Boulder opal

[physicsmajor]

Yoweh boulder opal in ironstone matrix, SW Queensland, Australia. This small specimen is presented along its natural cleavage face (irregular, no polish) showing significant multicolored fire throughout dark gray translucent opal in the shattered orange/brown host rock.


2.7x magnification with Minolta 5400 scanner lens, 8.5mm FoV:

And 3D stereo pair (to get a sense of the irregular surface and how some 'flakes' of opal fire are not at the surface):

At 100% with the 2.7x images I could see linear striation/texture in the flakes of reflected fire, so I decided to change systems and see if I could get more detail from the individual opal "flakes" of fire. And... the results surpassed expectations. Mitutoyo M Plan Apo 10x with Carmen Haas 200mm tube lens (set up as recommended by Mitutoyo, tube lens not reversed, actual 10x magnification with FoV 2.4 mm):

Stereo pair (didn't turn out ideally but it illustrates the depth of these different flakes of fire):

Different area at 10x:

Stereo pair (this one turned out better):

[physicsmajor]

It's come to my attention the stereo pair from the first 10x image is likely swapped; see if this works better.

[rjlittlefield]

At this moment, both the original of this pair and the update of this pair have the same ordering, and it's correct for crossed-eye view.

The problem is with the other two pairs, which are currently correct only for parallel view.

I have explained more fully in email.

--Rik

[physicsmajor]

I manually swapped because in the preview view, this looked incorrect. Then on revision I opened the incorrect version and manually swapped it again, resulting in the same output (with a bit of cropping on the edges to remove the lines from stereo artifacts).

This is the actually corrected version.

[rjlittlefield]

So, physicsmajor and I have worked this out by email.

It turns out there was a misunderstanding of what "crossed-eye" means, with the result that all of the stereo pairs he has considered correct are ordered as what we call "parallel".

All can be understood with enough patience. View accordingly...

--Rik