[interested_observer]

Usually folks evaluate a lens on paper and then save for and acquire it, take it out and shoot with it. Along the way they debate the merits of X vs Y. I have been fortunate enough to acquire a FA 31 a couple of years ago. Wonderful lens, I like it very much - essentially what is there not to like. Late last year I had an opportunity to acquire a Contax Carl Zeiss 28mm f2.8 Distagon T* lens. I needed to change mounts in order to use it on my K20. After reading about Zeiss here, this was an affordable way to try one out, since the old Contax glass was about 1/3 less than a brand new shiny Zeiss. With it being 28mm it begs a natural comparison with the 31. They are both top of the line premium lenses from notable optical design houses.

Where I am going with this is that in comparing the two lenses (in the field [landscapes] - not a brick wall), I have seen that with the only difference being the lens - same ISO (100), same aperture (f4), nearly the same time (a minute difference between shot sequences) same landscape scene, I am finding that the Contax/Zeiss lens has shutter speeds about twice as fast as the Pentax FA Ltd.

I have taken all of the images on a tripod, 2 second mirror up, external shutter release, AP mode on each lens, multi-segment metering same K20 body, really no differences - other than swapping lenses. The other thing is that I have been shooting 5 image brackets at 0, -.5, +.5, -1, +1 ev, and the shutter times are really staggering between the two lenses, with about a minute between sets (just swapping the lens).

lens....0ev....-.5ev..+.5ev..-1ev..+1ev
CZ 28 1/15...1/20...1/10...1/30...1/8 sec
PK 31 1/8.....1/10...1/6.....1/15...1/4 sec

This observation is just not across one set of images, but many sets of images. Overall, there is a consistency in the numbers, however its not absolute, there are instances where the difference is both larger and smaller, and I have observed instances where the shutter times are essentially the same. On the average though, the difference is on the order of twice as fast.

My observation here is that it appears to me that the Contax/Zeiss glass in terms of the amount of glass used and the design applied, appears to have a built in efficiency to it, of about 1 f stop. That is, it is able to transmit more effective light to the sensor through the same aperture.

Now I know that most folks do not have access to similar lenses, but is there something that I am not taking in to account? My single optics class was over 40 years ago, and I didn't even keep the text book.

[luftfluss]

Yes, I've seen this phenomenon, where the t-stops between 2 lenses do not correspond with the f-stops as we are used to seeing. I've not seen such a stark difference, though, as you've demonstrated here.

[Just1MoreDave]

On the cheap end, a lens with poor contrast can produce images that look brighter than a high contrast lens. But you can tell it's a contrast issue.

You probably want to be sure you're adjusting for the focal length differences, and looking at the brightness of the same elements in a scene. My guess is multiple causes and errors combining in the same direction.

[Rondec]

Not sure why there is such a big difference in these particular lenses, but it is related to the t stop versus the f stop.

[arpaagent]

Another thing to note is that different lenses can meter differently. I would like to see comparison shots between the lenses to make sure they are actually the same exposure. I would be willing to bet that the 31 Ltd made a significantly brighter exposure, simply due to the fact that the metering cause an overexposure when compared against the zeiss. It is not unusual for large maximum aperture lenses to either underexpose at the maximum f-stop or underexpose at smaller f-stops (usually caused by the fact that the maximum aperture is sometimes darker than it should be, given the aperture).

[Ron Kruger]

Zeiss long has been touted for its "light gathering capabilities," but I would suspect the small difference that might exist is exaggerated by the difference in FL. I know we're only talking about 3mm, but the wider one goes the more light gathering capabilities are multiplied.
It would be interesting to test the same way with the Zeiss 28 and the Pentax 21.

[newarts]

could a round-off error in labeling the lenses speeds be part of the explaination? Two lenses of nominally the same speed might differ by a 1/2 stop or more (I don't know what rules they follow when labeling lenses.) A half-stop is perhaps a 40% difference in actual aperture.

[interested_observer]

Yes, I've seen this phenomenon, where the t-stops between 2 lenses do not correspond with the f-stops as we are used to seeing. I've not seen such a stark difference, though, as you've demonstrated here.

Not sure why there is such a big difference in these particular lenses, but it is related to the t stop versus the f stop.

I too was surprised by the difference.

On the cheap end, a lens with poor contrast can produce images that look brighter than a high contrast lens. But you can tell it's a contrast issue.

You probably want to be sure you're adjusting for the focal length differences, and looking at the brightness of the same elements in a scene. My guess is multiple causes and errors combining in the same direction.

Zeiss long has been touted for its "light gathering capabilities," but I would suspect the small difference that might exist is exaggerated by the difference in FL. I know we're only talking about 3mm, but the wider one goes the more light gathering capabilities are multiplied.
It would be interesting to test the same way with the Zeiss 28 and the Pentax 21.

I wasn't expecting that 3mm at this focal length would be the sole cause. In an ultra wide lens, I could see this being a larger contributor.

Another thing to note is that different lenses can meter differently. I would like to see comparison shots between the lenses to make sure they are actually the same exposure. I would be willing to bet that the 31 Ltd made a significantly brighter exposure, simply due to the fact that the metering cause an overexposure when compared against the zeiss. It is not unusual for large maximum aperture lenses to either underexpose at the maximum f-stop or underexpose at smaller f-stops (usually caused by the fact that the maximum aperture is sometimes darker than it should be, given the aperture).

The Zeiss has much more contrast and in my opinion tends to under expose a bit, while the Pentax is a brighter lens. This probably is also helped by how they each individually render colors. That is very apparent from the shots - and I'll try to go back and dig them out for posting. Their characters differs substantially also in terms of rendering colors. I like both very much for their own individual qualities.

I tried to stay away from wide open, that the reason for f4 and it was near the maximum resolution for both lenses, so I thought that would be the most fair comparison. Plus, I want to haul these up to Sedona and see what I am able to do with them. I was playing around with how I want to apply them and have some idea as to how to approach the situation. Maybe this would just be pushing each lens, but the Pentax with morning light and the Zeiss in the evening. However, it will probably be both at every occasion. I would have the sneaking suspicion that the Zeiss in the morning and the Pentax in the evening would work just as well. Or that they would each excel in different situations depending on how the light and shadows fall within the composition.

Part of the purpose for doing this was to see if I need to bias the Zeiss lens by say a half or 1 whole EV, or just leave it alone. That is what started all of this....

could a round-off error in labeling the lenses speeds be part of the explaination? Two lenses of nominally the same speed might differ by a 1/2 stop or more (I don't know what rules they follow when labeling lenses.) A half-stop is perhaps a 40% difference in actual aperture.

Well the two lenses are 1 1/3 stops different in their wide open apertures - f1.8 vs f2.8...

[wlachan]

Without picture, I vote for user error because 1 stop difference is just too much. When testing lenses, you don't rely on Av mode, but have to try different shutter speeds for identical exposures at each f-stop. Another possibility is that the apertures of the Zeiss and the FA31 were calibrated to the opposite end of tolerance. I have seen aperture differed as much as 0.5EV, but not 1EV.

[Aristophanes]

Software?

Maybe the Pentax body reads an FA 31 and does some special, undocumented adjusting to reduce errors for what is normally a film camera lens by design. It stops down a tad to improve telecentricity and straighten the light path and reduce stray bounce. Pentax may be trying to keep the FA 31 "honest" to the needs of the APS-C sensor and improve the user experience, keeping the lens in its legendary place for a format for which it was not really designed.

That and the 3mm difference could add up to a sequence of changes causing this issue.

The CZ may be more designed as a resolution monster, and the FA 31 as a bokeh dream.

Just a thought. Interesting test.

[newarts]

...
Well the two lenses are 1 1/3 stops different in their wide open apertures - f1.8 vs f2.8...

What Zeiss called 2.8 might be 2.6 and what Pentax called 1.8 might be 1.9. That would lead to a consistent exposure difference when both lenses were at the same nominal f-number.

A way to test might be to use the same manual exposure settings for a uniform target like a sheet of paper and look at the resultant histograms with your PP software; then modify exposure settings for one lens so the histograms are identical to determine the actual exposure difference.

[interested_observer]

I just wanted to add before I write anything else, that this was not a real scientific test. The intent here was to shoot and understand the differences between these two lenses, essentially in terms of how they render and the coloration that each individual provides. Also, I wanted to understand the microcontrasts in rendering folks have been referring to. So I went up to my little vantage point and shot away, more for my understanding than anything else (also it duplicates the direction and to some extent the shot I had been thinking about up in Sedona). There is nothing wrong with the lenses at all - just trying to understand the differences. On previous evenings, I had observed the underexposure and was starting to wonder if I should push the CZ by some amount of EV, and that was the reason why I went to the 5 shot bracketing so that I could easily see the differences and line them up. That is where the exposure times started to stand out, and got me thinking - and asking.

Without picture, I vote for user error because 1 stop difference is just too much. When testing lenses, you don't rely on Av mode, but have to try different shutter speeds for identical exposures at each f-stop. Another possibility is that the apertures of the Zeiss and the FA31 were calibrated to the opposite end of tolerance. I have seen aperture differed as much as 0.5EV, but not 1EV.

Ok, I found the image sets and selected two images, one CZ and one FA. The sequence these images are from are...

lens......0ev..-.5ev..+.5ev..-1ev..+1ev
CZ 28......4......3........6........2........8 sec
PK 31 ....10....10......20.......8......30 sec

So the shutter time for these are CZ 8 seconds, FA 30 seconds.

User error - could be, although it could be just my error in interpreting what I am seeing. I too think that even with accumulation of error over say focal length, the CZ under exposing, possibly the body handling the FA or Pentax lenses a tad differently, etc. a whole f stop may be a bit much, even with some error/differences within the aperture of the two lenses. The difference in how T stops and F stops are handled could certainly be the largest contributor here too. But who really knows. Hence the posting for some expert advice.

Well to the images. I used Bible Pro 5 to convert to JPG, with just the appropriate resizing to upload. The EXIF should be intact. A bit about the images. They were taken in mid April at 7.30 in the evening, facing due west. The first image probably more accurately represents the actual lighting conditions. ISO 100, f4, the CZ is an all manual lens, so I manually indicated 28mm on startup. I shot in AP since with a all manual lenz I had only the choice of AP or manual, and I wanted the camera to compute the shutter time, hence the AP selection. The CZ image is the first, with the FA following up.

[wildman]

I have seen that with the only difference being the lens - same ISO (100), same aperture (f4),

Perhaps I'm being simple-minded about this but...

...is not aperture an absolute measurement of light?

If both lens' are set at the same marked aperture but they do not deliver the same amount of light to the sensor then they are not at the same effective aperture because they are not transmitting the same amount of light.

It has nothing to do with efficiency - f4 is simply f4 a certain absolute quantity of light. If for instance the Pentax when set at f4 is, in fact, a true f4 at that setting and any other lens transmits more or less than the Pentax then that lens is operating at some aperture other than f4.

If you want to test this just set up your camera on a tripod and plug in different lens' set at the same f stop under the same light. If using AV for instance they should all give you the same shutter speed all else being equal. I just did this with my 6 modern lens' and this was true.

BTW in my test. for instance, was included the Zeiss 50mm f/1.4 ZK and a version 1 kit lens and they both gave me an identical shutter speed of 125/sec with this particular test.

By definition two different lens' can not be at the at the same f stop and give two different levels of light.

[freewheeler]

So with the CZ on your camera your body computes shorter shutterspeeds
and hence delivers darker pictures ? Seems only logiacal to me.
When using the same shutterspeed and F4 on both, wat happens then ?

[RonHendriks1966]

...is not aperture an absolute measurement of light?

Yes it is, but it isn't always what is stated on the sticker of the lens:

DxO measures it with lenses. Non of the two lenses are tested, but here is one Zeiss: DxOMark - Carl Zeiss Distagon T 28mm f/2 ZE Canon Look at measurements and transmission.