Russian | Soviet | USSR Lens Survival Guide

[Malik Tigrine]

Greets, Malik.

I apperiate your thanks Plz post here if there's smth on the subject.

no no no, Sean meant SovScope-70 @70mm(65mm) lenses, 51,2x23mm frame size. They were not ever intended to be anamorphic despite the fact it has '-scope' in its name

But what's the frame size or image diagonal of 8K? Let's talk in millimeters, K numbers is of no use

29.90 mm x 15.77 mm (Diagonal: 33.80 mm)

Thanks I will check the book !

 

[Malik Tigrine]

Yes those lenses from USSR, damn hard to find, i have a set for sale)

Pm sent

 

[Ilya O.]

Malik, be extra cautious with Kinorussar shockingly wide for 70mm 28mm lens, it has common problem with the glass in front of the iris, it happens 80%, it simply goes white or white lines or matte glass degradation.
I'm now in the process of making this types of glass precisely polished to be as new so as to follow LOMO specs...

 

[Mike Quill]

Hi Ilya, any idea why the 70 OKC 10-100-1 100mm 1.4 is not discussed in this book?

Malik, I have a 70mm wide format USSR Book for ya But plz do not ask me where to get those babies and what mount or adapter you need

so plz welocme bilingual (in Russian and in English) 70mm Soviet gear Book, it has some info on the soviet 70mm mass-produced lenses:

http://drive.google.com/file/d/0B_9Rua4iG1MraDV5UVZWVHZFaDQ/view?usp=drivesdk

 

[Ilya O.]

Hi, Mike.
it's just another era. 100mm f1.4 had just been made far far later than 70mm Book was published, it's late 80-s calculation. Whilst the Book is from 60-s, golden age for wide format cinematography...

 

[Ilya O.]

Malik, why not try first to cinemod soviet still glass? Nice performance and character, full photo frame coverage, all metall barrels and focusing helicoids, if serviced properly, they are a joy to use, reliable and will last longer than any photo modern glass.

 

[Malik Tigrine]

Malik, why not try first to cinemod soviet still glass? Nice performance and character, full photo frame coverage, all metall barrels and focusing helicoids, if serviced properly, they are a joy to use, reliable and will last longer than any photo modern glass.

Hi Ilya, yes I have already the Mir 37mm and Helios 58mm, and I will take the Mir 20 mm and the Jupiter 85. But I'm looking for a zoom that can cover 8K WS @ 25 mm minimum (Lomo/Ekran). If you have any idea... Thanks Ilya.

 

[Ilya O.]

Malik, sorry, I had no strict answer about the 8K Helium coverage of LOMO zooms at 25mm. It's a subject for thorough investigation. Theoretically, they should NOT cover.

Now it occurred I can do some little test for ya. Here I have 35OPF15-1 25-250mm f3.2 LOMO zoom. I put it on Sony NEX-7 sensor, it's 23.5 x 15.6mm, image resolution is 6000x4000. So here how it looks like, crazy:

The OPF is at 25mm, you can judge the edges here, I placed it relatively to Helium sensor, 8k at right click - new tab:

I do not see any hard portholing here, but the upper right corner shows falloff. So one can only guess or imagine what happens to the image field if expanded to the red area... It's you who decide or risk buying whether it covers or not I highly doubt, 33mm is significantly larger and wider than 28mm circle of standard 35mm, but who knows...

small add-on:

I found my little powerhorse, 35OPF29-1 25-80mm f3 #850004 and put it on the same Sony NEX-7 sensor, check the coverage at 25mm:
(! the lens is not properly shimmed, or an adapter is off a bit, so out of focus at the wide, do not judge the sharpness here, it IS sharp !)

 

[Ilya O.]

Till some time before I have heard much about sample variation of the Soviet lenses (and all others due to the very mass production and economically reasonable tolerances nature). And I was very interested how the good sample renders versus a bad one.

Now at last I found an established way to properly determine a good and a bad lens and I found some examples to compare!

So what is a good sample? We here completely ignore the optical calculation / lens design IQ / lens character. This method leaves these untouched.

I here talk about mass production reality where tolerances take place and rule. That means the real lens would 100% have some spacing errors, decentering, tilt, not optimal rotational position, glass molding defects, lens pinch etc. All that leads to on-axis image degradation, which was NOT there back when lens was calculated. The typical mass-production problems are these aberrations: coma, astigmatism, spherical. How that look like? Plz go and browse my parallel thread.

Of course you can take 10 lenses, put them into test and choose the best. But would it be the absolute winner optically? Not in reference to the others? Would it still have above-mentioned aberrations? Would it be ideal for the particular design?

OK then, now I take two 50 yrs old Mir-1 37mm f2.8 lenses for the race:

How do I know which is bad and which is nice? I put them on the optical bench to see how each one renders tiny pinhole of light placed at infinity (projected by 1600mm collimator). This disc, so-called Airy disc, is the basis of all image formation, which in an ideal case should look like this in green light:

This is known as star test, an extremely sensitive one. It shows nearly all lens and glass defects.

So, I put both lenses on a bench test and here what each show:

Obviously the bad Mir-1 sample is on the left. What you see as a fireball / comet with little wings (left) and a nice point (right) is an Airy disc pattern, each lens optical certificate I would say.
Lower is the precise resolution measurement pattern. See how a dot / disc pattern somehow affect the character of the resolution chart? It just copies in the larger scale every aberration represented in an Airy disc.

What all that means to the on-axis resolution numbers? Viewed at 100% or more magnification, Good Mir-1, on the right, resolves 13th component with minimum contrast, but nevertheless. The Ugly can do only 6th component, and that's according to overly optimistic attitude Other 4-element squares you can not really count since huge astigmatism lets the Bad resolve 45Deg squares only...

So what are the optical on-axis resolution numbers of these lenses at minimum contrast for white light? 13th component of good disc Mir-1 means — shazam — ~535 line pairs per mm. Bad one gives 6th component, it is ~356 lp/mm, 1.5 times worse than the good one, and still much higher than sensor's own resolution of 128 lp/mm. The resolution of 535 lp/mm seems to be unbelievable for Mir-1 still lens, but it is not if we talk about per se lens resolution with no sensor or film as a limiting factor.

I hear you, guys, these numbers mean zero for the image makers Here's the Ugly, at f2.8, 6k on right click:

Well well well, you might think, — not so bad for a vintage 54 year old photo prime Plz do not make early decisions since you haven't yet seen good Mir-1

Ah, forgot to put disclaimer: all images here and farther were taken at Sony Alpha NEX-7 still camera cause it has superb sensor with very dense pixel package, the pixel pitch is 3.88µm (micrometers) or 0.00015 of an inch! Red Helium has 3.65µm, 50Mpix Canon EOS 5DS R has 4.13µm. NEX-7 is APS-C sensor camera, 23.5 x 15.6mm, 28.2mm diagonal, resolution is 6000x4000, sensor resolution is ~128 lp/mm. All shots done at ISO 100 (max DR for this sensor), white balance set to Daylight, with Front Curtain Shutter ON (less vibration), 10 seconds delay shot (to avoid after shutter button vibration), Picture Profile is Neutral (0,0,0), Noise Reduction is off.

Here's the Good, wide open:

Now let me dive more into pixel peeping so that you can see each in relation to the other within the same picture. Here you are, this is central, focused-at area, left is the Bad, right is the Good, mirrored (horizontal flip) for convenient outwards-the-center comparison:

Plz open it in new tab, zoom in, and just look at all this little details, bricks, even little tiles on the wall above the trees, etc. Now you can see that there's a Hebrew star in the center Mir-1, a good one, indeed delivers another level of high frequency details.

What I here find surprising is that corner/border performance is pretty CLOSE to each other! So you can not worry about the fact that Mir-1 looses its vintage charm and 3d fall off in case of having an ideal Airy disc pattern

Of course, the Good still beats the Ugly, but not so drastically as we saw at the image center. Frankly speaking I was expecting the Ugly to go into complete mash at the edge since if you have a bad on-axis Airy disc you will get it much worse at the edges...
and I was assuming the Good to be far better at the edge, besides, keep in mind that this is not even a full frame example...

Now lets see more night and day difference at the daylight shot. Mir-1 #632420, the Bad, wide open at f2.8, focused at the Hebrew star at the center:

Here's the lucky opponent from KMZ, same focus point:

Now the same shot, central region for head-to-head comparison:

High contrast of the daylight scene gives the Bad more chances, and you feel it's OK just till the moment you start looking at the right Now you see how much clarity (=high frequency resolution; notice that tiny square tiles above the tree line, a superb high res low con natural charts) gives the optically OK lens? That's wide open, infinity test, guys... That's how 50+ yrs old grandpa, if in the good mood and being fed and slept well, can one day kick some ass! This Mir-1 of KMZ is vibrantly sharp and seems to be indeed (as its creator Prof. Volosov mentions somewhere) very close to APO performance since zero CA or lateral color at this high contrast white buildings.

Let's move on and see how each performs at more realistic f5.6 usage scenario. Here's Mir-1 at f5.6, get a new tab for 6K:

Isn't it gorgeous for a tiny simple (6 elements) pre-computer calculated lens? ^-) Naturally sharp with slight resolution fall off which helps 3d feeling, nice color tone, nearly zero CA of any kind (I only see slight lateral brown-orange one, on the high contrast wall on the left white build.), pretty nice contrast handling with much room in the histogram for the color correction, no blacks or white crushing, nice healthy slope.

WAIT! Was is it a good or a bad one!? It was the BAD one, guys!!! ahaha ha h ah Can you believe, the bad?!

Rush for the good Mir-1's image at f5.6:

This one is even more optically perfect with crazy sharpness which Sony sensor can not handle, can you notice the light yellow 45 Deg. lines moire pattern and mosquito aliasing?:

OK, let's have both in front of each other, central region, left is bad and right is good, as accepted earlier:

Pretty indistinguishable, right? Well, as for large objects (low & middle frequencies) that is so, but I still see the difference in the high lp/mm (high spatial resolution) area — again, take a thorough peep at that tiny little square tiles right above the tree line, and then move your sight outwards: the good one resolves better and farther to the edge. The better sensor we have in the future the more we can get of a Mir-1 with nice Airy disc.

Plz note here, that here, at f5.6, Airy disc becomes bigger than at f2.8, that means — paradoxically to what we expect from the usual image taking perspective — that the lens optical resolution becomes much less! In the meantime nearly all Airy disc aberrations we saw at f2.8 disappear under the grown up big f5.6 disc. That's why all lenses are more or less equal at the middle f-stop openings like f5.6 or f8.
An ideal f5.6 lens, aberration free or as it's called diffraction limited optics, according to the conservative Rayleigh criterion would give Airy disc radius = 1.22 * 0.555µm (green light) * 5.6 (f-number) = 3.79 µm which equals to 264 lp/mm of optical resolution. Now do you recall that our NEX-7 sensor has the resolution of 128 lp/mm? See, that an ideal Mir-1 at f5.6 out-resolves the sensor by 2 times! So there Is a place for sensor evolution and getting even more potential out of Mir-1 or any vintage lenses with nice Airy disks!

Ok, further to f5.6 shots, let's check the edge performance:

Again here I'm amazed how the bad one rose its IQ at the edges. Though, good Mir-1 again rocks if we look at high spacial resolution, = all tiny details and accented object borders. Worth noting that the Bad still exibits some lateral light blue/pale orange CA (see white building at the left image edge), while the Good has literally none or very little.

 

[Ilya O.]

In the previous post I wrote that at f5.6 Airy disc becomes bigger than at f2.8 and that means that the lens optical resolution becomes much less! Besides, at f5.6, nearly all Airy disc aberrations we saw at f2.8 disappear.

Let's see what happens actually to the Airy disk and the resolution since the actual pictures we've just seen.

Again I take 37mm f2.8 primes for the show. A bad one, Mir-1 by ZOMZ, in the left corner, and a good sample, fine-tuned Mir-1 from KMZ is on the right. So let's see what both show us at f2.8 and what means good and bad optically speaking:

A calculation based on the resolution pattern gives this optical on-axis resolution numbers: good disc Mir-1 resolves ~535 line pairs per mm, and the Bad one gives ~356 lp/mm, 1.5 times less.

What then happens with the disk and resolution if we close the iris to 5.6? Diffraction eats the resolution fast Airy disk, on the contrary, feels itself much better in case of the bad Mir:

Notice how better becomes bad Mir's Airy disk (but still some coma and astigmatism). It's now much close to the perfect diffraction circle. It's because of the aperture cuts off each lens' periphery and thus now the lens works by its center which is much less prone to the aberrations and less sensitive do decentering.

The resolution is mainly dependent on the aperture size. The bigger diameter eye of glass you have the more you get. So what we have here? The good one, on the right, resolves ~318 lp/mm. The Ugly does ~ 283 lp/mm.

Here what happens to the details if you close the lens to f16:

Here we face Airy disk goes really big and its radius is simply larger than any line on this resolution reticle so the test shows zero resolution. That does not mean the resolution is really zero, it's just this pattern has its own limits and can not show us some little lp/mm numbers.

 

[Ilya O.]

What if we put Mir-10A 28mm f3.5 against OKS7-28-1 28mm f2.0!?

I understand, this is totally unappropriate comparison since Mir-10A is a full frame f3.5 still lens, whist OKS7-28-1 was made for the cinema frame, pretty fast f2.0...But isn't we free to compare?! Besides, it's interesting from the practical usage perspective.

We'll take both not in the top condition, but with medium Airy disk imperfections (=these are probably considered OK by the factory). Mir-10A is on the left, having a bit of coma, and OKS7-28-1 is on the right, showing coma and tilt:

The results above are wide open, just for the diagnose and fun. It also gives me the optical on-axis resolution numbers which are always measured at max aperture:
Mir-10A stops seeing lines at 17th component, that gives ~449 lp/mm at f3.5, OKS7-28-1 resolves 22th component, it's ~601 lp/mm at f2.0 (which is 1.33 times more). See how the f-stop directly relates to resolution both theoretically (Rayleigh criterion) and, here practically?

So called Rayleigh criterion limits any ideal lens' (with ideal Airy disc, aberration free) resolution and it is equal to the radius of an Airy disk (measured to its first minimum). The pretty conservative version of this formula is:
R = 1.22*lambda(=wavelength in nano meters)*F-stop
To know how much lp/mm this means, one should divide 1 by R.

By the way, I've got another OKS78-28-1 28mm f2.0 with nearly ideal Airy disk shown in this post http://www.reduser.net/forum/showth...oint-source)&p=1734051&viewfull=1#post1734051
I wonder how far that best OKS from the shown above Mir-10A and mediocre OKS. Well, it just OUTRESOLVES the resolution pattern :

Good OKS 28mm #2 sees all the lines of the last 25th component which equals ~714 lp/mm amd more, we can not count since there's no components left in this #5 reticle.

Now back to our Mir-10A and OKS7-28-1 #1. Let's see the 6K city landscape of Mir-10A first, wide open:

And here is OKS at f2.0:

Left image part is for some reason very soft, while the right one is OK, could be tilted element, so plz do not look at the left edge.

Separate pics are of little convenience for comparison, so here's 100% center region, Mir is flipped horizontally so that OKS represents its better edge:

First we see that Mir-10A is pretty neutral in color, at least not as warm as OKS is. Second, Plz open in new tab, zoom, and behold: Mir-10A tears OKS7-28-1 apart!!! A clear winner in the center and at the field a ha hah haaa :laugh:

Now calm down and take these facts into consideration: Mir-10 is at f3.5 hence the wider depth of field and contrast while OKS is at f2.0. If you look at the bench resolution patterns you'll see that OKS has clearly 1.33x more resolution wide open in the center when measured by itself, without sensor as a limiting factor. So what I'd say here is that 129 lp/mm of the NEX-7 sensor actively interfere with the lens and what we have here is photographic resolution which seems favor Mir-10A f3.5 Airy disk size and shape and simply can not show us high resolution details OKS is capable to deliver at f2.0.

Step two, shock number two, here I put Mir-10A f3.5 vs. good OKS7-28-1 #2, both wide open:

Plz take your time and zoom in. WOW! OKS7-28-1 sample #2, with very nice Airy disc, shows its teeth, isn't it? And do you keep in mind that OKS is at freaking f2.0 now!? At first I thought Mir-10A is worse and has some sort of white haze or spherical aberration halo. But in this case it's mainly not, it's just the contrast and exposure are bit different (while indeed Mir-10A has some sort of very delicate bloom). Here what both look like if Auto Color applied in Photoshop (does the auto contrast per channel and equates them for white balance):

Now can you tell the difference? As for me, I just can not see, both go head to head in the center. Of course good OKS #2 has a focus falloff since it's calculated for smaller image frame, even the edges here, at central region, are worse than of Mir-10A. So guys, what we have here again is a sensor (sensor resolution is ~128 lp/mm, pixel pitch 3.88µm) as a limiting point. Hurray to Helium and beyond with smaller pixel pitch!

Again let me go back to the Mir and mediocre OKS #1 we started with. How both render if compared the right way, at equal f-stop? Here you are, central region at f3.5 both:

Now, stopped down to f3.5, OKS7-28-1 strikes back and becomes on-par or slightly better than Mir-10A Why is OKS not very much better? Again, it's the sensor limits its horsepower so eats up the high spacial frequencies difference.
Still, I wouldn't say Mir-10A is a bad performer, it actually has nice subtle look, lesser contrast, and some sort of very delicate large area blooming and a bit of halo around superwhites, that's all nice and helps render the light with more dimension. Again, as with Mir-1, none or very little CA.

Now I close the contenders to f5.6, a useless test I usually say since all lenses are equal at this f stop :

Plz compare, am I right? I'd never shot f5.6 again for testing errrrr lost time errrr
At the same time f5.6 shows how clearer, sharper each one becomes compared to f3.5, and the focus field is now flat.

Here's the egde performance, images flipped in a way that the edge put in the picture center so that the eyes do not jump comparing, same f3.5:

Plz do not consider edge buildings scew as a distortion artifact. It's just the horizon is not set up so that I could position the Hebrew star in the very image center.
So, hard to decide the winner, yeah? Lenses have different look, edge character and field curvature. For me OKS7-28-1 is sharper at the edges, smoother just before the edges, but has more of lateral color aberration and ghosting of the high-contrast objects. Mir-10A has some sort of very delicate bloom.

Here's the edge at f5.6:

As usual at f5.6, both reach now clarity. It's seen very good if you open f3.5 edge at one tab, and f5.6 edge pic at another, and then click and chane the tabs rapidly, you'll get what I mean. Still, I see OKS is more clear and accented.

To sum up:

— Design IQ aside, it's all Airy disk / decentering / lens sample variation that rules it all in the real world
— never underestimate 'slow' unknown soviet still lenses in relation to the fast and pricey cine ones
— nowadays it's the sensor who limits lens' performance. Any lens with good IQ made in the last 70 years is 4K 6K or 8K ready, do not believe marketing adverts

 

[Vladimir Cazacu]

Interesting to see how similar (most of) the Lomo lenses and the soviet photo lenses are in terms of rendering and overall picture quality.

A while back a friend and I made a test comparing Mir-1 - 37mm f2.8 and Helios-33 - 35mm f2 (same as PO4-1M?) and even though the optical designs, aperture and coatings are very different, the overall rendering of both lenses was very similar.
A shame we didn't save any of the photos, as I would have loved to share them with you guys.

It would be very cool to compare all of the different focal lengths from the various Lomo OKS lenses to their photographic counterparts, to see the real advantages and disadvantages of both sets.
The Lomo lenses are usually faster (especially the EKRAN models) and (probably) have better optical resolving power but they were also designed to cover a much smaller format and as such they might vignette more and have a faster focus fall-off near the edges.

Cheers.

 

[Ilya O.]

Interesting to see how similar (most of) the Lomo lenses and the soviet photo lenses are in terms of rendering and overall picture quality.

I think that comes from the fact that the majority of Soviet lens designs are made by just a few scientists. Besides, the optical goals were similiar and put in GOST (state standatrds) and special industrial papers (lets say a good lens for cinematography must resolve on grainy film at least 50 lp/mm in the center and 25 lp/mm at the edge, must be maximized for resolution not for contrast (to make high magnification at theaters possible), must be corrected for spherical aberration, coma, and astigmatism, must have high price/performance ratio).

It would be very cool to compare all of the different focal lengths from the various Lomo OKS lenses to their photographic counterparts, to see the real advantages and disadvantages of both sets.

— that's for sure a nice goal I'm very interested in Vladimir, plz participate!!!

 

[Vladimir Cazacu]

I will if post test results if I ever end up buying a few Lomo spherical lenses.

Unfortunately, OCT19 lenses have gotten quite expensive recently and the mechanical quality seems to be quite poor.

 

[Ilya O.]

ahah aahah crazy prices, that's so true

Concerning the mechanics, I'd disagree. They just need proper care.

 

[David Jaure]

Hi guys,

I see you guys know a lot about russian lenses.
I just purchased the following lenses:

-Carl Zeiss Jena Flektogon 35mm F2.8 M42 wide angle zebra lens Praktica
-Carl Zeiss Jena Flektogon 20mm F2.8 MC Vintage M42 Zebra Electric

Can you recommend me which is the best converter/adapter that will fit these lenses?
thank you all!

 

[Ilya O.]

Hi, David.

For Canon, just go with 5 bucks (+free shipping) chrome plated brass adapter like this one:



For other mounts, make sure the adapter is from brass or steel, not aluminum, and, what is very important, has that inner shelf ring which stops auto aperture pin so that you can change the preset iris.

 

[Ilya O.]

Some shots from my beloved single-coated Mir-20M 20mm f3.5, make sure to turn on 4K so that at least it looks nice when watched at HD screen:

Frame grabs in HD:

 

[Ilya O.]

Just one shot that shows how Mir-10A reacts to direct sunlight and then then how it renders the landscapes of Ingushetiya:

 

[Ilya O.]

Same place, now shows Mir-1 37mm f2.8 look and flares:

And a bit of dynamic footage of Mir-1 showing how they drive at Caucasus, Ingushetiya