Focus Blending using Adobe Photoshop CS4

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Post by elf »

Here's a 100% crop of the lower 15 images:
Image

This shows the problem of hard edged masks with incorrect selection of in-focus areas. This was done with color blending unchecked.

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

We should be able to 'refine' the hard edged masks and instruct the software to use better segments where required as long as it doesn't mess with color blending or trip itself up in trying to be too clever. :smt079

Also, we need to supply the software with a 'deep' enough stack composed of 'fine' enough slices relative to the objective/lens/camera in use.

It should provide a good headstart to a manual composite. There is hope for it yet.

Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

The Adobe CS4 demo (30 day trial) is now available. 815MB

Just as a side note. Here is a quote from Graham Stabler. It was in another thread here and has relevance to my earlier question in regard to determining the focus state of a pixel.
You cannot tell if a single pixel in an image is in focus by looking only at that pixel, all you get is an intensity level, three for colour images. You must look at the variation of the image through that pixel. In focus parts of an image are generally much sharper, the spatial frequency is higher. That is one way to detect in focus information.
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

CS4 deserves respect!

For this shot I used the Olympus RMS Macro 38mm, wide open at f3.5.

One SB28 speedlight (@1/8th output), bare, 35cm from subject top-left; subject diffused with artist's acetate cylinder. Nikon D100.

This particular image was shot in 15 micron increments - there are some flat spots or banding, so I am intending to re-shoot at 10 microns - unless 5 is suggested? The image is sharper than the 200kb forum image indicates. I'll be more particular the next time 'round.

CS4 has some system related issues/requirements, most of which can be tickled into submission; but a 64bit OS with a truck load of RAM would be nice.

This particular stack consists of 22 image slices; basically just loaded into layers; aligned and blended. The stack could have been deeper to take into account some of the overlapping 'hairy' bits.

Papilio ulysses (it's for a project I'm involved in)

Full frame (D100 sensor), minus edges crop.

Image

Pixel crop
Image



Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

NikonUser
Posts: 2693
Joined: Thu Sep 04, 2008 2:03 am
Location: southern New Brunswick, Canada

Post by NikonUser »

It, CS4, and you have my respect! I'm still using CS2 so it's ppsibly a time to upgrade.

Very nice images. I've been following all the techniques discussions and was almost convinced that I needed some sophisticated fiber optics setup (currently using 1 flash). However, these images with 1 flash and a piece of plastic says to me that I don't need the fiber optics. Doesn't mean that "if you can do it" "I can do it" but it gives me hope.
NU.
student of entomology
Quote – Holmes on ‘Entomology’
” I suppose you are an entomologist ? “
” Not quite so ambitious as that, sir. I should like to put my eyes on the individual entitled to that name.
No man can be truly called an entomologist,
sir; the subject is too vast for any single human intelligence to grasp.”
Oliver Wendell Holmes, Sr
The Poet at the Breakfast Table.

Nikon camera, lenses and objectives
Olympus microscope and objectives

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

Here is a second version of the image in my previous post.

Spot the difference (I know it's not easy at this scale and resolution). One is the output from CS4 and the other from Helicon Focus - in that order.

32 images at 10 micron. I used a different diffuser this time and a small amount of fill flash from the right in addition to the SB28.

Playing with the iridescence on the wings and diffusers whilst attempting to keep the highlight blowout balanced - long way to go I know.

I'm running a 32Bit Vista System with 4GB of RAM, minus whatever the OS reserves. I've been receiving some ' cannot perform...out of RAM" error messages from CS4. I found a way around some of those hurdles. I expect those error messages would come much earlier in the process if I were working with larger DSLR files.

Image

Image

Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Post by elf »

The butterfly scales are impressive. I think CS4 can do a good job on relatively flat objects like this. It also seems to do a better alignment than CombineZM. An interesting test would be to align the images in CS4, crop to remove transparent areas, export layers as files, then stack in CombineZM and Helicon Focus.

My latest experiment was with a $3 Hong Kong postage stamp. It's a four frame pano and CS4 didn't have any trouble with the focus stacking or the pano stitch. I also didn't shoot the focus stacks with a small enough increment. Live View just isn't good enough to judge focus unless it's zoomed in.
Image

The eye is about 1mm, so this is was approximately 10x. At 250dpi this would print at 15"x40".

NikonUser
Posts: 2693
Joined: Thu Sep 04, 2008 2:03 am
Location: southern New Brunswick, Canada

Post by NikonUser »

augusthouse wrote: Spot the difference (I know it's not easy at this scale and resolution). One is the output from CS4 and the other from Helicon Focus - in that order.
Craig
HF image shows more fine detail; scales along the lower edge of that large vein and the scales adjacent to it lack definition in the CS4 image, much clearer in HF.
NU.
student of entomology
Quote – Holmes on ‘Entomology’
” I suppose you are an entomologist ? “
” Not quite so ambitious as that, sir. I should like to put my eyes on the individual entitled to that name.
No man can be truly called an entomologist,
sir; the subject is too vast for any single human intelligence to grasp.”
Oliver Wendell Holmes, Sr
The Poet at the Breakfast Table.

Nikon camera, lenses and objectives
Olympus microscope and objectives

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

elf wrote:
An interesting test would be to align the images in CS4, crop to remove transparent areas, export layers as files, then stack in CombineZM and Helicon Focus.


Will do elf and I'll post the results when completed.
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

NU wrote:
HF image shows more fine detail; scales along the lower edge of that large vein and the scales adjacent to it lack definition in the CS4 image, much clearer in HF.
Yep...What was interesting is that there was certain, fine detail in one particular area which both programs elected to ignore - identically. I'll go into more detail later.

CS4 interpreted the iridescence of the scales more acurately in this particular exercise.

Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

augusthouse wrote:CS4 interpreted the iridescence of the scales more acurately in this particular exercise.
Can you illustrate what it means, "interpreted the iridescence more accurately"?

--Rik

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

Yes Rik,

Surface texture definition.

CS4 applied less 'smoothing' to the overall surface texture of the iridescent scales.

Looking at the two images in my earlier post it is possible to see a difference.

Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Got it, thanks.

--Rik

augusthouse
Posts: 1195
Joined: Sat Sep 16, 2006 1:39 am
Location: New South Wales Australia

Post by augusthouse »

Here are two images.

Both images were stacked in Helicon Focus.

The second image was pre-aligned in CS4, edges cropped, layers exported as individual files and stacked in HF. I cannot see much difference with the full resolution files; but then I wouldn't expect to because the alignment in HF is built into its processing, so it would have re-re-aligned the files according to its own criteria.

I ran the files through CZP and bypassed the aligment proceedure, but the result was rather scary. Need to apply more time before uploading that result.

I was playing around with diffusers and lighting at the same time, and I expect you are tired of looking at blue scales; but that is what I am working with at the moment.


Image
Helicon Focus, Nikon CF M Plan MI 10/0.25 210/0

Image
Helicon Focus - pre-aligned in CS4, Nikon CF M Plan MI 10/0.25 210/0


Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Craig,

You can effectively bypass alignment in Helicon Focus by turning off all alignment options. On the Autoadjustment tab of the Preferences dialog, just un-check all of the Adjustment boxes. If you're using a version that doesn't have checkboxes, set the Maximum values to 0.

--Rik

Post Reply Previous topicNext topic