You are welcome.

There is a great collection of videos on lighting techniques by Joe McNally offered by Adorama TV, dubbed "Photo on the Go with Joe McNally".

Joe McNally is known for his elaborate use of many portable flashes for on location shoots. I find that he does not stray too often from the one technique he does so well (cold ambient light with warm flash accents) but watching at least a couple of his videos is very instructive indeed.

I cannot quite believe that he personally is into the Nikon CLS system to the extent he promotes it in these videos and his books and suspect that he is playing an ambassador role here. The things he does with it are indeed possible, but it seems to me that using radio triggers and manual flash settings would be quicker, more reliable, and more flexible.

Having said that, Joe McNally is a great photographer, very good teacher, and funny as well. I like that he stayed humble despite his many achievements and provides you with a very honest, down to earth perspective on photography.

Please see a respective discussion started by falconeye.

In a nutshell: The periodic sensor-based blurring movement is of course not instantaneous but takes time; it operates at 500 Hz. Hence, towards faster shutter speeds, a given region on the sensor will not be exposed to the full blur movement anymore but only to some fraction of it. Which phase of the periodic blur movement is "seen" by a pixel then depends on the pixel's position on the sensor. Some phases of the blur movement are more effective in terms of the desired AA-smearing effect than others (e.g., a full horizontal movement is better than half of it going forward and then backward, hence not covering as many sensels). As a result, the smearing effect will not be the same across the image, but will vary across it (in the vertical shutter travel direction).

In practice, this should not lead to any major issues, but I personally prefer the application of deconvolution to a uniformly blurred image. The images are sharp enough for me this way.

If you oversample then you avoid the aliasing in the first place, i.e., it is a hardware solution, not a posteriori software solution.

Your CD player analogy is not applicable because CD players use digital oversampling to simplify the analogue filter design of the output stage. The correct analogy are oversampling A/D converters which do exist but even those require (at least weak) filters to avoid aliasing unless you know for sure that the input frequency won't surpass half of the converter's Nyquist frequency.

Yep, that's the problem with having the option to use the filter. Unless you essentially use it all the time, you may get nasty surprises when reviewing the shots on a bigger screen. The option to use it essentially is the option to "not use it" in situations where you can make sure directly that you don't get aliasing artefacts.

It is furthermore not easy to anticipate moiré. It typically does not appear when you expect it and vice versa. Maybe that's what the new Bayer-AA-filter-less cameras are bringing to the table. They are teaching us to become moiré aware.

It is impossible to accurately deal with it in software.

You can mask/replace colour artifacts and/or moiré but you cannot systematically "undo" these artefacts.

The term "aliasing" refers to the fact that spatial frequencies that are too high for the sensor to record properly are still recorded, but incorrectly as lower frequencies.

The only proper way of dealing with a bandwidth limited sensor is to prevent it from "seeing" frequencies it cannot record. That's why manufacturers have been implementing Bayer AA filters in the past, despite the fact that they
a) add to the cost of manufacturing, and
b) slightly blur the image.

Note, however, that undoing the Bayer-AA-filter blur is possible in a systematic manner by a process called deconvolution. And regular capture sharpening is a good enough approximation.

As clever as the K-3's Anti-AA-simulator is, I deplore the fact that there is no conventional alternative with a Bayer-AA-filter. What if you get artefacts at 1/2000s? The Anti-AA-simulator won't help at this shutter speed.

Also, its effect varies across the image with shutter speeds higher than the sync-speed, i.e., it is not a uniform blur that is applied. It will be interesting to see how well it suppresses artefacts and responds to sharpening near its limits.

I like the trick he used with the axe shot. And the bird shot is also impressive.

I've just learned that the retail price will be US$ 89.95.

Sounds like a good idea at first, but a single device solution would pose the following challenges:

1. All batteries would have to fit into one device. That would be eight AAA in the case of the LV5. The latter is not power hungry at all (and can be run with four AAA in total only), but the higher amount of batteries is useful for multi-day, unattended use.
2. The alignment of the mirror would be very critical. You'd have to have very precise control and it would be hard to hit the device with the beam from a distance.
3. The maximum distance covered by the beam would be halved.

If they don't thieve your camera as well.

Has anyone tried using laser triggers?

For instance, the TriggerTrap?

I have just finished writing up my experience with the new Cactus LV5 and wonder whether others have done similar shots and how.

P.S.: If the blue K-30 user I met at Lake Tekapo reads this, please say "hello".

P.S.S: The K-30 was blue, not the user.