Interesting lens comparison

thecornflake

thecornflake

Messages
186
Name
Simon
Edit My Images
No
I know this has been done a lot of times before, but I do find these things interesting, and haven't actually done one myself before.

So without going into the reasons why, I currently own a Canon EOS 1100d and the following -

Canon EF 50mm f1.8 II prime lens - bought for around £70 new.
Canon EF 35-135 f4-5.6 (original old lens with metal mount etc, not the newer one with IS) - bought 2nd hand for around £40
Canon EF 24-105 f4 L IS USM - currently on hire for a couple of days but can be found for around £460 new

I decided to run a very quick test to compare the 3 lenses, with the zooms set at roughly 50mm. All shots were f5 1\100s ISO 1600, handheld. I shot in RAW, converted to JPG then took a couple of 100% crops from the same 2 spots from each photo, choosing a couple of good areas for comparison - 1 shows text on a bottle and part of a bar code and the other is part of a toy with hair and also smooth plastic that shows any grain well. I took 2 shots with the L series zoom - one with IS, one without, although it wasn't needed at these settings.

This wasn't that scientific - the shots aren't all exactly the same size (hence different size crops) and in fact although I thought I had the zooms both set to 50mm when I put the prime on I had to move back a half step to frame the same shot so I probably had them set slightly shorter. I'm sure I'd also see different results at different f stops etc but 50mm is a very common length for me to use (hence owning the prime). I set the ISO to the highest I would dare use for a shot that may be used professionally, and then set the f stop so I got the slowest shutter speed from the in-camera metering that I'd be happy with to get a sharp picture (usual rule I use is keep the speed one step above the focal length so I played it safe using 1\100 for 50mm).

The resulting pictures are a lot closer in quality than I expected. If you want to try and guess which shot is which I've put the answer below the image so scroll carefully :) There is one from the prime, one from the older zoom, and 2 sets from the L series - IS on and off.

View attachment 10683

1 - L series, no IS
2 - L series, with IS
3 - Older zoom
4 - Prime 50mm
 
Are the crops from reasonably central parts of the photo? If so, it's perhaps not that surprising that they're quite similar. At f/5 you were about a stop down from wide open, and that's pretty much in the sweet spot for each lens.

But there's an awful lot more to image quality, and lens quality, than just central sharpness. Comparing an expensive L series zoom with an inexpensive consumer zoom, you might reasonably expect to see:
* better central sharpness (though as noted, most lenses are pretty good in the centre)
* better corner and edge sharpness
* lower distortion
* better control of flare
* lower chromatic aberration
* less light fall-off in the corners
* better performance wide open
* more consistent performance across the whole zoom range (this can be a real issue for cheaper zooms)
* faster and more accurate autofocus
* better build quality
and probably some other things I haven't thought of too.

Testing just one dimension out of a dozen or so isn't terribly meaningful in my opinion.
 
PS
thecornflake said:
... the shots aren't all exactly the same size (hence different size crops) and in fact although I thought I had the zooms both set to 50mm when I put the prime on I had to move back a half step to frame the same shot so I probably had them set slightly shorter.
Click to expand...
This looks like focus breathing. Some zooms are very prone to it.
 
I'm glad someone here has done this test, I always wanted to see a comparison with the 24-105 and a consumer zoom lens as I was seriously considering buying a 24-105L to replace my 28-135 usm.
To be honest I thought the image quality produced by the 24-105 f4 L would be better than the 35-135 zoom by a huge margin but the difference is hardly noticeable.
Thanks Simon, you've saved me from making an expensive mistake.
 
redsnappa said:
I'm glad someone here has done this test, I always wanted to see a comparison with the 24-105 and a consumer zoom lens as I was seriously considering buying a 24-105L to replace my 28-135 usm.
To be honest I thought the image quality produced by the 24-105 f4 L would be better than the 35-135 zoom by a huge margin but the difference is hardly noticeable.
Thanks Simon, you've saved me from making an expensive mistake.
Click to expand...
Did you not read the post by Stewart?!

For a start, this test was executed on a crop camera, so corner sharpness / drop off doesn't even get a look in.

Believe you me, the 24-105 L is far better than the 28-135 in every area, having used both (and even more so on an ff body).
 
Last edited:
Did you not read the post by Stewart?!
Click to expand...
Yes I read Stewart's post, his comment regarding the crop sensor using the central part of the lens backs up my point.

24-105 L is far better than the 28-135 in every area
Click to expand...
In Simons pictures the 24-105 does not look any better than his cheap zoom.
 
thecornflake said:
1 - L series, no IS
2 - L series, with IS
3 - Older zoom
4 - Prime 50mm
Click to expand...
I picked my choices : far right jumped out with best contrast, great sharpness, predicted 50mm prime.
Third one was obviously unsharp and lacking contrast, predicted old zoom.
First two were fine in every way.
Then I checked which was which : I was right.
 
I thought the 24-105 with IS just about the sharpest, not the main criteria perhaps.
I would have expected the prime to be hard to beat at the one spot it can do.
 
redsnappa said:
Yes I read Stewart's post, his comment regarding the crop sensor using the central part of the lens backs up my point.


In Simons pictures the 24-105 does not look any better than his cheap zoom.
Click to expand...
And you state this from web sized images?

Stating a 24-105L not producing images better than a cheap zoom?

I dispair....
 
Last edited:
1/100 hand-held at 50mm on an APS-C? ISO 1600 on a 1100D?

I'm not sure what this is supposed to test - it's pushing the ISO hard for that body and I know I couldn't reliably hand-hold much below 1/250 regardless of focal length. Faster shutter speed and lower ISO would be far more useful for comparing the lenses without introducing quite so many opportunities for compromise from everything other than the lens.
 
I know this thread is about sharpness but there is a lot more to It when it comes to consumer (non L) and L lenses. It's not all about just sharpness.
The 24-105 isn't the best L lens but it's weather sealed, constant f/4, and does produce some very nice images.
If the first two aren't of interest then thee are non L lenses that can easily better it's sharpness.
 
dave_bass5 said:
I know this thread is about sharpness but there is a lot more to It when it comes to consumer (non L) and L lenses. It's not all about just sharpness.
The 24-105 isn't the best L lens but it's weather sealed, constant f/4, and does produce some very nice images.
If the first two aren't of interest then thee are non L lenses that can easily better it's sharpness.
Click to expand...
Personally, I think sharpness is another great strength for the 24-105. When first used mine, I was surprised by how sharp it was, even for an L. The detail it resolves on a 6d is remarkable.
 
Oh, and I shot most of today's wedding with one!
 
They are not macro lenses and their performance may be suboptimal at closeups. Just another reason why this comparison is anything but valid

If nicely centred bottle labels are your typical images then we may have a winner lens
 
dave_bass5 said:
I'm not saying it's a bad lens, just not up there with the best of the L series IMO, not even my old 17-55 ISS f/2.8.

Good all rounder though.
Click to expand...

Agreed - It is a good all rounder.

You say 'not even' your old 17-55 IS f2.8. That's a very good lens and not cheap. It may not have the build quality or the weather sealing of the L lens but for IQ on a cropped body it should be as good.
 
dinners said:
Agreed - It is a good all rounder.

You say 'not even' your old 17-55 IS f2.8. That's a very good lens and not cheap. It may not have the build quality or the weather sealing of the L lens but for IQ on a cropped body it should be as good.
Click to expand...

Agreed. I had two copies of the 24-105 during the time i had my 17-55. Cameras were 40D and 60D. The 24-105L was the better lens i suppose, due to its build quality and sealing, but the 17-55 produced sharper, more punchy images.
At the time the 24-05 cost more than the 17-55.
In fact ive had 3 copies of the 24-105 over the years, and while focusing was never an issue i felt it was quite a boring lens as far as IQ goes.
 
Alastair said:
1/100 hand-held at 50mm on an APS-C? ISO 1600 on a 1100D?

I'm not sure what this is supposed to test - it's pushing the ISO hard for that body and I know I couldn't reliably hand-hold much below 1/250 regardless of focal length. Faster shutter speed and lower ISO would be far more useful for comparing the lenses without introducing quite so many opportunities for compromise from everything other than the lens.
Click to expand...

I've generally used the rule of speed equal to focal length for hand holding. ISO 1600 on this camera has given me some great results before, especially on the 50mm prime. But it is probably the limit I would use unless I really had to. I also figured having some noise would be more useful for comparison purposes.

In general however I did say that I didn't really intend this as a fully scientific test. It was just a 'I wonder how it will turn out' thing. I'd hired the L series to shoot a field hockey match on Saturday. But I also took the other zoom. Quite a few shots needed copping down (first time doing sports photography and also know nothing about hockey) and in these cases I noticed slightly better quality in the L shots. Also bear in mind L series are weather resistant, and the focussing was faster and slightly more accurate (although to be fair I'd expect the older zoom to be slightly out if it's seen a lot of use).

I don't think this test is enough to say it's not worth spending the extra money on the L series. It depends on a number of factors. If I wasn't planning on moving into some freelance work and just shooting for myself and friends and family, I'd just keep the standard zoom and not bother with the L series. But for any photography where a certain standard of results was expected I'd go for the L. This is just my opinion however. Considering you can hire the L for £30-40 for 3 days it's potentially money well spent to give it a really good try if you were considering buying it (or any other higher prices lens for that matter)
 
I have a 28-135 USM, and have shot extensively with the 24-105L, and it's clear to me where the extra money goes on the 24-105. I've used both on a 550D and a 6D, so crop and FF.

Let me list the practical reasons why the 24-105L is better:
  • Much, much faster AF and IS
  • Keeps the rain out
  • Focus and zoom rings much smoother and more precise
  • Constant f/4 aperture
  • Better image quality - colour, geometry and aberration are all better
  • 77mm front diameter, not 72
  • doesn't look nearly as pervy when fully extended, matron!
There is one reason, and one reason alone, that I haven't replaced the 28-135 with the 24-105: cost.
 
Im also glad that someone else did this test, as it meant that I could leave the house and go and take some photos ... :whistle:
 
You must log in or register to reply here.
Back
Top