I am having a discussion with someone on another forum who says that really, at web size, the sample threads here don't sow much quality difference other than dof (and of course photog skill) because at web size, a pixel is a pixel and you really can't tell say, a real difference in these threads between the Canon kit 18-55 and say, the 24-70 L II, or the efs 55-250 and the 70-200 L IS F2.8 II
I'll quote his words below, and if he is right, and I fear he may be, then all these sample threads aren't really shwoing the qualty of these leses because the web limits them to being the same other than dof differences and photog skill. Which means many like myself using these sample threads to help dice on purchases are simply seeing what they want to see or under some kind of suggestion or something. I can admit if I have been seeing something in these threads that isn't there, I just want to know the truth of the situation. Here are his words. These words were specifically about the 85 1.2 L sample thread
"Sorry, but what is so amazing about any of those images and what does it prove about a lens?
They are all, at best, average. Images that size could have been taken with an iPhone, apart from the DOF, and they would still look OK.
Show me a 36X24" print of an open landscape that is sharp to the edges and free of CA and distortion, and you may be on to something. You may convince me that it is a good lens. But you can't prove anything from a small sized web image.
Or are you of the opinion that an image is only good when most of it is out of focus. In which case how come that is not a major feature of most of the worlds greatest photographs?"
Then, after I responded and said I could swear I see a difference in quality between that lens and the kit lens, he said this:
"At that size almost anything looks good. Even the 18-55 if someone actually does some decent PP.
A web size image is usually at least a 3X (linear) reduction form native size. A lens would have to be pretty bad before you noticed any difference in resolution, and everything else except bokeh could be down to PP.
Moreover, there are lenses which are a quarter of the price which are just as good in most instance. The 85 F1.8 is sharper in the borders and corners and only slightly less sharp in the centre. Certainly a better landscape lens. But how would you know if you didn't read a test?
But I can assure you that any image which is compressed by a factor of 2X or more is going to hide the majority of 'issues' that may have been visible at 100%. You can't see anything smaller than a pixel on your screen, and you screen resolution doesn't change.
So if I compress 4 pixels into 1 (2X size reduction) the blur radius of the 'poor' lens would have had to halve the overall resolution before the effect would be obvious. By the time you have sharpened it and processed it the difference will be even smaller.
If you are referring to the images on that website, then you are also comparing photographers, cameras, scene type, ISO and editing as well as the lens. It is also a very small sample set.
To be frank, if you can see the problems in a lens at that size, then it is a coke bottle or the photographer is the problem. In the case of these samples, the 18-55 images are mostly moving subjects taken at a distance, hand-held, sometimes in poor light. The are also over-sharpened.
The shots taken with the 24-70 are mostly static, taken by a more experienced photographer, often with lighting or on a tripod. They have also been well processed. All this proves is that 24-70 lenses tend to be used by people with better cameras and more experience."
It sounds as though he is right and this makes the sample threads lose a lot of impact for me. Is what he is saying true, and I only think I am seeing real, quality differences in these sample threads when I am only seeing dof, photog skills?
f2.8
