Leitz Lenses and design software eng. view

M

Mustafa Umut Sarac

Member
Local time
4:52 AM
Joined
Jan 18, 2008
Messages
14
View My Gallery
Hello ,

I used two summitar , one elmar and one summicron after selected them using some 50 years old magazines and the pictures at there.

I was believing that every leitz lens have a specific character , summitar is better at smooth skin tones , elmar for automobile , architecture or mountains , summicron is for powered colors.

I decided to analyze these lenses with lens design software and simulate their aberrations at pictures.

But a lens design software engineer - which has 40 years work on single software development - told me that he is not believing all of these and these told effects have no business with lenses but with films , developments and papers !

Some said , film is changing the image more than the lens.

May be Leicas are sharper because tighter tolerances and better mtf for their price.

Are they right ? or is there something they dont know ?

Best ,

Mustafa Umut Sarac

Istanbul
 
Every optical design is a compromise, a trade-off of corrections, price, and performance. Some lenses are "warm", ie the coatings and optics favor the warm tones. Others are "cold", have more of a blue cast to them. Load up a roll of slide film, shoot it with your various lenses, and look at the results.

"warm" lenses will do better with skin tones than one with a blue cast. Now- film and filters used to print color negatives make a huge difference independant of what lens was used. But- "in the day" that color slide film ruled, the color-cast of the lens had a much bigger role.
 
In any kind of recording/playback, the main, the first issue always is the front end of the sequence. What was not captured by the lens (because it was a Brownie camera one used) will never appear lateron in the chain. So, the lens is the first and governing factor in every picture you take. Never mind that afterwards we might send the film to a Wmart who ruins it ... This rules with optics, acoustics, health, ...

Just remember : garbage in, garbage out. Think processed food here, too
 
Dear Mustafa,

Everyone has favourite lenses for certain subjects. These preferences are a complex cocktail of equipment, materials, subject matter and technique (exposure, processing, printing...)

I too have talked to lens designers about this, and they too have their favourite lenses for different applications, though few would be as specific as you are. One Zeiss designer of my acquaintance has five 50mm lenses, all of which he uses.

The mistake, I think, lies in trying to tie it down too much. Do it your way; learn what you can from others; but be very hesitant indeed in suspecting that you have discovered specific and universal truths.

Cheers,

R.
 
Lens software designer says that color factor only changes with coatings not lenses ! I forgot to mention this , this mans have 40 years experience on lens design.
Good point , if you know how to remove and recoat your lens with different cast coatings , you hit the bulls eye.
May be you can use a uncoated lens and coated filters or or..
sitemistic is a guru 🙂
 
Mustafa Umut Sarac said:
But a lens design software engineer - which has 40 years work on single software development
Click to expand...

Hi Mustafa,

even though your friend is exagerating (he did not develop
software for lens simulation/design 40 years ago), there is some
truth to what he says.

Input and Output devices to a recording/replay flow are the
most important components. Like turn-table head and speakers
in an audio system. The amplifier / camera is less important.

Like in audio systems, the input device (lens vs turn-table/CD player) has
improved over the last 50 years - modern lenses are much harder
to distinguish than older lenses. And the wet printing process is still
basically the same (just as speakers in the audio world) and
therefore becoming more important (I'm not considering
digital output and photoshop).

For example, a classic 50/1.5 Zeiss Sonnar lens is easy to distinguish
from a classic Summarit. A modern Summicron is very hard to distinguish
from a modern Zeiss Planar, IMO.

Best,

Roland.
 
Last edited:
Ferider's answer makes sense to me. Today there are better optical glasses and there is greater insight into the subtleties of design. Designers have gone beyond the correction of the seven Seidel abberations, to correct higher-order ones that were not even know about 40 yeaqrs ago. I imagine most designers today strive for perfection in correcting and balancing these higher-order abberations. As they do so, they must be all converging on mathematically ideal solutions which make their various lenses more similar than different. So it makes sense that today's lenses all produce results that should not differ much from each other. I guess we can expect, from here on out, high resolution and low distortion, probably at the expense of some smoothness, creaminess, and bokeh.

The pleasing personalities of older lenses is a consequence of the undercorrected spherical abberation, as well as small amounts of astigmatism, coma, etc. Ferider's audio analogy is irresistable here. Second harmonic distortion, while technically a fault, is agreeable to the ear; so an old Marantz vacuum tube amplifier sounds more "musical" to the ear than a highly corrected transistor design: the audio equivalent of, perhaps, a Summar. Some solid-state amplifiers that measure perfectly on the test bench, sound grainy and irritating: the equivalent of some aspherical, or highly corrected lenses.

You might say that the preference of some photographers for the older, classical lenses, is parallel to the insistence of some audiophiles for tube gear. It reminds me a bit of impressionism arising in painting, as a reaction to what some considered excessive attention to details among certain artists of the time.
 
sitemistic said:
Universal truth: the lens you want but don't own is always better than the one you have. 🙂
Click to expand...
Don't think so. I'd rather have my 1,5/50 C-Sonnar than a Noctilux, but I'd like a Noctilux as well. I can't think of any other 50 I'd really like -- and I've tried quite a lot of them.

Cheers,

R.
 
ferider said:
Hi Mustafa,
even though your friend is exagerating (he did not develop
software for lens simulation/design 40 years ago), .
Click to expand...

I don't know exactly when E. Leitz started using computers for lens design. But I do know that Rodenstock first began using computers for this in either 1975 or 1976. This was a big deal in German optical circles at that time. The computer was a MODCOMP IV running the MAX IV operating system. I can't remember if it had 16K or 32K of memory.
 
> even though your friend is exagerating (he did not develop
> software for lens simulation/design 40 years ago), there is some
> truth to what he says.

I would not assume that. "Invent" as used here probably means developed his own custom code and algorithms. One of the optical engineers that worked for me did custom software for lens design 40 years ago. The package is still around, written in FORTRAN. It was not uncommon for a designer to write custom software. I worked with X-Ray optics software written in FORTRAN when I started working 30 years ago. It had been around for a while even then. One of the first lens design software appeared in the 1950's, I think it was for an IBM "7094" Stretch computer. Read about it in an old journal, probably Optical Society of America decades ago. This machine had 100KBytes of memory. The machine I used in 1979 had 4MBytes of memory. It was very big, used 256BITS per memory chip and was 100MFLOPS.

As far as optical design- correction for chromatic aberration plays a factor in rendering color. Leitz lenses have always tended to be highly corrected for chromatic aberrations. The IR Focus index is very close to the prime focus, ie very little color shift. My first job was optimizing the code that the Physicists wrote. The machine cost $960 per CPU hour. I re-wrote the code to make use of the vector/parallel instruction set on the TI-ASC. Most advanced instruction set of any computer I've ever used.
 
Last edited:
Yes computers became available some time before the advent of the home computer. An electronics & space company I worked for had an IBM System 360, circa 1970. I imagine an optical company could easily buy time on an IBM or Control Data system by that time, even if they didn't own one.
 
Brian Sweeney said:
"Invent" as used here probably means developed his own custom code and algorithms. One of the optical engineers that worked for me did custom software for lens design 40 years ago. The package is still around, written in FORTRAN. It was not uncommon for a designer to write custom software. .
Click to expand...
Dear Brian,

Very true. I remember Sir Kenneth Corfield telling me about using computerised lens optimation in the 1950s, with one of the University of London computers (Imperial College?). We were talking about wave-front design (not ray tracing) and as far as I recall -- it was a casual conversation, some years ago -- he said that wave-front design was also explored in the 50s but took too much computing power.

Cheers,

Roger
 
I was commenting on which has 40 years work on single software development. Yes there were computers back to the Zuse 1, but
in terms of easy access, there is the time before and after the Apollo
program.

Roland.
 
You should meet some of the old-timers that I work with. The Senior optical engineer that just retired used the same optical design package for decades. I'm still using software that I wrote in 1981, and getting paid for it! I'll still be using it in 10 years. It does the job, and is unique.
 
Last edited:
I would guess you are paid by big brother 😀
In my world big packages get re-implemented every 5-10 years ....
 
I love writing code. A good day is over 100 lines, a great day is 300 lines, and if I hit 500 lines I take the rest of the day off. I tell people that "in my day, everyone was like this. Then the Asteroid came crashing to Earth, wiping out the rest of the dinosaurs."

Mustafa's friend probably survived the hit as well.
 
Last edited:
I learned from this thread that
summitar , summaron or elmar was not computer generated.
there are many aberrations not corrected at them.
the corrected aberrations are even not perfectly corrected.
And all of these errors create a tube amp. taste, all the world liked.
Now ı will ask few questions in a row.
First one : is there any clue about some of these aberrations artificially added ?
May be a book , leitz scientific paper , letter etc ?
 
You must log in or register to reply here.
Back
Top Bottom