Enlarging lens performance

[trendland]

there are lenses with three, four, six, seven and more elements

in any focal length some of the quality factors that matter are:

maximum aperture (faster lenses can be used with shorter exposure times, also they are brighter wide open and easier to focus)
corner darkening (all lenses suffer from this to a certain extent, it makes the corners of a the print lighter than the center which is often aesthetically displeasing)
flatness of field (are the corners and the center of the negative in focus at the same time at all apertures?)
chromatic aberration (how much red and blue blur is there? regular achromat lenses are corrected for two wavelengths red and blue, apo-chromatic lenses are corrected for three wavelengths)
range of enlargement factors it was designed for (for example 2x - 6x, outside this range results may be slightly inferior)

most lenses improve with closing the aperture

the Rodenstock catalog describes the quality gradations of their lenses in clear language, see the attachment (with apologies for the loss of some line justification, this is an extract, the original is too large to post unfortunately)


PS: better quality lenses offer three advantages: they are a pleasure to use; they deliver high quality results; they hold their resale value.

Rodenstock (sorry my tabled has changed into Rosenstock because it is not familiar with any kind of lens manufacturers but it is familiar with all kind of botany.....?)

So - R O D E N S T O C K (hope tablet has learned this new vocabulary now) is a choice you can
not make anything wrong for sure!

But pricing is increasing more and more (small series) I remember highest price about
USD 4000,- so I should not go onto there side (for some month)!

with regards

PS : Doctor (possible heard attacks [just a joke no fear]), budget and other interists have restricted
any corespondence with Rosenstock to me!

Perhaps I will ignore restrictions!

 

[Ron789]

I did many comparisons between EL-Nikkor, Schneider (Componon-s), Rodenstock (Rodagon), Leitz and Computar lenses. At full opening there are differences. Stopped down, all these lenses give very good results. EL-Nikkor typically gives a bit more contrast, German lenses tend to be a bit softer but give better details.
For a recent exhibition I printed one photo that, much to my surprise, using a Componon-s 50 F/2.8 stopped down to F/5.6 or 8 did not produce the snappy image I had printed before, using a Leitz Focotar. I then switched back to a Focotar-2 and with that lens I was able to get the print that I wanted. So yes, there are differences and in some cases those differences can be significant.

 

[StepheKoontz]

Another positive vote for the Minolta CE, I have an 80mm f5.6 version that is great. That said, I've used a rodenstock 105mm f4.5 ysaron (it's in a polaroid/prontor self cocking shutter, no idea what it came off of) that works fine for enlarging 6X9 negs.

 

[alanrockwood]

Have a look in the book from Ctein who provides the criteria of a good enlarger lens and why a good lens should be used:
http://ctein.com/PostExposure2ndIllustrated.pdf
The book is currently free to download and it provides many details I was not aware.

I downloaded the pdf of the Ctien book and started reading it, particularly the section on enlarger lenses starting on page 57. From what I have read so far, from a scientific perspective this is probably the best and most informative information I have ever read in a photography-related work, particularly when considering the balance between practical and fundamental information and the balance between experimental work and the underlying physical theory.

I am extremely impressed. I wonder if the other chapters are of the same high quality. I hope so.

A note of apology. After making the post above I realized I misspelled Ctein.

 

[Deleted member 88956]

... I am extremely impressed. I wonder if the other chapters are of the same high quality. I hope so.
...

The short answer is an absolute yes. The long one ... there is a lot less to making quality photographs, but one thing needed that cannot put in a book of any length by any scholar is ... persistence.

 

[John51]

I did many comparisons between EL-Nikkor, Schneider (Componon-s), Rodenstock (Rodagon), Leitz and Computar lenses. At full opening there are differences. Stopped down, all these lenses give very good results. EL-Nikkor typically gives a bit more contrast, German lenses tend to be a bit softer but give better details.
For a recent exhibition I printed one photo that, much to my surprise, using a Componon-s 50 F/2.8 stopped down to F/5.6 or 8 did not produce the snappy image I had printed before, using a Leitz Focotar. I then switched back to a Focotar-2 and with that lens I was able to get the print that I wanted. So yes, there are differences and in some cases those differences can be significant.

I've noticed this with some camera lenses. I'd guess it's to do with resolution vs actuance and personal preference sets the priority. Interesting that enlarger lenses have similar differences.

 

[MattKing]

A note of apology. After making the post above I realized I misspelled Ctein.

I'm sure he won't mind. Now lets hear how you pronounce "Ctein".

 

[whojammyflip]

I suspect that individual sample of a given lens design will vary more than differences between competing brands of similar design.

This is a mathematical impossibility. If the designs were so similar between brands as to be the same, then the variance within that sample of two brands would be the same as the variance of a sample from a single brand. This phrase is just repeating something I have previously challenged on Photrio, which apparently originated with Ctein.

 

[reddesert]

This is a mathematical impossibility. If the designs were so similar between brands as to be the same, then the variance within that sample of two brands would be the same as the variance of a sample from a single brand. This phrase is just repeating something I have previously challenged on Photrio, which apparently originated with Ctein.

The standard deviation within a brand could be larger than the difference of the mean of brand 1 and the mean of brand 2, and that is probably what whoever said this first meant.

 

[RalphLambrecht]

Given that a hell of a lot of enlarging lenses are sold nowadays as second, third etc. hand and that, even when new, there are bound to be manufacturing and QC differences between a manufacturer's model and another, how useful are threads rating different enlarging lenses? Then of course, there is the case of lenses made by an unknown third party that may be sold under several different brands. There appears to be a standard answer of, "it has to be Schneider, Rodenstock or El Nikor", as if those companies were the only ones engaged in serious lens manufacture.
The second question is what differences could the average user notice even when printing up to a maximum size of, say, 20x24"?

a satisfying fact is however that most enlarging lenses are very good;it's rare to get a realy bad one;Hence the recommendation to stick to a name-brand and just go for it.

 

[RalphLambrecht]

This is a mathematical impossibility. If the designs were so similar between brands as to be the same, then the variance within that sample of two brands would be the same as the variance of a sample from a single brand. This phrase is just repeating something I have previously challenged on Photrio, which apparently originated with Ctein.

Ctein wrote the book on enlarging lenses and his advise is good.

 

[mmerig]

I suspect that individual sample of a given lens design will vary more than differences between competing brands of similar design. .

This is a mathematical impossibility. If the designs were so similar between brands as to be the same, then the variance within that sample of two brands would be the same as the variance of a sample from a single brand. This phrase is just repeating something I have previously challenged on Photrio, which apparently originated with Ctein.

What voceumana said is possible, but perhaps unlikely and difficult to measure. Quality control (lens grinding, centering, etc.) could account for more variation than just the design does, so variation across brands could be wider than variation between them.
Here is a crude graphical representation of this possibility:

From a more practical standpoint, closing down the iris a few stops could make up for a lot of the differences between lenses..

 

[Patrick Robert James]

I tested all the lenses in my drawer a few years back for kicks. I put my Saunders 4550 all the way up and went to town. At that size I think the print is like 21x or something, so pretty significant. The 60mm S-Orthoplanar took the prize followed closely by the 65mm Computar DL. Those two of course had less magnification than the 50s. The top of the heap of the 50s was the Minolta C.E. Rokkor-X which was almost as good as the Orthoplanar and the Computar. I've since sold some lenses and bought some more so i should probably run those through with the same neg. My general conclusion is if you want to make a big print and you don't want to buy an expensive lens, use a Minolta or Fujinon EX over the German non APO. Since the German APO lenses aren't going to out resolve the Orthoplanar the Minolta should be just as good for a fraction of the price.

At smaller sizes there really isn't much difference in resolution between lenses. Japanese lenses tend to have more contrast overall. Rodenstock lenses are in between and Schneider has less overall contrast. Those are generalizations.

 

[wyofilm]

A useful illustration. It highlights that there is no data out there to inform us about variation within one model of lens (manufactures might have made such measurements). Unfortunately, this means that without knowledge of variance within a single/brand model, any comparison between brands is largely useless for lenses of the same design. The only difference I would make to your illustration is to make the ranges gaussian curves as a first approximation.

A look at Ctein's table in Post-Exposure really tells me almost nothing because he provides no analysis of error. Mind, his book is very informative in other ways. It is just that table that everyone refers to that I find somewhat objectionable.

 

[mmerig]

The only difference I would make to your illustration is to make the ranges gaussian curves as a first approximation..

Thanks for the feedback.

Sure, bell curves would be more realistic. But I kept it simple (or as I said, crude). The Y-axis is a nominal variable, "Brand", so it would have been a more complex graph to show frequency distribution, quality, and brand, and I was just using key-board characters to make the graph.

 

[wyofilm]

Thanks for the feedback.

Sure, bell curves would be more realistic. But I kept it simple (or as I said, crude). The Y-axis is a nominal variable, "Brand", so it would have been a more complex graph to show frequency distribution, quality, and brand, and I was just using key-board characters to make the graph.

Your graph is great! But our exchange here now covers a bit more complexity.

 

[M Carter]

Interesting discussion. But, as someone who does 90% or so lith printing, mostly at 16x20... I use my lenses wide-open most all of the time. I'm just done with 60-second exposure times, and then throwing in all the burns and the mask exposures. Just kills any enjoyment for me.

But primarily using EL-Nikkors (80 and 105) and a Componon-S (135), I'm in the camp that believes enlarger alignment and neg flatness are bigger issues. Glass carrier, enlarger braced to the ceiling joists, micromega focuser and laser alignment (and alignable lens boards) give me the prints I want, wide-open. Yep, I do understand that the oddball contrast control (or contrast-magic) of lith printing may be making up for contrast deficiencies wide-open - but I get excellent print sharpness and no complaints (as far as glass goes anyway) - I get the prints I want after a lot of work. Sometimes throwing money and gear at a problem won't help if you don't sweat the details.

(If I owned a cheaper lens, I'd still be interested to do an A/B test and see if there's a marked difference though - but I only have the 50mm Beselars that came with my enlargers and I don't print 35mm).

 

[mmerig]

. . . . .our exchange here now covers a bit more complexity.

Not sure what you mean by "exchange here now", but the original statement that the graph addresses seemed to be restricted to just range or variance ("vary" was used). but I don't want to read too much into it. The actual distribution within each brand is not as relevant as the width or domain of the values, and for most analyses, the central limit theorem makes the sample distributions.unimportant.

But I think your comment about not having enough data, and ctein's table being too simplistic, are very good points. It's easy to lose sight of the basics in photography, and my graph and comments may have contributed to this problem.

Thanks again.

 

[wyofilm]

Not sure what you mean by "exchange here now", but the original statement that the graph addresses seemed to be restricted to just range or variance ("vary" was used).

I'm sorry that I moved the discussion off track. The 'exchange' I was referring to was our back and forth on a bell curve. Nothing to deep. Again, sorry for the confusion.

I think that Ctein in his book leading up to his lens table does a great job of discussing issues related to lens function and how those issues relate to print making.

Just drove through Teton Valley on the way up to W. Yellowstone. Pretty nasty weather!

 

[John51]

With camera lenses, sharpness isn't everything. Some lenses are liked for their character, eg. Tessars. Even triplets have their place as some like them for portraits.

Does similar apply to enlarging lenses? What type of print would be better with say a 4 element 50mm Ektar over a 6 element lens?

 

[jjphoto]

With camera lenses, sharpness isn't everything. Some lenses are liked for their character, eg. Tessars. Even triplets have their place as some like them for portraits.

Does similar apply to enlarging lenses? What type of print would be better with say a 4 element 50mm Ektar over a 6 element lens?

Kodak gave it's best lenses the designation Ektar. The Kodak Enlarging Ektar is a 5/3 Heliar design and was extremely well regarded until the proliferation of 'better' 6/4 lenses from the 70s and 80s. It's also coated, despite being from the mid 1940's.

Better lenses typically have sharper details across the frame at wider apertures, possibly with better contrast. A crappier lens might be better suited to portraits, at a guess (not that the Kodak Ektar is crappy).

 

[John51]

'Character' in a camera lens is flaws that create pleasing errors. I'm trying to think of how enlarging lens errors can be useful when printing.

Maybe a tack sharp portrait negative makes for a harsh print but when printed with a triplet has the edges a lot softer than the centre?

 

[DREW WILEY]

Life is easier printing large format film because you can choose from Apo barrel graphics lenses that optically exceed typical enlarging lenses, but now sell for bargain prices because the printing industry has mostly switched to scanners. Worrying about sample to sample consistency with this kind of thing, or even with any reasonably late top-end enlarging lens from any of the "big four" is ridiculous, unless it's a used lens with mold or suffering from careless use. Cheapo "student" lenses are a different story.

 

[wiltw]

I tested all the lenses in my drawer a few years back for kicks. I put my Saunders 4550 all the way up and went to town. At that size I think the print is like 21x or something, so pretty significant....
At smaller sizes there really isn't much difference in resolution between lenses. Japanese lenses tend to have more contrast overall. Rodenstock lenses are in between and Schneider has less overall contrast. Those are generalizations.

One needs to keep in mind in any such testing, that enlarging lens have a DESIGN MAGNIFICATION RANGE...and if the chosen enlargemen take it outside of that range, there is zero understanding about how that same lens will work compared to others, when all the lenses are inside their designed range of magnifications.
The earlier linked Rodenstock PDF clearly points out the design magnification range for a specific lens model. One lens says 2-8X, another says 2-20X, the first lens would suffer in a comparison at 21X, while neither woud suffer at 2X.

Then we also have to consider the limitations of human vision...at 21X our vision might well be sufficient to SEE defects that emerge at 21X, but without 10X magnifier to supplement our vision we might not SEE the same defect present in the 2X print.

 

[Patrick Robert James]

One needs to keep in mind in any such testing, that enlarging lens have a DESIGN MAGNIFICATION RANGE...and if the chosen enlargemen take it outside of that range, there is zero understanding about how that same lens will work compared to others, when all the lenses are inside their designed range of magnifications.
The earlier linked Rodenstock PDF clearly points out the design magnification range for a specific lens model. One lens says 2-8X, another says 2-20X, the first lens would suffer in a comparison at 21X, while neither woud suffer at 2X.

Then we also have to consider the limitations of human vision...at 21X our vision might well be sufficient to SEE defects that emerge at 21X, but without 10X magnifier to supplement our vision we might not SEE the same defect present in the 2X print.

I kind of feel like you are telling me how the moon is to visit in theory and I've already been there...