Standards seem to vary, can't find a definition (on a working page) on the 3M site, for what eg "1200" means.
http://en.wikipedia.org/wiki/Sandpaper gives around 10 microns.
Others quote grit sizes down to a few microns, with "polishing sheets" down to 1 micron.
The grains should have sharp edges!
new mfg Nikon 10X NA 0.25 finite objective looks very good
Moderators: rjlittlefield, ChrisR, Chris S., Pau
I have a resolution tester out there like your smoked glass.
It's a proof silver american eagle (mirrored fields). I was trying to clean a couple pieces of dust off of it for photography and used some canned air. Got the blast a little close and spattered the coin with the propellant. Ruined the coin, but it now has really nice and colorful patterns of fine details - well under 1 um.
It's a proof silver american eagle (mirrored fields). I was trying to clean a couple pieces of dust off of it for photography and used some canned air. Got the blast a little close and spattered the coin with the propellant. Ruined the coin, but it now has really nice and colorful patterns of fine details - well under 1 um.
-
rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
ChrisR, I just took a quick look. You're right about the amount of available fine detail. I'll post out some pictures later today.
Thanks for prompting me to take another look at sandpaper as a target.
The last two times I looked, I was not very encouraged -- what I saw struck me as not even interesting enough to post out.
But that was a couple of years ago -- different materials, different optics, different lighting, different camera, probably even a different me -- who knows what went wrong.
Today it's working fine.
It's interesting how much value there is in revisiting things you thought you knew!
--Rik
Thanks for prompting me to take another look at sandpaper as a target.
The last two times I looked, I was not very encouraged -- what I saw struck me as not even interesting enough to post out.
But that was a couple of years ago -- different materials, different optics, different lighting, different camera, probably even a different me -- who knows what went wrong.
Today it's working fine.
It's interesting how much value there is in revisiting things you thought you knew!
--Rik
I remembered which drawer in the attic I'd put the Wetordry paper I liberated from the last metallurgy lab I worked in ( and had horrid memories of the car paintjob it was destined for). The finest we used, before going to 1 micron diamond spray, was grade 400.
Seems a big gap. So I called in my local car accessory shop and asked for the finest grade they had.
Matador SiliciumCarbid 991A Softflex
grade 3000. Well gollygosh.
Everything gets miniaturised!
Seems a big gap. So I called in my local car accessory shop and asked for the finest grade they had.
Matador SiliciumCarbid 991A Softflex
grade 3000. Well gollygosh.
Everything gets miniaturised!
-
rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
-
rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
As promised, here are some sandpaper images.
On the left is "1500 Grit Mirror Fine Silicon Carbide" from the shelves of Ace Hardware some years ago. The stuff doesn't show on their web site. I don't know if it is still available in their stores.
On the right is "12000 grit Soft Touch Pad" from Micro-Mark. The picture doesn't look quite like the product that I have. Mine are all uniform colors, not the mottled coloring shown on their web site.
The images shown here are from the new Nikon Achromatic Finite Conjugate 10X NA 0.25 objective operating at 10X on a standard 150 mm extension. These images were shot at full camera resolution 4752x3168, then reduced to 70.7% (3360x2240) and USM'd 50% at 0.6 pixel to look sharp for display. The level of detail you're seeing here is what would appear in a 22.4" x 14.9" print at 150 pixels per inch.
--Rik
On the left is "1500 Grit Mirror Fine Silicon Carbide" from the shelves of Ace Hardware some years ago. The stuff doesn't show on their web site. I don't know if it is still available in their stores.
On the right is "12000 grit Soft Touch Pad" from Micro-Mark. The picture doesn't look quite like the product that I have. Mine are all uniform colors, not the mottled coloring shown on their web site.
The images shown here are from the new Nikon Achromatic Finite Conjugate 10X NA 0.25 objective operating at 10X on a standard 150 mm extension. These images were shot at full camera resolution 4752x3168, then reduced to 70.7% (3360x2240) and USM'd 50% at 0.6 pixel to look sharp for display. The level of detail you're seeing here is what would appear in a 22.4" x 14.9" print at 150 pixels per inch.
--Rik
-
rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
The 6.9X images are stacked. There is far too much curvature of field in the Edmund lens at this magnification to be useful without stacking. It is about 125 microns in depth from corner to edges.
Here is an individual frame, to give you an idea how bad the curvature of field is. Bear in mind the target is a microscope slide, perfectly flat.
Now, about the blurring...
The extreme streaking in the corner of the CF N image (second) is due to aberrations in the lens. The individual images look equally awful.
In the Edmund image, some blurring is due to aberrations and some is due to stacking. The difficulty with stacking in this case is that a) the lens exhibits a small but visible scale change from frame to frame, and b) with this type of target the alignment algorithm cannot figure out exactly what the scale change is and then correct for it. The result is that best results are achieved by turning off alignment altogether and just accepting the resulting radial blur.
In the image shown earlier, I allowed X and Y shift alignment and I used the PMax stacking method, which may combine pixel data from more than two frames. I just now reran the stack with alignment turned off completely and using the DMap stacking method, which is guaranteed to combine pixel data from at most two frames. The resulting image has slightly different alignment and is significantly sharper away from image center.
It's always interesting in these tests, trying to wring out the best possible performance from a lens et.al. There are a lot of interactions that can muck up the comparisons, and I learn more about them every time I tackle one of these jobs.
Thanks for the question.
--Rik
Here is an individual frame, to give you an idea how bad the curvature of field is. Bear in mind the target is a microscope slide, perfectly flat.
Now, about the blurring...
The extreme streaking in the corner of the CF N image (second) is due to aberrations in the lens. The individual images look equally awful.
In the Edmund image, some blurring is due to aberrations and some is due to stacking. The difficulty with stacking in this case is that a) the lens exhibits a small but visible scale change from frame to frame, and b) with this type of target the alignment algorithm cannot figure out exactly what the scale change is and then correct for it. The result is that best results are achieved by turning off alignment altogether and just accepting the resulting radial blur.
In the image shown earlier, I allowed X and Y shift alignment and I used the PMax stacking method, which may combine pixel data from more than two frames. I just now reran the stack with alignment turned off completely and using the DMap stacking method, which is guaranteed to combine pixel data from at most two frames. The resulting image has slightly different alignment and is significantly sharper away from image center.
It's always interesting in these tests, trying to wring out the best possible performance from a lens et.al. There are a lot of interactions that can muck up the comparisons, and I learn more about them every time I tackle one of these jobs.
Thanks for the question.
--Rik
-
rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Update...
After some weeks of being out-of-stock, this objective is now again available at Edmund, HERE.
Interestingly, the price has dropped. It is $73 USD today, with no indication that it's any sort of special offer.
Many thanks to Paul Furman for pointing this out.
--Rik
After some weeks of being out-of-stock, this objective is now again available at Edmund, HERE.
Interestingly, the price has dropped. It is $73 USD today, with no indication that it's any sort of special offer.
Many thanks to Paul Furman for pointing this out.
--Rik
What about everyone who is interested purchasing an agreed-upon subset of the Edmund (or other vendor) microscopy test objects and test patterns?
Just a thought.
I do like the richly complex and seemingly use-relevant character and texture of your test pattern, though.
Just a thought.
I do like the richly complex and seemingly use-relevant character and texture of your test pattern, though.
-Phil
"Diffraction never sleeps"
"Diffraction never sleeps"
-
Brodie Foster
- Posts: 59
- Joined: Mon Jul 05, 2010 3:31 am
Last edited by Brodie Foster on Sat Dec 17, 2016 11:55 pm, edited 1 time in total.
-
rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
For a finite objective, I still recommend this thing. The image quality and range of coverage is not as good as the new infinite design Nikon CFI Plan Achromat 10x NA 0.25 WD 10.5mm (part number MRL00102), reviewed HERE and HERE. But if you want something to stick on bellows or extension tubes, then the Nikon Achromatic Finite Conjugate 10X NA 0.25 160/- from Edmund works really well.Brodie Foster wrote:Rik, a year later does this still stand as one of the best solutions for 10x in this price range? It is still available from Edmund Optics although at a higher price. I am looking for something to eventually replace my Nikon E 10x 160/0.25 LWD.
I'm curious about your "higher price" comment. When I visit Edmund just now, the 10X NA 0.25 is still showing as $73, which is the same low price I saw back in July 2010. What are you seeing?
I have no idea how this lens would compare against your Nikon E 10x 160/0.25 LWD. We think that the "Nikon Achromatic Finite Conjugate Objectives" currently sold by Edmund are closely related to the now-discontinued Nikon E Achromat objectives shown in the brochure at http://www.krebsmicro.com/Nikon_CF.pdf (illustrated on page 12, tabulated on page 14). I'm guessing that your lens might be from the E Plan Achromat series (pages 11 and 14 of the brochure). But even if it is, I still don't know how the two lenses compare against each other.
What sort of deficiencies are you seeing in your current LWD?
--Rik
-
Brodie Foster
- Posts: 59
- Joined: Mon Jul 05, 2010 3:31 am
Last edited by Brodie Foster on Sat Dec 17, 2016 11:56 pm, edited 1 time in total.
-
rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
No idea on the paint, sorry.
About the E PLan 10X LWD, John Hallmén's report suggests to me that for focus stacking at 10X, there's probably not a lot of difference between it and the new Edmund lens. How the E Plan LWD holds up at lower magnifications, can't tell without running the test. It certainly came as a surprise to me that the CF N Plan Achromat 10X NA 0.30 falls off so quickly outside 28 mm diameter image circle. No idea if the E Plan does the same thing.
That 40X 0.55 LWD could be an interesting lens for somebody. I think that 0-2 notation means it has a correction ring that allows getting good images all the way from no cover glass to looking through a full 2 mm of glass. Judging from page 14 of the CF brochure, it might give up to 2.39 mm working distance with no cover glass. Pretty tight to get light around, but workable with some fiddling, and being able to see through glass would occasionally be useful. You might want to track eBay for a while to see what people are paying for them. (Price of this lens is yet another thing I have no idea about!)
--Rik
About the E PLan 10X LWD, John Hallmén's report suggests to me that for focus stacking at 10X, there's probably not a lot of difference between it and the new Edmund lens. How the E Plan LWD holds up at lower magnifications, can't tell without running the test. It certainly came as a surprise to me that the CF N Plan Achromat 10X NA 0.30 falls off so quickly outside 28 mm diameter image circle. No idea if the E Plan does the same thing.
That 40X 0.55 LWD could be an interesting lens for somebody. I think that 0-2 notation means it has a correction ring that allows getting good images all the way from no cover glass to looking through a full 2 mm of glass. Judging from page 14 of the CF brochure, it might give up to 2.39 mm working distance with no cover glass. Pretty tight to get light around, but workable with some fiddling, and being able to see through glass would occasionally be useful. You might want to track eBay for a while to see what people are paying for them. (Price of this lens is yet another thing I have no idea about!)
--Rik