28 'cron falls flat

[lars_bergquist]

Advertisement (gone after registration)

A couple of days ago, I did show some preliminary vignetting tests of the 28mm Summicron ASPH, with results that clearly did not shock just me, but others too. Here are some differents tests with the lens, all shot with the aid of an ExpoDisc, which furnishes an even and completely afocal light. These tests should be more fair than the preceding ones.

 

All three shots were taken wide open. The first (top) one with the lens automatically identified as itself, the second (middle) as a pre-ASPH 21mm Elmarit, and the last one as a 24mm Elmarit ASPH. All identifications were verified by info menu and subsequently by EXIF file.

 

I must say that it is something of a mystery to me why the lens, using proper factory identification, performs so badly, compared to the results when the lens masquerades as either of two adjacent focal lengths. If anybody could enlighten me, I would be happy to listen -- especially if this someone worked in Solms ... --So here we go:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[ArtZ]

A couple of days ago, I did show some preliminary vignetting tests of the 28mm Summicron ASPH, with results that clearly did not shock just me, but others too. Here are some differents tests with the lens, all shot with the aid of an ExpoDisc, which furnishes an even and completely afocal light. These tests should be more fair than the preceding ones.

 

All three shots were taken wide open. The first (top) one with the lens automatically identified as itself, the second (middle) as a pre-ASPH 21mm Elmarit, and the last one as a 24mm Elmarit ASPH. All identifications were verified by info menu and subsequently by EXIF file.

 

I must say that it is something of a mystery to me why the lens, using proper factory identification, performs so badly, compared to the results when the lens masquerades as either of two adjacent focal lengths. If anybody could enlighten me, I would be happy to listen -- especially if this someone worked in Solms ... --So here we go:

 

Lars,

 

I think there's a mistake. According to the pictures names, the second (middle) picture is the 24mm Elmarit ASPH and the third picture (buttom) is the pre-ASPH 21mm Elmarit. You said the opposite.

 

Anyway, I find the 3rd picture the best one. Yes, it's wierd that coding the Summicron 28 ASPH as an Elmarit 21 pre-ASPH gives better results.

 

Interesting, indeed.

 

Best

[lars_bergquist]

Lars,

 

I think there's a mistake. According to the pictures names, the second (middle) picture is the 24mm Elmarit ASPH and the third picture (buttom) is the pre-ASPH 21mm Elmarit. You said the opposite.

 

Right, you caught me!

 

But this ExpoDisc technique seems to work well. I will soon do some testing of the Zeiss 18mm Distagon, which also performs badly when coded as a WATE. Maybe that one too will do best as a 21mm Elmarit ...

 

The old man who now will wash his mouth.

[ArtZ]

Right, you caught me!

 

But this ExpoDisc technique seems to work well. I will soon do some testing of the Zeiss 18mm Distagon, which also performs badly when coded as a WATE. Maybe that one too will do best as a 21mm Elmarit ...

 

The old man who now will wash his mouth.

 

I won't tell BB -promised!

 

This is an interesting pick. In other words, Leica could maker better better profiles for lenses. It would be also interesting to see what happens with the M8 cropped sensor...

[Nitnaros]

Potentially the correction for the 21mm is stronger;

 

If the exposure and ISO values allow stronger correction, the 21mm yields more aggressive correction than the default 28mm correction....?

 

...Just a thought, based on the assumption that the default correction will be conservative to avoid overcorrection in all situations.

 

Peter

[Jamie Roberts]

Lars, I'm not convinced the third one is the best. It measures all over the place for colour. The seoond one is downright horrible.

 

If the the first one is for the 28 cron and you're also shooting with the 28 cron, that's the one that shows the least colour weirdness and in my opinion is the best correction of the three by far.

 

The vignette seems strong with the Expodisc, I admit, but that's partly the middle-gray exposure and from Leica's perspective probably keeps the "film" character of the lens and is nothing that couldn't be quickly corrected (or not!) in post.

 

Personally, I'd take colour neutrality of capture over an over-correction for vignette.

[zlatkob]

Advertisement (gone after registration)

Just a guess ... perhaps the 1st shot provides better correction for color, despite the greater vignetting. The 2nd (especially) and 3rd shots seem to have a mix of cyan and magenta, whereas the 1st seems to be more uniformly cyan. The 3rd shot has the least vignetting, but there is a shift to cyan in the corners.

 

ETA: Jamie, you had the same thought, but faster!

[Jamie Roberts]

Just a guess ... perhaps the 1st shot provides better correction for color, despite the greater vignetting. The 2nd (especially) and 3rd shots seem to have a mix of cyan and magenta, whereas the 1st seems to be more uniformly cyan. The 3rd shot has the least vignetting, but there is a shift to cyan in the corners.

 

Right--and those colour shifts compensate for the different angle of entry of light the wider 24 and 21 lenses need with regard to the M9 sensor and the microlenses, if I understand correctly. So they induce colour errors with the 28, which doesn't need the same amount of correction.

 

The first shot is way better for colour (we definitely cross-posted, Zlatko ). Could Leica me more agressive with vignette adjustment for the 28 cron? Probably. I like a light-hand on capture adjustments though

[sandymc]

Umm, be aware that the ExpoDisc is NOT a good way to measure vignetting. It vignettes in of itself, in unpredictable ways.

 

This came up when some CornerFix users were getting really bad corrections using the ExpoDisc. After I contacted the ExpoDisc's manufacturers they confirmed that the ExpoDisc vignettes. Not a problem in its intended application, but a problem when used this way.

 

Regards,

 

Sandy

[zlatkob]

Leica's perspective probably keeps the "film" character of the lens and is nothing that couldn't be quickly corrected (or not!) in post.

That seems reasonable ... vignetting is very film-like, whereas funky color shifts are just strange and harder to fix.

[ArtZ]

Oops,

 

You're right! I have checked in my calibrated monitor and I can see funky colors I couldn't see before in my laptop.

 

I think it would be interesting exemples in real life.

 

Cheers

[charlesphoto99]

First one looks best to me and I'm on an iPhone. I'm glad that Leica are maintaining some of the true vignetting quality of the lens. It's part of what makes wide open photography distinct.

[lars_bergquist]

If you like the factory correction best, you are welcome to it. To each his own and no arguing about taste and geese and ganders and all that. But I don't.

 

I don't like artificially applied fake 'film grain' either. Not that I was too enamored of the real article.

 

The interesting thing is that my old v.4 35mm Summicron, as retro-coded by Leica, shows excellent colour and practically no vignetting at all -- even through its on-film vignetting is on the same order as that of the 28. Actually, it is fractionally less. See the embarrasingly non-scientific picture below.

 

The old man seeing red (oh well, magenta)

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[brill64]

and i was thinking only the other day just how magical this lens is on my m9..must be something wrong with me..

[lars_bergquist]

and i was thinking only the other day just how magical this lens is on my m9..must be something wrong with me..

 

You are to be congratulated ... but I am too old and ornery to follow you. Yes I loved hat lens on the M8, but M9 -- no.

 

The same old man

[lct]

I see a strong cyan cast in the 28/28 sample, less so in the 28/21 but the latter shows a bit more green cast than the former IMHO. Very few magenta in both. (my monitor is not out hopefully). All in all i prefer the 28/21 personally.

[viramati]

Not one to normally complain about these things but I have noticed that my 28 cron is causing problems. on the auto setting I seem to be getting quite a bit of magenta edging (can be seen on post here http://www.l-camera-forum.com/leica-forum/leica-m9-forum/119369-sane-attitude-rededge-7.html ) and a certain amount of vignetting. The new FW seems to have made no difference to this even though they say it is included in the upgrade and in fact appears as if it may be even worse. Interestingly my zeiss biogon 24 now seems to work better on the auto setting (has coded milich adaptor) and shows less magenta and vignetting than the 28.

Now what really concerns me is that Leica have overstepped themselves with the full frame sensor and that it has intrinsic problems that will never be solved by firmware upgrades in which case were all in possession of a fantastic but again flawed camera. If this is the case and an M10 comes out or M9.2 with a different sensor I hope that all of us who have spent our hard earned cash will be first in line for at least free upgrade or rebate on new model. Lets hope I'm wrong and that this can be sorted in the FW and hopefully sooner rather than latter. last thought I am also worried that all of this manipulation of the file in the FW will be causing image degradation but not being a techno freak maybe someone can enlighten me on this

[UliWer]

... It would be also interesting to see what happens with the M8 cropped sensor...

 

I posted an example with the 28-cron on the M8 here:

f/2:

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/119369-sane-attitude-rededge-3.html#post1268579

 

and f/5.6:

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/119369-sane-attitude-rededge-4.html#post1268635

 

You see the red shift in the corner, though in practical use this very, very rarely becomes obvious.

 

The 28-cron has considerable vignetting if you look at the data sheet:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

Though I don't really understand the patch of light in the center of Lars' example with the coding for the 28-cron, which does not show when it's coded as a 21-Elmarit. The 21mm-example seems to me to have a slight red tint in the center, but none on the edges.

[ArtZ]

I posted an example with the 28-cron on the M8 here:

f/2:

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/119369-sane-attitude-rededge-3.html#post1268579

 

and f/5.6:

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/119369-sane-attitude-rededge-4.html#post1268635

 

You see the red shift in the corner, though in practical use this very, very rarely becomes obvious.

 

The 28-cron has considerable vignetting if you look at the data sheet:

[ATTACH]193933[/ATTACH]

 

Though I don't really understand the patch of light in the center of Lars' example with the coding for the 28-cron, which does not show when it's coded as a 21-Elmarit. The 21mm-example seems to me to have a slight red tint in the center, but none on the edges.

 

Uli,

 

I know the Leica graph. What I was saying to Lars it was it could be interesting to repeat his test with a M8.

[Advertisement]

Hallo,

Look here: Leica M

[ArtZ]

If you like the factory correction best, you are welcome to it. To each his own and no arguing about taste and geese and ganders and all that. But I don't.

 

I don't like artificially applied fake 'film grain' either. Not that I was too enamored of the real article.

 

The interesting thing is that my old v.4 35mm Summicron, as retro-coded by Leica, shows excellent colour and practically no vignetting at all -- even through its on-film vignetting is on the same order as that of the 28. Actually, it is fractionally less. See the embarrasingly non-scientific picture below.

 

The old man seeing red (oh well, magenta)

 

Yes! This picture seems good to me. The funky color on the first image I could see on a calibrated monitor, disapears completely here.

 

Thanks for posting, Lars