Johnny-Jump-Up, next generation coming up...

rjlittlefield · Post by **rjlittlefield** » Fri Apr 11, 2008 10:56 pm

It must be Spring -- Rik is ripping the flowers apart again!

These are the reproductive parts of Johnny-Jump-Up, Viola tricolor. You can see the whole flower here. The image shown above is with the bottom petal removed, looking up into what would normally be the back bottom side of the flower.

From a technical standpoint, the most interesting thing about this image is that it has had no manual retouching whatsoever. Note the absence of halo against the dark background and where there is strong overlap of foreground/background structures. All the tools necessary to do this are now available off the shelf, though at this moment controlling them is far from easy. With a bit of luck, all that will soon change. In the meantime, I bid you adieu 'cuz it's my bedtime now.

--Rik

Technical: Canon 300D with Olympus 38mm bellows lens on 97 mm extension, marked f/4, 28 frames at 0.005" focus step.
Aligned with PTGui, stacked with TuFuse -p 1 --wMode 1 --wExposure 0 --wContrast 1 --wSaturation 0 -e 10

MacroLuv · Post by **MacroLuv** » Sat Apr 12, 2008 12:42 am

Hey, it looks absolutely great!

Nice piece of software too. Thanks for the link.

Planapo · Post by **Planapo** » Sat Apr 12, 2008 3:13 am

Rik wrote:

It must be Spring -- Rik is ripping the flowers apart again!

Good Mornin' you odd gardener!

Yeah, this looks excellent. Did you report on the software earlier and I happened to miss it?

Thanks for the eye candy on this Saturday morning.
--Betty

Charles Krebs · Post by **Charles Krebs** » Sat Apr 12, 2008 8:28 am

It does look very good!

...comparison stack...comparison stack...comparison stack...

acerola · Post by **acerola** » Sat Apr 12, 2008 11:59 am

I tried the software also. It's good. There is no halo and no exposure differences. There is a little problem with overlapped things, as with all the others.

rjlittlefield · Post by **rjlittlefield** » Sat Apr 12, 2008 12:52 pm

Planapo wrote:Did you report on the software earlier and I happened to miss it?

Nope, this was the first image that I've posted.

Charles Krebs wrote:...comparison stack...comparison stack...comparison stack...

But of course! These things just take time...

See http://www.photomacrography.net/forum/v ... php?t=4721 .

Péter, the treatment of overlap is often improved if your images are well registered before running them through TuFuse. In some cases, this can be easily done with the align_image_stack tool distributed with hugin. It just depends on whether the tool can find proper control points.

--Rik

acerola · Post by **acerola** » Sat Apr 12, 2008 1:10 pm

Is it aligning the frames also? I tested Tufuse with the aligned frames from CombinezM. Can it be better than CombinezM?
I think overlap is and will be a problem because in the same place there is high contrast pixels from two plane.

rjlittlefield · Post by **rjlittlefield** » Sat Apr 12, 2008 1:39 pm

acerola wrote:Is it aligning the frames also? I tested Tufuse with the aligned frames from CombinezM.

Perhaps I've overlooked a capability. How does one tell CombineZM to export aligned frames? When I export after aligning, what comes out looks like the un-aligned version.

Can it be better than CombinezM? I think overlap is and will be a problem because in the same place there is high contrast pixels from two plane.

All of the current tools allow background detail to "show through" foreground structure, if the background detail is equal or higher contrast than the foreground. That problem is still on the to-be-solved list.

--Rik

rjlittlefield · Post by **rjlittlefield** » Sat Apr 12, 2008 2:10 pm

Oh, I see now.

Export after alignment does seem to export the aligned versions, but the alignment function itself was broken in the February 3 version of CombineZM (and for some months before that). It would shift, but not scale.

That problem is repaired in the April 1 version. If you have an older version, I suggest to get the new one and repeat your tests.

One caveat: in the April 1 version, Do Stack has difficulties with some stacks. Large fuzzy areas appear in places that used to be handled OK. This seems to have appeared at the same time as Do Soft Stack and the three-argument form of the Find Detail command.

--Rik

acerola · Post by **acerola** » Sat Apr 12, 2008 3:12 pm

Thanks Rik, there was no scale in the alignment. I was suprised. I will try with the newer version too.

beetleman · Post by **beetleman** » Sat Apr 12, 2008 3:45 pm

Lots of really great detail in the picture Rik. You can look at it for a long time and still see things you missed the time before.

Ken Ramos · Post by **Ken Ramos** » Sat Apr 12, 2008 4:50 pm

A very nice image there Rik.

I can't say as I have ever seen any around here but there probably are.

Seems that when I looked them up, I found them to go by a number of other names, mostly associated with violets, of which I have a front lawn full, well at least until I mow the lawn.

rjlittlefield · Post by **rjlittlefield** » Sat Apr 12, 2008 8:59 pm

Doug, that's the beauty of biological subjects -- no matter how deep you go, there's always more interesting detail to see.

Ken, if you had one of these, I guarantee that after a short while you'd have a zillion of them. They spread prolifically by seed, and they're self-pollinating so one plant is all it takes, not even any bugs. As for the names, well, there's apparently N+1 for any N. I just learned another one a couple of minutes ago when I happened to Google on viola tricolor pollination and turned up the Wikipedia article that starts out

Heartsease (Viola tricolor) is a common European wild flower

"Heartsease", eh? Never heard that one before!

Thanks for the comments, guys.

--Rik