Kit,
This is a great test and an important issue to address here. Thanks for taking your time to do this test and report.
Kit wrote, "The microprism allows just as much accuracy as the diagonal prism, IMO".
From your statement that I quoted you above, I find that in general folks who are very good at manual focus will have no problem using the the mircroprism mechanism in aiding their focus and you certainly have keen eyes. The split screen is also very popular and are well received by folks who are learning how to manual focus. However, I find that each of these screen are important to use under certain condition of light and the type of lens focal length. For me, I like to use the split screen when the light is bright or for longer focal length lens. While I like to use the micro-prism when the light is poor or for wide angle lens. If the ratio of the microprism collar to the whole screen is small and in poor light, I even prefer to use simple bright matte screen which offer me more surface area in poor light condition.
In the pass I find that Contax RTS 45 degree split screen combined with microprism collar to be a very advantage focusing screen to use. In fact, when mount this focus screen on the Canon DSLR full frame camera I find that the focusing screen is slighter thinner than the original Canon focusing screen. Thus, when you focus and the image look in focus but the film plane does not see the same thing and the result is out of focus. I think one of the way that you can judge whether a third party focus screen will work on your Canon Camera, before you actually test out the third party focusing screen, is to measure the thickness of the Canon focusing screen and the third party focusing screen. If they are similar in thickness then your chance will be good.
Son; I was actually thinking of measuring the thickness of the Brightscreen and compare it to the Canon. I can measure it down to 1/10th of a millimeters accuracy.
Kit- how close is the camera to that point you are focusing on? Am I correct in assumming that the closer you are, any focus errors will show up better?
John; when focusing on a ruler correctly and close up you will see if you focus is spot on or if it is in front or back. The focusing error is greater at longer distances, especially if you use telephoto lenses at medium distances where a minute twist on the barrel might throw it out of focus a lot.
In my case it was off by half an inch close up with a 90/2.8 and when focused a few metres away the error was proportionally greater and with the 90mm unbearable. On a wideangle the depth of field might cover it for the most part.
Update:
Brightscreen has kindly offered to send me a replacement screen if I don't get it working properly. I hope to be able to work it out though.
I will be testing with another lens, no adapters involved and possibly another body as well before taking the trouble to have it replaced.
John; that too. Close focus is always critical and I sometimes bracket focus for close ups. As depth of field always extends further beyond point of focus than in front of I usually bracket from slightly in front of to dead on. But with most lenses / objects I don't have to bracket.
On medium distances a telephoto lens can cause internal bleeding from popping a vein if your focusing screen is not spot on - a slight twist can mean 8-10 inches off (ok, I made up that number, it can be both more and less) focus and if it ends up behind the subject you are smoked.
Guy; I am watching those tabs, for sure.
I won't file anything down as I am in conversation with Brightscreen on the subject and if I don't get it right I'll start with a replacement. If there is no other way I'll start grinding it ;-).
(still looking for those lens test compendiums for you, but things are kinda closed down for the long weekend, so bare with me)
JJL and Henrik: on the second set of images, I was ~3 feet away from the rule. For the joke one (on bellows), about 2"! The closer you are to the subject, the shallower the DOF for any focal length and ƒ stop.
So, as I understand it, a particular focussing error will be easier to see in the final image the closer you are to the subject. Henrik, I understand what you mean with teles, but unless you pick your subject very carefully (the reason Guy recommended shooting a textured wall at an angle in an early post re. DMRs and focus screens) is that if it's off, you can see where it's off, and by how much. In a macro shot, it's easy to see; in a tele shot the same error (focus-wise) will produce a proportionally much larger shift in the zone of acceptable focus—but IMHO, the error is the same. I am in a bit of a rush now and can't go into this in more detail, but will take the laptop away with me and (hopefully) will be able to check in.
As well, I think Guy might be right re. tabs, and effective seating of the screen and so on. Worth looking at closely.
Finally, by my measurements just now, the Brightscreen is 0.005" thinner than both the standard matte screen (Ee-S) or the original screen by my vernier ecalipers. I will check with the factory to see what their take is on this.
When I was doing 4x5 work, replacement screens were often vastly different thicknesses, but there was some rule about finding the point of the screen 1/3 the way in from the lens side, and finding the 1/3 point on the new screen and making sure that point sat at the same 1/3 point of the original screen. Or somewthing like that.
5 thousandths of an inch thinner is much much less than the difference on any 4x5 screens I ever compared. I don't know how one could properly register and shim a replacement screen on a 35mm camera.
p.2 #10 · Microprism-only Brightscreen on Canon 5D
A thousandth of an inch? Eeeh, that ain't a lot or is it? I was gonna check the sides and tabs and see if anything is catching on the frame or differs from the Canon screen in that regard.
But a thickness variation of a thousandth of an inch seems within the mark.
About focusing errors; yes Kit, you are right - if it is off the mark it is off the mark.
What I meant is that in normal shooting conditions with say a 50mm or a wide you can sometimes have a hard time telling it is out of whack - it can be acceptably sharp, while still it is off the mark.
If you install a screen and it doesn't go in right - like the one I thought was installed properly - and you go out and shoot things and you wonder what the hell is going on with your aim when you shoot portraits with a tele and with the focus ending up on the ears or that end of the field behind ;-).
I thought I got this rare morning shot the other day with some ferries meeting in the morning mist and with three swans in flight inbetween them, all against a nice backdrop. That's when I found out it was out of whack; the pictures were "soft" with the focus slightly behind the ferries. With some strategic sharpening it looked almost ok, as it was just out of whack but still - I lost a great shot.
Bummer.
With this tale of misfortune I leave yee and bid ye good night.
Jan 05, 2006 at 07:31 PM
jjlphoto Offline • • • • • [X]
p.2 #11 · Microprism-only Brightscreen on Canon 5D
A Post-It note is .004" thick. Yeah- taking 1/3 of that .005" is .00015". Even a human hair is over .001 thick! Yikes. I guessing that properly seating the screen in the frame is the only issue that really matters here.
p.2 #12 · Microprism-only Brightscreen on Canon 5D
Just checked with the factory:
if the screen was not at the proper film plane the split images and/or microprisms would not register clear in the auto focus mode, just as the auto focus will confirm focus in the manual focusing mode.
A slightly less than factory screen thickness will not matter (with proper upward tension) and a slightly thicker screen would not matter either, so long as it fits to be pressed upwards to the flat focal plane.
The thickness of the screen does not factor into the focal plane as it is on the top (upward) position and is pressed upwards by the tray (screen holder) design into place with reasonable tolerance and space. ...Show more →
So—the main points here are that the focussing parts of the screen are on the top of the screen, and it needs to press closely to the glass layer that sits just under the prism.
Henrik: please make sure that when you latch the screen holder, the screen is pressed equally upwards by the two fine 'leaf' springs (on either short side of the carrier). I needed to press the carrrier frame quite firmly upwards to seat me screen; may I suggest that you examine this aspect and make absolutely sure that the top surface of the screen is seated home properly. That was the original problem that I had with my first Brightscreen.
p.2 #13 · Microprism-only Brightscreen on Canon 5D
Thanks Kit, I will do that today. I will also check the side tabs on the Brightscreen as one of the tabs was actually irregular in shape, a bit pointy, and not flat like the other tabs.
Still, I wonder how much trouble it should be - the Canon screen hits the ruler test dead on all the time.
p.2 #14 · Microprism-only Brightscreen on Canon 5D
Henrik,
If it's any consolation I've had similar mis-focussing problems with the Canon Ee-S screen for weeks, and it's driving me crazy!
The original Canon screen focuses spot-on every time, but the S is always about 3-4cm behind what appears in the viewfinder (using Zeiss 85mm f1.2 or 135mm f2 at minimum focus distance). I've tried 3 different Ee-S screens, all gave the same result.
Just to confuse the issue though, when I manually focus my Canon 300mm f2.8L IS at minimum distance, it's spot-on accurate with both screens.
I've spoken at length with Canon tech guys, and no answer could be found (well, apart from 'maybe you should buy the Canon 85/1.2 and 135/2' ha ha ha).
I don't really understand how a lens or adapter problem could be responsible though, and would welcome any explanation of why this might be the cause.
Surely if an image is focussed on the screen it should be focussed on the sensor, irrespective of lens/adapter, as long as the screen is correct.
p.2 #15 · Microprism-only Brightscreen on Canon 5D
deshojo,
does the autofocus confirmation work with the Ee-S screen and a CanonAS lens? If so, as I understand it, the camera is finding focus.
With my screens, and Canon AF lenses, if the AF confirmation lights (I have the centre one only programmed to lght) that point defeinitely is in focus—screen and sensor (and processed image) confirm this. I admit I cannot understand your version of the problem at all.
If you had not been able to get accurate focus with any lens, then we would have to consider whether the screen-to-flage distance is differenct to the sensor-to-flange distance (i.e., that the geometry of the carrier is incorrect or bent).
p.2 #16 · Microprism-only Brightscreen on Canon 5D
deshojo: it is much like my problem I'd have to say.
Also, I find it very very strange to blame a focusing error on an adapter. If you can focus on it on the screen it should be sharp at point of focus. Period. If the original screen works flawlessly every time you put it back on, you can't really blame the camera, the adapter, the original screen or the screen carrier. If there was a discrepancy between the changes back and forth there would be cause for suspicion. But if not, I'd put the blame on the third party screen.
If the screen carrier had been warped or bent surely it would have shown with the original screen as well.
Also, if I test several lenses, with different adapters and with original EF lenses as well and the results are the same I'd say it's a closed case - I have changed screens on at least eight or nine cameras over the years (preferred the prism only before) and I have never run into a change that wouldn't succeed in at the most two tries.
Example; I have a correct focusing screen and make my own adapter to adapt an old lens, which turns out to thick, which means I loose infinity focus.
However, at all distances up to the cut off I should still be able to focus the lens properly.
I once used Mamiya MF lenses on my Nikon F90 and FE2 and while they didn't focus on infinity they worked well up to the cut off.
After todays rigorous try out I have given up on the Brightscreen and I am wondering if it is slightly warped or so. In any case it is going back. I mean, I am no klutz and if it takes thirty tries to seat a third party screen and I'd have to do it all over again when it gets dirty and needs a bit of a clean up. Arrrgh... ...I am popping a vein. Two hours ago, I gave up, popped the old screen in and - presto! Accurate as can be... Sigh.
Thanks for all the help, much appreciated!
;-) Henrik
p.2 #17 · Microprism-only Brightscreen on Canon 5D
Kit,
Yes, the focus confirmation does light up on the selected AF point with the Canon lenses (300/2.8 and 70-200/2.8).
It is a very perplexing problem as other Zeiss lenses using the same adapter are ok. I have the 70-210/3.5 and 100-300/4.5, and both have a fairly narrow DOF at closest focus, and both focus accurately.
It just seems to be a problem with the 85/1.2 and 135/2 (though they do have the narrowest DOF).
As you say, if it was all lenses it would likely be a difference between focus screen and sensor distance. If it was just the Ee-S screen it would likely be the screen itself. But the fact that the standard screen is fine with everything, and the S screen is fine with all but those 2 lenses is really bizarre.
It is focussed a consistent distance (about 3cm) behind what's shown in the viewfinder with all 3 Ee-s screen tried. Yet every time I replace the standard screen it's right again.
I've been into photography for 27 years and have never been confused by anything like this before. I really am determined to get to the bottom of it, but I'm just glad that the standard screen is fine for everything!
I shall see what the Katzeye screen does when that comes out, as their 20D one worked superbly for me, but anyone's input on this problem would be appreciated.
p.2 #18 · Microprism-only Brightscreen on Canon 5D
Henrik,
You posted while I was writing, but I have to say you're absolutely right, it has to be a flaw with the screen, I can't see any other reason.
I can only think with mine that the DOF is so narrow with the 2 problem lenses that they show it up when others don't.
Your mention of a warped screen has given me the idea to maybe try focussing on a spot away from the centre. Perhaps if it focussed correctly at the edge of the frame this would prove a screen defect. I'll try that tomorrow (it's too late for that now..)
p.2 #19 · Microprism-only Brightscreen on Canon 5D
Deshojo: on my 90/2.8 at 2 metres away the focus ends up a few cms behind where I focused.
The odd thing (and does this ring a bell?) is that I thought I got it right one of the last attempts with the Brightscreen when focusing close up on the ruler - not using the split, I used the prism and the screen. But when I focused on a flat subject two metres away it was off.
With the standard screen of course I have neither prism or split but regardless of where I place the focus on the screen it ends up right.
p.2 #20 · Microprism-only Brightscreen on Canon 5D
Update:
Over the last week I have been in conversation with Brightscreen over the email to try and sort it out. They have handled it exemplary so far with follow-ups on the issue without me having to chase them down. After my second email to Brightscreen I did some final tests and when I realised I wasn't going to get it right I sat myself down to write them for an exchange or return.
But they beat me to it and I had already got an email from Brightscreen saying that a replacement screen is on the way, free of charge, and that they will contact me to arrange pick up and return on the first screen.
This is simply first class customer policy; just the way you want it. Hats off to that!
