This is a 4.5 cm wide specimen of Spessartine garnet from Navegadora claim, Penha do Norte, Conselheiro Pena, Minas Gerais, Brazil. This particular mine is known for its exquisite stepped terminated gemmy garnets. I got carried away on this particular piece and had great results experimenting with stereo pairs.
Entire specimen and stereo pair, Nikon Zoom Micro-Nikkor 70-180mm at f/5.6 stacked with Zerene Stacker
Magnified view at 3x (8mm FoV) with Minolta 5400 scanner lens, with careful balance of diffused front/overhead lighting with backlight via high-CRI LED point source. This is the top upper right of the specimen, from a shallower angle than the overall image. I may have to print this one! The stereo pair turned out excellent as well.
Spessartine garnet
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Re: Spessartine garnet
The backlit versions are wonderful!
- rjlittlefield
- Site Admin
- Posts: 23605
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Spessartine garnet
That's a gorgeous specimen, and the images came out great!
The stereo aspect is still giving you trouble, though.
The first pair (as posted at this moment) is not stereo at all. Both images are identical. Pull that pair into StereoPhoto Maker, hit the Swap Left/Right button, and there's no change to the appearance.
The second pair is stereo, and quite good in that there are very few stereo artifacts. However, it appears to be laid out for parallel viewing, not crossed-eye which is customary in the forum. I cannot be 100% sure about that because I don't know the actual geometry, and the geometry as viewed is possible both normal and inverted. However, in the pair as posted, I have to view in parallel to make the dark upper edge of the stone recede into the background as I think it should. When viewed crossed, those dark sections protrude forward like an unusually sharp lip on a bowl.
--Rik
The stereo aspect is still giving you trouble, though.
The first pair (as posted at this moment) is not stereo at all. Both images are identical. Pull that pair into StereoPhoto Maker, hit the Swap Left/Right button, and there's no change to the appearance.
The second pair is stereo, and quite good in that there are very few stereo artifacts. However, it appears to be laid out for parallel viewing, not crossed-eye which is customary in the forum. I cannot be 100% sure about that because I don't know the actual geometry, and the geometry as viewed is possible both normal and inverted. However, in the pair as posted, I have to view in parallel to make the dark upper edge of the stone recede into the background as I think it should. When viewed crossed, those dark sections protrude forward like an unusually sharp lip on a bowl.
--Rik
-
- Posts: 108
- Joined: Sun May 10, 2020 12:56 pm
Re: Spessartine garnet
The top image is definitely a stacked stereo pair but the differences are subtle. I think I had Zerene set to around 1% or less (maybe around 0.6%) instead of 3%; I got more artifacts if I tried to go higher probably because I overlapped focus in the stack too much.
The close-up backlit stereo pair was stacked in Zerene with conical path and I believe the default 3%; I think this is intended for cross-eye viewing though you would be the definitive resource! The dark shelf/point at the upper right does represent the actual correct geometry of the specimen. The overhead lighting was only slightly diffused so the rest of the shelf does not have a white reflection.
Edit: The specimen was also rotated for that lower image, which brought these protruding edge points closer to a similar plane so the stack didn't have to be too deep.
The close-up backlit stereo pair was stacked in Zerene with conical path and I believe the default 3%; I think this is intended for cross-eye viewing though you would be the definitive resource! The dark shelf/point at the upper right does represent the actual correct geometry of the specimen. The overhead lighting was only slightly diffused so the rest of the shelf does not have a white reflection.
Edit: The specimen was also rotated for that lower image, which brought these protruding edge points closer to a similar plane so the stack didn't have to be too deep.
- rjlittlefield
- Site Admin
- Posts: 23605
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Spessartine garnet
Photoshop disagrees. I tried layering the left and right halves, with blending mode "Difference", and no edges appear -- only random differences in pixel values, mostly in the range of 0-4 out of 255, none greater than 12. A shift of 0.6%, in an image size 512 pixels width, is still 3 pixels left/right. That much shift, while producing a weak stereo effect, would be very obvious with the layer-and-difference technique. The test of repeatedly clicking Swap Left/Right in StereoPhoto Maker is also very sensitive. If nothing visibly shifts anywhere in the image when the button is clicked, then there's no stereo.physicsmajor wrote: ↑Sat Aug 15, 2020 2:54 pmThe top image is definitely a stacked stereo pair but the differences are subtle.
As further confirmation, if I compare either of the views in the pair against your larger single frame (resized to match), then I do see a clear rotation, and laying out those two as a stereo pair results in a clear stereo effect. So whatever rotation you used is fine, it's just that both sides of the pair have the same rotation.
Thinking about what might have gone wrong, I wonder if perhaps you accidentally specified both shifts with the same sign. It's either that or you accidentally constructed the pair with two copies of one view.
Unfortunately, what I can definitely say is "it depends". What it depends on is which order the stack was shot (front-to-back versus back-to-front), combined with whether Zerene Stacker saw fit to reverse the order to avoid edge streaks. There's an item someplace on the "requested features" list to let the user tell the software about the shooting order, and have that automatically propagate through to determine the left/right order for stereo. But right now, the safest approach is to imagine that left/right order is random, and figure it out based on actual geometry versus appearance of the pair.The close-up backlit stereo pair was stacked in Zerene with conical path and I believe the default 3%; I think this is intended for cross-eye viewing though you would be the definitive resource!
--Rik
-
- Posts: 85
- Joined: Sun Mar 26, 2017 3:18 pm
- Location: Tennessee
- Contact:
-
- Posts: 1527
- Joined: Mon Jan 15, 2018 9:23 pm
- Contact:
Re: Spessartine garnet
Backlit minerals indeed generate some of the most beautiful "macro landscapes", well done!