Mother of All 50mm Tests

[raid]

I have just received from Kiu a beautiful Nikon camera with a 50/1.4 lens for the test. Thanks, Kiu. Now I better start put together a test. I still have two days off to do the work.

I am considering doing three parts to the test:

1. The portrait at the window: this will hope fully allow us to see how a person's face is rendered in good light.

2. I will create a piece of paper with letters and numbers in different sizes. Then, I will make copies of this paper. I will then attach these pieces of paper to the wall. Maybe such a test can show us how small a font can still be seen legibly. This test is new to me,so it may be faulty.

3. The typical bokeh test with objects in the foreground.

* I will take photos at 3.5 for all lenses since all lenses are at least as fast
as 3.5.

* I will take photos wide open for all lenses since this is an important part of
rangefinder use.

* I will use Fuji Reala for color and a slow B&W film.

This is your last chance to contribute with suggestions for the test before it will start.

Raid

 

[Sparrow]

Raid; will you use a tripod?

 

[raid]

Raid; will you use a tripod?

In my last test I used a sturdy Gitzo tripod and then also used a cable release to reduce vibrations and shake. Should I leave lens hoods off to see how flare resitsant the lenses are or is a lens hood a must?

Raid

 

[ampguy]

Hi Raid

Hi Raid

I think you should use a Digital rangefinder to isolate development differences and film batch differences.

I also think you should add a bunch of great lenses like a Hexanon M, I-22, I-50, and use multiple examples, especially of the older and FSU stuff.

I think you should use a 3d similar object at the point of focus in each image, so one can tell immediately where your point of focus was.

I have just received from Kiu a beautiful Nikon camera with a 50/1.4 lens for the test. Thanks, Kiu. Now I better start put together a test. I still have two days off to do the work.

I am considering doing three parts to the test:

1. The portrait at the window: this will hope fully allow us to see how a person's face is rendered in good light.

2. I will create a piece of paper with letters and numbers in different sizes. Then, I will make copies of this paper. I will then attach these pieces of paper to the wall. Maybe such a test can show us how small a font can still be seen legibly. This test is new to me,so it may be faulty.

3. The typical bokeh test with objects in the foreground.

* I will take photos at 3.5 for all lenses since all lenses are at least as fast
as 3.5.

* I will take photos wide open for all lenses since this is an important part of
rangefinder use.

* I will use Fuji Reala for color and a slow B&W film.

This is your last chance to contribute with suggestions for the test before it will start.

Raid

 

[raid]

I think you should use a Digital rangefinder to isolate development differences and film batch differences.

I also think you should add a bunch of great lenses like a Hexanon M, I-22, I-50, and use multiple examples, especially of the older and FSU stuff.

I think you should use a 3d similar object at the point of focus in each image, so one can tell immediately where your point of focus was.

ampguy: I do not own an M8, so using a digital rangefinder camera is out. I always test all lenses on each film, so there will be no development or film batch differences.

I have one of the two Industrar lenses you mentioned above. As for multiple examples, get them to me, won't you?

You must be joking, since there already are too many lenses in the test.
Maybe someone else will continue the testing?

Raid

 

[ampguy]

Raid

Raid

the first lenses tested on each roll will get more exposure to light and radiation than the last exposures which will have more time in the canister.

What about the 3d object, do you know what I'm talking about? Something that shows in every frame at exactly what inch the center of actual focus is.

ampguy: I do not own an M8, so using a digital rangefinder camera is out. I always test all lenses on each film, so there will be no development or film batch differences.

I have one of the two Industrar lenses you mentioned above. As for multiple examples, get them to me, won't you?

You must be joking, since there already are too many lenses in the test.
Maybe someone else will continue the testing?

Raid

 

[raid]

the first lenses tested on each roll will get more exposure to light and radiation than the last exposures which will have more time in the canister.

What about the 3d object, do you know what I'm talking about? Something that shows in every frame at exactly what inch the center of actual focus is.

ampguy: Take it easy; this is a test that can be reproduced and that will be applied to sixteen lenses. There will be limitations, but the past tests were quite consistent and the results were confirmed by many viewers of the test results depending on the lenses they owned. I had three replicates last time, and the results from replicate to replicate were consistent. The aim is to present general results on vintage 50mm lenses that are still widely used by RFF members. If you know of a more practical and more scienific way to compare sixteen lenses on a weekend, please let me know. I am working with what is available to me. I appreciate your input and your concerns.

In the last test, people commented that the focus is unimportant in such a test. The overall image qualities are moreimportant to the viewers.


Greetings,

Raid

 

[raid]

The test design is as follows:


Based on max aperture, we have two major groups (*@1.4~1.5 and @2.0) and also a minor group with max aperture 3.5.

In order to allow as many comparisons as possible with images taken on the same roll [for consistency of factors], I suggest the following:


Group 1:
1. Canon 1.2
2. Canon 1.4
3. Canon 1.5
4, J-3
5. Summarit
6. *** separately, the Nikkor 1.4 is taken with a separate Nikon camera.

All lenses will be tested at 1.4~1.5, plus 2.0 and 4.0


Group 2:
1. Nikon 50/2
2. Canon 50/1.8
3. Summar
4. Summarit
5. Summicron rigid
6. Summicron collapsible
7. Zeiss Sonnar 50/2
8. J-8

All lenses will be tested at 2.0 and 4.0.

Group 3:
1. Elmar 50/3.5
2. Industrar 50/3.5

The two lenses will be tested at 4.0 and 8.0 to see if anything improves.

Now we can compare all lenses at 4.0 and we can compare the lenses from Groups 2 and 3 at 2.0 and 4.0.

As ampguy is saying, there will be limitations to the results, but I still say that the results will provide useful information on general features of lenses.

I can still manage using one roll of film for alllenses [except theNikkor 50/1.4].
Randomizing the order in different rolls of film will hopefully eliminate systematic sources of variation.


Raid

 

[ampguy]

Hi Raid

Hi Raid

Raid,

I think you should call this "Raid's observations on some lenses lying around the house"

I just don't want you to get excited later and accuse me of going "dr. science" on you, as this "test" with n=1 for some known sample variants of dubious age and QC, and your inability to make note of the exact focus center means this will not be reproducible.

This is all I will say. Have fun, and I will enjoy observing, but the results will have very little meaning.

The test design is as follows:


Based on max aperture, we have two major groups (*@1.4~1.5 and @2.0) and also a minor group with max aperture 3.5.

In order to allow as many comparisons as possible with images taken on the same roll [for consistency of factors], I suggest the following:


Group 1:
1. Canon 1.2
2. Canon 1.4
3. Canon 1.5
4, J-3
5. Summarit
6. *** separately, the Nikkor 1.4 is taken with a separate Nikon camera.

All lenses will be tested at 1.4~1.5, plus 2.0 and 4.0


Group 2:
1. Nikon 50/2
2. Canon 50/1.8
3. Summar
4. Summarit
5. Summicron rigid
6. Summicron collapsible
7. Zeiss Sonnar 50/2
8. J-8

All lenses will be tested at 2.0 and 4.0.

Group 3:
1. Elmar 50/3.5
2. Industrar 50/3.5

The two lenses will be tested at 4.0 and 8.0 to see if anything improves.

Now we can compare all lenses at 4.0 and we can compare the lenses from Groups 2 and 3 at 2.0 and 4.0.

As ampguy is saying, there will be limitations to the results, but I still say that the results will provide useful information on general features of lenses.

I can still manage using one roll of film for alllenses [except theNikkor 50/1.4].
Randomizing the order in different rolls of film will hopefully eliminate systematic sources of variation.


Raid

 

[raid]

ampguy: OK. I give up. I will not do the test.

Raid

 

[ampguy]

Raid

Raid

Please do the test, and do it the best you can with what you have and what you know.

I simply want to provide input, and I will try to keep it positive. Here is why:

A lot of the responses from these tests will be subjective comments using adjectives such as "warm", "soft", "glow", "smooth", "sharp" and I believe that focus has a factor in some, if not all, of these observations.

A ruler at an angle showing how many inches in front of, and behind the subject are in focus is a simple thing to add to the test photos. Possibly the lens xyz is softer than lens abc because lens xyz is a little off focus?

Another simple thing to do, to increase credibility and reproducibility is to use the same paper of text, instead of copies. You are only taking 1 photo at a time, right?

I'm not expecting you to get hold of 10 Summarit or FSU examples, but I would love to see the simple things covered.

Please, go on and do it, whether you publish it or not.

ampguy: OK. I give up. I will not do the test.

Raid

 

[Magus]

Post deleted by posters request

 

[Bosk]

I had an another idea regarding the test I thought was worth sharing.

Raid it would be excellent if you could upload the full resolution .TIF files of each image to an FTP site, or make them downloadable somehow, so that we could download them and print each one at home or at a digital print kiosk.

I imagine one could learn more from studying prints than viewing the images onscreen.

 

[eIII]

For crying out loud! I can't believe how critical and demanding some are being about this simple test! Raid isn't a team of scientists analyzing glass for the Hubble telescope!

I think the few simple tests Raid has devised is more than adequate to compare the characterisitics of all the different lenses. Maybe not completely scientific, but who takes pictures in a vacuum?

As far as I can tell Raid is one guy about to test 16 lenses at his own expense, for fun, and we have to do is turn on the computer and view the results. The time involved is tremendous ! I for one am grateful to Raid for doing this and I look forward to checking out his work.

My one request? Please put the results in a new thread so we don't have to find the pics in the middle of some of this drivel!

e3

 

[raid]

Please do the test, and do it the best you can with what you have and what you know.

I simply want to provide input, and I will try to keep it positive. Here is why:

A lot of the responses from these tests will be subjective comments using adjectives such as "warm", "soft", "glow", "smooth", "sharp" and I believe that focus has a factor in some, if not all, of these observations.

A ruler at an angle showing how many inches in front of, and behind the subject are in focus is a simple thing to add to the test photos. Possibly the lens xyz is softer than lens abc because lens xyz is a little off focus?

Another simple thing to do, to increase credibility and reproducibility is to use the same paper of text, instead of copies. You are only taking 1 photo at a time, right?

I'm not expecting you to get hold of 10 Summarit or FSU examples, but I would love to see the simple things covered.

Please, go on and do it, whether you publish it or not.

ampguy: I am going to do the tests and I was just messing with you. If you had recalled what I had written in my last test write-up,you may recall that I said nearly verbatim what you said. This is a test of the lenses that I happen to own. I am fully aware of this fact. Note, however, that doing tests sequentially with the same lenses, will allow the tester to learn quite a lot about the lenses being tested. My lenses have been cleaned, adjusted, and re-adjusted by professionals after each test.If the RFF viewers of the test results comment on the findings andthink that a certain lens better gets checked again, I would do this.

My goals differ from the goals of someone evaluating a new lens. I amworking with lenses made in the 30's~60's at best. There is no way of having access to lenses that can be claimed to be "the norm". Each vintage lens has a history of its own, but I can show general feautures of these old lenses.

I will now spend some time thinking about ways to make the tests more "scientific", but I need the input from interested parties here.

ampguy, I have been asking for suggestions since this thread started. Where were you all this time? Anyways, the testing hasn't started yet, and any useful comment will be considered.


"A lot of the responses from these tests will be subjective comments using adjectives such as "warm", "soft", "glow", "smooth", "sharp" and I believe that focus has a factor in some, if not all, of these observations."

==> I usually don't comment on my test results. I post the results and let people decide for themselves what they oberve.



"A ruler at an angle showing how many inches in front of, and behind the subject are in focus is a simple thing to add to the test photos. Possibly the lens xyz is softer than lens abc because lens xyz is a little off focus?"

==> This would be difficult to do with a person as a model. Maybe a table is needed here?


"Another simple thing to do, to increase credibility and reproducibility is to use the same paper of text, instead of copies. You are only taking 1 photo at a time, right?"

==> All exposures will be made with the same targets. I take one photo at each chosen aperture in a given roll for consistency. Then I change rolls for another replicate to get check repeatability.


Greetings,



Raid

 

[raid]

For crying out loud! I can't believe how critical and demanding some are being about this simple test! Raid isn't a team of scientists analyzing glass for the Hubble telescope!

I think the few simple tests Raid has devised is more than adequate to compare the characterisitics of all the different lenses. Maybe not completely scientific, but who takes pictures in a vacuum?

As far as I can tell Raid is one guy about to test 16 lenses at his own expense, for fun, and we have to do is turn on the computer and view the results. The time involved is tremendous ! I for one am grateful to Raid for doing this and I look forward to checking out his work.

My one request? Please put the results in a new thread so we don't have to find the pics in the middle of some of this drivel!

e3

I will try to do what you have suggested above. I will start a new thread with the results so that people can find the results easier.

Raid

 

[raid]

I had an another idea regarding the test I thought was worth sharing.

Raid it would be excellent if you could upload the full resolution .TIF files of each image to an FTP site, or make them downloadable somehow, so that we could download them and print each one at home or at a digital print kiosk.

I imagine one could learn more from studying prints than viewing the images onscreen.

Bosk: I usually get the film scanned commercially, so I am unsure whether TIF files are included on the CD or not. If I dothe scans, this is possible. I will think about it later.

Raid

 

[ampguy]

Hi Raid

Hi Raid

I think the table is a good idea. Placing easy to distinguish focus objects, or markers, in front of, and behind the subject. This would really make the test great since we all know that focusing accuracy is most challenging wide open and close up, so here's a suggestion:

Use dofmaster.com or other tables to help determine optimum intervals in front of, and behind what you are actually focusing on (e.g. plane of face), then place objects or ruler at the optimum intervals, in front of, and behind the subjects face, so they'll need to be at a table or desk where objects or ruler can be placed.

Then when reviewing, and see softness (or rendering of skin tones, etc.) wide open, but not at f4, for example, we can see quickly if the subject was in the center of focus or at the edge or out of focus by looking at the markers or ruler focus reference.

ampguy: I am going to do the tests and I was just messing with you. If you had recalled what I had written in my last test write-up,you may recall that I said nearly verbatim what you said. This is a test of the lenses that I happen to own. I am fully aware of this fact. Note, however, that doing tests sequentially with the same lenses, will allow the tester to learn quite a lot about the lenses being tested. My lenses have been cleaned, adjusted, and re-adjusted by professionals after each test.If the RFF viewers of the test results comment on the findings andthink that a certain lens better gets checked again, I would do this.

My goals differ from the goals of someone evaluating a new lens. I amworking with lenses made in the 30's~60's at best. There is no way of having access to lenses that can be claimed to be "the norm". Each vintage lens has a history of its own, but I can show general feautures of these old lenses.

I will now spend some time thinking about ways to make the tests more "scientific", but I need the input from interested parties here.

ampguy, I have been asking for suggestions since this thread started. Where were you all this time? Anyways, the testing hasn't started yet, and any useful comment will be considered.


"A lot of the responses from these tests will be subjective comments using adjectives such as "warm", "soft", "glow", "smooth", "sharp" and I believe that focus has a factor in some, if not all, of these observations."

==> I usually don't comment on my test results. I post the results and let people decide for themselves what they oberve.



"A ruler at an angle showing how many inches in front of, and behind the subject are in focus is a simple thing to add to the test photos. Possibly the lens xyz is softer than lens abc because lens xyz is a little off focus?"

==> This would be difficult to do with a person as a model. Maybe a table is needed here?


"Another simple thing to do, to increase credibility and reproducibility is to use the same paper of text, instead of copies. You are only taking 1 photo at a time, right?"

==> All exposures will be made with the same targets. I take one photo at each chosen aperture in a given roll for consistency. Then I change rolls for another replicate to get check repeatability.


Greetings,



Raid

 

[raid]

ampguy: Thanks for the tips. Once I get started with the testing,I will take your suggestions into account. I am still looking for my Summar lens. I can't find it.


edited: I may do the test tomorrow without the Summar.


Raid

 

[Bosk]

Regarding DOF, I'm not sure how you planned on doing the tests but one possibility would be to perform them next to a brick wall.
I don't mean the typical front-on brick wall shots, but rather have the wall stretching from the right hand edge of the frame towards the centre, going as far towards infinity as the wall will allow. (depending on it's length)

This would make it a little easier to see and judge the transition between focus and OOF.