Attached Link: ARTICLE

This is useful Kort.  I did not know ISO effects Depth of field.  Just learn something

 It doesnt - if you read it it says that the shots were all taken at 1/250th and the aperture changed to match the increase in ISO... the aperture changed the DOF...

Nice link tho kort, explains the relationships nicely  (although the DOF - ISO relationship could have been better explained to be fair)

I remember reading an article which had a graph of noise to signal ratio (ie clarity of exposure) mapped against ISO, the graph started at 100, went up at 125 then down at 160 then up again at 200... So i use 160 as a default as i get 2/3rd or a stop extra light for only a fraction of the extra noise that 200 would cause.  This was only for a 30D tho - i'll try and find the article when i have a bit more time...

Here ya go Rich:
for the Canon 30D;
http://www.imaging-resource.com/PRODS/E30D/E30DIMATEST.HTM

By the way, Imaging Resource has Imatest results for most cams....

You always must have the needed light to get a balance of ISO and DOF and Stop Action.   So, there is a place for higher ISO, as we all know.  Each balance of ISO, shutter and aperture is a balance to do what we want of the scene.

What do you do when it is not sunny sixteen sorts of daylight?   What you can is what you do.

Thanks for the link...and for the comments thusfar.

Thanks for the link Kort..

I for one am not afraid of higher ISO

especially with band and billiards..

But I try to stick to as low as possible..

Hmm this is fairly well explained in laymans terms, and perhaps more relevant for film users than digital.. He does accept the advance of the latest digitals re noise, and yet then compares one of the older poor digital performers re noise, in the use of higher shutter speeds.

I would  also point out that there seems to be mounting evidence that the use of the in between ISO settings ie 125 or 160 perform worse than 100 or 200 and is even more  evident the higher up you go.. The faster the lens the more expensive and thus better noise performance.. 

The latest offerings from Canon and esp Nikon are remarkable and make the purchase of better optics less of a concern.  

My new camera has auto ISO from 100 - 1600 so for most of the time I now don't worry I just let it do it's own thing.  I'm sure it knows what it's doing better than I do.

Fred, it might not be as smart as you think...it knows what it is doing but that might not give the best image.    

Running a series of test shots as done in the article might help any of us to know better what is showing up in the file.  I have done this before using various variables of exposure and results are generally predictable but sometimes reveal peculiar characteristics of that camera used.

I agree Tom, from my own experience, I've found in some situations it works better to under expose..other times..better to over expose..

For example, I find better color saturation (when shooting sunsets) if I slightly under expose the image.

Me too. I shoot -1/3 or -2/3 EV for the majority of my images.
Only when trying to capture detail in a dark area of an otherwise bright scene to I go above 0 EV.

You've got some great shots David, but try experimenting with ISO. If you have bright conditions, try keeping it as low as possible.

David - you upgraded to a 50D and put it on what is technically an auto setting? almost blasphemy in my books. Set the camera free with full manual (/ISO 100)!!

Quote - Here ya go Rich:
for the Canon 30D;
http://www.imaging-resource.com/PRODS/E30D/E30DIMATEST.HTM

By the way, Imaging Resource has Imatest results for most cams....

Very interesting to see all those sets of data, thanks for the link!

The ISO vs noise/signal ratio graph only maps full stop jumps in ISO and not the intervening settings. This is the graph (but not the article) i was talking about after a quick search on google so you can see what i mean (admittedly it is only for blackframe noise, but considering most problematic noise is down that end anyway...) :

http://forums.dpreview.com/forums/read.asp?forum=1019&message=19721647