Zhongyi (Mitakon) Super Macro Lens (1 - 5x)

Macrero · Post by **Macrero** » Mon May 13, 2019 3:34 pm

I tried all those lenses years ago, I was not impressed honestly. I currently own the 80mm only, it works reasonably well in combos at lower magnification, and makes a nice short FL tube lens.

I am afraid we are hijacking this thread with this off-topic discussion. Sorry, Peter.

Lou Jost · Post by **Lou Jost** » Mon May 13, 2019 3:37 pm

Yep, save this for elsewhere....

ray_parkhurst · Post by **ray_parkhurst** » Mon May 13, 2019 4:27 pm

Lou Jost wrote:Each Scitex L-series scanner comes with three of those 80mm lenses built into an array for scanning and stitching long originals, plus one 67mm, one 89mm, and one 110mm lenses. The latter two lenses are really good. I had always considered the 80mm lenses as "filler material". Someone sold me two Scitex scanners and a set of spare lenses for them, so that's why I have so many. Maybe I should take a better look at them. The scanner apparently stitched the results from the three 80mm lenses automatically, so maybe they are specialized to have very low distortion.

I'd bet the lenses are 80/4 Rodagons with f5.6 stops, all 3 matched for focal length. The 80/4 Rodagon is a really nice lens. It did well in my 80mm shootout. Now, if it's more similar to the Beseler HD OEM version of the lens, then indeed it could be a real gem. That lens is superb.

Edited to add: sorry to extend the hijacking, did not see the last 2 posts when I posted this.

mjkzz · Post by **mjkzz** » Mon May 13, 2019 7:17 pm

Did it with Mitutoyo 5x 0.14 on a Raynox 250. I under exposed by -2/3EV because I noticed some area in previous images (both ZY and QV) were over exposed. This made the image look a bit better, to me at least. So this might not be fair

Pixel count is about 100MP

https://www.easyzoom.com/imageaccess/ad ... fec4a87fdb

mjkzz · Post by **mjkzz** » Mon May 13, 2019 7:18 pm

Next step is to use Mitutoyo 10x 0.28, pushed down to 6.25X with Raynox 250, to compare it with the 5x done with ZY, I can see Mitty to be better, but by how much

mjkzz · Post by **mjkzz** » Mon May 13, 2019 7:20 pm

Macrero wrote:I tried all those lenses years ago, I was not impressed honestly. I currently own the 80mm only, it works reasonably well in combos at lower magnification, and makes a nice short FL tube lens.

I am afraid we are hijacking this thread with this off-topic discussion. Sorry, Peter.

Well, if I had those lenses you mentioned, I would do comparisons. I am here, not to hype ZY lens, but to provide sample/comparison images for all of us.

So not a problem at all.

Macrero · Post by **Macrero** » Tue May 14, 2019 12:38 am

mjkzz wrote:
Well, if I had those lenses you mentioned, I would do comparisons. I am here, not to hype ZY lens, but to provide sample/comparison images for all of us.

So not a problem at all.

No need to compare it with all lenses out there either... You already did a great job testing this new lens. Seems a nice optics, although not of my interest.

Get a Mitty 7.5 if you have a chance, you'll love it.

mjkzz · Post by **mjkzz** » Tue May 14, 2019 5:36 am

Macrero wrote:
mjkzz wrote:
Well, if I had those lenses you mentioned, I would do comparisons. I am here, not to hype ZY lens, but to provide sample/comparison images for all of us.

So not a problem at all.
No need to compare it with all lenses out there either... You already did a great job testing this new lens. Seems a nice optics, although not of my interest.

Get a Mitty 7.5 if you have a chance, you'll love it.

Thanks, I will get the 7.5x if I can find one

mjkzz · Post by **mjkzz** » Tue May 14, 2019 5:42 am

OK, here we go, "10x" on Raynox 250, final image is about 440 MP

https://www.easyzoom.com/imageaccess/df ... 6d67129e5d

It is better than 5X done with ZY lens and this is noticeable, but just not as much as I expect. To be fair to Mitutoyo, all the obj's used are beat up ones, and the 10x might not even be a Mitutoyo as it is missing the outer shell. I got two of this 10x and I was told it is a Mitutoyo, I just took the vendor's word for it, though they ain't cheap in my eyes (roughly 200USD each) and this is the first time I use it.

Edit: forgot to mention, no sharpening, no enhancement, straight out of camera (after stack and stitch).

Edit 2: 66 stitched.

ray_parkhurst · Post by **ray_parkhurst** » Tue May 14, 2019 11:10 am

Wow, that's a big coin image! Very sharp as well. 66 image panorama. How are you managing the workflow of stitching 66 stacked images into the pano? I've been avoiding thinking about higher magnification due to the arduous workflow to process all those images. Do you have a batch solution?

mjkzz · Post by **mjkzz** » Tue May 14, 2019 6:47 pm

Thanks, for that image, it was relatively easy. It has 10x7 = 70 tiles, the 4 corners have 1 image each as they are basically blank. For the rest 66, it has 25 images each (total 1650 images).

Image acquisition took about 90 minutes, after that, I manually created 66 folders, putting all images belonging to the same tile into same folder. Then I drag and dropped 66 folders to Zerene's batch processing dialog. Then I just let Zerene do the magic (thanks Rik) and went to have lunch. I do not know how long it took to finish all stacking, but it was done when I came back.

So I think I will automate this process, even use some algorithm to automatically detect and group images, it should be relatively easy, sort of like scene detection in most video editors. But this is more about my system, so PM me if you want to know details.

Here is an image with DEEP stack, though it only has about 200MP, it took 5,000+ images to complete, just image acquisition took 4 hours, stacking took some lengthy time (did not pay attention). Some tiles are as deep as 7mm or more

https://www.easyzoom.com/imageaccess/cc ... fba94474cd

Why is it relevant to this topic? I think if I used ZY lens (it was done with QV), because ZY supports full frame and because ZY can change aperture to get more DOF (larger step size), and because ZY has less CA, and most importantly, ZY's near telecentric property, I think I could have done it in about 1/4 time, though with less sharpness, it should be good enough for client to view it, if approved, redo it with mitty.

So, that is what I will do next with this lens, I am just tying to evaluate it from user point of view.

mjkzz · Post by **mjkzz** » Wed May 15, 2019 12:03 am

OK, before I talked to ZY, I think maybe the aperture ring should be marked differently. Here are some analysis:

Take MP-E 65 as an example, it is marked as f/2.8 because at 5x, using NA analysis, it, in terms of light gathering capability, is equivalent to a thin lens with 15mm diameter, 41mm focal length, able to focus to infinity. If the real focal length is 45mm (which seems a little too long, because at this focal length, you can not get an image at 41mm, the WD at 5x), then the focal length matches very close to f/D, ie, 45/15 = f/3. This is close match.

Now, with ZY, it, in terms of light gathering capability, is equivalent to a thin lens with focal length of 100mm (WD) and diameter of 33mm, and is able to focus to infinity.

In both ZY and MP-E 65 cases, the equivalency, ie, the condition to use this formula, implies that the derived thin lens should be able to focus to infinity and both are "virtual" equivalency.

So since we are marking equivalency, I do not see any problem with marking the lens as f/0.8 because this is how a lens should be marked using N = f/D when its real focal length is 25mm and when considering focusing to infinity, particularly with photography lens.

However, it does sound "creative" as f/0.8 sounds ridiculously large and requires some explanation. So maybe the aperture should be marked as :

W 1, 2, 3, etc for Wide open, one stop down, two stops down.

This is because regardless what the calculated aperture is, for every mark, it will stop aperture down by one stop.

lothman · Post by **lothman** » Wed May 15, 2019 3:16 pm

do you know for what price they intend to sell this lens?

will there be a Sony FE-Version?

mjkzz · Post by **mjkzz** » Thu May 16, 2019 3:45 pm

yes, they said it will be 600USD retail.

mjkzz · Post by **mjkzz** » Thu May 16, 2019 3:50 pm

did the star anise again with this lens, it took about 3 hours, half on image acquisition, half on stacking, total 2070 images. This is about 3/8th of the time. I think I can make it even faster without losing image quality

https://www.easyzoom.com/imageaccess/10 ... 236f3f7c73

also did this beetle, about 500MP (half giga pixels), the worrying part for me was the tiles that contain legs and body -- those tiles are deep stacks, about 8mm deep, so if there were too much parallax error, they will not stitch -- tiles are connected only with legs, very little shared parts relative to whole image.

https://www.easyzoom.com/imageaccess/2b ... 39ed01484f