p.1 #1 · The Glass in the Path: Sensor Stacks and Adapted Lenses
As Roger wrote, this is a geek post, but if you aren’t interested in adapting lenses this article may be for you.
Here is an excerpt:
"Lens designers plan for the thickness of the sensor stack, and others have made corrective lenses to allow very expensive lenses developed for film to be used on digital cinema cameras.
The main purpose of this post is to serve as a demonstration of just how much of an effect the sensor stack has. I figured if I was surprised, then some of you would be, too." - Read full article
p.1 #2 · The Glass in the Path: Sensor Stacks and Adapted Lenses
I'd be interested to know more about the correction lenses that Roger mentions which are used on cinema lenses, designed for exposing film but used on digital cameras.
I'd also be interested to know if my images wouldn't be improved by removing my NEX-7's filter stack, permanently attaching my Speed Booster to protect it, and then screw on some UV/IR filters onto my lenses. I suspect that might be where Roger is going, though.
Personally, I'd like to buy Roger Cicala and Brian Caldwell a few pitchers of beer, sit back with a tape recorder (and a few physics textbooks and years to catch up), and watch the fireworks fly. Or to serve as either's research assistant, which would be more informative but less entertaining. Just sayin'.
p.1 #6 · The Glass in the Path: Sensor Stacks and Adapted Lenses
Fred Miranda wrote:
Here is an excerpt:
"Lens designers plan for the thickness of the sensor stack, and others have made corrective lenses to allow very expensive lenses developed for film to be used on digital cinema cameras.
The main purpose of this post is to serve as a demonstration of just how much of an effect the sensor stack has. I figured if I was surprised, then some of you would be, too."[/url]
The Leica 35mm ASPH result is as expected, but I am surprised by the large amount of spherical aberration with the Otus, even without focal reducer. The Leica MTF is not affected by SA, while the lens is only one stop slower. It would be interesting to see Otus at F/2.
Hmm. The 35mm ASPH actually improves slightly in the image center with increasing glass thickness. Roger, what is the focus criterion used in your bench tests?
p.1 #7 · The Glass in the Path: Sensor Stacks and Adapted Lenses
Toothwalker wrote:
Hmm. The 35mm ASPH actually improves slightly in the image center with increasing glass thickness. Roger, what is the focus criterion used in your bench tests?
Paul, we had set it to focus for best MTF at 20 lp/mm (cross-type reticle) at the center point. All other measurements were then taken at this focusing distance - it was not refocused at each point. We'll obviously be doing a bit more of this, so any suggestions you have would be most appreciated.
Brian made the same comment about the Otus and speculated that perhaps we were also seeing something right in the center that had to do with the rear aspheric element. But I'm not certain I followed what he said accurately (he lapsed into optical physics speak), so forgive me if I misquoted.
p.1 #8 · The Glass in the Path: Sensor Stacks and Adapted Lenses
RCicala wrote:
Paul, we had set it to focus for best MTF at 20 lp/mm (cross-type reticle) at the center point. All other measurements were then taken at this focusing distance - it was not refocused at each point.
Very good, but of course you refocused each time the glass situation changed?
We'll obviously be doing a bit more of this, so any suggestions you have would be most appreciated.
Brian made the same comment about the Otus and speculated that perhaps we were also seeing something right in the center that had to do with the rear aspheric element. But I'm not certain I followed what he said accurately (he lapsed into optical physics speak), so forgive me if I misquoted.
An optical flat introduces spherical aberration, but until today I didn't think the effect would be that big. On the other hand, the Otus is so well corrected that even small filter-induced aberrations may be noticed on the optical bench. I don't think the glass plate will have a noticeable effect in the image center with the Planar 1.4/50, for instance.
Thanks for performing these invaluable experiments. Personally, I would try to gain more insight first by repeating the experiment with a few different lenses, at various apertures. Out of curiosity I would also try glass plates (or regular filters) in front of the lens.
p.1 #9 · The Glass in the Path: Sensor Stacks and Adapted Lenses
So, my Zeiss ZF lenses, that I really like on film (Ex: 28/2; I don't see the CA that others say are there) may not be as good on a DSLR with a significant sensor stack. Say D800 vs. D800E.
p.1 #10 · The Glass in the Path: Sensor Stacks and Adapted Lenses
carlitos wrote:
So, my Zeiss ZF lenses, that I really like on film (Ex: 28/2; I don't see the CA that others say are there) may not be as good on a DSLR with a significant sensor stack. Say D800 vs. D800E.
BTW - thanks for the insight...
The fact is we don't know if the Zeiss engineers designed these lenses with or without a filter stack in mind, and if with, what thickness.
I am pretty sure my ZM lenses were designed without, because they were designed for the film Zeiss Ikon.
p.1 #11 · The Glass in the Path: Sensor Stacks and Adapted Lenses
Fred, Roger, thanks for developping and posting this info. It really helps bring some science into a debate that has, to me at least, sounded at times more like some kind of voodoo ritual.
Now I am going to really take a big plunge. Does this mean that lenses designed for film (i.e. older lenses) may well perform better than newer ones on thin-stack cameras? For example ZM designs over modern Leica ones? That is, of course, all other things being equal, which they never are between different lens designs...
p.1 #12 · The Glass in the Path: Sensor Stacks and Adapted Lenses
edwardkaraa wrote:
The fact is we don't know if the Zeiss engineers designed these lenses with or without a filter stack in mind, and if with, what thickness.
I am pretty sure my ZM lenses were designed without, because they were designed for the film Zeiss Ikon.
Agreed. But, also, does this have some bearing on DxO Mark's ranking of lenses with respect to camera sensor?
p.1 #13 · The Glass in the Path: Sensor Stacks and Adapted Lenses
philber wrote:
Now I am going to really take a big plunge. Does this mean that lenses designed for film (i.e. older lenses) may well perform better than newer ones on thin-stack cameras?
Yes.
For example ZM designs over modern Leica ones? That is, of course, all other things being equal, which they never are between different lens designs...
I don't know whether modern Leica designs take the glass cover into account. The FE Sonnars are designed to compensate for the astigmatism of the A7(r) filter stack. I doubt that the ZE/ZF lenses and Otus have such a correction, since they are SLR lenses with sufficient distance between the exit pupil and the sensor. What remains is the apparent spherical aberration with the Otus at F/1.4, which deserves more investigation.
p.1 #15 · The Glass in the Path: Sensor Stacks and Adapted Lenses
I appreciate the investigations and reports!
As a former µ4/3 user (since November 2008 when the original G1 reached the market, later followed by the OM-D with some FF models in-between) I'm not surprised. The smearing and bad border and corner (on that small sensor!) performance by some usually top notch lenses were discussed back then and the glass stack in front of the sensor was one of the possible reasons (the main one actually) behind the findings.
The A7/A7R glass stack has been discussed in the same way.
Seeing these more detailed measurements confirms what we expected back then. It will be interesting to follow further findings and here is hope for coming articles!
p.1 #16 · The Glass in the Path: Sensor Stacks and Adapted Lenses
You can still use nicely all kind of third party lenses including many Leica rangefinder ones on the Sony A7 series mirrorless camera without negligible performance loss.
By experience and from studying a lot of test images since the release of the Sony marvels I think some can be used and some not depending on how picky you are. The Summilux-M 50/1.4 ASPH - what do you think about that lens combined with an A7(r)? Not to mention many of the WA lenses, including some for SLR cameras.
p.1 #18 · The Glass in the Path: Sensor Stacks and Adapted Lenses
Jonas B wrote:
I guess that is literally?
By experience and from studying a lot of test images since the release of the Sony marvels I think some can be used and some not depending on how picky you are. The Summilux-M 50/1.4 ASPH - what do you think about that lens combined with an A7(r)? Not to mention many of the WA lenses, including some for SLR cameras.
Yep - this was discussed in detail in a recent Leica post here in this forum. Some Leica lenses work better than others as I said. I personally use the Summicron-M 2/50 as my only Leica lens on the A7R which works fine.
p.1 #20 · The Glass in the Path: Sensor Stacks and Adapted Lenses
Toothwalker wrote:
I don't know whether modern Leica designs take the glass cover into account. The FE Sonnars are designed to compensate for the astigmatism of the A7(r) filter stack. I doubt that the ZE/ZF lenses and Otus have such a correction, since they are SLR lenses with sufficient distance between the exit pupil and the sensor. What remains is the apparent spherical aberration with the Otus at F/1.4, which deserves more investigation.
In fact I suspect that at least earlier ZE/ZF designs actually did not take the sensor stack into consideration in their designs, either because they expected the lenses to be used on film too, or maybe it was because the sensor stack influence wasn't considered at all.
Earlier ZE/ZF reached infinity way before the infinity hard stop, which may indicate the flange distance at least was calculated for film.
I would be surprised though if newer designs do not take the glass thickness into account.
