Hi,
i just got my 5D back from Son, and i replaced the focusing screen with Hi-Lux screen.
The screen looks very nice and bright. focused area really pops up.
However, even after many trials of adjusting the diopter, the sharpness of view finder does not match the actual output.
What I mean is...when i see the sharpest focused target, I press the shutter, but the actual photo is out of focus.
when I trust/rely on AF confirmation, the picture seems better. But what I see on the viewfinder when beep goes on is not sharp at all.

What am I doing wrong here?
Is it possible that I may have install the screen either upside down or flipped over?

Upside down may do this.
Although iirc, when it once happened on my 30D it was too obvious to even wonder and post on the forums, such a slight difference you describe i had to deal with when i tried to shim my Haoda screen for my 30D (w/o success) a little too much shim or too little shim and i front or back focused.
Do you notice this with your original screen?

If the focusing screen is installed correctly and the autofocusing works "better"; the other possibility could be the mirror is mis-aligned somehow. Manual focusing relies on the reflection of the image from the mirror and with autofocusing, the sensor relies on the pass-through image of the subject; there could be a slight difference between the two as to the effectiveness/accuracy of focusing when the mirror is mis-aligned.

Few things you can check:
Put the original Ee-A screen back, do you have focus issue?
Check the Ee-S screen, did you installed the wrong side?
When things are in focus with AF, can you actually see it focused on the viewfinder?
I never had issue with my focus screen. I shaved the mirror myself so I don't know much about mirror mis-aligned issue.

I tried everything; 3rd party focusing screens, magnifiers, focus confirmation adapters...did I miss anything? Well, long story short, I gave up trying to do MF with anything except WA stopped down; everything else came in OOF. I don't know why, it must be something about the body, but at any rate I would have to say you are not alone. I hope you solve it. I gave up and sold all my beautiful CZ glass (right, I could have kept the 21mm).

There have been a number of threads on the issue, but the 5D is infamous for not tolerating ANY focus screen that it didn't ship with. The prism box or whatever you want to label it has wide tolerances.

Every 5D when shipped from the factory has it's focus screen is custom-shimmed to illustrate accurate focus when appropriate. In some 5Ds (mine was one of them) swapping out that factory screen for ANY focus screen even one from Canon (no, I'm not kidding ) results in an inaccurate indication of focus. Mine refused to indicate accurate focus with the Canon Hi Lux screen.

The only solutions are (assuming you want to keep the 5D, I didn't)

1. Ship the body to Canon with the Canon screen you want installed and request they shim it for proper focus indication. Obviously it MUST be a Canon screen. Also, every time you want to change screens, you'll need to follow this route.

2. Try it yourself with a shimming material like metallic duct tape (consistently 0.001" thick or some such). If the camera back focuses vs. what you think is in focus - add shims vs. current set up. If it front focuses - remove shims or replace with thinner ones.

3. Go back to the stock screen.

Like the mirror box issue, it varies by 5D. Yours could be the 5D from hell re: FC or mirror, your buddy's could be a poster child. The 5D has a nice sensor, but it's a sloppy camera re: manufacturing/moulding tolerances. They had to save $$ somewhere.

What I learned from here for 5D is: Once you find a camera which can MF correctly, don't change the focus screen, then you will be fine". For DMR, I remember Leica R9/DMR had the same issue

If you are using split image, getting sharp output - but with the lower part of the split showing a shift to the left means that you will have add shims or to the right subtract shims.

Recently fitted a split and microprism ring screen to my 5D and it took quite a while with various thickness of shim (I used narrow strips of tape on the vertical ends of the screen) to get MF, AF and split image to agree with one another. Possibly I was just lucky with this particular 5D.

Slightly OT: How are people finding finding AF confirm adapters compare to split image/microprism screens?

You can order focusing screen shims from Canon parts in pretty much any thickness for about a buck each...buy one of each and try them until it focuses correctly.

Kevin - good input. This is always a controversial topic (for any bored EOS 'experts' out there - I'm not going to get into a debate about it and get the OPs thread off track), but I've found that a FC adpater, when chipped for the exact lens to be used (e.g. 100mm f2.0) gives more accurate FC than a generic chipped unit. Also tends to meter better. Unlike Nikon, EOS cameras require contacts on the lens to indicate focus which implies it uses the encoded lens data in it's FC circuitry to some extent.

I usually use a split screen to check my adapters. If the split shows exact focus when the adapter does, bon. As but one example: My 100 Zf 2 using a custom coded adapter (from happypagehk on ebay) showed proper 'aligned' focus. When the same lens was used with a Haoda generic chipped adapter (generic meaning are typically encoded 50mmf2.0 or F1.8), focus confirm was way off proper focus. Been thru dozens of lenses and the same number of adapters from various mfg's, including generics from HPHK and a custom encoded always works better.

If you're willing to play with shims of some form yourself, getting your 5D tuned is perfectly possible. Using tape as shims works better if you switch screens because the shims are always attached to the FC and you know it will stay calibrated.

Ahhh..this is so much headache.
anyway, I'm going to try to install the screen flip over, and see what happens.

Conner999 wrote:
...but I've found that a FC adpater, when chipped for the exact lens to be used (e.g. 100mm f2.0) gives more accurate FC than a generic chipped unit. Also tends to meter better. Unlike Nikon, EOS cameras require contacts on the lens to indicate focus which implies it uses the encoded lens data in it's FC circuitry to some extent.


Thanks for the info,
Partly guessed it wouldn't be as simple as just getting an adapter with a generic chip attached. Haven't tried one yet - but I can see what you are getting at....

Apologies to Paul for wandering a bit off topic.

Kevin

WOW - who would have guessed it would be so hard

who would have guessed it would be so hard

It isn't, in my experience. The screen either needs to be move closer to the pentaprism or further away. It is as simple as that.

The difficult part, for me, is how to understand that this is so difficult.

With the following dSLR bodies, I have had no unsolvable issues getting proper manual focus:

300D x2
350D x2
10D x2
20D
1D
5D

So, if the 5D is so shoddy in its mirror box tolerances, imagine how difficult it was for me to get those dRebels shooting right. Sorry to be sarcastic here guys, but this is a mountain made from a mole hill if I've ever seen one.

If you are using matte screen, adjusting the diopter would help. If you are using split/prism screen shimming would do ( but painful).

From my experience, different lens behaves differently under the same screen ( wide open - so focus shift should not be a problem).

I am still puzzled with the WYSIWYG SLR focusing....

bluelight wrote:
If you are using matte screen, adjusting the diopter would help. If you are using split/prism screen shimming would do ( but painful).

From my experience, different lens behaves differently under the same screen ( wide open - so focus shift should not be a problem).

I am still puzzled with the WYSIWYG SLR focusing....



Mine is hi-lux matte screen.
And I tried so many times already with diopter.
I also tried with Planar 100/2.0, 35-70/3.4, and the problem is same.
My only option right now, it seems, is to flip it over and try it again.
Did anyone solved the problem this way?


Also, I'm curious that when you install a screen for the first time, how does anyone know which side is a correct side of the screen to begin with?

The surface of the s screen should be dull on one side and shiny on the other. The dull side faces the prism and the shiny side faces the mirror.

As Kevin said, shiny side towards mirror. A can of compressed air, etc is always handy to give the screen a quick squirt BEFORE it goes in. Careful, they scratch easy as well.

Don't use compressed air directed up towards the pentaprism - it can drive dirt, etc., into the viewfinder area beyond your access to clean. A gentle puff with a Rocket Blower of a quick pass with a Visible Dust brush will remove any dust without resorting to canned air towards the pentaprism 'box'.

Of all the bodies I've owned (5d, 30D, 1Ds, 1D2, 1Ds2, couple Nikons) the 5D is the only one that's give me any issue re: focus screen - and mine had NO, I mean NO mirror box issues (thankfully).

The shim, test shoot, shim again, test shoot, etc., process is a PITA, and should you have to do it with a Canon (vs Katzeye, etc) screen? Hell, no. That said, once you get it dialed in, your good.

Kevin M wrote:
The surface of the s screen should be dull on one side and shiny on the other. The dull side faces the prism and the shiny side faces the mirror.


Thanks!
That helps me a lot.

Taking pictures is my hobby. (I mean just pressing the shutter button)
That's all I knew.
Who would thought that I'll be digging inside a camera, changing focusing screen and finding out these informations.
On top of that, I'm seriously thinking of trying to modify the mount my 58/1.2 rokkor.
I've come a long way since I got into manual focusing ( and that was just last month).
I'm not sure whether I should be thanking you guys or not......

Stuff like this gets easier with experience mixing/matching lenses and bodies.

Not much point having these forums if you can't ask questions and get civil answers from people who've (likely) been there already. Over time you'll be the one answering more and more questions - just remember you were there once yourself.

You can thanks folks -your wallet/wife/girlfriend might not be so charitable however ;>

cogitech: which brand screen do you use for all of those cams? Aside from Ee-S for 5D?

Daniel Goller wrote:
cogitech: which brand screen do you use for all of those cams? Aside from Ee-S for 5D?

Well, it has been a while (only have a 10D, 20D and 5D, now), but I basically used Haoda and other random HK or Chinese stuff from e-bay. As cheap as I could get it, as usual. Once shimmed right, they all ended up working fine.

Interesting you would say haoda, since i had no luck with mine. (I still have it packed up over here somewhere.)
what amterials did you use to make shims?
Even the smallest differences in material thickness seemed to throw me from back to front focusing or vice versa.
What's even funnier is that this haoda was designed to be shimmed, unlike the older ones that were designed to be installed w/o any shims.

Doesn't really matter now, i have a big bright viewfinder that allows me to nail focus w/o any split prism, and i can join you in smiling at those who think they can't manually focus ;P

Kevin M: Oh, focus confirm adapters work prety good if you don't have a VF to judge focus from, just keep your eye on the actual green confirmation dot, as it's a good indication if you already passed proper focus or not. (I say passed since you may not stop in time after hearing the beep/seeing the red indicator flash, but the green focus confirmation is what you really want to look at)

What do you know....
I reinstalled the screen as soon as I got back home, and focusing almost matches the AF confirmation beep.
I think I didn't press the screen holder hard enough last time.
what a relief......
I'm going to try this at a daylight just to make sure.

I guess Paul's suggestion fitted my situation after all.

Thank you guys....