20х24" enlarger

[Pere Casals]

From the available evidence, the O-2 you mention seems to be an apochromat - which makes me suspect your testing methods rather more than the lens - especially as the design mentioned should be more than capable of focusing green & blue light accurately.

With the LOMO I made informal tests, yes... it seems apochromat to me as there are no colour fringes, so lateral chromatic aberration looks corrected, but I was suspecting that longitudinal chromatic aberration is not very well corrected, an ucranian user told me that they were using monochromatic light to improve results. Perhaps Apochromatism has degrees of performance...

The link states 35 Lp/mm in the center and 12 in the corners... I'd said less, I found LOMO less sharp than a Symmar converted to 620, for 8x10, still the O-2 600 has some 1m circle (945mm?). Then there is the lack of coatings, so if enlarger is diffusion a lack of microcontrast may be expected. Also sovietic lenses had a lot of QC issues, so perhaps my copy is not very good... Still I'm eager to make deeper formal tests with the O-2, this is using a 1951 pattern with CMS 20, and making sure the target is in perfect focus.

From the available evidence

Heiland's LED systems are considerably more sophisticated (and usefully matched to peak paper sensitivities) than the bizarre mish-mash of cheap components you are suggesting. They are expensive for a reason.

Well, that ebay source delivers 100w color LEDs for some $3 each, this is equivalent to hallogen 800w, so 3 leds are like 2400w . Even in the case those leds are not completely optimal they may be very cost effective. For the moment I've only replaced the Philips Photocrescenta bulb of an OPEMUS with a cheap RGB LED bulb (like the depicted one) to test that system, preliminary tests were very succesful, I'm to do extensive calibration to check it well.

Even in the case those LEDs do not match the peak sensitivity of the paper perhaps it is not a problem (more light/time to see what we are burning...), but it should be tested if 00 to 5 grades can be obtained.

[Pere Casals]

Heiland's LED systems .... ....They are expensive for a reason.

As you pointed it, I've been reading that. http://heilandelectronic.de/files/do...t_Brochure.pdf

It is a good product, well finished, and straight, a good option. It cannot be cheap... Still diffusion only is available.

I'm asking myself how CRI (color reproduction index) of the light source may impact fidelity in color work. I bought a box of 8x10 Fuji CDU II transparency copy film and I plan to copy Velvia/Provia transparencies, and enlarging MF slides to 8x10 slides, perhaps this can be known by making contact copies of an IT8 slide target, under different light sources...

For what's WB I think that if blue channel is able of rendering grade 5 and green channel can deliver grade 00, I think that it has to work well. A problem would be a too greenish blue channel or a too bluish green channel , then perhaps only 1 to 4 grades would be available.

One interesting thing to test it would be the VC paper spectral response... it may depend on the aniline used for green sensitization. Perhaps it can be done by testing different green LEDs, of some 30nm difference in their emision peaks...


From ilford information http://www.ilfordphoto.com/Webfiles/...0201152306.pdf I undertand that there are 3 emulsions, 1) with no green sensitization added, 2) with some green sensitization added, 3) with a lot of green sensitization added.

I understand (guessing) that 2) and 3) emulsions have the same specific green sensitization dye, but in different ammounts. They could use a different sensitization dye (green-bluish) for emulsion 2), but I guess that would give more inconsistent results under different light sources. If it is like this, it simplifies a lot light source for VC paper, I think. (also knowing cold cathode issues with VC...)

In my DIY LF enlarger I don't want to give up the possibility of condenser type, not of the filament type, but of the bulb type. I plan to use the Fujimoto concept, that can switch from condenser to diffuser. I think that the today's solar energy fresnels and powerful LEDs availability changes design rules...

[interneg]

With the LOMO I made informal tests, yes... it seems apochromat to me as there are no colour fringes, so lateral chromatic aberration looks corrected, but I was suspecting that longitudinal chromatic aberration is not very well corrected, an ucranian user told me that they were using monochromatic light to improve results. Perhaps Apochromatism has degrees of performance...

The link states 35 Lp/mm in the center and 12 in the corners... I'd said less, I found LOMO less sharp than a Symmar converted to 620, for 8x10, still the O-2 600 has some 1m circle (945mm?). Then there is the lack of coatings, so if enlarger is diffusion a lack of microcontrast may be expected. Also sovietic lenses had a lot of QC issues, so perhaps my copy is not very good... Still I'm eager to make deeper formal tests with the O-2, this is using a 1951 pattern with CMS 20, and making sure the target is in perfect focus.

I think there's more going on here - the drawing suggests it follows the Artar design which is a true apochromat with very low levels of other aberrations & high resolution - that is if it was assembled correctly! Though it is worth noting that Soviet/ Warsaw Pact era optics for military & industrial uses can be of exceptionally high quality.

As you pointed it, I've been reading that. http://heilandelectronic.de/files/do...t_Brochure.pdf

It is a good product, well finished, and straight, a good option. It cannot be cheap... Still diffusion only is available.

I'm asking myself how CRI (color reproduction index) of the light source may impact fidelity in color work. I bought a box of 8x10 Fuji CDU II transparency copy film and I plan to copy Velvia/Provia transparencies, and enlarging MF slides to 8x10 slides, perhaps this can be known by making contact copies of an IT8 slide target, under different light sources...

For what's WB I think that if blue channel is able of rendering grade 5 and green channel can deliver grade 00, I think that it has to work well. A problem would be a too greenish blue channel or a too bluish green channel , then perhaps only 1 to 4 grades would be available.

One interesting thing to test it would be the VC paper spectral response... it may depend on the aniline used for green sensitization. Perhaps it can be done by testing different green LEDs, of some 30nm difference in their emision peaks...


From ilford information http://www.ilfordphoto.com/Webfiles/...0201152306.pdf I undertand that there are 3 emulsions, 1) with no green sensitization added, 2) with some green sensitization added, 3) with a lot of green sensitization added.

I understand (guessing) that 2) and 3) emulsions have the same specific green sensitization dye, but in different ammounts. They could use a different sensitization dye (green-bluish) for emulsion 2), but I guess that would give more inconsistent results under different light sources. If it is like this, it simplifies a lot light source for VC paper, I think. (also knowing cold cathode issues with VC...)

In my DIY LF enlarger I don't want to give up the possibility of condenser type, not of the filament type, but of the bulb type. I plan to use the Fujimoto concept, that can switch from condenser to diffuser. I think that the today's solar energy fresnels and powerful LEDs availability changes design rules...

At least for the Beseler 45 I was asking about, Heiland can make a lightsource that uses the condensors, or a diffusion head. They also do one that does a similar thing on the Leitz Focomat. I quite strongly dislike condenser enlargers (with very limited exceptions) - I feel diffusion heads offer a vastly better quality of light (less dust too!). Furthermore, by using a flat panel you circumvent issues with beam angles - as nice as it would be to simply substitute an LED bulb, they have serious issues in this regard.

The significant thing about LED choices is picking ones that match the peak sensitivities of the paper - if you look inside an Ilford MG500 head you'll see that it uses a pair of narrow bandpass dichroic filters, one for blue & one for green - I suspect that following the peak bandpass wavelengths of the classic tricolor separation filter set (25,47, 58 or 29, 47b, 61 or 70, 98, 99) would probably get you on the right track. Obviously, care needs to be given to get adequate LED density/ coverage & the control interface will need a degree of calibration/ sophistication to be properly useful. With regards to graded paper, the MG500 instructions simply state that the machine should be set to grade 5 (ie maximum blue).

[Pere Casals]

Though it is worth noting that Soviet/ Warsaw Pact era optics for military & industrial uses can be of exceptionally high quality.

Of course, LZOS (Lytkarino Optical Glass Factory) were not amateur people: They invented Sitall / Astrositall glass... Also, in 1945, soviets seized all Carl Zeiss technology when Jena ocupation happened.

Still there are very bad LZOS glasses, and very good ones.

LOMO O-2 600mm can be obtained for some $55, what is really a bargain (http://www.ebay.es/itm/Soviet-lens-r...0AAOSwnHZYSTQF)

Here I was trying some macro with the O-2: https://www.flickr.com/photos/125592...posted-public/

A 1200mm rail (with two CAMBO in a row) was comming short




My impression is that it cannot be compared with western glasses, and this can be seen (my impression) even in the GG with a x8 magnifier, but still it is an impressive glass with an impressive circle. It's a lens I love. I made a 3D printer lens board for it and I'm planning a shutter. Also I'll measure well its LP/mm in the air, by using a diopter, soon... (this is straighter than using CMS 20)

At least for the Beseler 45 I was asking about, Heiland can make a lightsource that uses the condensors, or a diffusion head.

The significant thing about LED choices is picking ones that match the peak sensitivities of the paper - if you look inside an Ilford MG500 head you'll see that it uses a pair of narrow bandpass dichroic filters, one for blue & one for green - I suspect that following the peak bandpass wavelengths of the classic tricolor separation filter set (25,47, 58 or 29, 47b, 61 or 70, 98, 99) would probably get you on the right track. Obviously, care needs to be given to get adequate LED density/ coverage & the control interface will need a degree of calibration/ sophistication to be properly useful. With regards to graded paper, the MG500 instructions simply state that the machine should be set to grade 5 (ie maximum blue).

My theory (to be proved) is that green sensitizaton of VC paper is far enough (in wavelength nanometers ) from blue, so if used Green source peak do not match exactly top paper green spectral sensitivity then it will happen that more time or power will be required, but same result may be obtained.

I agree, careful calibration has to be performed, this is what it will tell the truth. As I said before, I placed an RGB LED bulb in an OPEMUS enlarger, in place of the Photocrescenta. The remote command of the bulb is IR, like with TVs. For the moment What I do to change the bulb color is to place the remote command under the enlarger lens, pointing up, in this way the IR signal arrives to the bulb IR receiver. Then I can switch off the bulb, prepare the Stouffer contact copy, and make the exposure. Those calibrations will tell me if it will match curves I obtain with an LPL VC6600...

Calibration... nothing else will say if it is good.

[Andrew O'Neill]

wow

[ridax]

there are no colour fringes, so lateral chromatic aberration looks corrected, but I was suspecting that longitudinal chromatic aberration is not very well corrected

If longitudinal CA is poorly corrected, color fringes are unavoidable. Also, the lateral CA just can't be corrected well when longitudinal CA is not.

they were using monochromatic light to improve results.

Yes monochromatic light improves sharpness and microcontrast with ANY lens. ;)

Perhaps Apochromatism has degrees of performance...

Yes it is. Strictly speaking, there are no true apochromatic lenses among the photographic ones - in the sense the word is used in microscopy, for example. The best of our "APOs" would be considered semi-apochromats (or even lower graded) by a microscopist... And there is more to it. At 1:1, any symmetrical lens is automatically apochromatic. And an unsymmetrical one can easily be made to be an APO... at 1:1. But not at any other magnifications.

Apo-Ronars use special glass types to make them "more apochromatic". Thus Apo-Ronars are sharper, at any magnification, then CZJ Apotessars and Induatar-11's and O-2's. That glass also makes Apo-Ronars much more expensive - and much more yellowish, too. (And the yellowish tint itself also adds to sharpness quite a bit. Put a yellow filter on an Apotessar or an I-11 or an O-2 and see an improvement in sharpness... though the Ronar will still be sharper than those.)

So one gets what they pay for. But for wetplate and for blue-sensitive film like many X-ray ones (or for an ordinary camera film + Wratten 39 filter), you get an additional bonus of much more lens transparency and shorter exposure times with that cheap glass!

still the O-2 600 has some 1m circle (945mm?).

No it hasn't. Celor-type lenses are narrow angle, and their edge performance does not improve with stopping down. The O-2 is no exception. It's good within about 50° at most, which corresponds to about 22" (56cm) for the 600mm focal length. Go further, and you get a total mess.

I found LOMO less sharp than a Symmar converted to 620, for 8x10

With all respect, I find it too hard to believe in such results. Mine are quite different. I've found a whole Symmar to be way sharper then the O-2 but a half of it to be just as much worse then the O-2. I've really no clue why your experience differs that much.

Also sovietic lenses had a lot of QC issues, so perhaps my copy is not very good...

A lot of people disagree with me but decades of usage and formal testing of the Soviet glass made me believe that's a myth. That's not QC problems. The matter is, those lenses are cheap. And they were cheap from the start. The goal of mass-production of anything exceptional just wasn't there at all. And BTW the O-2's were not mapping lenses. They were common process lenses for book publishing, etc.

Then there is the lack of coatings, so if enlarger is diffusion a lack of microcontrast may be expected.

Unlike I-11's, the O-2's were later versions of process glass, and they were never made uncoated. Though the coating was definitely poorer then on any German and Japanese lenses.

But the coating has no influence on microcontrast (aberrations do). Coating deficiency decreases the overall (macro-) contrast. And if we try to recover the overall contrast by using a more contrasty paper or/and processing, the microcontrast gets stronger indeed. In fact, the common technique of preflashing the paper (or film) leads to the same results.

... Some of my first-hand data on Russian and East German process lenses:
http://www.largeformatphotography.in...=1#post1035389

P.S.: I'd suggest two bigger tripods, or better a table, with 1200mm bellows extension for any sharpness to be visible....

[ridax]

process lenses .... I don't want to repeat what has been pointed out on this forum many times before; but these lenses are even better at enlargement than regular enlarging lenses if your priority isn't sheer speed, typically just a stop down from their f/9 maximum.

Never ever heard such an opinion - except that high-end process lenses have less distortion and are thus preferable for multi-piece murals that have to be stitched together. The sharpness story is quite different though.

Even at 1 to 2 magnification, the paths of the light rays are closer to their paths at infinity then to those at 1 to 1 so even a taking lens works nearer to its native range than a process lens. And an enlarging lens beats both of those hands down.

One of the best (and most probably the best) process lenses ever made are 240, 300, and 360 mm f/5.6 Rodenstock Apo-Rodagons. I have the 300mm one, and it's magnificent at 1:1. But I'd never use it for enlarging purposes as it's far inferior to my 300mm f/5.6 El-Nikkor that is optimized for 1:2 from the start. And at 1:4 the El-Nikkor is as much inferior to the regular 300mm f/5.6 Rodagon that's optimized to 1:4. And for stronger enlargements, my Schneider Componon easily beats all the above....

That's not an opinion speculated out while reading datasheets. That's my own practical experience.

[Pere Casals]

Thanks for all that information.

Let me discuss some points...

No it hasn't. Celor-type lenses are narrow angle, and their edge performance does not improve with stopping down. The O-2 is no exception. It's good within about 50° at most, which corresponds to about 22" (56cm) for the 600mm focal length. Go further, and you get a total mess.

Well... going beyond you obtain some 12lp/mm, (Yakolev catalog 1970), or less deopending on the Modulation %, but for a very big image this is a lot of graphic information.

With all respect, I find it too hard to believe in such results. Mine are quite different. I've found a whole Symmar to be way sharper then the O-2 but a half of it to be just as much worse then the O-2. I've really no clue why your experience differs that much.

The converted symmars are much better than it is said, near as good in the center than the whole lens, and with not that bad figures in the corners if stopped, see the Perez measures, not lab test... but good for comparing some samples:

http://www.hevanet.com/cperez/testing.html

Symmar 150
f/11 42 48 38
f/16 42 48 42
f/22 48 64 42

Symmar 150 converted to 265mm
f/16 48 48 23
f/22 48 48 33
f/32 42 42 38

The converted configuration has focus shift, so focus has to be checked with lens stopped. Not every one knew that in the past...

Unlike I-11's, the O-2's were later versions of process glass, and they were never made uncoated.

Yes... six months ago I was thinking it was uncoated, but mine has a coating without a particular cast, so I was thinking it was uncoated. I was told that there were 3 types of coatings

But the coating has no influence on microcontrast (aberrations do).

Here I disagree, abscence of coating delivers more flare, and this destroys microcontrast in shadows, imagine you have a litte spot with 0.01Lux, next to another 0,02 Lux spot, you have a microcontrat 1:2. If you add 0.1Lux flare then you have a 0.12 vs 0.11 microcontrast, way less.

P.S.: I'd suggest two bigger tripods, or better a table, with 1200mm bellows extension for any sharpness to be visible....

A good single tripod is more stable than it looks, I usually fix a toy laser pointer to the front standard, so I see the true vibration projected far, and when it stops vibrating after inserting the film holder. Sometimes I used a light monopod attached to the rail front end...

... Some of my first-hand data on Russian and East German process lenses:
http://www.largeformatphotography.in...=1#post1035389

Thanks, I've read it, it is very good information

Regards.

[ridax]

... going beyond you obtain some 12lp/mm, (Yakolev catalog 1970)

Yakolev (1970) states 41° coverage for the O-2's so his 12 lp/mm are for the even smaller field then my liberal 50°. 41° corresponds to 45cm (17.5") coverage for the 600mm focal length at infinity. (Though the published data was in fact not for infinity but for 1 to 1 process work.)

absence of coating delivers more flare, and this destroys microcontrast in shadows, imagine you have a little spot with 0.01Lux, next to another 0,02 Lux spot, you have a microcontrast 1:2. If you add 0.1Lux flare then you have a 0.12 vs. 0.11 microcontrast, way less.

All the above is about the overall (macro-) contrast, not microcontrast. Adding flare, we get lower contrast no matter what size the darker and brighter fields are. That's macro-contrast decrease. Adding aberrations, we get no change in contrast between large fields but much lower contrast between small spots and lines (that would be of the same contrast as the big ones without the aberrations). That's micro-contrast decrease - with the overall contrast unchanged. Adding edge effects in developer (or digital "sharpening"), we get small details contrast (=microcontrast) increase with the overall (macro-) contrast unchanged. Of course in the real wold both micro- and macrocontrast usually get changed simultaneously but I believe that's not a reason to mix the concepts themselves.

[Pere Casals]

(Though the published data was in fact not for infinity but for 1 to 1 process work.)

OK, my confusion came from that... of course coverage do vary a lot from infinite to 1:1 !!!

All the above is about the overall (macro-) contrast, not microcontrast. Adding flare, we get lower contrast no matter what size the darker and brighter fields are. That's macro-contrast decrease. Adding aberrations, we get no change in contrast between large fields but much lower contrast between small spots and lines (that would be of the same contrast as the big ones without the aberrations). That's micro-contrast decrease - with the overall contrast unchanged. Adding edge effects in developer (or digital "sharpening"), we get small details contrast (=microcontrast) increase with the overall (macro-) contrast unchanged. Of course in the real wold both micro- and macrocontrast usually get changed simultaneously but I believe that's not a reason to mix the concepts themselves.

Let me demonstrate that flare decreases microcontrast or (better said) acutance, substantially in the shadows.

Well... microcontrast (or acutance) is related to density difference, see here "Acutance Definition" section https://en.wikipedia.org/wiki/Acutance

So Acutance is density difference. A = D2-D1. Let's say these are densities of two neighbour areas, supose that D2 has more density than D1.


Then add the same amount of light (flare) on both areas, new A will ve lower, because D2 won't increase density as much as D1.

You can say... but film has a linear response !!!

Yes... but this is when you plot light irradiance (abscisse in sensitometric plot) in a logarithmic scale. Imagine that you had the irradiances L1 and L2 that delivered D1 and D2 densities.

Compare Log (L2)- Log (L1) to Log (L2+flare)- Log (L1+flare), being flare a constant for that scene+lens. In the second case D2 - D1 is lower, and much lower when flare is big compared with image forming irradiance. (Note we can consider D linear dependent from Log(L)).


This can be adressed with SCIM masking in the darkroom, unsharp masking for the border, or with photoshop... but it is difficult to get as "good" acutance as with a competent coating. I say "good" becasue for a portrait it can be bad...

This is what I understand, I can be wrong, please correct me if I'm mistaken.