I have been doing some more experiments with my FZ330, creating stacked images from post-focus videos using Helicon Focus. I haven't done any more with flowers, berries etc, because it is already clear to me that this technique has great potential with those subjects (see here for example) and the usability seems fine, for both capture and processing.
For the past week or so I have been working mainly with smaller subjects. Insects etc are difficult to find at the moment but I did find two that I could repeatedly photograph over the course of several days. This let me compare different setups, settings and techniques. One was a bee, which had died and I brought indoors. The other was a wasp, which I thought must be dead because it stayed, unmoving, on a stone for several days. However, when the weather warmed up a bit it started moving slowly, but still on the same stone, so I could carry on photographing it.
During the course of these experiments I have also photographed some man-made objects, for example coins and batteries. I use these to test flash setups because they are very prone to having nasty highlights and that lets me check the effectiveness (or not!) of my diffusion arrangements. It was instructive to see how stacking coped with these tricky subjects.
Indoors and outdoors the light levels were problematic. It was heavily overcast most of the time. Post-focus requires at least 1/30 sec shutter speed, and that is too slow anyway for small subjects as the magnification magnifies hand-shake. (I am primarily interested in working hand-held.)
You can't use flash with post-focus so I bought an LED light that fits on the camera hot shoe. I had to make a snoot for it so as to concentrate the light on the subject, and even then for darker subjects I could only just get to 1/30 sec even when using f/2.8 with ISO 100.
Raising the ISO turned out to be not a good idea as fine detail becomes a formless sludge in the stacked images as the ISO goes up. I bought an LED light that was twice as powerful and that gave me about a one stop improvement in shutter speed, but the shutter speeds were still rather slow.
However, I was surprised at how usable slower shutter speeds were. I suppose that is because the software picks out what is sharp enough to use and ignores the rest, and with 20 to 150 or so images to pick from, some sharp and some not, it may quite often have enough information to produce tolerably decent results.
As the magnification increases and the scene size decreases camera movement becomes increasingly problematic. Even with the LED light shutter speeds were often very low, so at higher magnifications some sort of stabilisation for the camera became necessary. Because of where the wasp was I could kneel down and rest the camera on my thighs. Indoors I had the fly on my desk and for most of the shots I rested my hands on the desk, but for a few I used a tripod. For a few of the wasp shots I used my steadying stick, which is intermediate in both flexibility/speed of use and effectiveness as between unsupported hand-holding and using a tripod.
Working with the wasp suggested that subject movement is going to be a much bigger problem than slow shutter speeds – or it will be when I get more live subjects. The problem isn't that the subject is for example on a leaf that is blowing around in the wind. The software can, up to a point, cope with that, as it did sometimes with flowers, when the flowers moved as a whole rather than relative to one another. And insects etc move around, but that obviously isn't the time to be attempting stacking. The problem was that even when the wasp appeared to be still it sometimes wasn't. When it was really cold the wasp was completely still, and the stacking worked fine. But when the weather warmed up just a little, even though the wasp appeared to be motionless its antennae were moving slightly and its abdomen was expanding and contracting. (I only noticed this when I captured a video of the wasp).
Helicon Focus has retouching facilities and one of the Helicon videos makes it seem that it is easy to correct things like moving antennae. I suspect their example was rather carefully chosen, using an insect against a plain background, and choosing one with the antennae parallel to the plane of the sensor and captured with a single image, in which case you can simply choose which of the antenna positions you want to use and paint it into the stacked image in one, simple move. In my experience insect images are not normally like that and correcting a stack can be a much more involved process, and sometimes impossible. And worse, a subject doesn't have to be moving for these complications to arise.
When a subject is walking around you have to capture single shots; stacking is impossible. To maximise image quality and cropping potential of single shots I use base ISO. To maximise the DoF I use a small aperture. To get this to work, with a small aperture and base ISO, with a fast enough effective shutter speed to avoid blur from subject movement and/or camera movement, I need either bright ambient light or flash.
For a stack I need better than very dull ambient light. It doesn't have to be as bright as for single shots because I can use f/2.8 rather than f/8. Alternatively, for a stack, I can use an LED light.
These approaches of flash and LED light don't fit together well. It is time-consuming to switch between flash and LED light, and it would be extremely awkward to carry around both; one could be on the camera, but how to carry the other one? – they are awkward pieces of kit.
However, if the ambient light is strong enough to do post-focus without needing to use the LED light, the flash can stay on the camera and using camera setting presets I can quickly switch to post-focus when a subject is still and back to flash when it starts moving again. I was doing this with the wasp on the warmer day, switching between flash and natural light post-focus depending on whether the wasp was moving around or not.
One of the disadvantages of post-focus is that it produces JPEG images rather than RAW. This means that DXO Optics Pro Prime noise reduction can't be used. Also, since the easiest approach (when using Helicon Focus) is to feed the MP4 video straight into Helicon Focus, it isn't practical to apply any DXO corrections or enhancements to the base images. It might be possible to apply DXO corrections to the stacked image (although Prime couldn't be used), but I didn't get round to testing this as I found that using Silkypix and Lightroom on the stacked images, or often Silkypix by itself, produced results that I didn't feel the need to use DXO on.
The following posts go into some of this stuff in more detail and/or provide some examples to look at.
Part 2 - LED lights, snoot and steadying stick
Part 3 - Helicon focus stacking methods
Part 4 - Controlling the focus boundaries
Part 5 - Retouching methods
Part 6 -The retouching boundary problem
Part 7 - Batteries, coins and breakfast cereal
Part 8 - Wasp
Part 9 - Dead bee
Part 10 - Conclusions
