available Test chart for resolution sharpness

lothman · Post by **lothman** » Sun Mar 01, 2009 6:13 am

Hi together,

in order to compare different lenses at different system/magnifications it would be nice if we all had a similar test chart.

So is there something out there acessible for everybodey what could be taken to judge the quality of a lens, like poppy seed, banknotes or something like that.

Best regards from Germany
Lothman

NikonUser · Post by **NikonUser** » Sun Mar 01, 2009 6:52 am

How about a Blowfly?

Green bottle and Blue bottle flies are found worldwide. There are several species but for macrowork all with essentially identical properties. Perhaps the commonest is Lucilia vicina and it is cosmopolitan. You can never run out of detail, even all the way down to the electron microscope level.

Man-made objects lack further detail at some, usually low, magnification.

see Charles' Blowfly as an example
HERE
and the linked images

Planapo · Post by **Planapo** » Sun Mar 01, 2009 1:21 pm

Yeah, obtaining better comparable images to judge for lens performance is a good idea. And a blow fly being easily available everywhere, appears to be a suitable "test chart". For higher magnifications that do not show the whole fly or larger parts, we could agree upon a suitable, defined area of the fly's body.

--Betty

lothman · Post by **lothman** » Sun Mar 01, 2009 2:32 pm

the target should be something flat, would make it easier to judge image quality over the frame.

DaveW · Post by **DaveW** » Mon Mar 02, 2009 3:21 am

See:-

http://www.pbase.com/iangreyphotography/image/82910480

Click on "Library" in the link below to download USAF test charts:-

http://www.darkroomagic.com/

DaveW

rjlittlefield · Post by **rjlittlefield** » Mon Mar 02, 2009 9:14 am

Choosing a good standard target involves some tradeoffs.

Man-made targets are available, but very expensive. The ballpark price is about $1 per lppm (line pair per mm) of resolution. For example a target that includes 250 lppm will cost around $250. This is a serious issue because high magnification macro work begins at somewhere around 100 lppm on the subject and goes up from there to 500-1000 lppm with microscope objectives.

There is a trick for cheaply producing high resolution virtual targets using a microscope objective as a tiny projection lens. Basically you set up the microscope optics "in reverse" to turn a large resolution target into a tiny image of a resolution target, which you look at using the lens being tested. This works OK, but it's not exactly easy to set up.

In microscopy, many people use diatoms as resolution targets because they're cheap, widely available, and consistent enough to be usable. The "dots" of Amphipleura pellucida are famous. See for example the discussion and images HERE. BTW, I strongly recommend to read that article and pay attention to the careful treatment of illumination. Macro photography is not quite as sensitive to lighting as micro is, but still the lighting can make a big difference in detectable detail and thus apparent resolution.

The blowfly has a lot going for it. Worldwide availability of similar targets is a huge advantage. The blowfly also may have the advantage that its structures are less "twitchy" about the lighting. I'm not sure about that aspect with the blowfly, but I do know that the amount of visible detail on butterfly and moth scales can change a great deal depending on illumination.

Having a flat subject is good for a couple of reasons, but you don't want "flat" to impact repeatability. Moth and butterfly wings are nicely flat, but there's a lot of variation from one species to another, and also from one part of the wing to another. Same problem with wood grain, and there again, it's not easy to prepare even a single suitable specimen. Nicely sheared or polished end grain can't be found lying around in windowsills and wasp traps.

I'll be interested to hear more of this discussion.

--Rik

g4lab · Post by **g4lab** » Mon Mar 02, 2009 9:33 am

How about EPROM memory with a window for ultraviolet erasure.
For incident light illumination tests.

Joseph S. Wisniewski · Mon Mar 09, 2009 6:04 pm

g4lab wrote:How about EPROM memory with a window for ultraviolet erasure.
For incident light illumination tests.

The transparent window will throw all the tests off. Fast lenses look at a pretty wide cone of light, and the shallower the angle, the longer a distance through the window light has to travel and the more it gets shifted directionally. The end result is a dramatic increase in spherical aberration.

That's why microscope objectives are marked as working with particular thickness of cover glass, and the fat high magnification ones have "compensating collars" that let you adjust the objective's SA to compliment a range of cover glass thickness.

That aside, I was going to propose the use of "popped top" microprocessor chips. Their dies are huge, they're wonderfully flat, and they have features down to insanely small levels. Homomorphic deconvolution can compute resolution from the patterns of a chip.

elf · Post by **elf** » Mon Mar 09, 2009 6:45 pm

Joseph S. Wisniewski wrote: That aside, I was going to propose the use of "popped top" microprocessor chips. Their dies are huge, they're wonderfully flat, and they have features down to insanely small levels. Homomorphic deconvolution can compute resolution from the patterns of a chip.

How do you 'pop a top' ?

g4lab · Post by **g4lab** » Mon Mar 09, 2009 9:18 pm

Joseph you make an excellent point. When I get a chance I will see whether it is possible to punch out the quartz window with out rendering the chip useless as a target.

And if you have a good method for popping the tops of big chips I would be interested too.

I suppose you could find something with chip on board technology and
use methylene chloride to dissolve the epoxy.

Joseph S. Wisniewski · Tue Mar 10, 2009 5:45 pm

rjlittlefield wrote:There is a trick for cheaply producing high resolution virtual targets using a microscope objective as a tiny projection lens. Basically you set up the microscope optics "in reverse" to turn a large resolution target into a tiny image of a resolution target, which you look at using the lens being tested. This works OK, but it's not exactly easy to set up.

I'll have to look into that. I have the absolute perfect lens for the purpose: a 28mm Ultra-Micro Nikkor, designed for photolithography in IC fabrication. I can't imagine a lens with a flatter field or higher resolution than an UMN.

Joseph S. Wisniewski · Tue Mar 10, 2009 5:50 pm

elf wrote:
Joseph S. Wisniewski wrote: That aside, I was going to propose the use of "popped top" microprocessor chips. Their dies are huge, they're wonderfully flat, and they have features down to insanely small levels. Homomorphic deconvolution can compute resolution from the patterns of a chip.
How do you 'pop a top' ?

For ceramic packages, I've found that taking them up to about 250 deg C will weaken their seals to the point where a vice with wedges in the jaws will open them easily. Literally, they go "pop" and the lid sails away.

rjlittlefield · Post by **rjlittlefield** » Tue Mar 10, 2009 7:07 pm

Joseph S. Wisniewski wrote:28mm Ultra-Micro Nikkor, designed for photolithography in IC fabrication. I can't imagine a lens with a flatter field or higher resolution than an UMN.

Great lens in general. From what I read, I gather it's rated at f/1.8. Assuming P=1, that's max NA=0.28, more like 0.25 at 10X. So it'll work great for evaluating most macro lenses, but it won't do full justice to microscope objectives.

When I suggested this scheme to the Yahoo Microscope group last summer, as a means of generating test targets to compare stereo microscopes, one fellow ended up using a 20X NA 0.65 as the projector. That would be wide enough for anything I care about right now.

But the blowfly is probably a lot cheaper and easier to get!

--Rik

lothman · Post by **lothman** » Wed Mar 11, 2009 6:24 am

g4lab wrote:Joseph you make an excellent point. When I get a chance I will see whether it is possible to punch out the quartz window with out rendering the chip useless as a target.

And if you have a good method for popping the tops of big chips I would be interested too.

I suppose you could find something with chip on board technology and
use methylene chloride to dissolve the epoxy.

Eproms are a great idea. I was curious and just took one with ceramic top and base plate, clamped it in a bench vice and put a chisel on the glued gap between the two ceramic plates. On littel blow with a hammer and the top plate break off and the chip can be accessed without window. Perhaps later I'll make some pics. So we could even decide for our Standard Eprom

For resolution it is OK, contrast and color are very strong depending on lightning and difficult to reproduce.

g4lab · Post by **g4lab** » Wed Mar 11, 2009 7:53 am

Now that I think of it, often the pictures of these Eproms shot by the companies that build and analyze them, often use incident light interference contrast. Polarization would probably help too. It would affect insulating layers that might be present.

I don't know what the effect on polarized light when incident on semi conducting material. Probably depends on the presence or absence of bias voltage.

I have seen these chips used as demo targets by several major microscope manufacturer's at shows. But the scopes in question were either stereos or makros of similar range. The long working distance allowed focusing through the window. I suppose there was distortion but it was not noticeable.
The NA was probably no higher than about 0.100, if even that high.