The Canon Ee-s may just be the bees' knees, the best design, the universal screen we have all been looking for. At this point I would especially like to thank Toenis for pointing me at this screen.

I am talking about the Canon Ees replacement screen for digital SLR cameras. As far as I know it is laser cut with a very, very fine microlens structure. This does the job of a normal sandblast with one major advantage: The lens structure can be artfully designed down to the nanometer to make sure ony out of focus areas are diffused and the other rays pass through virtually unaffected (and hence nice and sharp).

As a result of the laser cutting, the grain is tiny and very uniform. Great for a static design. On the other side of the screen, a tiny fresnel structure is cut to eliminate hotspot (which worked well with the 40 mm lens I tested it with today).

Light loss is not a problem if you stay with lens with wide apertures. I have found a diagram of the light loss compared to the very bright ee-a here:
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00GrRY

The bokeh is good, I have looked at a light bulb completely out of focus and it showed as a nice and evenly lit disc with very littly diffusion around the edges.

So to summarise the attributes of this screen:
- very fine grain structure
- grain shows only in out of focus areas (microlens design)
- build in and invisible fresnel
- light loss very manageable
- and it is at $35 quite affordable (but you will need two, as you will destroy the first one while building your adapter, they are quite fragile)

Now, I would love to bombard you with pictures. Problem is, I did not manage to get the image to fill my FX1 sensor. I tried a Letus 58mm acromat (without the rest of the Letus, obviously) and I tried a Raynox DCR 250. The Raynox was almost impossible to handle, but I managed to get 1 handheld shot (I held the lens-screen construction). The attachment shows something like a 3/4 crop of the screen, vignetting comes from the Raynox lens, the unsharpness because, as it was handheld, the optical planes did not align. I the middle of the image, where the raindrops are sharp, you can get an idea of the screen's quality. The image is deinterlaced (as I said, handheld ;-) and rotated 180°, the black spots are dust.

Next I will try to build a simple static adapter with this screen for my FX1. Which brings me to the achromat question: Has anybody got a suitable achromat / close up lens suggestion that is able to enlarge a 36 by 24mm screen to cover the whole FX1 image? Has anybody got an idea how many dioptries / how much magnification I'll need?

Wayne, does your SGPro design require any more lens behind the GG apart from the achromat? If not, how much would you ask for one of your achromats?

Ok, now I hoped to inspire some of you to look more closely into this screen option and we can all do some more testing to verify the claims I made above, which are based on two days of fondling with the screen and doing most evaluation with magnifying glasses and the naked eye. So for now, you may take my conclusions with a pinch of salt.

Thanks for sharing the information. I would however like to see moving footage. The grain and dust is much more visible on moving footage than on stills. Also, what lens did you use? F-value?

Thank you for doing the test. Although here is a quote from the link you provided, it has me thinking....

"The Ee-S screen makes it easier to focus by reducing the depth of field in the focusing screen. It has nothing to do with brightness, though the effect is to make the screen darker as you stop down. For example, I see no difference in screen brightness between the two screens at f:1.4. But at f:2.8 there's a very strong difference, the Ee-S screen is noticeably darker."

Maybe the other original screen will be brighter, the Ee-A. Is there any reason you chose te Ee-S over the Ee-A?

... out of focus areas are diffused and the other rays pass through virtually unaffected (and hence nice and sharp).
This statment is very confusing.
All rays are diffused in screening glass. The only artifactial exception are rays those what construct hotspot. Theres no difference for screening glass what is focused image, what is not.
It's easy to explain with experiment. Look at Your screen at axis going through from center of the lens - You are seeing image. Move Your eye away from the center - You still see image(maybe with some screens not so bright).
Why? Not because those rays realized You about moving Your eye another place and they quickly refract themselfs to new place. Rather because all rays what fall to diffuser will splitted to many less intensive lightrays.
Now if You experienced darker image viewing at the angle then it means that every diffuser has directional pattern. That means if You enlarge Your viewing angle then You'll see less intensive rays.

- grain shows only in out of focus areas (microlens design)
This is because Your eye can't find them from area which is filled up with fine details. Out of focus areas are smooth it's easy to see there even smallest deflection.

The problem with moving footage is, that I did not finalise my adapter design yet and have currently no means of mounting screen, lens and camera properly together. Unfortunately, I am off all week (duty travel to the UK) but I will continue to build and share next weekend.

Frank, sorry for my lack of precision. What I meant to say was that the canon screen uses laser cut microlens to diffuse and not a sandblasted surface with random structures. Hence the in-focus areas are not unsharp (blurred by the random screen surface like with a diffusion filter for a camera). That's what I meant to indicate, when I said they are not diffused, which in fact they are (for the reasons you stated above).

I am aware that the test pic I have shown is insufficient to prove anything. I will try to get some footage next weekend, although it may at first be SD (I'll use my raynox converter on my little Pana MX300, that should work to make the screen cover the entire image).

Once I finished the design, I'll buy a new screen and there will be no more dust :-)

The reason I used the EE-S over the EE-A is the finer microlens structure and better bokeh.

I am aware (as mentioned in another thread) that there may be an issue with a circle appearing when the lens is stopped down. I will investigate next weekend.

All in all I just wanted to point out a really interesting screen. Maybe I am not at the stage of a finished adapter yet, but I'll promise I'll get there soon. I was just so excited that this screen combines (may combine) a good, light and sharp and almost grainless image with good bokeh.

Brought the Canon microlens screens up years ago (along with a few others) but people were interested in only talking DIY GG. Good to see progress has finally happened and the price and it is available cheap price now.

I think these screens are well worth it, they even come with a dedicated tweezers kind of tool to handle them. Which DIY GG does :-)

Anyway, it's very good that we already third indication of a good experience with those screens.

what about Canon Ee-s against Beattie?

We are only day(s) away from testing Beattie, Nikon D100 and canon EE-S and EE-A screens.
It has been confirmed by Ben Winter that a $25 Nikon D100 replacement screen is as good if not better than Beattie when used in vibrating design.

One sure thing I`ve examined is that when your lightning allows f3.5 or more the bokeh issues fade away but noise issues start to kick in.


T

Ok, first an apology for everyone waiting for more info. My laptop PC broke, hence the delay with more material.

Toenis, congratulations on your little oscillator.

So, here we go with more screenshots. With the broken PC I am currently unable to be put a video clip together.

My biggest problem is that I need a better acromat. I tried an aspherical 12 dpt bi convex lens, but it has different focal planes, as illustrated in the first picture attached (hillside view). In the centre, this screenshot also illustrates the maximum sharpness possible with the GG and the FX1, unfortunately the rest of the image is very much distorted by the lens.

As a result, I had to change to using my Letus achromat combined with a 50mm achromat I had laying around (I'll post pictures of my "gaffa tape prototype" later).

The results are almost as sharp in the centre (it was darker that time of the
day, so maybe that's the reason why they are not as sharp as the first shot). However, the vignetting and barrel distortion has decreased significantly as well as the hot spot problem - however, it's not gone :-(

I think the "Garden Bokeh" shot nicely illustrates the bokeh capabilities of the screen.

So now I am back at square one. I could buy an awfully expensive (by DIY standards!) HD achromat, such as the Brevis 35 one, or I can go on to try and error with cheap lens. Any suggestions?

PS: I still think the Canon EES screen produces an excellent picture, well usable for static design. I wish I had a ready made adapter where I could just drop the screen in to show the gorgeous quality. Unfortunately, I am not there yet.
Maybe Toenis (as I understand he has got the screen lying on his desk) can post a screenshot from his adapter with the oscillator turned off.

Kind regards,

Thomas

All shots taken with a 40mm 1.8 lens:

Maybe Toenis (as I understand he has got the screen lying on his desk) can post a screenshot from his adapter with the oscillator turned off.


That would be very helpful. Is it possible, Toenis?

Yes, it is possible to post a screenshot during this week.
We are producing a music video to be aired on local MTV and I`m quite occupied with that too.

thomas, FX1 should be able to focus that screen without vignetting/hotspotting without acromat and so on as HVX200 could almost do that.
Try different camera to screen distances and so on. EE screens have focal distance of ca 50mm I quess and camera should not be placed only 10mm away from focusing screen as that can produce some serious hotspotting.

Best regards,
T

this screen seems horrible than optosigma gg

Toenis, the focus of the fresnel on the screen is a very good point indeed! I'll try moving further away and focussing directly on the GG.

Martin, I dont have the optosigma, so I cannot compare. However, I find the screen itself (in a static setting) better than my letus screen if used static (I know it is not supposed to).

What I am intending to build is a static design to avoid moving grain issues. I may fail, but lets wait for Toenis' shots, which will be taken in a real professional environment.

Please disregard the dust for now. I'll buy a new, untouched screen once the constructing is done.

Just wanted to show some 100% frame grabs in low light at max gain.
Bokeh sample 1. (http://www.cinedof.com/test/EE_S_bokeh_max_gain.jpg)
Bokeh sample 2. (http://www.cinedof.com/test/EE_S_bokeh_max_gain2.jpg)
Bokeh sample 3. (http://www.cinedof.com/test/EE_S_bokeh_max_gain3.jpg)

Some raw wide angle fun here! (http://www.cinedof.com/helenanova/wide_solo039.avi) 19mm f4 in that low light ;)...
It`s deinterlaced and lock your player`s aspect to 4:3 if it doesn`t to that automaticly.

I hope never to use anything else than EE-S screen as it does what the name says - super precision focusing.
Since my 8" LCD died on set I was forced to focus and frame everithing on camera LCD for 4 hours and I was absolutely blown away that it was so easy with this screen plus that nice bokeh.

(Unfortunately) CINEDOF™ vibrator / noise reductor was turned on all the time.

Cheers,
T

Thanks Toenis for sharing the screengrabs.

I think you pointing at the EE-S was a great contribution to the comunity here (see Toenis' bokeh vs low light loss thread).

And your oscillator unit looks very well designed, do you take preorders for the unit yet?

I wrote enough praise for the EE-S at the beginning of the thread, so I'll cut it out here :-)

I'll continue to build on my adapter - hopefully I'll have some more (better) samples within the next few days. I've found a promising new way to align the many lenses I currently play around with.

Hopefully we can get this great value screen into an easy design most non-pro people can afford and DIY.

Toenis,

Im having trouble playing your videos. Anyone knows what could be the problem? My Windows Media Player keeps saying it encountered an unknown error and Quicktime doesnt play it.

My videos might require this XviD codec (http://www.xvidmovies.com/codec/)

T

Attached you'll find a screengrab of a new setup of lenses I tested.

Still far from the goal, but the middle of the screen displays the potential of the screen - the rest is distortion due to bad achromat setup.

I have just seen what Wayne's revised SGPro can do with the FX1 which makes me utter the frequently quoted statement: "I wish I had never started my DIY project".

It definitely adds a note of caution to further DIY experiments (and the expenses that go with it).

Thomas, there are some are seen (at least) three rays of segmential curves. They seem to have an interferencial nature. Can You explain what it is. And thank You for starting Your own DIY project :) It engages worth of many people here.

Those must be the frensel rings showing up.
One thing to do is to reverse the screen so frensel is at the another side but I`m not sure if that wouldn`t introduce some other problems.
Light spill from the sides of screen makes frensel rings visible sometimes and if those are not visible equally throughout the screen your screen might be out of alignment.
See my STATIC example here: 19mm ultrawide at f4 (www.cinedof.eu/test/EE_S_static_19mm_f4.jpg)

And some oscillating ones:
Sample 1. (www.cinedof.eu/test/EE_S_oscillating_19mm_f4_01.jpg)
Sample 2. (www.cinedof.eu/test/EE_S_oscillating_19mm_f4_02.jpg)
Sample 3. (www.cinedof.eu/test/EE_S_oscillating_19mm_f4_03.jpg)
Sample 4. (www.cinedof.eu/test/EE_S_oscillating_19mm_f4_04.jpg)
Notice that all the dirt is gone.

Short hand held clip can be downloaded here (www.cinedof.eu/test/car_max_gain_handheld.avi), it was shot at max gain unfortunately as sun was set already.
And this is almost as good as it gets (www.cinedof.eu/test/car2_handheld.avi)

I quess 19mm ultrawide doesn`t take P series filters well as you can see the vignetting (cokin`s gradual ND2 was on).

Please bear in mind that it was shot with tiny Panasonic GS150, 180deg flip and deinterlace in vDub.

Best regards,
T

We just measured EE-S screen`s light loss and it is virtually ZERO given that HVX200 didn`t change it`s aperture from f11 without or with adaptor.
HVX was in shutter priority mode and feild of view was equalized to test chart not to allow any errors. HVX registered 1/3 stop light loss only when SLR lens were stopped down to f2.8 from f2 (it changed from f11 to f8)!
From f1.2 to f2 there was no visual lightloss that HVX could register but as far as I know HVX registers light loss in 1/3 stop increasements.

Weird thing was that when tested with tiny GS150 there was 1/3 stop light loss with adaptor attached, SLR lens at f2 or f1.2 again didn`t matter to camera`s light meter.
GS150 changed from f4 without adaptor to f2.8 with adaptor when in shutter priority mode.

As 1/3th stop of light loss isn`t much I quess that Canon`s microlens technology rules big time.

Cheers,
T

Can you re-post the parts required and cost overall? Thanks!

Thanks Toenis,

seems your little oscillator is the way to go.

The fresnel pattern showing on my screengrab could be both misalignment and light spill. I am still using prototypes (early prototypes).

The Canon EE-S is, as far as I can tell (and also Toenis' tests have shown), the best ready made screen you can get.

Now, building my own adapter around it has proven far more difficult than I expected. With a smaller cam, you may get by without achromat etc. A smaller lens and sensor may well allow to focus directly on the screen. In this case a DIY setup is very easy. Just Camera -> Toenis oscillator with EE-S screen -> macro tube with FD or nikon mount -> lens

Cheap and excellent for cams like the Pana GS series

However, if you have a beautiful beast like the FX1, you will need a really good achromat. And a more complex setup. Which does require a significant investment ... and extra money to play around and experiment.

I spent roughly 150 - 200 EUR experimenting. I don't regret it and I will continue to do tests with the lens and screen I now own. But enough spent on test equipment, I may go for one of the ready made solutions you can find in this beautiful forum.

Thomas, tomorrow I`ll take our HVX200 to local Canon facility to test out their premium achromats like 250D and 500D.
HVX200 (http://www.panasonic.com/business/provideo/hvx200microsite/index.asp) has lens diametre of 82mm but my quess would be that it will still work with less powerful but premium grade achromats like 250D ($65.00).

I know that Z1/FX1, DVX100A and XL series worked without an achromatic CU lens with one frensel focusing screen oscillator at least but I just cann`t recall the settings those were using. HVX`s closest framing settings for example are zoom77 focus0 and that results in lens to focusing screen distance of ca 50mm.

Regs,
T

Thanks a lot for the info, I'll try varying the distances between screen and camera. I'll post if I find out anything new.

Kind regards,

Thomas

PS: I was not able to get the screen to fill the format without close up lens. The problem is the opening pupil at the exit point already exceeds the diameter of the screen up to ca. 30mm focal length (manual zoom). And at 30mm focal length the closest focus is aound 50 cm.

This is at aperture wide open. However, if you stop down, it may be possible.

Thomas, how does FX1 macro function work? As far as I know it`s only 1cm from camera lens to subject, can you comment on that one?

Dang, I almost forgot I had uploaded two quick samples:
Sample footage from f1.2 to f16 (http://www.cinedof.com/test/f1_to_f16_deintelaced.avi) Dirt on the screen is still not noticeable.
And OFF/ON sample with lens set at f2 and camera shutter at 250 (http://www.cinedof.com/test/shutter250_f2_on_off_deinterlaced.avi) You can see the dirt oscillating but noise is still quite organic, I quess it`s at max gain also. I was experimenting with additional direct mount options and noise reduction possibilities so it`s kinda shaky.


T

FX one macro function is at max wideangle. The exit pupil is around 55 mm in diameter, the screen has a diagonal of 43 mm. At one cm the picture hence covers roughly 60mm diagonal, screen is only 43. Around 70% coverage in full scan - plus the hotspot and the barrel distortion is really bad then.

I think you have written an excellent suggestion for an easy DIY solution here (http://www.dvinfo.net/conf/showthread.php?t=74719&page=3) , so I will not try to imitate it here.
However, I will continue experimenting with the achromats I got, maybe I can find a good setting.

I also noticed the curved lines in the screen grabs "fresnel pattern". They are also on Toenis' footage of the motor turned off.

The thing is, you cannot see these when the screen is in an SLR camera. My cameras use these same type screens with the fresnel and you cannot see the lines. So why do you think they are showing up when the screen is not vibrating?

(EDIT)
Okay I have just checked with one of my nikon screens. And it turns out that you only see these lines if your screen is in the wrong way around. I dont know how you have yours, but it must have the Fresnel side facing the SLR lens, and you view the Smooth side.

I just tested HVX200 and EE-S screen with Canon`s 58mm 250D and 500D achromats.
Both of those achtomats were small, it means with 58mm threads.
First off 500D doesn`t let you focus as close as 250D.
With 250D macro HVX was able to fill the full 35mm frame but image had little bit soft corners.
See a frame grab here. (http://www.cinedof.com/test/250D_58mm_achromat.jpg) I was holding that printed office paper sheet by hand in front of the lens so maybe some of the softness is because of that, centre to bottom right corner is quite sharp.

One more grab here. (http://www.cinedof.com/test/250D_58mm_achromat_star.jpg) Still little bit soft.
I really don`t know how those $400 achromats can stand such tests? As far as I know even the Redrock`s achromat is soft from the corners when used wide open.

One other thing to mention that with 19mm ultrawide it has some vignetting but I haven`t seen any adapter footage that hasn`t it with full 19mm coverage.

BTW, HVX was wide open and at max gain.

Best Regards,
T

One other thing to mention that with 19mm ultrawide it has some vignetting but I haven`t seen any adapter footage that hasn`t it with full 19mm coverage.

You mean Canon 19/3.5 ?

I have Tokina 19-35mm f3.5

T

One update on the focussing on the ground glass issue with the Sony FX1. With a 3 dioptries achromat I can get a full frame image of the GG at 30cm closest distance. This is fully zoomed in (focal length of 54mm).
The hotspot is gone ... but unfortunately this distance is completely impractical for building a short setup, which was my intention. Even if I fold the path of light with mirrors twice to rotate the image, it still is too long.

One unfortunate thing with EE-S screen is that it has some kind of microlens pattern showing up in bokeh and lens flare areas fom time to time.

>> See the lens flare sample here (http://www.cinedof.eu/test/EE_S_microlens_lens_flare.jpg) <<

I quess it`s because this screen doesn`t have random microlens construction as EE-A does. I quess most of us doesn`t shoot straight up to sun tough.

>> See how bokeh acts here (http://www.cinedof.eu/test/sauna_bokeh_f2_8_deinterlaced.avi) <<

Best regards,
T

Yeah, there is drawn out that microlens arrays act as any other GG. You can experience such thing with optosigma and nikon screens as well but just pattern is random kind as there is random grinding done.

Movie with out of focus image isn't that bad.

I just wonder Toenis, did You use there Your biaxial vibra system?

Frank, Yes, I did use my vibra system there but forgot to clean it as you can see...

T

what is approximate frequency You use there?

Frank, It`s somewhere between 60-100Hz.

T

OK, I have some quite good static EE-S screen SD samples here for evaluation: http://www.dvinfo.net/conf/showthread.php?t=78492

T

Just wanted to show the worst case scenarios with static EE-S screen (www.cinedof.com/static/worst_cases_static_ees.wmv) taken with one of the cheapest manual camcorders from Panasonic.

Low light bokeh in metro station, back lighted subject with direct sunlight to lens and low sun shining into the lens from the side. All taken with camcorder too close to the focusing screen and 28mm f2 lens without lens hood. Added compression to all that and this is how worse it became. Not too bad.

When camcorder lens was at widest angle fully open and gain at max setting the footage became softer as as seen in the metro station part.

One aspect that needs fruther study is that vibrating EE-S matte side facing the SLR lens isn`t that soft at the same settings.

Cheers,
T

It has been confirmed by Ben Winter that a $25 Nikon D100 replacement screen is as good if not better than Beattie when used in vibrating design.
It should be noted that through my experiences, these focusing screens have absolutely horrendous bokeh, and anyone desiring anything resembling a normal-looking image should avoid the D100 and Beattie.