Focus stacking

[Mica]

This is my first try at focus stacking using the Nikon 105mm f/2.8 Micro.

[KeithH]

A cracking image Mica, perhaps crop off the white bit on the left hand side because it catches the eye (mine) too easily.

[Shadowman]

Nice effort.

[Mica]

Thank you for the suggestion. I have completed another try with better cropping technique. See attached.

[Geoff F]

That one looks fine.

[DanK]

Good start. Both look free of obvious stacking artifacts. My one suggestion is more images in the stack. You don't have all of the anthers in focus in the second one, for example. In my experience, out-of-focus areas in the foreground are usually more bothersome to viewers than OOF areas further back in the flower.

How did you stack these? The exif information looks like it is from one image. My stacked images lose all of that.

[William W]

Both are very good first attempts.

I agree with the comment made by Keith H regarding the Cropping of Image 1. A SQUARE Crop would look good I think.

And I also agree with Dan's comment about OoF areas in the Foreground, being more distracting than OoF areas in the Background

***

Matters that you might consider for improving Focus Stacking technique:

1. Shutter speed. I'd be interested to know the Shutter Speeds that you used. Interrogation reveals 1/640s and 1/50s respectively, but I seek confirmation that data is correct. The (large) file of the second image appears on my studio monitor to exhibit SUBJECT MOVEMENT (esp. top petal 11 o'clock).

2. Number of exposures. I'd be interested to know the number of exposures you applied, especially for the first image. All the petals are not complete in focus and artistically it would be nicer for them to be so (esp. petals bottom 8'oclock and 5 o'clock).

3. Subject Distance (or Working Distance). I'd be interested to know the distance from Subject to Film Plane (the letter phi ϕ on your camera) or the Subject to Front of Lens Distance. Approximates are OK. The reason being that creating a longer distance can leverage having fewer stacks.

4. Aperture Used. Would you please confirm F/8 was used for both shots, that information goes to comments 1,2, and 3 above & also will promote the discussion/consideration of the value of using a smaller aperture.

WW

[William W]

And please answer Dan's question also . . . "How did you stack these? The exif information looks like it is from one image. My stacked images lose all of that."

WW


PS

haha!
Dan, you beat me this time!

[DanK]

haha!
Dan, you beat me this time!

Doesn't happen often

[Mica]

Doesn't happen often

The aperture was correct at f/8, 1/250, 200 ASA. The stacking was carried out by varying the camera to subject distance rather than focusing ring, using a Novoflex Castel rail, starting at the closest distance and moving to the back of the subject. A total of 8 images were made at fixed intervals and processed using Helicon Focus.

[Mica]

And please answer Dan's question also . . . "How did you stack these? The exif information looks like it is from one image. My stacked images lose all of that."

WW


PS

haha!
Dan, you beat me this time!

You are absolutely correct: the information is from a single image from the stack which was inadvertently uploaded instead of the 8-shot composite. Here is how it was originally assembled using Helicon Focus. I apologize.

[DanK]

Thanks. This makes sense. This composite has fairly severe stacking artifacts, unfortunately. They are apparent if you follow the edges of the petals. They are of two types. One is an echo of the red edge. The other, which is more typical, is a simple blur. you can see this, for example, between the red echo and the petal in the lower left.

Artifacts can arise for a number of reasons, but the most difficult to handle stem from parallax, when an edge is quite far, front to back, from the surface behind it. That may be what is happening here. This happens regardless of whether you change focus on the lens or use a rail.

I have never used Helicon for stacking--I use Zerene--so I don't know what options you have. I believe Helicon has two stacking methods, and if so, you might compare stacks using both. In the case of Zerene, one method is better for preserving color, while a second produces fewer and less severe halos. You might be able to repair some of this in Helicon after creating the composite. In Zerene, you can paint from any image in the stack onto the composite, and you can also paint from a composite done with one method onto a composite done with the other. Where the front-to-back distance is large enough, there are no individual images that can solve this problem--paint for the back, and you ruin the front, and vice versa--so painting from a second composite is often the best option. It isn't always possible to clean it up entirely, at least with my level of skill. The other option is to try to touch up the composite directly using tools in your main editor, like the clone and healing tools in photoshop.

This doesn't always happen to a noticeable degree, at least with Zerene, but it happens often enough to be a problem. I often blow up the composite while still working in Zerene and methodically go over the entire image looking for artifacts so that I can use tools in that software before moving the image out to my regular software.

With enough work, one can usually get rid of most of it.

[Mica]

Thanks. This makes sense. This composite has fairly severe stacking artifacts, unfortunately. They are apparent if you follow the edges of the petals. They are of two types. One is an echo of the red edge. The other, which is more typical, is a simple blur. you can see this, for example, between the red echo and the petal in the lower left.

Artifacts can arise for a number of reasons, but the most difficult to handle stem from parallax, when an edge is quite far, front to back, from the surface behind it. That may be what is happening here. This happens regardless of whether you change focus on the lens or use a rail.

I have never used Helicon for stacking--I use Zerene--so I don't know what options you have. I believe Helicon has two stacking methods, and if so, you might compare stacks using both. In the case of Zerene, one method is better for preserving color, while a second produces fewer and less severe halos. You might be able to repair some of this in Helicon after creating the composite. In Zerene, you can paint from any image in the stack onto the composite, and you can also paint from a composite done with one method onto a composite done with the other. Where the front-to-back distance is large enough, there are no individual images that can solve this problem--paint for the back, and you ruin the front, and vice versa--so painting from a second composite is often the best option. It isn't always possible to clean it up entirely, at least with my level of skill. The other option is to try to touch up the composite directly using tools in your main editor, like the clone and healing tools in photoshop.

This doesn't always happen to a noticeable degree, at least with Zerene, but it happens often enough to be a problem. I often blow up the composite while still working in Zerene and methodically go over the entire image looking for artifacts so that I can use tools in that software before moving the image out to my regular software.

With enough work, one can usually get rid of most of it.

Thanks for your input. I just started working with Helicon and as I moved up from Photoshop CS, the workflow is much improved. Helicon also provides copying from the original to the composite and vice versa but I have not used yet. I was trying to see how Helicon would manage the merging of the frames without much post manipulation and touch up. I would look into it. I will also run comparisons between adjusting focus using the rail and the focus ring; based on what I know, however, parallax errors using these two methods are almost inevitable.

[DanK]

Thanks for your input. I just started working with Helicon and as I moved up from Photoshop CS, the workflow is much improved. Helicon also provides copying from the original to the composite and vice versa but I have not used yet. I was trying to see how Helicon would manage the merging of the frames without much post manipulation and touch up. I would look into it. I will also run comparisons between adjusting focus using the rail and the focus ring; based on what I know, however, parallax errors using these two methods are almost inevitable.

yes, parallax problems are often unavoidable, and you can't avoid them by opting for adjusting focus with the lens or with a rail. There was a thread discussing this latter point some time ago, but I haven't searched for it. However, stacking algorithms differ in their ability to handle this, both in terms of how bad the problem has to be before it affects the composite and how severe the resulting artifacts are. Using Zerene, a fairly large proportion of my images have no visible artifacts at all, although the proportion--and the severity--differ between the two algorithms it offers. I'll be interested in your findings, as I don't know Helicon.