How to test lens transmission

[kanishka]

I was wondering what is the best way to test the transmission of a lens.
I will receive in a few days an old lens and I want to know how to expose correctly with it.
Can you share with me what you think would be the best method?

Thanks

[Nodda Duma]

Assuming the glass is clean..

The best method is relative measurement on a spectrophotometer, but I'm guessing you don't have access to one.

The next best is to calculate based on number of air-glass interfaces using the formula 0.96^x, where X is the number of air-glass interfaces. That calculation is for uncoated optics. For coated, use 0.99^x. So an uncoated singlet with two air-glass interfaces has a transmission of 0.96^2 = 92%. An old uncoated Rapid Rectilinear or something like a Dagor has 2 cemented groups, or 4 air-glass interfaces, so 0.96^4 is ~85% Transmission. An uncoated Cooke triplet has 0.96^6 = 78.3%. Incidentally you can see why older designs minimized the number of air-glass interfaces as much as possible and also why coatings are so important. Any scientific calculator will do this calculation.

The next best method is use a 35mm camera's exposure meter to read the change in readings with a known lens vs. with the unknown lens in front of it. This is trickier to set up, since you'll need to minimize differences between the two measurements: block stray light with a fabricated tube, maintain a similar FOV, and also factor out the f/#. Once all is said and done, the results should be close to the calculated values above. If not, something went wrong with your setup.

As a lens designer, my recommendation is to calculate the transmission using the formula above. Assuming you count the number of surfaces correctly, the result will be very close to truth.

Keep in mind that if your lens has aperture markings, this transmission loss may very well have been factored in when the markings were scribed on the lens barrel.

[Tin Can]

All good until the last caveat.

However, big thanks OP and Nodda for addressing an obvious question considering recent threads.

:)

[Nodda Duma]

All good until the last caveat.

My attempt to be generous to the original designers. :)

[Jac@stafford.net]

Any scientific calculator will do this calculation.

And if you are online, just type .96^2 in the search window. Very handy! The iPhone will do the same without its browser in the search window. (And U*X wizards know how to do it at the command line.) :)

[ic-racer]

The most practicle is a FP meter. I use the Horseman focal plane meter.

[Tin Can]

I use the Horseman focal plane meter.

I thought most of those failed long ago.

No matter, basicly too rare and old for most of us to find and use.

The question could be, do you find variation when changing glass and apertures? Do you read significant differences in lens transmission?

[ic-racer]

Perhaps even more practicle would be a Zone I test. That would give the correct exposure for the lens/shutter system.

[ic-racer]

Randy Moe posted the words "too rare" :)...this guy collects and uses the rarest of rare equipment like a madman...

[ic-racer]

BTW that Horseman meter is a great tool, especially if you ever need io expose sheetfilm under an enlarger.

[Tin Can]

Randy Moe posted the words "too rare" :)...this guy collects and uses the rarest of rare equipment like a madman...

Ha, you found my old nickname from my early motorbike years.

'Madman Moe' LOL :) AKA '10 Grand' as I always shifted at 10K revs. New Hondas were something.

I dislike old electronics. They alway fail. I collected tube radios since the 50's when I was a kid. I have plenty of non-working old electric gear. I made my own pirate radio station in 1966.

I'll trade a you Wire Recorder, which precedes Tape recorder.

As for collecting. I have less than 5% of my former collections.

Now will you answer my question?

Do you see significant, more than 20% variation, in lens transmission from lens to lens? As measured by the old Horse?

[desertrat]

Several years ago I saw a post at another photographic forum, I think the poster was Richard Knoppow, stating that the optical glass made before world war two was "filthy" by modern standards.

Later on I downloaded and read a book on lenses in the early 20th century, I think it was by Hans Harting, translated into English. He stated that thick lenses like cemented quadruplets and quintuplets absorbed an appreciable amount of light and slowed the whole lens system down somewhat.

I rigged up a test fixture using an old American Optical research style microscope illumination lamp with condenser and iris, and a Weston Master exposure meter for the sensor. The light pencil was made small enough so all of it would reach the sensor.

My test subjects were a Wolly RR, an old uncoated Dagor, and a Turner-Reich triple. All of these lenses have 4 air-glass interfaces. The meter readings were taken before and after the lens was inserted in the light path.

The RR absorbed a small amount of light, a small fraction of a stop. The Dagor absorbed a little more. The T-R triple absorbed a whopping stop and a half, roughly.

Unfortunately I can't find the piece of note pad paper I scribbled the readings on.

[Tin Can]

Several years ago I saw a post at another photographic forum, I think the poster was Richard Knoppow, stating that the optical glass made before world war two was "filthy" by modern standards.

Later on I downloaded and read a book on lenses in the early 20th century, I think it was by Hans Harting, translated into English. He stated that thick lenses like cemented quadruplets and quintuplets absorbed an appreciable amount of light and slowed the whole lens system down somewhat.

I rigged up a test fixture using an old American Optical research style microscope illumination lamp with condenser and iris, and a Weston Master exposure meter for the sensor. The light pencil was made small enough so all of it would reach the sensor.

My test subjects were a Wolly RR, an old uncoated Dagor, and a Turner-Reich triple. All of these lenses have 4 air-glass interfaces. The meter readings were taken before and after the lens was inserted in the light path.

The RR absorbed a small amount of light, a small fraction of a stop. The Dagor absorbed a little more. The T-R triple absorbed a whopping stop and a half, roughly.

Unfortunately I can't find the piece of note pad paper I scribbled the readings on.

Yes, paper disappears, but your memory is good. :)

I like empirical science.

[Willie]

Manual or automatic shutter? Need to know before recommending ATF or MTL, dinosaur or synthetic. ;-)

[EdWorkman]

So back in the 1930s or so movie days, how did those folks determine T stops ?
I understand that T-stops were used to get consistent exposure account/despite the problems discussed above
I'll bet that the web would give up a technical article or book on the subject, maybe

[Dan Fromm]

Ed, I expect that they measured transmission photometrically.

I had some cine cameras with zoom lenses that seemed a lot slower than marked (Beaulieu 4008ZM with Angenieux 8x8B, marked f/1.9; Nikon R-10 with 7-70/1.4 Cine-Nikkor). The cameras auto exposure systems worked well, gave me properly exposed Kodachrome. To find out what the lenses t-stopped at I aimed them at an evenly exposed blank wall, saw what aperture the cameras selected and then metered the wall reflected with a Lunasix 3. I made sure than only the wall was visible in the cameras' finders and metered with the Lunasix from close enough that the meter's cell saw only the all. The f/1.9 Angenieux t-stopped ~ t/3.3, the f/1.4 Cine-Nikkor ~ t/2.8. And then I understood why both cameras were useful at highish framing rates only around mid-day.

Similar tests with other Beaulieus and 6-66/1.8 and 6-70/1.4 Schneider Cinegons found that they t/stopped at t/1.8 and t/1.4 respectively. These surprised me considerably.

To answer the obvious question, yes I knew the cameras' shutter speeds at 18 fps and I'd tested to make sure they ran on speed.

[EdWorkman]

thanks Dan
I found a reference, a 1949 conference of The Society of Motion Picture Engineers, but no details whatever

[Jac@stafford.net]

Several years ago I saw a post at another photographic forum, I think the poster was Richard Knoppow,

I traded some literature with Richard long ago. With trepidation I ask: what became of him?
.

[Jim Andrada]

Video lenses are usually marked in T-stops instead of F-stops Maybe that's what makes them so expensive.;)

http://www.bhphotovideo.com/c/product/839226-REG/Canon_6142B002_CN_E_30_300mm_T2_95_3_7_L.html

I'd have this one on my video cam if it weren't for 44,650 little reasons.:(

[swmcl]

I once used a Sekonic meter on incident mode with the lumisphere retracted and placed it flat onto the centre of the ground glass under the darkcloth. I then compared the light hitting the ground glass with one old petzval lens to the other known lens - a Sironar S 180mm. Shoot the camera towards a nice bright even light source or take a photo of the sun through a white plastic sheet or something.

When the lightmeter reads the same for both lenses the amount of light is the same. ie the aperture is the same.

(I think I'm right ...)

:-)

[Nodda Duma]

.

(I think I'm right ...)

:-)

Yep, and it's a really easy way to get the answer, too.

[desertrat]

Several years ago I saw a post at another photographic forum, I think the poster was Richard Knoppow, stating that the optical glass made before world war two was "filthy" by modern standards.

Later on I downloaded and read a book on lenses in the early 20th century, I think it was by Hans Harting, translated into English. He stated that thick lenses like cemented quadruplets and quintuplets absorbed an appreciable amount of light and slowed the whole lens system down somewhat.

I rigged up a test fixture using an old American Optical research style microscope illumination lamp with condenser and iris, and a Weston Master exposure meter for the sensor. The light pencil was made small enough so all of it would reach the sensor.

My test subjects were a Wolly RR, an old uncoated Dagor, and a Turner-Reich triple. All of these lenses have 4 air-glass interfaces. The meter readings were taken before and after the lens was inserted in the light path.

The RR absorbed a small amount of light, a small fraction of a stop. The Dagor absorbed a little more. The T-R triple absorbed a whopping stop and a half, roughly.

Unfortunately I can't find the piece of note pad paper I scribbled the readings on.

Post in haste, repent at leisure...

The RR had beautiful glass, like new, and the cemented doublets were pretty thin. Most of the difference in meter readings was probably due to loss from reflections at the air/glass interfaces, and very little due to absorption.

The triplets in the Dagor were a little thicker, so the difference in readings after inserting the lens in the light path between it and the RR were probably due to absorption. Most of the absorption was probably occurring in one of the cement layers in the front group which is a bit hazy. A pristine uncoated Dagor might have given readings almost the same as the RR.

The T-R was a late uncoated one, the cemented quintuplets were very thick, and the light was passing through a lot of glass. I think most of the stop and a half loss was due to absorption.

[kanishka]

Thanks for all the info!

[Jim Andrada]

When the lightmeter reads the same for both lenses the amount of light is the same. ie the aperture is the same.

(I think I'm right ...)

:-)

I think it would be the same aperture as measured in T-stops, not F-stops. Not much difference if the lenses are similar, maybe a bigger difference if one is a 13 element zoom lens for 35mm/digital and the other is a Landscape Meniscus.

[ic-racer]

Ha, you found my old nickname from my early motorbike years.

'Madman Moe' LOL :) AKA '10 Grand' as I always shifted at 10K revs. New Hondas were something.

I dislike old electronics. They alway fail. I collected tube radios since the 50's when I was a kid. I have plenty of non-working old electric gear. I made my own pirate radio station in 1966.

I'll trade a you Wire Recorder, which precedes Tape recorder.

As for collecting. I have less than 5% of my former collections.

Now will you answer my question?

Do you see significant, more than 20% variation, in lens transmission from lens to lens? As measured by the old Horse?
My brother still has his Atwater Kent collection he put together in the 1970s.
You may not know this but the Horseman exposure computer has no aperture markings. It indicates only the appropriate shutter speed for the light that would fall on the film. Utility of the device is realized best with bellows extension, unmarked lenses, improper or missing aperture scale and sloppy aperture mechanisms.

[Tin Can]

My brother still has his Atwater Kent collection he put together in the 1970s.
You may not know this but the Horseman exposure computer has no aperture markings. It indicates only the appropriate shutter speed for the light that would fall on the film. Utility of the device is realized best with bellows extension, unmarked lenses, improper or missing aperture scale and sloppy aperture mechanisms.

I'll add one to my wish list.

Not kidding. :)

Thanks!

I just got a pair 60's Shure Vocal Master 6 foot tall speakers. The amp is solid state, NG, but my old system now sounds fantastic with 12, low watt speakers. $20, well spent.