How do I test a lens for focus?

[atlcruiser]

Hi all,
Been having some issues with some lenses. I was shooting velvia 100 the other day and when I got it back i noticed things just were not that sharp on either roll. Items were in focus but not sharp. One lens is a 28/2.8 elmart and the other is a 90/2 'cron. The camera is an M6. With B+W i seem to get a bit sharper but difficult to tell with the greyscale.

Exposure does not seem to matter and f stop makes a difference but not that great of a difference.

I look at the slides on a light table and they are OK but I seem to lose some of the sharpness when I scan and see them larger. I thought scanner was an issue but it has always worked correctly in the past and this issue is new.

Part of the frustration is that I have been shooting MF and getting razor sharp results...difficult to compare to 35mm.

I thoguht a simple lens test would be a good starting point.... just needome direction 🙂

thanks
david

 

[Sonnar Brian]

If the focus is off, "usually" the RF of the camera needs to be set.

for a focus test, I've used a fence post or a ruler/tape measure. I would suggest the latter. focus on a point on the ruler, and see if the best focus is in front or behind the spot you focused the RF on. Try multiple lenses. If all of the lenses are off by about the same amount, the RF of the camera is off. If only one lens is off, and the majority agree with the camera- the lens needs to be shimmed.

With the lenses you are using, focus shift caused by changing F-Stop should not be a big contribution to the error. With a Sonnar lens, it is noticeable. SO: I would suggest shooting one series of shots wide-open, and another at F4. With a 5cm f1.5 Sonnar, or Jupiter-3, the shift can be seen and is about 1inch or so at 3ft.

(and... when your M8 gets home from the hospital... that is the easiest way to check, pixel-peeping)

 

[J. Borger]

Where your 35mm scans alright in the past??
Did those lenses perform well on the M8??
If you are comparing your 35mm scans with MF scans or M8 files you will be disappointed, especialy since i read "a real cheap scanner" in your signature and you state the slides look sharp on your lighttable.

I bet it is a scanner issue.

 

[atlcruiser]

The scanner is cannon 8800F...I guess it is cheap but not crap!

Scans were good in the past.

Both lenses did well on the M8 but I cant really dbl check as the M8 is away for a while.

It is amsot as if the velvia is the issue...makes no sense as whn I over/under expose I still get the muddied look.

Stuff is in focus just not sharp. I am not sure that makes sense at all!

 

[Sonnar Brian]

I would try another roll of film.

Off-hand, it sounds like the film or its development could be an issue.

 

[atlcruiser]

Would out of date velvia cause this? I am confident on development as the same lab, same day did some 120 provia for me that was perfect.

the velvia is from 9/09 so it is really not that old.

 

[elmer3.5]

Ruler

Ruler

Hi, always use a 15 ft meter, so i stretch it and then i put some clear standing up elements at 1 mtr, 3 mtr and so.
Then i scan negatives.
It´s very useful use nore than one lens.
To get proper rf alignment i wait until it´s dark, and use an antenna far away from my house to check infinity if i can i use the moon as reference.

Bye and luck!

 

[Sonnar Brian]

I end up using this fence post for most of my tests. It gets me outdoors, and gets a "bokeh" test in at the same time. The slats of the fence give a depth to the image that is easy to measure. I end up working on a lot of lenses, so the grain of the wood also gives a comparison for sharpness of the lens.

'56 ZOMZ J-3, at F1.5:

100% Crop:

I ended up optimizing this lens for "F2", so the F1.5 is ever-so-slightly front focusing.

The Sonnar focus shift is toward infinity as you stop down, this is at F4:

100% crop.

You can see the rearward shift. If I had set the optimum for F1.5, that shift at F4 would put the point of focus fairly far out.

So- get outdoors, some contrast to the object, and a lot of out of focus areas. I like my wooden fence.

 

[Sonnar Brian]

A pre-war Sonnar 5cm F1.5 converted to LTM test.

Wide-Open

At F4:

I've seen some Japanese "Mooks" that are just full of "mundane" shots. Rare lenses, fairly boring shots that are essentially lens tests. Shots of kids toys, sidewalks, street signs, etc. I could probably publish a book of wooden fence posts as seen by 80 different lenses in Japan.

 

[Ezzie]

Wouldn´t it be possible to use an SLR with a long lens to check focus? Attach a piece of cellotape (clear tape) across the film gate and add cross with a pen to what would be the "emulsion side". Set lens to infinity. Open the back to a strong light source. Then set your SLR lens to lens, also set to infinity and see through the SLR viewfinder if the cross comes up in focus? I´ve used this method for lens collimation.

Better description here: http://elekm.net/zeiss-ikon/repair/collimate/

P.S Beware of SLR lenses that focus beyond infinity as one of mine does. Focus on something far away before using for this test, instead of relying on infinity marking being infinity.

P.P.S Of course this will not give you an answer as to if the RF coupling is calibrated. But tell you if the lens flange to film plane distance is correct.

 

[Sonnar Brian]

That "should work"- but I find that some "fudge factors" are often required for the RF.

On the FSU and Zeiss lenses modified for Leica, it does not work as the focal length is different from the Leica standard. On other cameras, some of the adjustments are just a little off.