Home

Please support mflenses.com if you need any graphic related work order it from us, click on above banner to order!

SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

how to check infinity focus
View previous topic :: View next topic  


PostPosted: Sat Dec 26, 2009 5:45 pm    Post subject: how to check infinity focus Reply with quote

Hi all,
in the latest days I was around taking some photos with my new MIR-1 37mm f/2.8.
I'm really happy of this lens but I can't understand if the lens focusing ring or the M42 adapter allow focus to infinity.
That's what happened: I focused a subject that was about 400 meters far and when I checked the focusing ring it wasn't at the infinity end. I took the photos and checked it: the focusing was ok.
So I tried placing the focusing ring to the infinity end, I shot another one and checked it: the focusing isn't correct.
I didn't expect this kind of behavior. Is it caused by the M42 adapter? Or is it a normal (or not) behavior of this lens?

Happy holidays,
Jenner


PostPosted: Sat Dec 26, 2009 6:47 pm    Post subject: Reply with quote

At infinity focus, the lens should be at its closest to the sensor/film. If the M42 adapter is too thick, it is possible that the lens cannot reach this position due to the adapter acting as a very short extension tube. Similarly, if the M42 adapter is too thin, the lens will focus beyond infinity at the marked infinity focus position (but it can still focus at infinity properly at some other position, at the cost of some, typically small, amount of close focusing distance).

Likewise the lens itself may be adjusted to go past infinity, perhaps even deliberately (but that is more typical with long lenses where the manufacturer may intentionally leave some room for thermal variations and manufacturing tolerances), or the lens may be incorrectly set so as not to reach infinity.

In your case it seems like the lens and/or the M42 adapter cause it to focus beyond infinity. The typical way to test this is to focus at some object "at infinity" (i.e. at something several times farther than the largest marked finite distance on the lens), like you did at the 400m distance. If there's still focusing range left, your set-up goes beyond infinity. Use another lens to check whether it is the lens or the adapter that is at fault. If it's the adapter (and this really bothers you), get another one. If it's the lens, I'd probably leave it alone unless you are losing a lot of close focusing distance as a result, but of course it's good exercise in lens repair to adjust the distance.

(For some of my notes about testing M42 adapters that are too thick, i.e. not reaching infinity focus, see this picture on my flickr.)


PostPosted: Sat Dec 26, 2009 7:08 pm    Post subject: Reply with quote

Most of my adapters go beyond infinity. It's not really a problem for me as I don't use the distance scale when mounted on DSLRs.

Last edited by martinsmith99 on Sun Dec 27, 2009 4:08 pm; edited 1 time in total


PostPosted: Sat Dec 26, 2009 8:37 pm    Post subject: Reply with quote

Hi Jenner - the important thing is that what is sharp on the focusing screen is sharp on the image out of the camera. I guess the ultimate test of whether there's some kind of "fault" with the lens would be to try it on an M42 body; then the question of variation in adapters is avoided. Well, in theory it might be the ultimate test - but the body could still have a misaligned mirror or wrongly seated focusing screen - Wink


PostPosted: Sat Dec 26, 2009 9:48 pm    Post subject: Reply with quote

Take a LiveView Cam out of the store for test reasons and use the function.
Than you can really see, if the infinity is really infinity or not.

It shows you in real, without any problems, if the lens reach infinity or behind that.

Not more and not less.

Cheers
Henry


PostPosted: Sat Dec 26, 2009 9:51 pm    Post subject: Reply with quote

scsambrook wrote:
Hi Jenner - the important thing is that what is sharp on the focusing screen is sharp on the image out of the camera. I guess the ultimate test of whether there's some kind of "fault" with the lens would be to try it on an M42 body; then the question of variation in adapters is avoided. Well, in theory it might be the ultimate test - but the body could still have a misaligned mirror or wrongly seated focusing screen - Wink


Sorry, not only in theory... lots of lenses are not really aligned to infinity on the old bodys.

Each lens has his own optimized working range.. not all are developed to give best and correct results at infinity..
otherwise there would be no difference between M42 Lenses in the terms of INFINITY as we realized today as wrong. Different M42 Lenses showing different behaviors on the same adapter-ring !!

Cheers
Henry


PostPosted: Sat Dec 26, 2009 10:03 pm    Post subject: Reply with quote

Hi Jenner,

I suppose you are using a digital camera.
If you have an analog camera with a 50mm, this trick could help you checking infinity (warning: i have read this somewhere, never had the need or the analog camera for testing) ...

Ar first, cut a sheet of paper for technical drawings a bit bigger than 24x36mm and draw some lines using a very fine pencil (eventually, you can use a liner for technical drawings). Then, open the body of your analog camera and fix the drawing where should be the film, beyond the shutter.
Focus the lens on the camera to the infinity, and, leaving the back of your camera open, put a lamp to back-light the drawing.
Set the shutter to "B" and using a cable release lock the shutter, as if you were taking a picture.

Now, put your main camera with the lens to test in front of the analog one and focus to infinity (take care to the alignment of the two cameras and try to keep the two lenses as close as possible each other). If the infinity focus on your lens is correct, the drawing backlighted should appear correctly focused. You can even test if your lens focuses beyond infinity this way ...

Hope this helps ...

Cheers


PostPosted: Sat Dec 26, 2009 10:30 pm    Post subject: Reply with quote

I also found many lens what is not reach infinity exactly at infinity position even on genuine film camera bodies.


PostPosted: Sun Dec 27, 2009 12:55 am    Post subject: Reply with quote

hinnerker wrote:

Sorry, not only in theory... lots of lenses are not really aligned to infinity on the old bodys.


Henry


That's my experience as well, quite a few are in focus in the viewfinder just before the mechanical stop. Mind you my youngest 35mm camera is about 28 years old.


PostPosted: Sun Dec 27, 2009 5:47 am    Post subject: Reply with quote

I've got M42 adapters that work fine with some lenses and not at all with others. Extreme wide-angles are the most sensitive as they have a very short focusing throw between infinity and close-up compared with telephotos. If a ring is a tiny bit too thick a 20mm lens can become unable to focus further than a few feet away.


PostPosted: Sun Dec 27, 2009 6:44 am    Post subject: Reply with quote

I tested by focusing to infinity on a Helios 44-3 until it popped the mirror out of my 5D.

FYI this is, apparently, the incorrect way.


PostPosted: Sun Dec 27, 2009 9:17 am    Post subject: Reply with quote

haley wrote:
I tested by focusing to infinity on a Helios 44-3 until it popped the mirror out of my 5D.

FYI this is, apparently, the incorrect way.


LOL =)


PostPosted: Sun Dec 27, 2009 9:56 am    Post subject: Reply with quote

haley wrote:
I tested by focusing to infinity on a Helios 44-3 until it popped the mirror out of my 5D.

FYI this is, apparently, the incorrect way.


Did you want to see reverse if all parts of the sensor are alright? Very Happy

But i did have the same Problem some month ago with my Carl Zeiss HFT Planar 1.4/50mm on my old 5D MKI..

The mirror hits the back part of the lens.. looks like this after the crash.




After this malheur and the repairing of my 5D, i did try to find out on the old mirror a good method to reduce the size of the mirror.. to file down.. and this happened ... so this is the mirror glue fault, Canon fixed on some models like your 5D MKI




After seeing this result on the old mirror, i decided not to file down the new mirror in the 5D after repair.

Cheers
Henry


PostPosted: Sun Dec 27, 2009 4:10 pm    Post subject: Reply with quote

hinnerker wrote:
haley wrote:
I tested by focusing to infinity on a Helios 44-3 until it popped the mirror out of my 5D.

FYI this is, apparently, the incorrect way.


Did you want to see reverse if all parts of the sensor are alright? Very Happy

But i did have the same Problem some month ago with my Carl Zeiss HFT Planar 1.4/50mm on my old 5D MKI..

The mirror hits the back part of the lens.. looks like this after the crash.




After this malheur and the repairing of my 5D, i did try to find out on the old mirror a good method to reduce the size of the mirror.. to file down.. and this happened ... so this is the mirror glue fault, Canon fixed on some models like your 5D MKI




After seeing this result on the old mirror, i decided not to file down the new mirror in the 5D after repair.

Cheers
Henry

Shocked Shocked Shocked Shocked Shocked Shocked Shocked Shocked Shocked


PostPosted: Sun Dec 27, 2009 4:24 pm    Post subject: Reply with quote

Henry, I had the same thought process. I intended to file it down, but once it came back from Canon (free fix b/c of the glue issue), I ended up leaving it alone. I only have two lenses that hit, and I've no need to even use infinity on them, anyway.


PostPosted: Sun Dec 27, 2009 5:39 pm    Post subject: Reply with quote

Thank you all for your replies.
I'm using a Pentax K200D with a non-original M42 adapter. The adapter falls completely into the bayonet and, if I look closer, the adapter stands less than an half millimeter inside the bayonet.
I also made some more test. The latest three marks in the MIR-1 focusing ring are 10m, 20m and infinity. It seems that the infinity focus is reached by placing the focusing ring on 20m... if placed on the infinity symbol I got a blur just like if I place the focusing ring on 10m.
The good thing is that what I see in the focusing screen (I use a split-screen made by focusingscreen.com) is what the sensor records... so if I find the focus correct on the viewfinder I will find it correct on the photos.

I wasn't aware that the behavior at infinity focus could change between the lenses. Does this means that, for example, a portrait lens could not be good when focused to infinity?

If the problem is the adapter, which falls a bit inside the bayonet, could I fix it by placing a cardboard shim between the lens and the adapter?
BTW I didn't expect that a so small difference in register length could get a so big difference in the focusing ring values.

Thank you all again! Smile

EDIT: I prepared some shot to show what happens. All the photos taken at f/8 - 100% crops.

Focusing ring set at 10 meters:


Focusing ring set at 20 meters:


Focusing ring set at infinity:


(sorry for the fence, my house is close to a football field so I have 5 meters tall fences behind my windows)
The house at the bottom is about 80 meters far, the bell tower is about 380 meters far (the fence is 5 meters far).
As you can see at 20 meters I get the maximum sharpness for that distance.

Bye
Jenner


Last edited by ntx on Sun Dec 27, 2009 6:23 pm; edited 1 time in total


PostPosted: Sun Dec 27, 2009 6:10 pm    Post subject: Reply with quote

I think a shim might work.

A portrait lens is normally 80mm or 90mm, so it is much less likely to suffer badly from an adapter than a 20mm lens.

My 20mm lens extends about 1cm between close focus (a few centimeters) and infinity. My 90mm lens extends about 2cm between close focus (1 meter) and infinity.

So a 1mm mismatch on the wide angle is a 10% error in focus distance, on the 90mm it is only a 5% error. And I'm pretty sure the physics (optics) will further multipy the error on the short lens, though I can't explain why that seems clear to me. Someone here will be able to say if that is right or not.

I think long lenses are much less likely to suffer from small errors in adapter manufacture than long lenses. So a portrait lens will probably be able to cope OK with a less than perfect adapter (allowing for DoF at infinity).


ntx wrote:
Thank you all for your replies.
I'm using a Pentax K200D with a non-original M42 adapter. The adapter falls completely into the bayonet and, if I look closer, the adapter stands less than an half millimeter inside the bayonet.
I also made some more test. The latest three marks in the MIR-1 focusing ring are 10m, 20m and infinity. It seems that the infinity focus is reached by placing the focusing ring on 20m... if placed on the infinity symbol I got a blur just like if I place the focusing ring on 10m.
The good thing is that what I see in the focusing screen (I use a split-screen made by focusingscreen.com) is what the sensor records... so if I find the focus correct on the viewfinder I will find it correct on the photos.

I wasn't aware that the behavior at infinity focus could change between the lenses. Does this means that, for example, a portrait lens could not be good when focused to infinity?

If the problem is the adapter, which falls a bit inside the bayonet, could I fix it by placing a cardboard shim between the lens and the adapter?
BTW I didn't expect that a so small difference in register length could get a so big difference in the focusing ring values.

Thank you all again! Smile

Bye
Jenner


PostPosted: Sun Dec 27, 2009 7:09 pm    Post subject: Reply with quote

My experience is, the wider the lens, the minor changes in register distances in all angles taking effekt. I cognized this on my wall in front of my balcony..
i use this wall for checking the edge to edge aspect of sharpness and barrel-distortion of my test lenses.

Often i realized faults with my Leica R to EOS adapter. They are often wrong aligned. Only parts of a mm did give me horrible results in terms of edge to edges sharpness for general and wideangle lenses. The effect was, that the center sharpness was not in the center.. instead of this, i did see the sharpest part out of center more to the right.. on another Leica adapter, the sharpest part then moved in the left part... with the same lens.
Indifferent sharpness spreading from left to right.

After correct the Adapter (Leica Adapters have 2 parts, assembled and fixed with 6 screws) with on or two strips of TESA tape between the two parts, the sharpness was going to be in the center..

It means, that the wider the lens, the stronger minimum misaligns of the Adapter will take a bad effect.

Cheers
Henry


PostPosted: Sun Dec 27, 2009 7:20 pm    Post subject: Reply with quote

PaulC wrote:

So a 1mm mismatch on the wide angle is a 10% error in focus distance, on the 90mm it is only a 5% error. And I'm pretty sure the physics (optics) will further multipy the error on the short lens, though I can't explain why that seems clear to me. Someone here will be able to say if that is right or not.


The simple lens approximation (see Wikipedia) might help

1/f = 1/S1 + 1/S2

where

f = focal length of the lens
S1 = distance between the lens and the sensor/film plane
S2 = distance between the lens and subject

Using this equation we can work out what happens when the adapter is too thick

20 mm lens, excess thickness followed by "infinity" focus

0.1 mm 12.3 m
0.3 mm 4.1 m
1.0 mm 1.3 m

90 mm lens

0.1 mm 81.1 m
0.3 mm 27.1 m
1.0 mm 8.2 m

For the Mir 1 37 mm lens, "infinity" focus at 20 m suggests that the adapter is too thin by 0.07 mm.

Calculations are based on the assumption that the simple lens approximation works, and are subject to checking


PostPosted: Sat Nov 05, 2022 2:33 pm    Post subject: Wrong wrong wrong Reply with quote

hinnerker wrote:
scsambrook wrote:
Hi Jenner - the important thing is that what is sharp on the focusing screen is sharp on the image out of the camera. I guess the ultimate test of whether there's some kind of "fault" with the lens would be to try it on an M42 body; then the question of variation in adapters is avoided. Well, in theory it might be the ultimate test - but the body could still have a misaligned mirror or wrongly seated focusing screen - Wink


Sorry, not only in theory... lots of lenses are not really aligned to infinity on the old bodys.

Each lens has his own optimized working range.. not all are developed to give best and correct results at infinity..
otherwise there would be no difference between M42 Lenses in the terms of INFINITY as we realized today as wrong. Different M42 Lenses showing different behaviors on the same adapter-ring !!

Cheers
Henry

to shorten the endless discussion: I tested endlessly P67 lenses and FF-MF-lenses for infinity accuracy but on digital since 1. i dont want to shoot film now and have no P67 cam. its a complicated story which can be followed in two threads on pentaxforums. Its about the new P67-M68/P6 and Kiev 60-clamp-adapters.
We didnt know there are two versions of M68 camera-threads involved. so i mounted wrong P68 adapter with pitch 1M to pitch 0.75M camera here Kiev 60. Pentacon six also has longer thread.
so i had a macrotube-adapter. lenses reaching 10(135mm) and 15.5m (200mm). i was never sure if lenses were not deadjusted. have now found a collimate-solution(no yet done)
http://www.elekm.net/zeiss-ikon/repair/collimate/
go it from here where is also link to this discussion
https://www.talkphotography.co.uk/threads/how-to-check-infinity-focus.358707/
Had also deep discussions with novoflex mr grahl and Zeiss Mr Franz Wagner.
They both confirmed that old lenses on digital dont reach infinity. Not 500m like mentioned in Novoflex Books by Dieter Gabler. also not 2.7km but 7.8km TV tower here with sharp edges on building and nearby chapel.
I tested 200mm lenses: 1. P67 new on K+F Comfort EOS and Nikon-with niko-EOS-adapter 2. Zeiss C/Y 200/3.5
at best apertures f8 and f11 they both looked identically unsharp. of course i wasnt yet sure if it was the max out of eos 60D 18MP. then i remember another 200mm. Noflexar 3.8 lens follow focus. And yes i could focus sharper and lens is going beyond infinity. only this cannot be adjusted like on B,C and D-system. Grahl naturally confirmed this behaviour. i am still unsure if P67 are not a bit deadjusted. will do collimate-test very soon.
Now comes RAFCamera-adapters into play. He made some mistakes which cannot be discussed here. one can read in pentaxforums, FB (Kiev 60, Pentacon Six and Pentax 67). I have now a nonanodized perfect(reg. 84.95 register of P67, about 0.1mm shorter) adapter. he is about to finish it and also offering a shorter(minus 0.6mm) adapter.
Now comes another problem which i should have mentioned earlier:
Both K+F comfort adapters are too long clearly. measured also with analogue vernier caliper. adapter on camera upside down on desktop at B testing register. Nikon + 0.65 EOS 0.8mm too long!!!!!
but they must be even shorter than register. if 0.6mm are taken almost 1.6mm shorter.
2. I cannot go now deeper since i have some secrets how i can test rafcameraAdapter anyway on digital.
One must know Novoflex new P67-adapter system Set 300 euro. additional mounts: 70 euro is going a bit beyond infinity. for obvious reasons. and zeis confirmed there new lenses batis otus etc are all going beyond infinity.
analoge lenses are calculated for film which is sensitive to IR which sensor are not.
I shot roche-towers at 2.7km focus was at center of laying 8. at infinity not perfectly sharp due to too long adapter or also deadjusted lenses. will soon get 200mm and shorter P67-M68 adapter. its not possible to check focus really on GG of 6x6 Kiev 60. i found a solution how to fuse it to cam-part of another adapter. have also adapter with variable register.
So if lenses are not very sharp at real infinity. here are the reasons.


PostPosted: Sat Nov 05, 2022 8:58 pm    Post subject: Re: Wrong wrong wrong Reply with quote

europanorama wrote:
to shorten the endless discussion: I tested endlessly P67 lenses and FF-MF-lenses for infinity accuracy but on digital since 1. i dont want to shoot film now and have no P67 cam. its a complicated story which can be followed in two threads on pentaxforums. Its about the new P67-M68/P6 and Kiev 60-clamp-adapters.
We didnt know there are two versions of M68 camera-threads involved. so i mounted wrong P68 adapter with pitch 1M to pitch 0.75M camera here Kiev 60. Pentacon six also has longer thread.
so i had a macrotube-adapter. lenses reaching 10(135mm) and 15.5m (200mm). i was never sure if lenses were not deadjusted. have now found a collimate-solution(no yet done)
http://www.elekm.net/zeiss-ikon/repair/collimate/
go it from here where is also link to this discussion
https://www.talkphotography.co.uk/threads/how-to-check-infinity-focus.358707/
Had also deep discussions with novoflex mr grahl and Zeiss Mr Franz Wagner.
They both confirmed that old lenses on digital dont reach infinity. Not 500m like mentioned in Novoflex Books by Dieter Gabler. also not 2.7km but 7.8km TV tower here with sharp edges on building and nearby chapel.
I tested 200mm lenses: 1. P67 new on K+F Comfort EOS and Nikon-with niko-EOS-adapter 2. Zeiss C/Y 200/3.5
at best apertures f8 and f11 they both looked identically unsharp. of course i wasnt yet sure if it was the max out of eos 60D 18MP. then i remember another 200mm. Noflexar 3.8 lens follow focus. And yes i could focus sharper and lens is going beyond infinity. only this cannot be adjusted like on B,C and D-system. Grahl naturally confirmed this behaviour. i am still unsure if P67 are not a bit deadjusted. will do collimate-test very soon.
Now comes RAFCamera-adapters into play. He made some mistakes which cannot be discussed here. one can read in pentaxforums, FB (Kiev 60, Pentacon Six and Pentax 67). I have now a nonanodized perfect(reg. 84.95 register of P67, about 0.1mm shorter) adapter. he is about to finish it and also offering a shorter(minus 0.6mm) adapter.
Now comes another problem which i should have mentioned earlier:
Both K+F comfort adapters are too long clearly. measured also with analogue vernier caliper. adapter on camera upside down on desktop at B testing register. Nikon + 0.65 EOS 0.8mm too long!!!!!
but they must be even shorter than register. if 0.6mm are taken almost 1.6mm shorter.
2. I cannot go now deeper since i have some secrets how i can test rafcameraAdapter anyway on digital.
One must know Novoflex new P67-adapter system Set 300 euro. additional mounts: 70 euro is going a bit beyond infinity. for obvious reasons. and zeis confirmed there new lenses batis otus etc are all going beyond infinity.
analoge lenses are calculated for film which is sensitive to IR which sensor are not.
I shot roche-towers at 2.7km focus was at center of laying 8. at infinity not perfectly sharp due to too long adapter or also deadjusted lenses. will soon get 200mm and shorter P67-M68 adapter. its not possible to check focus really on GG of 6x6 Kiev 60. i found a solution how to fuse it to cam-part of another adapter. have also adapter with variable register.
So if lenses are not very sharp at real infinity. here are the reasons.


Are you sure about the measures I underlined in bold? They are huge discrepancies, even for medium format!!

For 35mm format collimation tolerance as per the manufacturer's calibration procedures are typically in the 0.02mm region (in the image space, i.e. focus register), somewhere between one and two orders of magnitude more accurate than the discrepancies you are talking about...

Also, in the film-body days, some lenses came out of the factory with the centre of the image focusing slightly beyond infinity on purpose, to compensate for residual field curvature at infinity. The service manuals for a number of specific Minolta Rokkor MC lenses clearly show that those particular models need to be calibrated to focus slightly beyond infinity (paraxial, in the centre of the image) to accommodate the field curvature (from memory, about 0.01 to 0.02mm focus register, depending on the particular model lens). Of course once you focus using the central focus aids in the viewfinder of an older Minolta film camera body that correction is negated entirely. But for e.g. astro photography where one might set a wide-angle lens at the infinity stop, that original out-of factory compensation for field curvature would be beneficial to attain a best compromise overall sharpness across the entire frame (assuming the camera itself is 100% in spec as far as register, mirror and focus screen positions are concerned).


PostPosted: Sat Nov 26, 2022 1:38 pm    Post subject: Reply with quote

on YT there are videos how to correct M42 adapter to make it shorter. by removing ring inside.
I have a variable register P67-Ni, EOS, M42 adapter made myself. with that i can not only reach infinity or go beyond but also correct lenses which are deadjusted not reaching infinity also on film.
on digital old MFocus-lenses dont reach infinity since sensor is not sensible for IR like film is and lenses are built for. zeiss confirmed. Leica confirmed infinity is at 10^km here i have 7.8km but also extraterrestrial is needed. some lenses were explicitely made for reaching beyond like noflexar 200/3.8 novoflex newest follow focus lens. old but their latest. i was always unsure if my 200mm lenses reach infinity. but now i know how to collimate. not yet done.
and: some adapters dont reach infinity e.g. K+F comfort P67-NI i can prove with image showing scale shooting at 100m target. eos-version touching 100 on scale. nikon between 100 and infinity.
with mentioned variable adapter its easy telling by how much nikon-version is too long. thread inside has size incl. pitch. pitchsize= lenght-size/rotation.