In this sequence, I've normalized the images so that the amount of information in each is the same. This compares the same (to the extent that the seagulls cooperated) subject using the four different settings for capturing the data.

In other words, say I wanted to create a 16:1 shot by increasing the image size in Photoshop after capturing it at 10:1. How does Photoshop's resampling compare to just grabbing the data via the teleconverter?

Similarly, how does Photoshop's resampling compare to the resampling done by the camera?

These questions need to be answered in two different scenarios: publishing to the web; publishing to print. For this first pass through, let's look at the images assuming we're publishing to the web. Indeed, the images you're looking at have been published to the web to illustrate this discussion.

I confess to being somewhat surprised by the results. But let me not influence your analysis. Take a close look at these four shots and I'll offer my conclusions in the summary afterwards.