affinity photo Focus stacking - 160 and 25 frames

[unni]

My first test on focus stack. The speed of processing was impressive since the working space was set to prophoto-16 bit. It took only a few minutes to complete
25 images alignment and stacking. To compare, one single frame and the final stacked image are enclosed. Only minimal editing  (curves and USM) is done on the stack. The output
is very good. Soon I will be testing with 70 to 160 images taken with microscope objectives. 

Camera- Nikon D3100, Lens Tamron 90mm Macro, f18, 1/3 sec, ISO 200.

This is the final stack.

The image below is one of the 25 frames.

[John Rostron]

Very impressive. The individual image shows a fair depth of field. I suspect more than is necessary for a 25-frame stack. A stop of F/18 could well produce some burring due to diffraction, which would be much less with F/8 or F/5.6 and you would still be able to get a good focus stack with 25 frames.

John

[unni]

Thanks,john. yes,the diff limit would be in around f8. This is an earlier stack of 2015. I now keep the f between 8 and 13 for this lens after examining test shots for optimal f. The overlap of focus is kept around 50 to 20% while shooting and then I eliminate in-between images from stack before stacking if needed. The work flow is mostly that is followed for images taken when using microscope objectives where the dof is a few microns. in this case, alternate images could be skipped but some of the near horizontal elements will show blotches. From this result I expect very good results for stacks of 150. I have to understand how AP will respond to such high stack numbers due to buffer load when using 16 bit and 8 bit space, processing time etc. In case of issues, I plan to bring in Rawtherapee for batch processing and ImageJ for aligning before using AP for final processing of aligned stack. Maybe tonight I get time to try out !

[unni]

I am adding another image made with AP Focus stacking . @John Rostron

Subject: A very small moth with total head size less than 1mm.
Stack: 160 images
Lens: Microscope objective-PLAN Achromat-10X Magnification
Camera: Nikon D3100
Field of view : 2.5mm x 1.5mm

Software: Only Affinity Photo used.

Remarks: The performance and rendering of AP is very good. The halo and general quality is related to original image. Use of an APO objective will make the image much better. No special care has been taken during shooting . The shooting setup is completely homebrew and not commercial macro rails which has electronic controls and stepper motors.

Note: To view full size, please click on the image. When a + sign cursor appears, click again to open the image in a new window. In the new window, press once again. The image will go full size and cursor will become - . Pressing again will bring back to "fit in window size". Its a toggle.

 

For giving an idea on the DOF, one image of the stack is below.

[firstdefence]

Never ceases to amaze me the complexity of nature at such small dimensions.

[unni]

very true! the symmetry, proportions and perfection is all there, howmuch ever we zoom in. 

[John Rostron]

Very impressive stack. I no longer have access to the microscopes I used to have when gainfully employed. I would enjoy creating such composites.

The sample image you showed has no sharp areas as far as I can see. The bristles on the antenna are just nearing coming into focus. Where did this come in the stack?

John

[unni]

Hi john,

The sample image is the first one of the stack. Sample image uploaded is as-is, opened and saved as jpg using fastone image viewer. So it looks dull and smeared. The camera moves forward by approx 8 microns per image. I have used a small milling machine table on which the camera is placed. On the handle of the table, a large disk of 10 inches dia is attached. The disk is pasted with a graduated print out from 0 to 360. Due to the large dia, rotating the disk through each division is easy. One division rotation will give me 8.5 microns. This depends on the pitch of the table worm gear. The table is about 5 kgs and approx 6 inch square. I just keep it on the dining table or somewhere else when taking the photos. A small scissor bench lab jack is used to keep the specimen. Fotidox extension tube combination is used to get the correct distance between the objective and camera flange as I do not use the infinity objective. The setup is so solid that there is no room for even the slightest vibration. One or two flash is used to light up.

[cosack]

Very cool, thanks for sharing!  How did you deal with the wind movement?

[BnBGobo99]

"Wow, that is amazing" were the words that escaped my mouth from the results of the second image.  I have some scopes and macro filters around the house, I want to go try this!

[unni]

23 hours ago, cosack said:

Very cool, thanks for sharing!  How did you deal with the wind movement?

The specimen was not a live one. It was very small, something like a rice grain. Shot inside the house, so no wind issues. Holding the specimen and getting the right view in the camera is time consuming. Some form of manipulator arm will help. Using small needles, super glue etc may be required. Moving the camera in very small step without any axial or rotational error is very important. Linear bearing tables of 4 inch sizes with verniers can be used. They are relatively expensive if new. So I went for a milling machine table which is very robust but bulky and very heavy. Flash with diffuser in manual mode is a must to freeze any motion induced blur from shutter, etc. Around 1/32 power is sufficient as flash will be very close to subject.

[Roger C]

You've gone to great lengths to capture an astonishing image, especially when viewed full size.

Maybe you could edit your post to remind people to click on the posted image to view it as a (much sharper) pop-up, then wait until the mouse cursor turns to a plus symbol before clicking again to open it larger in a new window, then click on that image to finally view your eye-popping full size photo.

It's certainly worth it!

[unni]

Thank you @Roger C for the appreciation. I have added a comment in blue, as suggested in your above post.

I never knew that it is possible to view 1:1 in the posts !

[Roger C]

The size of the 1:1 image we can see obviously depends on the pixel dimensions of the image you choose to upload. You'll notice a variation of this around the forum depending on how much detail people are happy to share.

You have generously posted a wonderful hi-res image for us to enjoy. The detail on the coils of the proboscis is amazing.

R

Looking again at your photo,  don't you think it also looks like a parrot wearing super-bling designer shades?

(No disrespect to you intended).

[unni]

@Roger C : The original sensor resolution of Nikon D3100 is 4608x3072 pixels. To keep the upload file size low, I reduced it to 1800 on the longer side.
So what is on forum is about 40% of original. The proboscis has come out well. There is a light golden feature inside the dark brown boundary of the  coil. Probably the hollow section of the coil. I think the smallest recognizable feature is about 2 or 3 microns.

[unni]

The front view stacked from 150 frames. The image is not cropped and uploaded at downsized longer side 3200 pixels instead of the original 4608 pixels.

It took 12 minutes for the "Align merge sources" to complete and 38 minutes for the next step "Focus merge". Total 50 minutes for completing the stack.

 

[Roger C]

Superb. Your best effort yet.

The image, the lighting: exquisite.

This iconic image would be at home in National Geographic magazine. Yes really.

Seriously well done. (And that's not taking into account the technical limitations of your equipment).

Bravo!

(If I wasn't such a nice guy I'd be feeling jealous.)

[unni]

Many thanks @Roger C !

The above moth images were taken 3.5 years back. I was trying various software to make the composite. Combine ZP
was the only free one available. Did not want to go for commercial ones as I do photography as a hobby and experimentation.
Then I came across ImageJ which is free and used for medical work. I now forgot the workflow with those different softwares !
ImageJ needed some readings and trials because documentation is different. Affinity Photo output is excellent and fast. None
of the other options processed RAW files. Had to convert to 8 bit TIF before inputting to the align and stack software. 16 bit TIF
would work in ImageJ but takes time . Dont know how AP works under the hood but from RAW to final edit which includes aligning,
merging and editing is a great feature of AP. I will be posting this image in photomacrography.net  mainly to give exposure
Affinity Photo software .

 

[Roger C]

That'll make a great advert for Affinity. They should pay you!

[TheEarnestBunbury]

It's too easy to forget just how amazing the world is - work like this is a wonderful reminder. I'm blown away and really don't have the words to express my feelings, so I'll have to settle for Beautiful, and Thank You!

[unni]

Many thanks for the appreciation and nice words, @TheEarnestBunbury. I had put macro work with scope objectives at the back seat due to long
processing methods. Nowthat AP is available, I can do more macro work.

[BrianHermelijn]

Would be interesting to see what kind of results you would get with microscopic view of an ant, and stacking together. Because right now
there is some stunning results happening with your bees. So many details coming together, which turns into something 
wondrous at the same time. Beautiful work @unni

[unni]

Thank you , @BrianHermelijn.

I browsed my repository to check for ants and there was one small stack. This is a small common red ant seen around the house.
It could be about 2 or 3mm long and head may be 1mm or less. The angle of the shot is not proper. This was taken sometime in early 2015.
Same 10X objective and 30 shots in the stack.