Focus stacking workflow: Lightroom and Zerene

[DanK]

There have been a number of threads recently in which people asked about focus stacking. I thought this example might be useful.

First, my basic stacking workflow:
1. import the stack into lightroom
2. change WB, exposure if needed.
3. Examine at 100% to see which images in the set are needed to get the desired DOF
4. export as 16-bit TIFFs, still in proPhoto RGB
5. Stack in Zerene, retouch if needed (more on this below)
6. Import the stacked image back into LR
7. Edit in LR

I rarely need to use a pixel editor for these, but if I do, I move the image to photoshop.

A few days ago, I created this image, which is a stack of 20 images. I wanted the entire open flower in focus, and I thought I should keep a few of the star-shaped unopened ones in focus too. I used Zerene's retouching in a few spots where the stacked image had halos from parallax. You pick the image from the stack that is "correct" for each area and paint from it to the composite.

Here it is:



I didn't like it. The detail outside of the main flower is distracting. I wanted more blur outside the main flower. However, I looked at the images, and I found that I needed almost as many just to get the farthest point in the flower in focus. I could only shave off one or two, so the stack would still have the background too much in focus.

So, I stacked again (I still had the TIFFs), dropping one or two from the back of the stack. Then I used Zerene's retouching tool to paint the entire background from the top-most image, to get it more blurred.

Now I had the basic image I wanted, but I did not have the edits. Because LR is a parametric editor, there is no harm in redoing any adjustments at any time. You can return to any adjustment and adjust again. As a result, the history was 29 steps long!

However, LR makes this trivial. I just selected the first image and synched all settings to the new image. It took just a couple of mouse clicks and a few seconds.

Here is the result:



It's all a matter of taste. I like this better. Iowever, that's not why I posted. I thought this is a good example of why Lightroom and Zerene make a very powerful combination for focus stacking.

The original images were ISO 200, f/9, with a 100mm macro lens. At f/9, I probably could have made do with fewer images, but I always try to err on the side of too many, because too few means starting all over.

[Cantab]

Dan, thanks for posting this. One of my projects during the next couple of months is to try my hand at stacking.

[Glenn NK]

Dan, thanks for posting this. One of my projects during the next couple of months is to try my hand at stacking.

Bruce:

Don't put it off - you'll be surprised at how easy it is.

Dan:

Very good tutorial. I also use LR adjustment brush to blur the BG stuff.

Glenn

[Glenn NK]

"export as 16-bit TIFFs, still in proPhoto RGB"

Dan:

I forgot to ask - why proPhoto?

Another question - what setting do you most often use for Contrast in Zerene for the DMap?

Glenn

[DanK]

Glenn,

I stay in prophoto (actually, Melissa RGB) for two reasons: (1) it's LR's native space, and I am taking the images back into LR; and (2) my general procedure is never to discard information from an image until I have to.

Re the contrast slider: Zerene's author has put up a new tutorial for DMAP here. I just found it in response to your question, and it suggests that I have been using it incorrectly. In an earlier tutorial, he pointed out that you can increase the constrast to eliminate some types of artifacts, and I had been using it that way, keeping the contrast as low as I could. In the new one, he points out that you can use it for the same purpose that we have been discussing: you can maintain out-of-focus areas by increasing the contrast until those are blacked out. One more reason to use Zerene.

Dan

[Glenn NK]

Dan:

Well that was and interesting lesson (DMap) and many thanks for the replies on colour space and Zerene. BTW, what is Melissa (too many questions for a Saturday morning - it's still am here - we're three hours behind you)?

I've always struggled to get anything useful from the DMap process no matter what the contrast setting was, and have essentially resigned to using the PMax output (I probably should just to the one stack).

I shot an eleven image stack yesterday of a yellow dandelion-like flower (about the size of a small dandelion), and both methods produced junk, especially the DMap. Mostly I use PMax because it has less artifacts.

After reading your post and the link, I tried the same images again using both 15 and zero settings for contrast - the zero version had slightly fewer artifacts in the small petals although the difference was so small that I didn't notice until I switched back and forth several times.

BUT - the DMap image was essentially not usable - well neither was very good - the PMax was the least bad.

When I increased the contrast on the DMap, it seemed to "attack" the flower details first and had little effect on the BG until I really cranked it up. (the BG had very little detail - shot at f/2.8 - admittedly not enough DOF to get overlap for a stack). I'm using the pre-L version of the Canon 100 mm f/2.8 and I would say that achieving very small increments of focus distance is extremely difficult. It's been suggest to use LV, but even at 5:1 zoom-in, the flower is so much larger than the LCD view that I'd have to scroll all over the place to know where the focus was. Someone mentioned somewhere about a software/computer driven gizmo that does the focusing automatically in preset increments but it sounds very expensive.

I have a focusing rail but it seems to be a bit flimsy so I haven't used it (bought it at a consignment sale).

I'm afraid that I still have a lot to learn, particularly in the field - must try smaller apertures, say f/5.6 or f/8, and deal with the BG in Lightroom.

And I'm going to study the tutorial more to see if that helps.

Thanks for the help.

Glenn

[DanK]

Glenn,

Melissa is the LR variant do ProPhoto.

Re the two methods: PMAX supposedly preserves more detail, although I haven't noticed this difference. I have noticed several differences in flower shots. DMAP does a better job of preserving colors and textures, and it is a bit less vulnerable to false specular highlights. And it has the advantage I just discovered of using the contrast control to preserve blur. However, it has two disadvantages: it is slower, and it seems more vulnerable to haloing from parallax. At any event, these are my impressions, I now almost always use DMap for flowers.

Re focusing: I doubt the issue is your lens. I have both the 100L and the EF-S 60. I assume the latter is probably similar to yours in BQ. I switch between them on the basis of FL only and have not noticed a difference in focusing. However, I rarely shoot wider than f/8 when stacking flower images. One can bring back blur, but there is nothing you can do if your stock has too big a focusing gap.

For the most part, I only have obvious failures in stacking when I have bad parallax issues--usually, deep images with edges that are a far (front to back) from the surface behind them. Most do my failures are the raw materials I create, not the stacking as such.

Re rails: I have never used one. Some day, I may try closer than 2:1, but for what I do, I haven't seen the need, and the really good racks are expensive.

Dan

[Cantab]

Dan and Glenn, I think my question has been addressed elsewhere but between Helicon and Zerene, which do you think is better at least for someone just starting in focus stacking?

I have demo versions of both of them on my hard drive but haven't actually done anything with them.

My initial photo subjects will almost certainly be flowers and I'll use either the Canon 50mm or 100mm macro lenses.

[DanK]

Unfortunately, indont have any information. I've never used Helicon. Most of the macro photographers I know use Zerene, and I haven't been tempted to switch.

[Glenn NK]

Unfortunately, indont have any information. I've never used Helicon. Most of the macro photographers I know use Zerene, and I haven't been tempted to switch.

Bruce:

When I started, I used CombineZM, which has been superseded by CombineZP.

This is a free stacking program (donations are welcome) by Alan Hadley from Britain:

http://www.hadleyweb.pwp.blueyonder....stallation.htm

I used the CombineZM version for a few years before buying Zerene.

I think the new version is improved, but frankly, Zerene is more user friendly - a plus for me at least.

I did do a brief test of Helicon, but when I asked a question about something, it took several days for a response, and the reply was curt and pretty well useless.

When I asked a question during my trial of Zerene, the author replied very quickly with a meaningful response.

That turned the tide for me.

It's my impression that there are more users of Zerene than of Helicon, and that could possibly mean something too.

Then factor in this; Dan taught me something, and all the other Zerene users are potential sources of information - the large user base of Zerene is useful.

Putting it all together, this would be enough information for me to make a choice.

Glenn

[Cantab]

Glenn and Dan, thank you for your thoughts. I'll give Zerene a whirl.

How forgiving is the software of flower movement in a slight breeze? Presumably it's best wherever possible to shoot the images in a controlled environment -- such as indoors?

[Glenn NK]

Glenn and Dan, thank you for your thoughts. I'll give Zerene a whirl.

How forgiving is the software of flower movement in a slight breeze? Presumably it's best wherever possible to shoot the images in a controlled environment -- such as indoors?

The truth - wind is my nemesis. It's the only real problem.

In effect, wind movement creates parallax shift. If two objects (such as small petals) are in alignment in one image and the wind rotates or moves the flower so that they aren't aligned in the next one this would be same as moving the camera sideways - it causes a parallax shift.

I doubt that any software can handle this.

This article on Wikipedia about parallax is quite good, particularly the image on the top right:

http://en.wikipedia.org/wiki/Parallax

As an aside, if one is doing a panorama with stitched images, parallax shift can be problematic especially with an object that is close to the camera. With a tilt/shift lens, shifting the lens one way and moving the camera the same amount in the other direction eliminates parallax errors.

Glenn

[DanK]

Bruce,

How forgiving is the software of flower movement in a slight breeze?

Not very. Zerene does automatically align the images, but there is a limit to how much it can do. I do the majority of my stacked images, and all of the ones that entail stacking more than a few images, indoors. It gives a wind-less environment and more control over lighting. However, unless you have a very solid floor, floor motion is enough to cause problems. I often use some variation on the setup I'll post below, either holding the flower in the "plamp" (the plastic clamp), which is very prone to vibration, or putting it in a vase on the pile of miscellaneous stuff in the middle. I use a remote release, and mirror lockup, because with this amount of lighting (125 W of halogen), I need slow shutter speeds. I take a shot, move the focus slightly, move away, stand until the vibration has stopped, and take another.

If there is virtually no wind, I will sometimes take shots outdoors that require stacking a few images. this allows more interesting backgrounds. I'll post one below. If you look closely, you will find a few areas with slight blurring. This is because of motion for which the software did not fully compensate.

I don't know if you get much in the way of coastal onshore breezes, but we do. It's very pleasant, but a real pain for macro photography.

Dan

[Glenn NK]

Started looking at the tutorials on the Zerene site:

http://zerenesystems.com/cms/stacker...tutorialsindex

I alluded to a computer method in post #6, end of sixth paragraph; it's covered in the "More tutorials designed to be read" section. Sounds interesting.

Glenn

PS: I typed this last night, but then sidetracked myself checking out the ML information (I'll have to update my firmware first).

Dan: Neat setup; mine looks totally different (except for the Plamp which I use). My setup has grass and other vegetation all around; the sun is my light source (often with a hand held diffuser). In my setup, wind is almost always present hence my experience with parallax.