Godox V1 For Focus Stacking

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Macro_Cosmos
Posts: 1511
Joined: Mon Jan 15, 2018 9:23 pm
Contact:

Godox V1 For Focus Stacking

Post by Macro_Cosmos »

I recently picked up a Godox (God-ox) V1 and some accessories from the same company during China's capitalistic 6-18 Taobao promotions. There's about 3-5 of these a year. Spent about $80 on shipping, so all the money saved went there instead.
Image

I purchased the unit mainly because I was lacking a portable flash and setting up lights in the studio for the food photos I love so much can get annoying due to restricted space. To put the light to test, I decided to run a stack.

Here's an image sequence of the stack, illustrating the flashlight's output variations:
Fast (6 seconds): https://www.youtube.com/watch?v=ittapvLIyIo
Slow (2 minutes): https://www.youtube.com/watch?v=vh_WEI4s2Fo

I took the first exposure as the reference (0) and attempted to equalise the rest using sliders in Capture One Pro. My sanity stopped me from processing the entire stack this way. I only processed 16 images, should be a pretty good sample from the 140-image-stack. The reason why I didn't use the data generated by Zerene Stacker are:
1. No idea how the software attempts to equalise brightness
2. I have that option unchecked
3. For this specific butterfly, ZS generates weird results and produce large amounts of fogging.
The variation is within +/- 0.1, but certainly visible when a sequence is displayed. Not the worst I've seen from Chinese flashlights but far from the best. Either way, it posed no negative impact on my final stack. I can recommend that V1 for focus stacking myself.
Image

Specs: 1/4 flash power, Mit 20x NA/.42, ITL-200 tube lens, Dino810 camera, 1/200s shutter speed. 1um/step using stackshot.

The result:
Image

Constant lighting:
Image

Lighting angle accounts for colour differences.


Lou Jost
Posts: 5948
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Very nice. I'm curious, what happens if you do check "Brightness" in Zerene and run the stack? It would be interesting to see the weird results. That might tell us something interesting.

Macro_Cosmos
Posts: 1511
Joined: Mon Jan 15, 2018 9:23 pm
Contact:

Post by Macro_Cosmos »

Lou Jost wrote:Very nice. I'm curious, what happens if you do check "Brightness" in Zerene and run the stack? It would be interesting to see the weird results. That might tell us something interesting.
This is what happens:
Image
Image

Brightness unticked:
Image

This issue doesn't occur with any other moth/butterfly scales as far as I am aware, blue ones are just fine. Only green would somehow confuse the program and make it go wild.

Lou Jost
Posts: 5948
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Robert OToole has stacked that same butterfly species in one of his tests; I wonder how he did it? Unfortunately I don't remember which of his tests.

zzffnn
Posts: 1896
Joined: Thu May 22, 2014 1:25 pm
Location: Houston, Texas, USA
Contact:

Post by zzffnn »

Have you tried stacking a few sub-slices (for example, 20 sub-slices) in PMax, then DMap those sub-slices together?

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Macro_Cosmos wrote:The reason why I didn't use the data generated by Zerene Stacker are:
1. No idea how the software attempts to equalise brightness
It adjusts the scale and gamma of each image so as to make the mean and variance of the luminance channel be the same as that the first image in the list. In other words it's actually trying to equalize brightness and contrast simultaneously.

With stacks where most of the contrast is in fine details and the first frame is mostly OOF and hence low contrast, this strategy results in incorrectly reducing the contrast of all other frames too. The result is an image that looks fogged.
This issue doesn't occur with any other moth/butterfly scales as far as I am aware, blue ones are just fine. Only green would somehow confuse the program and make it go wild.
Hhmm...

The algorithm calculates all of its adjustments based only on the luminance channel. It doesn't care about hue.

So, I'm thinking that observed differences between blue versus green must be due to some correlation with other properties of the stacks that were tested.

--Rik

Macro_Cosmos
Posts: 1511
Joined: Mon Jan 15, 2018 9:23 pm
Contact:

Post by Macro_Cosmos »

Lou Jost wrote:Robert OToole has stacked that same butterfly species in one of his tests; I wonder how he did it? Unfortunately I don't remember which of his tests.
Not sure about that, maybe he had brightness unchecked or something. I use constant lighting, so that option is always unchecked.
zzffnn wrote:Have you tried stacking a few sub-slices (for example, 20 sub-slices) in PMax, then DMap those sub-slices together?
Sounds like an interesting method, never tried that. I'll give it a run to see what happens.

rjlittlefield wrote: It adjusts the scale and gamma of each image so as to make the mean and variance of the luminance channel be the same as that the first image in the list. In other words it's actually trying to equalize brightness and contrast simultaneously.

With stacks where most of the contrast is in fine details and the first frame is mostly OOF and hence low contrast, this strategy results in incorrectly reducing the contrast of all other frames too. The result is an image that looks fogged.
Interesting, the first frame of my stacks are always mostly OOF smudge, with just one band of details. Since the software is agnostic towards the hue, maybe there's some kind of correlation going on with these specific green papilio butterflies. I'll run some more stacks as an attempt to pinpoint the main cause.

In addition to that, if I'm correct, the output of ZS doesn't tell us definitively the amount of output variation in flashlights, since both contrast and luminance are matched, right? Luminance can change throughout the stack due to lighting angles, which means even theoretically stable constant lighting would have some sort of falloff throughout the stack due to the properties of the specimen.

To get around this, blasting the flash on some kind of flat plane, ie a piece of paper, then letting ZS run the stack should produce more reliable results? I do think manually adjusting the exposure to equalise the histograms is a pretty solid method as well, but it takes time.

Also with regards to the fogging, if I change the first (reference?) frame to something that has detail, the issue should be solved? Only testing will show, I'll get myself busy.

Lou Jost
Posts: 5948
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Those Papilios are unusual in that the green scales are surrounded by black, while the typical blue Morphos are more solid blue. Out of focus, I suspect the two would look very different, even apart from their color.

Macro_Cosmos
Posts: 1511
Joined: Mon Jan 15, 2018 9:23 pm
Contact:

Post by Macro_Cosmos »

I placed an exposure with a clearly defined focus band in the middle as the first image and ran a Pmax stack, the problem with fogging is gone.

Macro_Cosmos
Posts: 1511
Joined: Mon Jan 15, 2018 9:23 pm
Contact:

Post by Macro_Cosmos »

Lou Jost wrote:Those Papilios are unusual in that the green scales are surrounded by black, while the typical blue Morphos are more solid blue. Out of focus, I suspect the two would look very different, even apart from their color.
Besides colour, they do look quite similar.

Image
Image

Some have both colours:
Image
Only the green ones with the first frame as mostly out of focus smudge will cause ZS to produce weirdness.

I can even manipulate the hue and pass it off as a blue papilio, I'm certain most won't be able to tell.

Image

Lou Jost
Posts: 5948
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

I placed an exposure with a clearly defined focus band in the middle as the first image and ran a Pmax stack, the problem with fogging is gone.
Good job solving this mystery! Then the whole stack could be done by running "Add files" and perhaps re-ordering....

I think the green papilio does have more black than the second shot, the blue morpho. But it is weird that the third specimen doesn't have the same problems in stacking as the first specimen.

So if you changed the green to blue in the first shot, does the fogging go away?

Macro_Cosmos
Posts: 1511
Joined: Mon Jan 15, 2018 9:23 pm
Contact:

Post by Macro_Cosmos »

Lou Jost wrote:
I placed an exposure with a clearly defined focus band in the middle as the first image and ran a Pmax stack, the problem with fogging is gone.
Good job solving this mystery! Then the whole stack could be done by running "Add files" and perhaps re-ordering....

I think the green papilio does have more black than the second shot, the blue morpho. But it is weird that the third specimen doesn't have the same problems in stacking as the first specimen.

So if you changed the green to blue in the first shot, does the fogging go away?
Haven't tried that, but I suspect it will go away. I deleted all the tif and raw files though, I do have the batch of jpegs which I generated the sequence with, I'll give that a shot.

The blue butterfly isn't a morpho, it's in the papilionidae family. Morphos are different, belonging to the "Morpho" family. The "blue" that we commonly see are:
“Those iridescent colours are related to photonic band gaps structures that
are like gratings where you reflect back some light depending on the angle
which matches the periodic underlying structure inherent in them. The blue
ones are interesting because they are single colour at different angles and
that suggests more of a omnidirectional structure.”
The butterfly my supervisor was referring to is:
Image
(Morpho Achilles Lima)
I'm personally done with mophos, no matter how I try, the results are not satisfactory. Each attempt drains 2000+ actuations and many hours. Before I acquire better skills, I'm gonna leave it there. They are also extremely fragile, oh and some are also expensive.

The blue papilio would be something like this:
Image
(Papilio Ulysses)

Macro_Cosmos
Posts: 1511
Joined: Mon Jan 15, 2018 9:23 pm
Contact:

Post by Macro_Cosmos »

Also to add to the morphos, the only "satisfying" one I've found online is from here:
https://petapixel.com/2014/03/22/beauti ... fly-wings/

Image
(Exclude the oversharpening)
https://www.flickr.com/photos/13084997@ ... 467961844/
A technique apparently called "epi-cross-polarisation" was used. Those with an adequate polarisation setup could give it a shot.

No offence to anyone else, these blue morphos, the Morpho Didius in particular is IMO the most difficult to get right. All the others online including my own aren't close to satisfying.

The hardest one to photograph would be butterflies of the Riodinidae family, but that's more of an optical issue, dealing with interference waves is hard.

Lou Jost
Posts: 5948
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Morphos are native to my area, and I sometimes have them in my yard. I have set out to photograph all the native species of the ecological reserves we own, and I'll post the learning process here. Some of them are crazy, like Morpho sulkowskyi, semi-transparent. I have specimens of all of them now, and I really like them. Each species-group of Morphos has a distinctive micro-structure.

Post Reply Previous topicNext topic