I'm trying to decide which version to order. I've already checked whether CS3 Extended supports the HDV format used by videocams such as the Canon HV20 which record on mini DV tapes. It doesn't.

The diagnostic and analysis tools are not really needed for my purposes, which are to get the most out of photographic still images, so I'm favouring CS3 standard.

However, it's not clear to me whether or not there are improvements in CS3 Extended with regard to features and processes that are common to both programs.

For example, both programs boast the feature of auto-alignment of similar images.

Has Extended extended this feature or improved it in some way? Is there any useful feature in CS3 Standard that might appeal to photographers of still images, which has been improved in CS3 Extended?

I recently returned CS3 and bought CS3 extended because of the noise reduction capabilities that can be done with blending layers in stacks
http://www.photoshopforphotographers.com/p...d/sample-04.pdf
Marc

As posted above, the only real advantage is stacks. It's kind of neat and interesting, however, not sufficiently so to justify on its own the difference in price between CS3 and CS3E.

CS3 covers off 100% of what you are likely to need - provide you don't suffer top-of-the-range-itis (a case of must have the best so as not to feel inferior or that you are missing something).

If you want to get a feel for the differences between CS3 and CS3E go through Russell Brown's video tutorials on his website. This will give you a feel for the difference in capabilities between the two packages.

Russell Brown Tutorials

Thanks for the advice. I guess I'm a bit disadvantaged in downloading video tutorials with my slow 56k connection. However, the Martin Evening pdf provided a good demonstration of a feature which I wish I'd known was coming when I was last in Cambodia trying to photograph ruins with tourists wandering around.

The noise reduction feature of stacking seems very worthwhile. I get the impression this process is analagous to making multiple scans when scanning slides. I can see there could be situations when a slow shutter speed for merging to HDR purposes might not be the best approach.

Has anyone done an evaluation of the degree of noise reduction possible, in terms of ISO equivalence? For example, would 6 rapid shots at ISO 3200 with a Canon 5D, when merged using the Mean stacking process, have as little noise as a single shot at ISO 1600 or maybe even ISO 800?

Ray - why are you making work for yourself? If instead of taking 6 shots at ISO3200 and spending an age merging the images why don't you drop the ISO and just take the one image?

You would be better off using the median rather than mean to reduce noise in the image for multiple exposures. This would have the effect of minimising the impact of any 'shot' type noise influencing the noise reduction process. I would suggest that the maths is not as straightforward as you would hope as neither median filtering or the noise/ISO curves are linear.

If you want better performance you could always perform median filter on each layer in the stack and then median filtering of the stack itself. This would give you 3-D noise reduction.

David,
That's what I'd do when shutter speed is not an issue. I'm thinking here of shots at night or in very poor light without flash or tripod, where reasonable DoF is desired.

The example in the pdf link from Marc demonstrates 5 such hand-held shots being stacked applying median filtering, resulting in a noticeable reduction in noise which looks to me the equivalent of at least one stop lower ISO. But Martin Evening also makes the comment that mean stacking would produce even more noise reduction and would be the choice when stacking still life images.

I think I'm going to opt for CS3E. There seems to be some price rigging going on, though. The Aussie dollar is now worth around US$0.85. I don't know how the upgrade price of US$340 for CS3E on Abobe's US online store translates to A$619 on their Australian site. We have a modest 10% GST over here which could not account for such a huge mark-up.

Having done a bit of experimenting, taking indoor shots with my 5D in the evening, of smooth surfaces that show noise, I'd say that stacking in mean mode is equivalent to a 2 stop ISO advantage.

In other words, 5 hand-held shots at f4, 1/50th and ISO 3200, after stacking and auto-aligning, produce an image with the same degree of noise as a single shot at 1/13th and ISO 800.

Considering that ISO 3200 on the 5D is equivalent to ISO 1600 underexposed one stop, I find this result impressive.

Within the shutter speed limits I consider adequate for a particular shot, I can now stop down 2 stops for greater DoF, without getting the increased noise I would expect without stacking.