PDA

View Full Version : unscientific! LF vs 35mm lens test



thafred
3-Jul-2009, 06:32
Hi

I had one sheet of film left after a photo day and had the idea to waist it on a comparison of my newly aqired Super Angulon 90 f8 MC (ca. 1991) to the very good Tele-Elmarit 90mm on my Leica M. the plan was to cut the 4x5 sheet down to scan it at 4000dpi on my Coolscan V

I allready knew that the SA90 is a cracking lens (my first Slides were turning out amazing) but I thought that the TE90 (beeing a 35mm lens) will smoke it even on 400asa film.

the results are below. Super Angulon 90 f8 @ f16 was shot on T-Max 400 4x5" film and the Tele Elmarit 90mm f2.8 @ f5.6 shot on TriX. (both two stops down) all films developed in Rollei RHS developer (Amaloco AM74)

I know itīs flawed because the films are different and the focal lenghts do not match completely (or is it because the film plane of the LF camera was further back than the 35mm??...tripod position didnīt change)

anyway I think itīs an impressive performance for a 90mm lens that is covering 220mm!

hope you enjoy

http://farm3.static.flickr.com/2676/3684507212_a69a53c61d_o.jpg

http://farm3.static.flickr.com/2605/3684507124_3d65d63a8f_o.jpg

Tele-Elmarit on the left, Super Angulon on the right

Ivan J. Eberle
3-Jul-2009, 07:27
It'd been a much better comparison at the same f/stop. You're easily reaching the limits of film resolution with the Tele-Elmarit, while the SA 90mm shots appear to show softening due to the f/16 diffraction limit... at least to my unscientific eye, anyway.

ic-racer
3-Jul-2009, 09:33
Should have tested them both at f5.6. It does, though, show that one of the biggest factor in across-format lens comparisons is aperture.

Toyon
3-Jul-2009, 12:47
I have seen your findings replicated under more rigorous scientific conditions. It indicates that the chief advantage of large format is the small grain size, not the acuity of the lenses.

sun of sand
3-Jul-2009, 14:22
The bottom comparison is much more telling
Top LF moderate improvement
35mm appears nearly as sharp but is quite jagged/soft relying on contrast

Ivan J. Eberle
3-Jul-2009, 16:26
The micro-contrast-is-inherently-smoother argument between LF and 35mm has no merit in 100% crops of the film area using the same focal length lens (and the comparison would have zero point if they're not).

Athiril
3-Jul-2009, 18:26
There is plenty of merit in 1:1 comparison of focal length shot from the same distance, it allows you to compare the optical resolution of each lens.

The diffraction limit for 4000 dpi worth of detail is f/11.6, so I would have liked a f/11 vs f/11 and f/8 vs f/8, and on the same film stock.

Eric Leppanen
3-Jul-2009, 19:22
This is not really a true LF vs. 35mm test in that a wide angle LF lens is being pitted against a slightly longer-than-normal 35mm lens (advantage 35mm on that basis alone). IMO the main benefit of the test is that the LF lens holds up reasonably well despite this disadvantage.

Many LF lenses are optimized for around f/22, and opening up the SA beyond f/16 would not significantly benefit its performance (see Chris Perez's lens tests at http://www.hevanet.com/cperez/testing.html).

walter23
3-Jul-2009, 21:18
This is quite remarkable. I've never done a side by side comparison like this but my intuitive sense was also that a good 35mm lens would smoke an LF lens when compared over the same film area. This shows I may have been wrong; the lenses may be very good as well. But hey, there is the aperture variable, and I don't know the MTFs for my lenses (and I don't care either). This is close enough to be compelling for me.

Athiril
3-Jul-2009, 21:46
If LF lenses are optimised for f/22, then a 35mm lens is going to destroy it in resolving power on a 1:1 comparison basis (over the same area size, or per square mm, or per square inch, whatever you want to use) as f/22 is going to severely limit the resolution by comparison.

walter23: there are much sharper ~90mm lenses around than the elmarit, it also has room to be stopped down as its well under the diffraction limit for 4000 dpi.

thafred
4-Jul-2009, 01:33
interesting points everyone!

the reason of shooting the SA 90 at f16 is that I wanted to use my widest working aperture for the test. since the widest stop is f8 I highly doubt that it would have fared better at that stop...maybe f11 would be sharper than f16 in the very center of the frame but Iīd never use it that way ...

I shot a frame with the Elmarit at f16 too but itīs much sharper at f5.6 (wich I think is the sweet spot) at f16 the SA smokes it!

@athiril...wich lens is *much* sharper than the Tele-elmarit at f5.6?? puzzles me..I used to shoot a APO ASPH Summicron 90 (allways borrowed it from a more wealthy friend) and itīs probably the best 90 around (insane MTF)...at f5.6/8 the sharpnes advantage in practical situations was gone and I prefer my Tele Elmarit for itīs weight.

Athiril
4-Jul-2009, 06:59
The Canon 70-200mm f/4L at any focal length at any aperture, the Canon 85mm f/1.2L, and most likely the Korean Made Polar (rebadged as Samyang and Vivitar) 85mm f/1.4.

thafred
5-Jul-2009, 01:00
Athiril, I donīt doubt that the lenses you specified are very good glas but to claim that they are much sharper is just bold. do you actualy have first hand experience with all the lenses discussed?

hereīs an example of what Iīm talking about. much sharper would only be possible with a better scanner and film....the lens certainly isnīt the limiting factor (Acros in T-Max developer)

thafred
5-Jul-2009, 01:26
This is quite remarkable. I've never done a side by side comparison like this but my intuitive sense was also that a good 35mm lens would smoke an LF lens when compared over the same film area. This shows I may have been wrong; the lenses may be very good as well. But hey, there is the aperture variable, and I don't know the MTFs for my lenses (and I don't care either). This is close enough to be compelling for me.

Hi Walter! exactly what Iīm thinking...close enough to be compelling :-)

fact is...Iīll never have a scanner capable of resolving that much detail in a 4x5 sheet..a V700 is on my shopping list and Iīve had amazing results with that scanner from LF slides but itīs not comparable to the resolving power of the Nikon Slide scanner...

Beeing more interested in real world results I will not waste another 4x5 sheet on a test like that...but this short test comforts me that the lenses might be much better than I hoped!

Athiril
5-Jul-2009, 01:53
Some not all, but objective indepedant testing is superior to subjective first hand experience.

The 70-200mm f/4L IS resolves past 4000 dpi (20D, 30D, 350D, 5D Mk II) wide open, and Im sure Ive seen it doing the same on a 50D (5400 dpi), itd be one of the lenses to test CMS 20 with.

But this was never really the point.

boris
5-Jul-2009, 02:18
thank you for this interesting comparison.
i never qustioned the fact that 35mm lenses were sharper. i have to admit that it is even more amazing, that a lf lens at f16 can compete with a leica lens at f 5.6 :eek:

i love the realword results!
thanks again, boris

thafred
5-Jul-2009, 02:33
Some not all, but objective indepedant testing is superior to subjective first hand experience.

The 70-200mm f/4L IS resolves past 4000 dpi (20D, 30D, 350D, 5D Mk II) wide open, and Im sure Ive seen it doing the same on a 50D (5400 dpi), itd be one of the lenses to test CMS 20 with.

But this was never really the point.

OMG...do you compare the output of a bayer matrix sensor to a RGB film scan? I wonīt go there..no way!

IMHO Having first hand experience will allways be preferable to repeating "independent" lens test from the internet...

thafred
5-Jul-2009, 02:34
i have to admit that it is even more amazing, that a lf lens at f16 can compete with a leica lens at f 5.6

glad you like the test Boris! exactly my thoughts and the reason I posted this here :-)