lens (resolution) testing and application to the real world.

Hi

I'm not very experience at this, so if someone would perhaps comment on my methods I'd be interested to hear. Also, I would like to be able to compare results with other camera systems than my own (humble) setup.

I thought I would test this camera a Bessa RF with the Norman Koren chart (available here). Following the instructions I find that the results are something like 23 lp/mm



Being a 105mm lens the camera was (measured with a tape measure 5.3 meters) from the target. I used f16 for the test and the negative was scanned at 3200dpi on my Epson 3200 (yes yes ... I know, but its definitely sharper in black and white than it is in colour).

Now, I thought that (since I photograph scenes not just test my gear) that it might be interesting to put my 10D (my normal digital camera) on the tripod with a focal length lens more or less equal in field of view to that of the Bessa RF's 105 (24mm),

This is the overall scene:



and this is the result from a RAW conversion of the image



Which seems to support why I get such better results in landscape photographs with my (ancient?) 120 folder than I do with my 10D. I know that this seems to run counter to the sorts of things which may be found on sites like LL.

I'm quite interested to know how much better I might get with other MF equipment. I realize though that any 'step up' in gear will probably mean that my current scanning arrangement will be the weakest link in the imaging chain. Still ... I can always send out for scans ;-)

If you want to do lens resolution tests it's better to use the USAF1951 charts. Google them and the instructions for them.
If you set up properly you get a resolution in lp/mm on film (i.e. take the magnification into account).

I do this always with a new to me lens to get an impression if its a dog or not. The link to real life? That's hard to say. Resolution is only one component of the picture.

Comparing your results with others? That's difficult because film, developer and observation come into play here.

Today 100ASA bw films deliver something like 90 or 100 lp/mm (on film). Quality lenses deliver this as well, at least in the center and / or stopped down.

Hi

ath said:

If you want to do lens resolution tests it's better to use the USAF1951 charts. Google them and the instructions for them.
If you set up properly you get a resolution in lp/mm on film (i.e. take the magnification into account).

Click to expand...

well, I don't know for sure, but Norman Koren says of USAF charts:

The USAF 1951 lens test chart and its variants have been a standard for half a century. But it is unsuited for computer analysis because of its fragmented arrangement, and it is poorly suited for visually estimating MTF. It is often misused. People strain their eyes to find highest resolution at which bars can be distinguished. This frequently results in impressive resolution numbers (over 100 lp/mm) that are poorly related to visible image quality.

Click to expand...

I used to have an excellent post script file which had not only vertical + horizontal + diagonal bars in squares going from 20 through to 100 lp/mm but instructions on how to place it relative to the camera.

wouldn't you know I've lost that file!

ath said:

Comparing your results with others? That's difficult because film, developer and observation come into play here.

Click to expand...

I agree ... which is why I posted a scan of the film so that others could decide from the same evidence I used. Standards are possible though, and as long as everyone is behaving more or less in the same way then the figures will be more or less comparable.

Was rangefinder focussed by scale? Best to take extra shots each side of indicated distance.
Best aperture would be around f8, diffraction will reduce resolution at f16
Is a microscope or enlarger with a good lens available?Old scanner probably only gets around 30 line-pairs/mm.
A high resolution film, T-max 100 or Delta 100 or just the resolution of the film may seriously affect the result.
With a microscope I get ~100 lppm on T-max 100 in Xtol with a Canon ef 50mm f1.8.
In practice, if focus on infinity, a valid test but often depth of field needed for a shot makes this less relevant,but still interesting.

Alan

Alan Johnson said:

Was rangefinder focussed by scale? Best to take extra shots each side of indicated distance.

Click to expand...

rangefinder was focused by the matching of top and bottom image in the range finder. I've confirmed with a GG that its pretty bang on.

I'll scan this section when I get another scanner (meaning a 35mm scanner and I'll take scissors to the film)

ath said:

I do this always with a new to me lens to get an impression if its a dog or not. The link to real life? That's hard to say. Resolution is only one component of the picture.
Comparing your results with others? That's difficult because film, developer and observation come into play here.

Click to expand...

I've done a fair amount of lens testing before Fourier Transfer + Laser/ Computer analysis. ALWAYS on the Optical Bench, and always considering an aerial image - NO film or ground glass - with a microscope.

It might be possible to duplicate this system, roughly; Mount the subject lens in the enlarger, the resolution target in the negative stage, and look at the image with a Grain Focuser.

I recognize the probable difficulty in mounting the lens - and a LOT would depend on the quality of the grain focuser - you might want to adapt a microscope - but it should work.

IMO it is preferable to view the negative through a good microscope or enlarging lens (look at the chart image with a grain magnifier),as these should have high resolutions and will not give an erroneous result by being the controlling factor.
1/Rmeasured=1/Rfilm +1/R lens +1/Rviewing device

USAF chart

If you'd like to see the world's biggest USAF chart, go to Google Earth and view the USAF Museum in Dayton, Ohio. To the left of the main building, there is an old runway that has the test pattern painted on it. I imagine that this was for testing the resolution of spy satellite lenses.

Just go to google and search for the museum, then click on the maps button, then after it draws click on the satellite link. The chart is south west of the building and shoows good resolution all the way to the highest magnification that they have. Pretty neat.

For a good PDF of the 1951 USAF test chart go to Dead Link Removed

Jim Jones said:

For a good PDF of the 1951 USAF test chart go to Dead Link Removed

Click to expand...

Where I was "Forbidden"...
The Medium Format site have really deterioated from that really nice library of assorted (but not sorted ) information, down to a mess in dire need of attention.

//Björn

edtbjon said:

Where I was "Forbidden"...
The Medium Format site have really deterioated from that really nice library of assorted (but not sorted ) information, down to a mess in dire need of attention.

Click to expand...

It opens without a problem for me. The restriction is 'on your side'?

Apropos the state of the site: anyone knows how and where Robert is?

Q.G. said:

It opens without a problem for me. The restriction is 'on your side'?

Apropos the state of the site: anyone knows how and where Robert is?

Click to expand...

I'm with Bjorn ... I'm forbidden here too ... could be a non usa IP address issue?

pellicle said:

I'm with Bjorn ... I'm forbidden here too ... could be a non usa IP address issue?

Click to expand...

Nope.
I'm not in the U.S.A. either, and have no problem accessing the file.

In my case the link does not work but it will work to bring up the chart if the address is typed in manually.

Yup, the direct link seems to block it, but cut and paste into the address and it works.