Is there any database online of legacy lens T-stop ratings?

See my remark two posts above.

Dan's figure is total out of proportion and so far I got no idea how it could be reasoned. To evalute the transmission, have a look at the schematic of a lens, count the air/glass surfaces and make a calculation of the reflection loss. Any other loss could only be caused by the absorbtion of the glass.

AgX said:

See my remark two posts above.

Dan's figure is total out of proportion and so far I got no idea how it could be reasoned. To evalute the transmission, have a look at the schematic of a lens, count the air/glass surfaces and make a calculation of the reflection loss. Any other loss could only be caused by the absorbtion of the glass.

Click to expand...

Umm this is WAY beyond my understanding, I don't know how to do any of that or even what it all means in terms of "figuring it out" I'll just keep an eye out for the statistic, but thanks for trying to help the uneducated...


~Stone

Mamiya: 7 II, RZ67 Pro II / Canon: 1V, AE-1, 5DmkII / Kodak: No 1 Pocket Autographic, No 1A Pocket Autographic | Sent w/ iPhone using Tapatalk

AgX said:

See my remark two posts above.

Dan's figure is total out of proportion and so far I got no idea how it could be reasoned. To evalute the transmission, have a look at the schematic of a lens, count the air/glass surfaces and make a calculation of the reflection loss. Any other loss could only be caused by the absorbtion of the glass.

Click to expand...

Well, I measured. With K40, daylight filter, known good exposure meter. Understand, the 4008 has a variable shutter, I used Beaulieu's table of shutter speeds given fps with the shutter full open. Yes I shot the test footage with the filter full open. FPS measured by shooting a couple of hundred frames of a digital watch. My 4008 ran on speed at 18 and 24 fps, a couple of frames slow at 70. I got good exposure only when I overexposed (using the aperture, not fps, to control exposure) by a stop and a third.

I did the same with a Nikon R-10 whose nominal f/1.4 zoom t/stopped around t/2.8. And this is consistent with published tests. Its really usable only at high noon.

Surprisingly, with the same tests the 6-66/1.8 Schneider on my ZM2 and 6-70/1.4 Schneider on my 5008 both t/stop at their published f/stops, i.e., are faster than claimed. This is also consistent with published tests. And believe me, after a catastrophe with the 4008 (exposure system failed on a shoot and I didn't know what it t/stopped at) I tested every one of my cine cameras with manual aperture setting so I could, if necessary, use an external meter.

If you haven't done the testing, don't grumble about my results. If you did parallel tests and got different results, that's very interesting, please tell me more.

Your calculations depend on assumptions about loss/air-glass interface. Why do you believe your assumptions?

Dan, I said

Dan's figure is total out of proportion and so far I got no idea how it could be reasoned.

Click to expand...

Both is still true.
I did not know about any measuring you did, nor would I have doubt them, nor doubt I them now.

I beg your pardon that my statement/wording caused same aggrevation, that was not my intent.

Your calculations depend on assumptions about loss/air-glass interface. Why do you believe your assumptions?

Click to expand...

That hint of being out of proportion as well as at self calculating was directed at Stone and his doubt about his modern Canon lens. So far rough calculations based on published typical reflection losses per surface ended in magnitudes of loss coherent with T-values of modern lenses. Furthermore there are statements from the industry including an old one by a Angenieux designer stating a loss factor for the geometrical f-stop of 1.1 .

I take sort of the same position as AgX does - I'm astounded that two lenses Dan mentioned have t-stops implying that only about 30% of the light entering actually contributes to the image.

I don't doubt that Dan has to do what he says, based on his tests; my reservations are in the presumption that it is strictly a t-stop issue. Dan says, "If you haven't done the testing, don't grumble about my results." My counter to this is, I know how the standard (effective 1964 and 1969) says to test, and he didn't do it that way.

I can't say with any certainty what might be the reason for the large difference between f-number and t-numbers, but if I had to find out, I'd start by measuring (roughly) the focal length, then the actual aperture diameters. A discrepancy here could go a long way toward explaining the issues.

I'm glad I don't have one of those lenses, 'cuz I'd have a hard time sleeping 'til I found out why they work that way.

Bill, I can tell the difference between a lens whose aperture doesn't open fully and one whose aperture is stuck. That wasn't the problem.

Angenieux and Nikon have long and good histories. When either specifies a lens' geometrical aperture, I believe what they say.

The R10 might bleed a lot of light off into the viewing optics, since it has an integral beam-splitter, but other S8 cameras with beam-splitters in the light path don't have this problem. Beaulieu S8 cameras (except the 1008 and 1028, which are rebadged Japanese cameras) have mirror shutters, lose no light to the viewing system; with the shutter open the film gets all of the light, with the shutter closed the viewer gets all of the light. As a result, when the camera is running the image in the viewer flickers.

Re what the standard says (and I haven't gone looking for it), given a 4008 ZM with an Angenieux 8x8B, what would you do to get good exposure when the onboard TTL metering system dies? I don't see a better way than testing with narrow latitude film. If there is one, I need to know about it.

Dan Fromm said:

Re what the standard says (and I haven't gone looking for it), given a 4008 ZM with an Angenieux 8x8B, what would you do to get good exposure when the onboard TTL metering system dies?

Click to expand...

Hi Dan, I'd probably do what you did, if I knew exactly how (I'm not a cine guy). Whatever it takes to get good exposures.

The old T-stop standard (1960s) considers the lens alone, measuring light in the center of the field with some sort of photo cell vs a lensless standard. (I don't quite follow the method, they say it's explained in the referenced patent, I linked to similar text earlier this thread.) I doubt it will be any help to you for your purposes.

My only quarrel is with calling the exposure adjustment a T-stop if you don't know for sure which components are causing the issue. If you think the lens is solely to blame, then I still have a hard time believing that the glass in a commercial lens might only let ~30% of the light through.

Bill, I agree with you that a nominal f/1.9 lens that t stops around t/3.3 seems very odd. But K40 doesn't lie. I was very shocked by the disagreement between my 4008ZM and, e.g., my Nikkormat.

As for the R10, well, as I said I'm not the only person to have been surprised by its lens' poor transmission. On the other hand, the lens has several gazillion elements and light is bled for the viewer and for the meter. See http://imaging.nikon.com/history/cousins/cousins19-e/index.htm , the lens' cross-section is at the bottom of the page.

Coming back to the original question of this thread:

"Is there any database online of legacy lens T-stop ratings?"

Les Sarile said:

The old lens reviews listed the T-number of the lenses. In this page a Rokkor 50mm f1.4 with a T-number 1.53, 28mm f2.0 with a T-number 2.25 and a 135mm f2.8 with a T-number 3.11. In these examples it doesn't seem significantly off.

Click to expand...

This is interesting. I can't remember ever seen T-values indicated for still photography lenses.

What is also interestening is that these figures indicate the same T-factors as given for much more groups containing zoom lenses.

This is common on all the lens reviews I have seen in these older magazines. The first one I posted was from the January 1979 Popular Photography magazine as part of the Minolta XD-11 review. This one is from a newer December 1982 Popular Photography as part of the FM2 which shows a Nikkor zoom. As you can see, the actual T numbers are very close to f numbers marked. I can only imagine that the superb Canon 70-200 f2.8 L lens is as good or better in this regard. Unfortunately, the review of it doesn't show this info nor in Canon's specs. Since the older lenses are for all practical purposes close enough, they probably are no longer concerned about this bit of data anymore.

Dan Fromm said:

As for the R10, well, as I said I'm not the only person to have been surprised by its lens' poor transmission. On the other hand, the lens has several gazillion elements and light is bled for the viewer and for the meter. See http://imaging.nikon.com/history/cousins/cousins19-e/index.htm , the lens' cross-section is at the bottom of the page.

Click to expand...

It's quite an elaborate design, for sure. I would not normally think of a beam splitter as part of the lens, but in these cases it definitely is, so should properly affect the T-stop. I'd guess that the beam splitters are the predominant reasons for light loss, but perhaps only Nikon knows for sure.

Bill, most Super 8 cameras have beamsplitters in the light path. The Nikon R10 stands out among them for light lost between the front of the lens and the film.

Beaulieu S8 cameras (except the 1008 and 1028) have reciprocating shutters with a mirror. With them, when the shutter is open all of the light that the lens passes goes to the film. When the shutter is closed all of the light goes to the viewer; light for the meter is taken from the viewer by a beamsplitter. Y'r explanation doesn't work for these cameras. My dim Angenieux 8x8B was on a Beaulieu 4008 ZM.

Canon's 8.5-25.5/1.0 zoom, as fitted to the Canon 310XL, 310XL-S and AF310XL cine cameras was probably the fastest photographic objective made in (relatively) large numbers. Canon were very serious about making these cameras as usable as possible in dim light. To this end, the 310XL had a non-TTL meter, a relatively slow shutter speed for an S8 camera, and fed light to the viewing optics with a tiny fixed mirror (not beam splitter) behind the lens. If you look at the specs here http://www.canon.com/camera-museum/camera/cine/chrono_1971-1982.html?lang=us you'll see that they assert the 310XL has a TTL meter; this isn't so, the meter is external, with the cell located above the lens.

The search for a T stop database for still photography seems like an answer in search of a question. After you found the information you would then need to find out how accurate all of yor cameras are at the standard shutter speeds and variable speeds. Then you would need to test all of the meters to see how far off they might be. A teacher in High School once told me that if you considered the difference between T stops and F stops and then added in shutter speed variables and chemistry measuring variables and thermometer accuracy variables you would think it's just about impossible to get a good end result. In practice these variables tend to cancel each other out. If you have a lens which was marked with a maximum aperture of f/1.4 but measured T 1/.53 in a particular test, your example might measure T 1.59 or T 1.37. It is probably not practical to test every single lens for T transmission. If this were a serious problem we would have known about it a long time ago. There is the additional issue of the uniformity of the transmission. A fast standard lens will typically have some light fall-off at the corners close to or at full aperture. Should we then measure T transmission at different points of the field of view?