Out of the shadows....

Every so often I find a 'new' feature in Photoshop. Which is odd - since the feature in question usually isn't new and I've been using the program rather heavily for the better part of 10 years.

The most recent discovery is the Shadow/Highlights adjustment. It became a part of Photoshop in CS(1) - as of writing CS5 is almost due - and is also available in the more recent Elements versions.

While there are many ways to adjust pictures values in this way - brightness/contrast, levels, dodge/burn, exposure - Shadow/Highlights is both easier to use successfully and more dramatic in effect.

Quite simply, it takes the extreme areas of an image (without having to pre-select them) and drags them back to a normalized value. Which is probably best shown with pictures.

Unaltered image:

Shadows 50%

Highlights 50%

Highlights 50% + Shadows 50%

The only slightly confusing bit is that you're reducing the appearance of the said regions - so as shown above 50% Shadows makes the image brighter overall and 100% Shadows is even less 'dark'.

There's also an "Show More Options" version of the toolbar (at least in the full Photoshop) that will let you tweak things even more. Which probably wouldn't be a bad idea - as nice as the defaults are at bringing out 'hidden' detail - they tend to make the pictures more than a little 'flat'.

Original image inside red circles, 50% Shadow outside - Note how the water doesn't change much

More dramatic example - Shadow 100% on the right side of the image

How to Buy a Digital Camera - Part II

Card Type
Most cameras today take SD / SDHC cards. As a result, you have the most choice in card size, manufacturer and the likelihood of multicard deals increases. The HC addition (under the letters SD in the graphic) apply to cards 4GB and higher - these cards will likely not work in pre-2007(ish) cameras.

While they are the most common type - there are definitely exceptions:

Older Sony cameras take Memory Pro Stick Duo cards - same type as the PSP uses. They are more expensive than their SD counterparts and can also be harder to find. Newer Sony cameras take both SD-type and Memory Pro Stick Duo cards - still worth double checking though.

Older Olympus/Fuji cameras take the xD type card which - along with being more rare than either of the cards mentioned above - are also limited to 2GB. To get around this, some Olympus cameras come with an adaptor that takes a MicroSD card and puts it in a xD 'housing'. Newer Olympus/Fuji cameras are using SD though.

Some SLR cameras use the deceptively named Compact Flash card (it's actually the largest in physical size). Again rarity is an issue, though price is generally comparable to a similar capacity SD card. It can also have a speed edge on the SD type cards - but this only a factor if the camera itself isn't the bottleneck.


All of these cards have a rather long lifespan (10k-100k write cycles) and even the smallest commonly available (2GB) hold over 500 (jpg) pictures. As such, it might be tempting to dismiss the above information, get a few cards with the camera and be done with it. Of course, cards do get lost and not all external/internal card readers can accept every type and shape of storage. While memory cards are coming down in price, it's nice not to have to buy a new set of cards when/if you change/upgrade/replace cameras. Go with SDHC compatible.

Features
A lot of devices today go quite a bit beyond simply taking a picture. Often these features have a varying degree of sophistication from model to model. A few of the things you might want to look out for:

Image Stabilization
It is what it sounds like. As covered in a previous post, the feature really helps with indoor shots. Be careful to read how the camera is accomplishing this though - if it mentions 'digital' don't bother.

This is a very common feature of most new point and shoot cameras. In some SLR cameras it is lens based and, as it absence reduces the price by a large margin, is still fairly common to find them without it.

Face Detection
At it's most basic, it determines if there is a person looking at the camera and focuses on them (as opposed to the background). It may also adjust the exposure/white balance and determine if a flash is necessary.

Smile Detection
A rather cool, if somewhat creepy, feature of a camera to detect if someone is smiling - often to a variable threshold - and automatically take the picture.

Blink Detection
Still somewhat more rare than the above, this detects if someone has their eyes closed and either notifies you, or automatically takes a second picture.

Creative Effects
Black and White, Sepia, Fish Eye, Miniature Effect, Selective Colour, Red Eye Reduction and many, many others may be present on an increasing number of devices. Personally I prefer applying these effects after with software as it allows for greater flexibility - and undoing it if it doesn't work. On the other hand it does make picture taking more 'fun' and less workflow if it's on-camera. I wouldn't recommend getting a specific product for any of these features - but if you do have them, at least try them out.

Photomerge

Fortunately, not everything in Photoshop is labour intensive. Some of the coolest stuff is automated.

While some newer cameras have image stitch (panorama mode or another name depending on brand), they are somewhat limited in what they do.

Essentially these lock the cameras setting and let you 'weld' a few pictures together. It can also show you a bit of the previous image to help guide you on how much overlap you need. Also, it doesn't alter the perspective and the fixed camera settings aren't always helpful.

This is where Photoshop (full or elements) comes in. Open the pictures you want to merge and tell it to go to work.

I'm told that not everyone likes the music to this. Tough. Turn it off if you want :p



A couple of things to keep in mind:

1) It will only alter the perspective to about 120º - after that it does a cylindrical mapping - in plain English, straight lines become curved.

2) You *might* want to scale down your pictures first. Most cameras output a very large file and you're asking the computer to load them all into memory, find where they join, match perspective and do brightness/contrast calculations. That's a lot of work, and if it can't do it all, it usually skips the last things in that list or possibly not even match it at all. You can still manually align the pictures and have it more likely to merge that way, but it's best to let the computer do it. The combined file will still be huge and more detail than you need - especially if you do the trick in #4

3)You still need a bit of overlap. The Photoshop help file recommends 25-40%. Having a notable object at the extreme right of a picture, then extreme left on next shot is usually sufficient.

4) Unlike cameras, you aren't limited to horizontal or vertical stitching.