Photo Stacking Part 2

[John 2]

I have tried to find a test subject more in keeping with Dan K's recommendation and I have taken the test one step further. I have put the same image set through both the Franzis software and through Photoshop CC and compared the output. The results are below. The first image is the first exposure of the set for comparison purposes. The two that follow are the focus stacks. I should add that the lack of focus at the top of the wire loop results solely from the inability of the lens plus extension tube combination to cover that end of the range. All other differences however are down to the relative performance of the two software packages.

1. Original first exposure.


2. Stack Processed in Photoshop.


3. Stack Processed in Frazis Focus Projects 3


To my eye:

• The Photoshop Stack is too warm but nearer the actual scene as lit. The Frazis version too cool but unlike PS, their software does have quite a powerful built in adjustment tool to correct this by allowing you to specify the type of light source in use.
• The PS version is has definitely redered the image more accurately. It is sharper throughout (see the far end of the circuit board) and doesn't have the ghosting displayed by the Frazis version on the green wire lettering (extreme right).

I would have to say that if you have PS CC, it is the best option. If not, based on the results of the model car, the Franzis software will certainly do a reasonable job but might struggle in processing more demanding subjects.

Hope this is of interest.

[Mike Buckley]

This is really informative, John. I have absolutely no experience doing focus stacking, so much so that I didn't even know Photoshop CC provides that capability. I would be interested in learning from Dan and the other experienced users of focus stacking to learn how Photoshop CC's capabilities compare with the standalone software applications most commonly used.

[DanK]

John,

Very clear results. Thanks for doing the test and posting. This still doesn't fully get at the haloing issue, but it seems to me that it is enough to clearly show that Frazis isn't nearly as good as CC.

The color difference looks like a color space issue to me. The Frazis one looks a lot like ProPhoto not converted to sRGB for web display. I see this routinely with Zerene. I do all of my editing in ProPhoto, so when I export TIFFs from LR to Zerene, I leave them as 16 bit Prophoto TIFFs. Zerene, unlike photshop or LR, doesn't re-map to display on a screen, so they look weird When I am done and import the composite into LR, they look fine.

Mike, I have only used Photoshop for stacking once. It worked fine that one time. I don't use it because it is slower and less flexible than Zerene. Among other things, Zerene has two different stacking modes; it offers excellent retouching capabilities that allow you to paint either from one image in the stack or from the composite created with the other method; and for one of its methods (DMap, which is most similar to photoshop's), it offers considerable control over several parameters that can improve the quality of the stack. For example, I usually use DMap for flowers because it does better with color and some textures. However, it captures very fine detail less well than the other method, PMax, and more important for me, DMap is more prone to haloing from parallax. So if I have a deep flower with edges positioned such that haloing is a problem, here is what my workflow would be:

--examine the images in LR to delete any that are focused farther back than I need.
--adjust WB on a shot with a whiBal in LR, and sync the white balance to the entire stack.
--export the entire stack directly into Zerene from LR using Zerene's plug in.
--stack with DMap, adjusting contrast to avoid artifacts
--stack with PMax
--examine the DMap composite carefully for halos, and when I see them, use the retouching tool to paint from the PMax composite for only those small areas.
--save the retouched DMap composite and open it in LR to continue editing. Move from there to photoshop if necessary.

Dan

[Mike Buckley]

Thanks for all of that, Dan. I have added your post to the only other post I have bookmarked regarding focus stacking. After seeing your focus-stacked photos, it should come as no surprise that the other post is also yours.

[chauncey]

FWIW...I have an advantage over most of you guys in that I shoot Canon. It comes with a Canon
Utility Software which allows one to shoot live view tethered. Changing focus points is one of the
options available, which makes stacking a no-brainer...then it's off to PS CC.

There can be artifacts present when more than 10 images are used...easily correctable though.

[John 2]

Glad the test was of some interest. As Dan says, the disadvantage with CC is that what you get is what you get but of course if you have CC then PP is an option. My work flow was to convert the stack to 16bit TIFFs before processing it. The test wasn't a fair comparison wrt colour temp and colour space. Franzis has a prep screen which frankly, I didn't bother with i.e.:



There are also various brushes and adjustments available later in the work flow but the fact remains that the basic result for anything a bit more testing, seems not to be up t scratch.

William, Canon is not alone in that. Fuji provide a free App. which allows you to connect the camera to a smart phone or in my case, an 11" Tablet and you can control all camera functions with the touch screen. I used that in this test but I have also used the joystick focus point adjustment on the camera and a remote to perform the same function and I think of the two, the latter is perhaps the more accurate if you have a complicated subject; fat fingers being the problem.