, or alternatively look better on the XL1s.

Bob, so what is your final configuration for the agus35 ? you using a plastic GG, discman motor, are you correcting the image or recording it upside down ? post a picture :)

Hello again Agus.

Re: your question above, I have not developed the Agus any furthur and won't until I receive my glass disks. It is my original "plumbers version" made from sewer pipe caps and other plastic plumbers parts. There is an image of it in www.dvinfo.net/media/hart.

I replaced the potplant saucer CD spindle motor mount, which clipped into the housing with a flexible plate style mount which is adjustable at three points for lens back-focus and tilt. There is enough compliance to enable to disc to run true but the mount is stiff plastic which does not slump like the potplant saucer material did.

I have not done any more work on the mirrors or prisms. I'll see if I get my glass disks sorted first. Except for having to invert the image in post it is quite useable the way it is.

With careful adjustment of the motor mount and re-setting the objective lens back-focus which gets thrown out, the image through the telescope lens set can be made so that there is no barrel or pincushion distortion. But in the best position, the camcorder has to be zoomed in almost to the point where its own close-up focus crashes.

There is an image from the final set-up in media/hart.

The image path for the music video and for the still titled Agusprob was :-

Objective Lens = CIMKO 28mm - 80mm zoom.

Nikon Mount = home made from plastic pipe cap.
(This mount allows some backfocus adjustment as it slides on a pipe which runs right through the case and has the camcorder bayonet mount on the other end. There is a cut-out clearance for the disk.)

Rotary groundglass. = CD-R pack clear spacer disk prepared by the pressed method, furthur dressed with aluminium oxide and partially polished back with a soft flannel cloth to improve light transmission.

Groundglass is rotated by a CD player spindle motor, simple 1.5v to 12v DC type (Mabuchi style). The motor is fixed to a compliant plastic plate which is retained by three nuts on pillar bolts and supported from beneath by stiff springs on the pillars. This arrangement provides a second back-focus, camcorder focus and tilt correction adjustment.

Power is from two x 1.5v pen cells = 3v DC in pack fixed inside the case.

The close-up/macro lens set remains the SW5042 Tasco telescope eyepiece lens set, reversed with the composite element itself reversed. This set is in a custom brass mount, home-made to screw into the 58mm filter mount of the camcorder. Distance from front element to groundglass is about 12mm.

The Agus assembly fits around the close-up/macro lens and fixes to the lens hood bayonet mount.

My chosen image frame size is the 4:3 academy motion picture frame of 24mm x 18mm. (Brett Erskine's template). There was no hot spot problem with the image frame this size with the CIMKO lens.

Camcorder settings were DVCAM with electronic 16:9
selected. Shutter speed was 50fps, the image interlaced. Camcorder exposure setting was manual, ND notch 1, f5.6 selected. There is no progressive scan on the PD150P faster than 12.5fps

Inversion, colour, gamma, brightness, contrast and letterbox was adjusted afterward. The image in the clip does not represent the raw image quality.

The camcorder is a Sony DSR PD150P.

When I get some time, I'll break down the Agus35 and photo the current assembly method.

Thanks Rob for putting up the media pages, it answered some of my questions and even answered one I was expecting to research myself. Thank goodness I did not read all the 1000 posts before before seeing the pdf's.

It is interesting that ARRI is offereing a custom ground glass making process (for those that want the softer feel). The glass will probably produce a more subtler image than the variouse holo and reflexite screens I am looking at. So now just to research the optical quality of these holo and Reflexite screens. Anybody allready know if they are highly suitable for HDV, and where ever they distort or discolour the image, I assume they don't do contrast any harm?

Great work guys, thanks and hope to see all your designs when your done.

Hi all, I experimented with static GG adapter (aldu35) but it did not work out for me even with "grainless" Bosscreen material as GG because of the visible grain.

I'm making it for JVC HD10 high def camcorder, so unfortunately (or otherwise) the cam's high resolution imposes higher requirements than the SD cams'.

WMoving on to agus35, rotating GG version.

Looking for a custom-made quality CD-shaped ground glass.

Suppliers?

Just a thought here - at HD resolution, if even the paraffin structure of the boss screen is too visible, I think you'll be bothered by some facets of the inherent 'swirling' grain structure of the rotating glass too. One of the issues with the original Agus is the fact that the imaging happens up on a 'corner' of the spinning GG - which means that the 'grain' is always moving in the same direction. Some people complained about this even at SD. I would think at HD rez, you'd want to think about alternative methods of moving the GG, so as to not have this problem. Possibly a 'vibrating' GG - or one the moves not in a spinning motion but in an elliptical path (like the new version of the P&S Mini does). Whatever you choose, I'd think you'd want the grain's movement pattern to more or less center itself around the imaging area. I've only built a static adapter, so my knowledge here is purely theoretical.

I only mention it because you seem quite serious about building a very high-quality adapter for HD.

Jonathon, I did experiment with the VERY roughly ground blank CD as rotating GG, with a lot of really bad scratches on both surfaces, and still while rotating, it looked so much better than static GG of even highest quality "grainless" Bosscreen.

I did see *some* pattern, but I thought it was because of the rough scratches on CD.

Thus I'm intending to have a quality-made GG in shape of CD and try it. I'm sure with 3-5 micron ground, there will be no more visible pattern on the rotating GG.

Do you know of any optical labs who make such small order items?

And yes I'm damn serious about making a good quality adapter :)

Will report the progress back to the group for everyone to employ it if successful.

With any moving groundglass, the groundglass texture will be "seen" by the video camcorder as a fixed groundglass if the shutter speed is not slowed down, desirably to 1/50th of a second.

This is contrary to one recommended practice of video at 1/100th of second be it right or wrong, I was advised to follow.

In none of my tests have I had any texture from the groundglass showing up except when I forget to turn it on. However where there are large density defects in the groundglass, creating areas of varying transparency, this will show as a strobing effect, especially if the disk rpm is close to the camcorder frame rate or multiples of it.

This effect can be made to emulate the variable density of motion film if the disk surface travels from bottom of the frame to top of the frame as seen in the Agus35 itself, not the camera image, in which it will travel from top to bottom.

As for glass disks, you'll have to polish them locally but Ohara in Japan are prepared to make glass disks of CD dimensions. At a rough guess, going on exchange rates, ten unfinished precision disks will cost you about US$420-00 landed in the US.

If you want the raw disks and a quote for your requirements, contact Hiro at Ohara Glass in Japan. His email is Hiro@ohara-inc.co.jp