Do as I say, not as I do

Sometimes I don’t listen to my own advice. In my Focus Stacking Part 1 post I said that it was better to have too many images than too few and it’s true. But I got lazy when I found this gorgeous little mushroom on a log. Instead of doing a proper focus bracketing session, I manually chose different focus points in the scene. At the time the light was changing drastically because of wind in the canopy and I thought that if I timed things right I would have a better chance of blending the photos together.

Not really. After many months of more focus stacking experience, I’ve discovered that managing varying light is much easier than I thought, but missing critical focus is impossible to fix. To wit –

 

Not a good look. It doesn’t work at all. Ruined picture. It’s a stack of about 8 images that I thought covered the whole range when in fact it didn’t. I was so disappointed that I decided to see what I could do with fewer shots included in the stack, but none worked until I eliminated the stem images entirely plus the immediate foreground. Those were the areas with focus gaps and had to go. The result, while definitely shallower in depth, is more pleasing.

 

Lessons learned? Absolutely. Don’t be lazy. Do a bracketing session to make sure I get everything I want. Take the time! Slow down. It’s worth it in the end.

---

Another lesson is that variations in light, especially dappled sunlight on the forest floor, are easier to fix than I originally thought. Due to the fact that it can’t be avoided sometimes, and that I also really like its effect in photos, I have done bracketing where each image has a slightly different light pattern. When stacked and retouched, it’s often not a problem and can usually be finessed to even things out or paint in that bit of perfect light exactly where I want it. Here’s an example where light changed a lot and it doesn’t look weird –

 

I’d always thought it would be a deal-breaker, but it isn’t, so if the lighting conditions are variable and you can’t mitigate that with a diffuser, flash or LED panel, don’t sweat it too much. It’s easier to deal with that than missing critical areas of focus that absolutely can’t be fixed and ruin the photo.

One way to manage changing light in your scene is to work with the Contrast Threshold slider in Zerene Stacker.

It’s junk basically – not essential for the stack since out of focus parts don’t contain any information you need – no sharp bits. Sharpness is largely determined by contrast, so excluding low contrast areas from the calculation makes for smoother backgrounds and if you bring the black bits in very tight to your main subject you will reduce or eliminate annoying haloes. In my experience, even some sharp areas that are covered by the black mask render properly and with a good amount of detail. Even if they are a bit soft, you can use a PMax stack as the source image for retouching.

Try different settings until you get the cleanest DMap image you can. Most likely it will still need work, but try for one that needs the least amount. Most of the time I will use a PMax image to fill in fine detail. In general PMax produces a more contrasty result with more distinct haloes, but often manages overlapping details much better than DMap. It’s just the nature of the way each type is calculated, but the retouching tool in Zerene makes it easy to have the best of both worlds.

In the past I started with PMax images and retouched those with original photos to eliminate noise and haloes. Very time consuming and fiddly because you have to match the sharp detail in the PMax to the original image it came from in order to get the background to paint on without ruining the sharp part. The more depth in the shot, the more original images to sort through and it takes a lot of time. It’s MUCH easier to start with the clean background and rich color found in a DMap and fill in details or fix muddled overlaps with a PMax as your source. That you can use output images as sources for retouching is one of the things that makes the process worth the time. The key is to create a target image that needs the least work.

Oh and if you have the horribly named ProSumer edition of Zerene, the various brush types that are labeled Pro in the dropdown menu, do work as advertised. The standard brush is the best in my opinion because it automatically calculates luminosity and color when applying the source pixels to the target image. You’ll get better matching and smoothness by leaving it at the default Details. At least that was my experience when I tried the others to fix some haloes and other odd looking areas where there was too much contrast.

So that’s an update on things I’ve learned since I started focus blending with Zerene Stacker. I’ve also discovered Slabbing, but don’t really have to use it much so far. It’s a method of automatically creating Sub-stacks to reduce the number of source images you have to sort through when retouching. It takes a little more time on the front end, especially if you edit your Sub-stacks in Lightroom and then send them back to Zerene, but in the end it will save you lots of retouching time. Instead of having to sort through dozens of source images to find the part you want to paint on your target image, you now have only a handful. When I go over 20 shots for a stack, I use it, creating PMax Sub-stacks or Slabs. For one job I went from 57 source images to 6. These are output images and need to be re-imported as source images, but by sending them into Lightroom for this process, I could denoise and do other things like adjust color and contrast so they would look their best to be stacked into the final shot.

It sounds crazy, but if you do really big stacks, it’s something that sets Zerene Stacker apart from Helicon Focus. Check out some of Allan Walls’ videos about it; he’s the real pro and gives a lot of great information seasoned with very dry humor. Just search for Slabbing. Great stuff.