1:20 to 1:5 (0.05x to 0.2x)

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Lou Jost
Posts: 5987
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by Lou Jost »

"The CA control at f/22 at infinity looks incredible. Almost unreal."
chris_ma wrote:
Wed Feb 15, 2023 9:48 am
well, if you're shooting at aperture f11 then just take any reasonably good lens and results will be very similar.

it seems to me that you want zero distortion over a broad magnification range, plus high-resolution, plus large image circle, plus corner to corner sharpness, plus APO correction, plus low cost..

not going to happen!
Chris is right, many decent lenses will have good CA control at f/22, though this can just be because everything is a little blurry at that aperture.

But you can't rely on small apertures to control CA, because you want high absolute resolution, and that is physically impossible with small apertures (large f values), because of the laws of physics.

You might re-think your resolution needs. If you are only going to print the images at a size similar to the original art, and you are going to look at them at normal viewing distances, then I think you could get more than enough resolution with a pixel-shifting full frame sensor.

If you were more concerned about relative resolution (lines per picture width), then I bet the Apo-Nikkors would check all your boxes. Cheap, zero distortion, zero CA, flat field, largish image circle. Mine are amazingly good, though limited by their working aperture of about f/11-f/16.

jvanhuys
Posts: 49
Joined: Sun Jan 22, 2023 9:24 pm
Location: Seoul, South Korea

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by jvanhuys »

simplejoy wrote:
Wed Feb 15, 2023 2:36 pm
jvanhuys wrote:
Tue Feb 14, 2023 8:50 pm
Also, not sure if these are your images, but I found some sample pics taken with the Dokumar 47 f/5.6, which are unbelievably, sharp, but then this one image suffered from quite a bit of CA. Was this because it's a blade of grass, and thus, so much closer than the Dokumar's magnification range of 1:75 to 1:30? You were probably down to 1:1 on this shot right?
https://flic.kr/p/2mgGwJm
Glad you find the Dokumars interesting - however a word of caution: Even if they were among the most impressive lenses for certain magnifications when they were created, I doubt that many of the samples still around are up for the challenge. Don't get me wrong - I like my Dokumar 47 mm f/5.6 and I'm completely aware of the fact that the main point for not being fully convinced about its qualities might be caused by me using it outside of its sweet spot, but I just never felt like it was among the best lenses for anything. I'm not a very artistic person, so I don't have any real artwork to show, but I took a quick shot of one of my recent attempts at motivating my child to color something with wax crayons: Dokumar 47 mm (Stack of 6 images, unedited)
The size of the paper is 420x297 mm.

And here is an image showing the CA correction of the lens on a more challenging subject: Dokumar 47 mm (single shot, SOOC) I'd say it's quite okay actually, but there are better lenses in this aspect as well. The S-Planar 120 mm I just compared it with as well did significantly worse though, so if there are a lot of elements in your artwork which might lead to CA problems, I'd call this a significant flaw.
Lou Jost wrote:
Wed Feb 15, 2023 7:16 am
Just want to re-emphasize that field flatness doesn't matter much if you stack. Robert's tests often show this, as he often has to use different frames for his center and corner best shots, but these stack seamlessly.
I don't have a lot of experience with things like that, but I'd agree that stacking works pretty well in those cases.
Hey!

I can't tell you how valuable this is for me. Thank you so much.
It seems this lens is possibly showing its age due to worn out coatings or probably single-coating? Upping the contrast within reasonable levels seems to have a negligible filter hit compared to matrix operations like convolves, blurs, distortion/CA correction. Those things for my work, soften the image too much. When I increase the contrast by 25% on my Symmar-S 180 (single-coated) it is as good as my Micro Nikkor 105 in the centre... and it definitely beats the Nikkor in the corners with significantly better CA control and a touch less barrel distortion (possibly none?). On top of that I have the movements to do my stitches.

I'm trying to say, the lack of micro-contrast in the Dokumar doesn't worry me. My mortal enemies for flat artwork are distortion and CA. Even sharpness is lower priority. I noticed that the extreme shadows and highlights in ferrule of the brush, are very well handled by the Dokumar. That's a distance I would shoot too. The CA seems almost nullified at that . This is very promising. My Micro Nikkor 105, for example, is unusable with those extreme values due to CA. If we were looking at a 32-bit exr outside of Photoshop, in say, Nuke with a camera capable of ACES colorspace capture, with that contrast ratio, that same highlight's value would be between 30 and 50 and the shadow would be close to 0. So, again, I'm impressed.

I hear what you're saying about it being good for it's time but modern lenses being better. I get that it's not quite as sharp as the S-Biogon 40mm, or the Nikkor 55 Micro in terms of lines per millimeter, but I'm noting multiple things that I love. I looked at some very high resolution images of the S-Biogon on Flickr, and despite its remarkable sharpness, I noted that it has fairly bad distortion, of the pincussion-type. I had the Nikokr 55 for years, which is famously sharp too, but I sold it because it had too much barrel-distortion for flat artwork. The CA control was not bad, but nothing near a true Apochromat.

I think I'm gonna pull the trigger on the Dokumar and post some results once it arrives a few weeks from now. Thank you so much for spending time on this. Let me know if you ever want me to run a test setup for something you're curious about.
  • For th

RobertOToole
Posts: 2627
Joined: Thu Jan 17, 2013 9:34 pm
Location: United States
Contact:

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by RobertOToole »

Just in case you havent seen Marco Cavina's site:

Dokumar 47mm f/5,6

http://www.marcocavina.com/articoli_fot ... nglish.htm

Excellent coverage of the Dokumars near the bottom of the page.

Apologies if this was already posted but I didn't see the link in the thread.

Best,

Lou Jost
Posts: 5987
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by Lou Jost »

"My mortal enemies for flat artwork are distortion and CA. Even sharpness is lower priority."

In that case the Apo-Nikkors are even more attractive for this than I thought. They are essentially perfect defenses against your two mortal enemies. See
https://redbook-jp.com/redbook-e/rfap/ap.html
Nikon's official spec for the distortion is "0%".
Vignetting also 0% with a 400mm diameter image circle. You may be able to stitch seamlessly from a single aerial image, moving just the sensor (on a view camera back or something similar), keeping the camera and lens fixed, if your artwork is not too big.

Ah, but I forgot you wanted shorter lenses. The shortest of these are 150mm and 180mm, as far as I know.

jvanhuys
Posts: 49
Joined: Sun Jan 22, 2023 9:24 pm
Location: Seoul, South Korea

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by jvanhuys »

Lou Jost wrote:
Wed Feb 15, 2023 10:32 pm
"My mortal enemies for flat artwork are distortion and CA. Even sharpness is lower priority."

In that case the Apo-Nikkors are even more attractive for this than I thought. They are essentially perfect defenses against your two mortal enemies. See
https://redbook-jp.com/redbook-e/rfap/ap.html
Nikon's official spec for the distortion is "0%".
Vignetting also 0% with a 400mm diameter image circle. You may be able to stitch seamlessly from a single aerial image, moving just the sensor (on a view camera back or something similar), keeping the camera and lens fixed, if your artwork is not too big.

Ah, but I forgot you wanted shorter lenses. The shortest of these are 150mm and 180mm, as far as I know.
Hi Lou, hope you doing well.

I already mentioned in this thread that my Apo Nikkor 360mm f/9 was very soft, but it was pretty well corrected chromatically. Wasted $700 USD fo S.K. Grimes to re-mount the barrel with a shutter. I sold it. I think that's when you/ someone else pointed me to the Repro-Nikkors, which I then realized is a different beast entirely. Perhaps the Apo Nikkors were meant for 8x10 or larger? I wonder if there was an Apo Nikkor made for 6x7 up to 4x5 size... but yeah, left a bad first impression.

Are you happy with yours?

jvanhuys
Posts: 49
Joined: Sun Jan 22, 2023 9:24 pm
Location: Seoul, South Korea

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by jvanhuys »

RobertOToole wrote:
Wed Feb 15, 2023 7:49 pm
Just in case you havent seen Marco Cavina's site:

Dokumar 47mm f/5,6

http://www.marcocavina.com/articoli_fot ... nglish.htm

Excellent coverage of the Dokumars near the bottom of the page.

Apologies if this was already posted but I didn't see the link in the thread.

Best,
Hi Robert, yup. 'simplejoy' pointed me to that article, where I discovered it. Talk about a fascinating read... wow. At any rate, that Dokumar lens just ticks all the boxes for me on paper. If anyone's selling one, please give me a shout. The 47mm version if possible, it won't stitch like my 4x5 lenses, but I feel like I won't have to as it seems so ideal for 42mm diagonal sensor. I could maybe rig up a new tripod setup just for it.

chris_ma
Posts: 571
Joined: Fri Mar 22, 2019 2:23 pm
Location: Germany

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by chris_ma »

Lou Jost wrote:
Wed Feb 15, 2023 10:32 pm
"My mortal enemies for flat artwork are distortion and CA. Even sharpness is lower priority."
In that case the Apo-Nikkors are even more attractive for this than I thought. They are essentially perfect defenses against your two mortal enemies. See
https://redbook-jp.com/redbook-e/rfap/ap.html
Nikon's official spec for the distortion is "0%".
Vignetting also 0% with a 400mm diameter image circle.
Note that this page lists this specs are for a mag of 1x, so at 0.05x to 0.2x it will most likely do rather poorly in terms of distortion and corner sharpness.

In fact, I doubt it will be possible to find a lens with 0% distortion over that magnification range.

Why don‘t you make a test:
Shoot your artwork with your target magnification, then also shoot a lens distortion grid either according instruction of adobe or the foundry.
Then use Adobe Lens Profile creator or Nuke to undistort the image, and compare it to the original.

I doubt there will be a significant different other then noise and false details from debayering (but at f11 the later will not be happening anyway).

Also keep in mind that for final print or web display, there will be a resampling of pixels anyway.

And finally, if you are still concerned, then consider shooting at slightly higher mag. A 700MP file with lens correction applied will most certainly outresolve a 500MP file of one shot with no correction, even ignoring the fact that you can choose a lens with better sharpness and CA control in the first case.
chris

jvanhuys
Posts: 49
Joined: Sun Jan 22, 2023 9:24 pm
Location: Seoul, South Korea

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by jvanhuys »

chris_ma wrote:
Thu Feb 16, 2023 12:53 am
Lou Jost wrote:
Wed Feb 15, 2023 10:32 pm
"My mortal enemies for flat artwork are distortion and CA. Even sharpness is lower priority."
In that case the Apo-Nikkors are even more attractive for this than I thought. They are essentially perfect defenses against your two mortal enemies. See
https://redbook-jp.com/redbook-e/rfap/ap.html
Nikon's official spec for the distortion is "0%".
Vignetting also 0% with a 400mm diameter image circle.
Note that this page lists this specs are for a mag of 1x, so at 0.05x to 0.2x it will most likely do rather poorly in terms of distortion and corner sharpness.

In fact, I doubt it will be possible to find a lens with 0% distortion over that magnification range.

Why don‘t you make a test:
Shoot your artwork with your target magnification, then also shoot a lens distortion grid either according instruction of adobe or the foundry.
Then use Adobe Lens Profile creator or Nuke to undistort the image, and compare it to the original.

I doubt there will be a significant different other then noise and false details from debayering (but at f11 the later will not be happening anyway).

Also keep in mind that for final print or web display, there will be a resampling of pixels anyway.

And finally, if you are still concerned, then consider shooting at slightly higher mag. A 700MP file with lens correction applied will most certainly outresolve a 500MP file of one shot with no correction, even ignoring the fact that you can choose a lens with better sharpness and CA control in the first case.
I've had to shoot distortion grids for on-set work on a BBC show I was VFX supervisor for (just one episode though). It was so painful and costly, because it had to be done for each lens in the cinematographer's kit and at different magnifications. Don't even start with zooms... that's a whole different nightmare to deal with. We ended up running the grids through PFtrack for the match-move department. Worked a charm. BUT Those things are much larger than you think, especially for wider lenses. Also, getting them flat, is another battle as I found out the hard way. Not a bad idea, but I'm gonna trust the lens specs first, then purchase, then test.

I agree about higher resolution stitch potentially making up any significant filter hits. Still like to get most out of the lens first. If I can do less work and pay more money, I'm up for that. Cheers!

chris_ma
Posts: 571
Joined: Fri Mar 22, 2019 2:23 pm
Location: Germany

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by chris_ma »

well, as mentioned before, PTGUI does all the lens undistortion automatically but you seem to have made up your made to go the large format lens way.
chris

JKT
Posts: 424
Joined: Fri Oct 28, 2011 9:29 am
Location: Finland
Contact:

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by JKT »

chris_ma wrote:
Thu Feb 16, 2023 3:13 am
well, as mentioned before, PTGUI does all the lens undistortion automatically but you seem to have made up your made to go the large format lens way.
How does it handle stitching without rotation these days?

simplejoy
Posts: 46
Joined: Mon May 02, 2022 12:28 am
Contact:

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by simplejoy »

jvanhuys wrote:
Wed Feb 15, 2023 5:54 pm

Hey!

I can't tell you how valuable this is for me. Thank you so much.
It seems this lens is possibly showing its age due to worn out coatings or probably single-coating? Upping the contrast within reasonable levels seems to have a negligible filter hit compared to matrix operations like convolves, blurs, distortion/CA correction. Those things for my work, soften the image too much. When I increase the contrast by 25% on my Symmar-S 180 (single-coated) it is as good as my Micro Nikkor 105 in the centre... and it definitely beats the Nikkor in the corners with significantly better CA control and a touch less barrel distortion (possibly none?). On top of that I have the movements to do my stitches.

I'm trying to say, the lack of micro-contrast in the Dokumar doesn't worry me. My mortal enemies for flat artwork are distortion and CA. Even sharpness is lower priority. I noticed that the extreme shadows and highlights in ferrule of the brush, are very well handled by the Dokumar. That's a distance I would shoot too. The CA seems almost nullified at that . This is very promising. My Micro Nikkor 105, for example, is unusable with those extreme values due to CA. If we were looking at a 32-bit exr outside of Photoshop, in say, Nuke with a camera capable of ACES colorspace capture, with that contrast ratio, that same highlight's value would be between 30 and 50 and the shadow would be close to 0. So, again, I'm impressed.

I hear what you're saying about it being good for it's time but modern lenses being better. I get that it's not quite as sharp as the S-Biogon 40mm, or the Nikkor 55 Micro in terms of lines per millimeter, but I'm noting multiple things that I love. I looked at some very high resolution images of the S-Biogon on Flickr, and despite its remarkable sharpness, I noted that it has fairly bad distortion, of the pincussion-type. I had the Nikokr 55 for years, which is famously sharp too, but I sold it because it had too much barrel-distortion for flat artwork. The CA control was not bad, but nothing near a true Apochromat.

I think I'm gonna pull the trigger on the Dokumar and post some results once it arrives a few weeks from now. Thank you so much for spending time on this. Let me know if you ever want me to run a test setup for something you're curious about.
I'm glad it seems like a viable option for you. You're right - the Dokumar 47 mm seems more capable than my memory would have let me believe. Tried it again today and the main drawback seems to be a certain lack of contrast (besides the somewhat harsh bokeh in certain situations). Here are two SOOC shots:
https://www.flickr.com/gp/simple_joy/3688r0B2yG
https://www.flickr.com/gp/simple_joy/G7xh4hX09z

I also did two additional CA test shots in direct sunlight today and I think it did okay:
https://www.flickr.com/gp/simple_joy/321L9HE2oS
https://www.flickr.com/gp/simple_joy/3WwG24FCqd

All that being said, while I think this lens might absolutely be worth a try, I'm still wondering if there aren't a number of modern macro lenses which might have benefits that outweight the slight flaws... What about the Sigma 150mm f/2,8 EX DG Macro for example or the 105mm Art F2,8 DG DN Macro?

Lou Jost
Posts: 5987
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by Lou Jost »

What about the Sigma 150mm f/2,8 EX DG Macro for example or the 105mm Art F2,8 DG DN Macro?
The 105 Art is an extremely sharp lens with virtually no CA, but it has (correctable) distortion, a deliberate decision by Sigma to optimize the lens for CA-resistance and leaving distortion to be corrected by the automatic lens profile.

jvanhuys
Posts: 49
Joined: Sun Jan 22, 2023 9:24 pm
Location: Seoul, South Korea

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by jvanhuys »

simplejoy wrote:
Thu Feb 16, 2023 6:20 am
jvanhuys wrote:
Wed Feb 15, 2023 5:54 pm

Hey!

I can't tell you how valuable this is for me. Thank you so much.
It seems this lens is possibly showing its age due to worn out coatings or probably single-coating? Upping the contrast within reasonable levels seems to have a negligible filter hit compared to matrix operations like convolves, blurs, distortion/CA correction. Those things for my work, soften the image too much. When I increase the contrast by 25% on my Symmar-S 180 (single-coated) it is as good as my Micro Nikkor 105 in the centre... and it definitely beats the Nikkor in the corners with significantly better CA control and a touch less barrel distortion (possibly none?). On top of that I have the movements to do my stitches.

I'm trying to say, the lack of micro-contrast in the Dokumar doesn't worry me. My mortal enemies for flat artwork are distortion and CA. Even sharpness is lower priority. I noticed that the extreme shadows and highlights in ferrule of the brush, are very well handled by the Dokumar. That's a distance I would shoot too. The CA seems almost nullified at that . This is very promising. My Micro Nikkor 105, for example, is unusable with those extreme values due to CA. If we were looking at a 32-bit exr outside of Photoshop, in say, Nuke with a camera capable of ACES colorspace capture, with that contrast ratio, that same highlight's value would be between 30 and 50 and the shadow would be close to 0. So, again, I'm impressed.

I hear what you're saying about it being good for it's time but modern lenses being better. I get that it's not quite as sharp as the S-Biogon 40mm, or the Nikkor 55 Micro in terms of lines per millimeter, but I'm noting multiple things that I love. I looked at some very high resolution images of the S-Biogon on Flickr, and despite its remarkable sharpness, I noted that it has fairly bad distortion, of the pincussion-type. I had the Nikokr 55 for years, which is famously sharp too, but I sold it because it had too much barrel-distortion for flat artwork. The CA control was not bad, but nothing near a true Apochromat.

I think I'm gonna pull the trigger on the Dokumar and post some results once it arrives a few weeks from now. Thank you so much for spending time on this. Let me know if you ever want me to run a test setup for something you're curious about.
I'm glad it seems like a viable option for you. You're right - the Dokumar 47 mm seems more capable than my memory would have let me believe. Tried it again today and the main drawback seems to be a certain lack of contrast (besides the somewhat harsh bokeh in certain situations). Here are two SOOC shots:
https://www.flickr.com/gp/simple_joy/3688r0B2yG
https://www.flickr.com/gp/simple_joy/G7xh4hX09z

I also did two additional CA test shots in direct sunlight today and I think it did okay:
https://www.flickr.com/gp/simple_joy/321L9HE2oS
https://www.flickr.com/gp/simple_joy/3WwG24FCqd

All that being said, while I think this lens might absolutely be worth a try, I'm still wondering if there aren't a number of modern macro lenses which might have benefits that outweight the slight flaws... What about the Sigma 150mm f/2,8 EX DG Macro for example or the 105mm Art F2,8 DG DN Macro?
simplejoy, Thanks so much for this, I really appreciate it. I have to say this lens just blows me away. How did you mount it? What mount does it use? Please share with the rest of the group.

Lastly, can you sell it to me already? I did send you an email to that effect, but didn't get a response. Shall I take that as a "I'm keeping this gem"? Please answer yes or no, so my soul can rest. I need to pursue this lens as in yesterday:)

jvanhuys
Posts: 49
Joined: Sun Jan 22, 2023 9:24 pm
Location: Seoul, South Korea

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by jvanhuys »

To Chris Ma, Lou Jost and everyone else who keeps recommending the Sigma; are you guys working for Sigma or something? Coz you sound like salesmen:)

I did some tests tonight disregarding the back standard of my camera and mounting only to the front standard, so I'm still keeping my geared X,Y and Z movements, which are really important to me. I popped on the Micro Nikkor 105mm into a M39 adapter, then the M39 into a special adapter Rafael from RAF made for me to mount M39 to Copal 1 lensboard. It's not ideal, but it works. Because I can't control the back standard and the IC is small, I can of course only see straight. No more wiggle room anymore. So in a sense, my movements/ angle of view (sorry it's hard to explain) is now limited.... so are my stitches....(my tripod is mounted to a camera dolly on the floor, so I can only move forward and backwards, not left/right (only my LF camera does those)

BUT I'm willing to give the Sigma 105/ 150 a go. Is anyone on the thread selling theirs? Which one has less distortion and CA? I'm hoping that the lens is so sharp, that it might outresolve (with 4 stitches) an image taken with a top large format lens, doing 6-12 stitches.

Also, any honest comparison to the Laowa 90mm or longer focal appreciated. I shoot strictly manual, so hope that helps too. See this is why I come here guys, to learn from those who know more and change (if it adds up of course)... if I had all the answers, I wouldn't have bothered (and who really has all the answers anyway). Stay safe and chat tomorrow. It's now 1:11am in Seoul. Neon everywhere. Puts me at ease.
Last edited by jvanhuys on Thu Feb 16, 2023 6:50 pm, edited 1 time in total.

chris_ma
Posts: 571
Joined: Fri Mar 22, 2019 2:23 pm
Location: Germany

Re: 1:20 to 1:5 (0.05x to 0.2x)

Post by chris_ma »

I don‘t think I ever spoke about the sigma, but about the laowa 100mm, which is absolutely stunning and an incredible value.

I never tried the sigma art 105mm macro but from what I‘ve seen and read from people whom I trust it‘s pretty much comparable to the Laowa and a great choice if you prefer autofocus.

You can also take the latest macro lens from Sony E mount or Canon RF mount and spend more money - I‘m sure they are also very good.

The main point is: ALL of them are MUCH better then your Nikkor 105mm, which is a horrible macro lens imo.
They are also all MUCH better at F8 in their corners then any vintage large format lens will be at F11 in the corners.

You say you don‘t care about sharpness, but what I don‘t get is: why put up with the hassle of creating a 600MP file anyway when you use a soft lens?

Seems much more practical to do a 400MP image with a sharp lens, or even a 200MP single pixel shift image with a really really good lens (which Lou already suggested twice)
chris

Post Reply Previous topicNext topic