Diatom Aulacodiscus

[Smokedaddy]

Trying to get back into Microscopy. I haven't done much Diatom imaging for a while. Middle image Zerene, the other two Helicon.

-JW:

[rjlittlefield]

I do not know the actual structure of that diatom, but I am always worried about Helicon's treatment of them. Helicon Focus produces 3D by fitting a surface model. That approach allows rendering at angles that are very far off axis, which is good, but unfortunately it also tends to smooth out detailed shapes, sometimes to the point of being totally unrealistic. See for example viewtopic.php?p=55619#55619 .

Since you have Zerene Stacker, I suggest to use its stereo capability to let you see the actual structure, and then you can decide whether the Helicon rendering is close enough to meet your needs.

--Rik

[Smokedaddy]

Okay, I'll give it a shot. Number 7 below.

[Smokedaddy]

Here you go. Parallel with Zerene.

[rjlittlefield]

Thanks. As I suspected, each of those ball-on-stalk structures actually sticks out quite a long ways, essentially perpendicular to the smooth wavy surface of the disk. Helicon's version shows them bent over and plastered down against the surface -- totally misleading.

--Rik

[Macro_Cosmos]

I agree, the HF rendering is completely off. The hands of the frustules are supposed to stick up, much like the legs on a crab. To illustrate what I am trying to say, a simple search of "aulacodiscus SEM" brings this:

(Not the exact species, same genus.)

Those hands (called processes) are facing up like thorns!

[Smokedaddy]

BTW, I did search for SEM "Aulacodiscus sollittianus" and never found an SEM image of it specifically. The only thing I found was item #7 as I posted above. Maybe I have it identified incorrectly too, dunno. I did find this colored SEM labeled as Aulacodiscus oregonus.

-JW:

[Smokedaddy]

Here is another parallel attempt with Helicon.

[rjlittlefield]

The problem is not the rendering, it's the model. Helicon's 3D model of the surface is not tracking the actual structure. In this case it could certainly be doing a lot better than it is right now, but as long as they derive their model from the depth map, it will never be able to correctly handle 3D structures where there are front/back overlaps like the five of these projections that still have their bulbous tips.

--Rik

[Pau]

To illustrate what I am trying to say, a simple search of "aulacodiscus SEM" brings this:
...image...
(Not the exact species, same genus.

I did the image search and it seems to be Stephanodiscus, a different genus.
(And, please, quote the source of the image. You are allowed to post it because it is licensed under Creative Commons, but citation is needed. At PMG.net we are pretty strict with authors rights)

BTW, I did search for SEM "Aulacodiscus sollittianus" and never found an SEM image of it specifically. The only thing I found was item #7 as I posted above. Maybe I have it identified incorrectly too, dunno. I did find this colored SEM labeled as Aulacodiscus oregonus.

I would say the same than to MC, although in this case the image seems to be in stock pages

[Macro_Cosmos]

To illustrate what I am trying to say, a simple search of "aulacodiscus SEM" brings this:
...image...
(Not the exact species, same genus.

I did the image search and it seems to be Stephanodiscus, a different genus.

Whoops, wrong diatom! Not sure why Gettyimages called it Aulacodiscus, those are not processes. I was supposed to link this one:

Found here:
https://fineartamerica.com/featured/70- ... ssner.html

Source of the one linked above: https://www.sciencephoto.com/media/617806/view
That one is Stephanodiscus, my bad. The arms are called spines, I think.

Okay, this one is better:
https://www.researchgate.net/publicatio ... gures?lo=1
Free to download as well.


Anyway, the 3D rendering is totally off.

[rjlittlefield]

For whatever it's worth, the projections ("processes") that I see in the Zerene stereo look very much like the inset numbered 8 in that last panel, except for the apparently broken-off one at the 2 o'clock position, which looks like the inset numbered 9.

Macro_Cosmos, great find on that paper. I see that the number of processes is variable, which answers one question I have long had about this group.

--Rik

[Smokedaddy]

Thanks for finding the SEM's and input. For some reason, I wasn't able to find the proper SEM. Maybe that's a DuckDuckGo problem, dunno. As far as the LAST stereo pairs I posted, meaning the updated Zerene and Helicon, I personally can't see the difference in the projections. Maybe I'm not understanding. Then again my eyes are failing rapidly.