Minolta DImage Scan dual F-2400 - scanner lens
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Minolta DImage Scan dual F-2400 - scanner lens
Thanks to Robert O'Toole's excellent series of tests with the DImage Scan Elite 5400 lens, I put a note on our camera club forum to see if anyone had one they didn't need (specifically the Elite 5400 model). At the club meet last night, a chap rocked up with a plastic bag containing an old DImage Scan DuaL. I explained there might be a decent lens in there, but sadly not the one I was looking for. He said I could have it anyway as he was going to throw it away, and to just do the same if it was of no use to me.
So I pulled it apart this morning and got a nice little first-surface mirror out of it, along with the lens pictured. On first inspection, using it like a loupe, I was immediately struck by how sharp and contrasty the lens is (pitch black blacks) and also how far the image circle extends when you look at an angle to the edges. It actually looks quite promising.
This site is great - but flippin' distracting at times Today's plans are a washout due to weather but my current POL illumination experiments can wait a bit longer too. I'll be spending the rest of today mounting and testing this (hopefully) little gem. I'll report back here when done.
So I pulled it apart this morning and got a nice little first-surface mirror out of it, along with the lens pictured. On first inspection, using it like a loupe, I was immediately struck by how sharp and contrasty the lens is (pitch black blacks) and also how far the image circle extends when you look at an angle to the edges. It actually looks quite promising.
This site is great - but flippin' distracting at times Today's plans are a washout due to weather but my current POL illumination experiments can wait a bit longer too. I'll be spending the rest of today mounting and testing this (hopefully) little gem. I'll report back here when done.
Last edited by Beatsy on Fri Apr 27, 2018 5:31 pm, edited 1 time in total.
Thanks Saul. I've not done the test stacks yet, but thought I'd post an update.
I found an old Wild 40/0.65 objective that I never use. Unscrewing the back cap let the main optical assembly and a big spring fall out - so there's no need to completely trash the objective. The minolta lens was wrapped with a bit of PTFE and then slid nicely into the tube with a slight friction fit. I didn't put the spring and end cap back on because the minolta is nearly halfway up the tube and I'm sure the small hole in the back plate would cause vignetting. I was worried that the objective case itself might do that too, but it didn't. Parts pic here with the black minolta lens on the left - all the rest are parts from the Wild objective.
I raided my bits box and found an M42 focus mech and flat plate RMS to M42 adapter. These were fitted on an M42-NEX adapter that went on my A7rii. This arrangement gives a useful zoom range of 1.1x - 1.5x with no vignetting on full frame.
I also want to test the lens at mags approaching 2x. I have a larger focus mech, but it's too large. I found some fixed tube that will do for now - but will search some more for other bits and pieces. I can mount the objective backwards (so it extends inside the tube) which will give other mag ranges. As it stands though - this arrangement gives 1.85x on sensor (FF).
I couldn't resist a quick handheld shot. This is the result (no post processing other than levels adjustment and creating the 100% cropped inset). I think it looks quite promising. More in-depth tests later...
I found an old Wild 40/0.65 objective that I never use. Unscrewing the back cap let the main optical assembly and a big spring fall out - so there's no need to completely trash the objective. The minolta lens was wrapped with a bit of PTFE and then slid nicely into the tube with a slight friction fit. I didn't put the spring and end cap back on because the minolta is nearly halfway up the tube and I'm sure the small hole in the back plate would cause vignetting. I was worried that the objective case itself might do that too, but it didn't. Parts pic here with the black minolta lens on the left - all the rest are parts from the Wild objective.
I raided my bits box and found an M42 focus mech and flat plate RMS to M42 adapter. These were fitted on an M42-NEX adapter that went on my A7rii. This arrangement gives a useful zoom range of 1.1x - 1.5x with no vignetting on full frame.
I also want to test the lens at mags approaching 2x. I have a larger focus mech, but it's too large. I found some fixed tube that will do for now - but will search some more for other bits and pieces. I can mount the objective backwards (so it extends inside the tube) which will give other mag ranges. As it stands though - this arrangement gives 1.85x on sensor (FF).
I couldn't resist a quick handheld shot. This is the result (no post processing other than levels adjustment and creating the 100% cropped inset). I think it looks quite promising. More in-depth tests later...
OK, I'm going to have to alter my lighting setup for these larger subjects (at 1.8x). Too uneven here which kinda messes up the comparison between corner and centre 100% crops. Tomorrows task. However, for now, I think this still shows the performance of the lens pretty well, hence posting this first wafer test. The image quality is rather better than I expected it to be, especially in the corners!
Whole FoV, full frame, 1.8x magnification. No adjustments or sharpening for any of the images - straight out of PMax stacking (13 images, 30 micron steps). I purposely made the steps too small to start with, but the lens will easily support 60 microns and perhaps a bit larger.
100% crop from near the centre (offset a little so as to image the same chip pattern as the corner crop). Bear in mind, this is onto a 42 megapixel sensor with 4.5 micron pixels. Darn good resolution IMO (especially for the price ).
And finally, 100% crop from the top left corner. Sorry about the lighting change. The resolution is a little lower I think, but still pretty close to what was achieved in the centre.
I assume this lens is designed for 1x reproduction ratio or less, so it's being pushed quite a way out of spec at 1.8x. I'm impressed with it, but interested to hear others views on the performance. Good, bad or middling? Cheers.
Whole FoV, full frame, 1.8x magnification. No adjustments or sharpening for any of the images - straight out of PMax stacking (13 images, 30 micron steps). I purposely made the steps too small to start with, but the lens will easily support 60 microns and perhaps a bit larger.
100% crop from near the centre (offset a little so as to image the same chip pattern as the corner crop). Bear in mind, this is onto a 42 megapixel sensor with 4.5 micron pixels. Darn good resolution IMO (especially for the price ).
And finally, 100% crop from the top left corner. Sorry about the lighting change. The resolution is a little lower I think, but still pretty close to what was achieved in the centre.
I assume this lens is designed for 1x reproduction ratio or less, so it's being pushed quite a way out of spec at 1.8x. I'm impressed with it, but interested to hear others views on the performance. Good, bad or middling? Cheers.
-
- Posts: 1965
- Joined: Sat Oct 07, 2006 10:16 am
- Location: Bigfork, Montana
- Contact:
I'm done with testing. This lens works eminently well enough for me and fills a useful magnification gap in my objective-based setup (below a 5x Mitty which I can only push down to 2.6x on a crop sensor or 3.5x on FF). It's easy to switch between this (finite) setup and (infinite) mitties as both use M42 fittings - so I only have to unscrew stuff from the M42 adapter with no need to (re)move the camera on my rig.
A few last notes before trying an insect stack with it.
1) No aberrations or vignetting on a FF sensor from 1.2x to 2.5x. The "extended" usable range on FF is about 0.8x to 3.3x with a little bit of vignetting on the low end and diffraction kicking in at the top. It can be pushed down further on crop sensors but the lens will be *very* close to the camera and might cause issues with "reach" on some rigs. I intend to use the lens primarily at 1.8x.
2) Working distance is ~50mm from the front of the lens. It's nearer 45mm on my setup because the lens is recessed a little inside the RMS objective I used as a housing for it. That hasn't caused any issues.
3) The lens is sensitive to off-axis light hitting the front element, causing flare and general lack of contrast. Not surprising given the lens's intended use - the coatings aren't anti-reflective enough. A tube of flocking paper slipped over the barrel and extending 10-15mm past the front element completely fixes the issue (with a negligible loss of working distance). I flocked the inside of the objective housing "holder" too as bright features reflected off the inside and caused similar problems
4) Depth of field is around 100 microns-ish, but out-of-focus features at distances of 100mm-200mm+ behind a specimen remain clearly discernible. Unless you use pure black, or a colour gradient, be prepared to take more care arranging your backgrounds (than needed for higher mag objectives).
I'm very pleased with the lens. Needless to say, the gentleman who gave me that old scanner will receive a nice "gratuity" when I see him next week!
A few last notes before trying an insect stack with it.
1) No aberrations or vignetting on a FF sensor from 1.2x to 2.5x. The "extended" usable range on FF is about 0.8x to 3.3x with a little bit of vignetting on the low end and diffraction kicking in at the top. It can be pushed down further on crop sensors but the lens will be *very* close to the camera and might cause issues with "reach" on some rigs. I intend to use the lens primarily at 1.8x.
2) Working distance is ~50mm from the front of the lens. It's nearer 45mm on my setup because the lens is recessed a little inside the RMS objective I used as a housing for it. That hasn't caused any issues.
3) The lens is sensitive to off-axis light hitting the front element, causing flare and general lack of contrast. Not surprising given the lens's intended use - the coatings aren't anti-reflective enough. A tube of flocking paper slipped over the barrel and extending 10-15mm past the front element completely fixes the issue (with a negligible loss of working distance). I flocked the inside of the objective housing "holder" too as bright features reflected off the inside and caused similar problems
4) Depth of field is around 100 microns-ish, but out-of-focus features at distances of 100mm-200mm+ behind a specimen remain clearly discernible. Unless you use pure black, or a colour gradient, be prepared to take more care arranging your backgrounds (than needed for higher mag objectives).
I'm very pleased with the lens. Needless to say, the gentleman who gave me that old scanner will receive a nice "gratuity" when I see him next week!
Last edited by Beatsy on Sat Apr 28, 2018 8:47 am, edited 1 time in total.
As it happens, I had it the other way round for all the wafer shots in this post. The "back" of the lens, without the groove, was pointing at the subject. This is recessed a little more than the grooved end, so I was mainly testing for vignetting issues given the (overly) deep objective tube the lens is housed in. But it doesn't seem to make any difference to image quality as far as I can tell, though I've done no controlled tests.lothman wrote:The side of the lens with the groove is orientated to the camera - correct?
I've now got it mounted as you stated (groove to the front) and used it to do the bug stack I just posted. That came out fine with regards to sharpness and overall quality IMO (1.8x)
Note: I tried to post a link, but kept getting a blank reply when I included a URL. The new image is in 'technical and studio photography'
All complete guesses from eyeballing pictures. Maybe 25mm-30mm FL because of the perspective. And perhaps f/11 or f/13 or so - because of the way it made a sensor dust spot show up quite small and clear - though still a bit blurred. These guesses are based on my experience with regular lenses on my cameras so it could be utter tosh too. Do the numbers sound at all feasible?ChrisR wrote:Have you worked out/guesstimated a FL and F number for this lens?
- rjlittlefield
- Site Admin
- Posts: 23608
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Estimates based on appearance of sensor dust would be for effective aperture, not lens aperture. At 1.2x, f/11 to f/13 effective would make the lens aperture be about f/5 to f/6.
If you want accurate numbers, then the simplest way I know is to calculate FL = (ext1-ext2)/(m1-m2) for two different extensions and the corresponding magnifications. Pupil diameter can be measured with a ruler or calipers, best at arm's length, and then F-number is FN = FL/D.
--Rik
If you want accurate numbers, then the simplest way I know is to calculate FL = (ext1-ext2)/(m1-m2) for two different extensions and the corresponding magnifications. Pupil diameter can be measured with a ruler or calipers, best at arm's length, and then F-number is FN = FL/D.
--Rik