How reliable are online lens comparison tools

sploo wrote in post #17646135
Agreeing with your post, but of course it's worth noting that sample variation may well hinder "I always try to borrow one and take a few pictures myself".

You're right, it's hard to judge on one lens alone which is why you have to look at all the various sources. Look at the theoretical numbers, forum reputation, sample images, and review sites then make your best guess. There are a few lenses you can't go wrong with though.

sploo wrote in post #17646130
Come on man, try harder

Well, I have made some 8"x10" prints (at Walmart) that came from Fuji 800 film. They look good to me.

gjl711 wrote in post #17646160
You're right, it's hard to judge on one lens alone which is why you have to look at all the various sources. Look at the theoretical numbers, forum reputation, sample images, and review sites then make your best guess. There are a few lenses you can't go wrong with though.

Funnily enough; Roger's blog has a timely article on lens measurements (http://www.lensrentals​.com …ents-for-telephoto-lenses) and includes data on sample variation. Take home message seems to be: buy the Canon 100L Macro - it's great, and it's consistently great

GeoKras1989 wrote in post #17646185
Well, I have made some 8"x10" prints (at Walmart) that came from Fuji 800 film. They look good to me.

That's a great example of something I think many people (myself included) loose sight of sometimes; a great 8x10" print is a very realistic size. Pixel peeping to 100% on a large monitor could quite easily be the equivalent of a 60" wide print, which is a wholly unrealistic standard to judge a lens for the vast majority of shooters (professional landscapers aside).

I never shot that much print film (though I definitely used some ISO 800 in my time). Mostly it was Fuji Velvia. Ahhh. ISO 50

sploo wrote in post #17646364
I never shot that much print film (though I definitely used some ISO 800 in my time). Mostly it was Fuji Velvia. Ahhh. ISO 50

Me I really liked Kodachrome 25, but you did really need good light for it.

Alan

GeoKras1989 wrote in post #17645917
When perusing review sites, please remember that any reviewer, for a huge range of reasons, can post shots worse than a lens is capable of. It is not possible to post shots better than a lens is capable of.

This is why I defend (well-executed) lens tests. A good resolution/MTF test, which takes sample variation into account, will show you the best a lens is capable of, and how likely you are to get an acceptable copy. Sample shots only show what one person with one sample can do, and additionally they may have gotten a really sharp copy - or a dud. If you only rely on tests of someone else's sample, you may be disappointed when your copy arrives. I'm lucky to have a local store with a good return policy, even though I haven't needed it yet. I always test a lens after buying it, and if it seems to be close to the typical online samples, I will run it through a sample shoot to confirm that I get good results. That way, I at least have something to compare and use as a basis for keeping or rejecting the lens.

Yea, I prefer to see what people are saying about the lens, preferably folks who also have lenses to compare - which is why I also like that people post their gear.

I've seen a bunch of folks say how 'tack sharp' the 24-105 is. Well, you look at what lenses they are comparing it to and what photos they are taking with it and draw your own conclusions.

If you look at the 55-250 (I've never used it), it has a good reputation from both beginners and people with experience of much more expensive lenses. Doesn't mean it is going to have the IQ of a $3000+ lens, but you can bet that whatever any review site says about it, it is going to be good value for money.

If you then buy the lens and it is really not sharp compared to say some other 'decent' lenses, then perhaps you look for another copy. Folks who just way 'well it's not as sharp as I expected based on the review sites', I ignore. IMO it's almost impossible to make that kind of comparison unless you have exactly the same setup as the review site. Much easier to compare with a lens you own under controlled conditions.

For instance, when I bought my 135L, I tested it at f4 and f5.6 against my 70-200 f4 IS expecting it to be sharper than that. Then I compare f4 with f2.8 and f2 and see a gradual softening (which I expect) and not sudden jumps in IQ. If the corners are all equally sharp (or soft) wide open, then it passes and I call it good.

BTW, I do regularly make 20"- 30" prints of my shots, so I can tell the difference between say a 18-55 and a 16-35 f4L. I would say on even an 8x10 print most modern cameras and lenses are good enough that with a bit of extra processing/sharpening it could be hard to tell the difference. Of course there are many other aspects to a lens than just sharpness anyway.