Canon 28mm f/2.8 vs. Canon 28mm f/1.8 - Test Results!

[sheldonnalos]

I know that this is a FAQ and a hotly contested issue, so I felt

that I had to share this with you all. I own a Canon 28mm f/2.8 and

just recently was able to borrow a friend's Canon 28mm f/1.8. I

couldn't resist running a little test to compare the two against

each other.

 

The general consensus has been that the 28mm f/2.8 is a better

(sharper) lens than the 28mm f/1.8. However, every time someone

brings up the 28mm f/1.8, the issue is debated by many who have (and

have not) used the lens. Many people have acknowledged that it is

soft wide open, and many others have stated that their copy is as

good as their 50mm f/1.4.

 

So, here are my test results, and some disclaimers...

 

Methodology: Canon 10D, tripod mounted, self timer, mirror lockup,

Large Fine JPG, Standard Paramenter, ISO 100, Daylight WB, No UV

Filters, front of lens shaded from any sunlight, No post processing,

100% center crops. The subject is about 150 ft away. There are *NO*

focus issues, since I can see other items in the full frame shot

that are nearer and farther from the crop, results are the same

regardless of whether you crop closer or farther away. There are no

motion blur problems, since both lenses visibly improve as they are

stopped down (and the shutter speeds correspondingly slow down). No,

I did not smear vaseline on the front element to influence the

results. ;-)

 

Disclaimer: This is a sample of ONE copy of each lens. YES, sample

variation exists (I have just seen it comparing two 50mm f/1.4

lenses). This does NOT settle the issue once and for all. I am NOT

saying that if you own a 28mm f/1.8 you are a bad person. Yakim

Peled is NOT always right, even if he's been right in this case. ;-)

 

Hope this is helpful to you all!

 

Sheldon<div></div>

[jon wilson]

2.8 obviously seems better at sharpness and contrast, weird.

[ky2]

So you're shooting an outdoor shot in broad daylight, and expect the available light (f/1.8) lens to perform better? Take both lenses indoors and try to shoot them handheld, and then compare.

 

My point: it doesn't really matter which one is sharper. The f/1.8's key purpose is extreme light gathering, not ultimate sharpness...

[jbq]

Very helpful, thanks. Makes the 28/2.8 look like a decent choice for a "normal" lens on the 10D.

[yakim_peled1]

Hummmm.......... I feel pretty silly right now :-)

 

 

I have recently decided to buy the 28/1.8 over the 17-40/4 (slow aperture) and 28/2.8 (slow AF) for my 1D. Yes, I know that it is optically inferior to the other two but I was thinking that it might (just might) make no difference for me.

 

 

At this point I advise you to sit down, have a glass of cold water and breath deeply. Are you relaxed now? Good, now I'll explain :-)

 

 

 

You see, I never print over A4 and most of my prints are half that size. I was thinking that for these parameters, the optical difference between these lenses will be negligible. If that happens, I will only enjoy the fast aperture and fast AF. If it won't, well, I tried that once with the 70-200/4 and it didn't work so I sold it. If it doesn't work again, I will sell that one as well. No harm done.

 

Happy shooting,

Yakim.

 

 

P.S. Let the stoning begin...... :-)

[yakim_peled1]

<p> BTW, if you look at the center performances of both lenses in <a href="http://www.photozone.de/8Reviews/index.html">http://www.photozone.de/8Reviews/index.html</a> you can see that the 28/1.8 is actually a bit <b>better</b> from f/2.8 onwards. </p>

<p> Sample variations?</p>

 

<p>Happy shooting, <br>

Yakim.</p>

[andrew robertson]

It's not weird that a f/2.8 lens would perform better than an f/1.8 lens. It's actually pretty

predictable.

 

It's much harder to design an f/1.8 lens, and compromises must be made. However, you

can't shoot the 1.8 at 2.8!

[panos_voudouris]

I think that copy is not good. I had the 2.8 and have the 1.8 and the difference is nowhere near this. Mind you, I don't pixel peep, I just look at the prints and the (mid-res) scans. In any case, even if the 1.8 is softer, it has better bokeh, colour, construction, has USM, looks good and makes shots with DOF and bokeh like the one attached at f/2 possible.

 

I think, for all the people who worry about it too much, get the f/1.8 and if you don't like the results then go back to the f/2.8. I won't even if you paid me.

[lester_wareham]

Good test.

 

They look about the same at f8 which agrees with the Canon MTF data for the APS-C sensor. For full frame the f2.8 MTF falls off faster than the f1.8 at the edge. It would be interesting to see full frame edge results.

 

I would have expected the lens to sharpen up withing 2-3 stops of wide open however.

[lotsawa]

Seems the 28/1.8 is better for short distances. That's what it is probably mainly designed for -- indoor low light shots.

 

@ Panos: cute!

[marknagel]

I never had results that bad on my 28/1.8 unless I accidentally left it in MF and forgot to focus. Mine is almost as sharp as my 50/1.4 and you really have to peep to see that.

 

m

[sheldonnalos]

A couple other side notes...

 

I think that the 28mm f/1.8 does have better color and contrast, especially when you look at a full frame shot.

 

For this test on an APS sensor, the corners for the f/2.8 lens remain better than the f/1.8 lens, all the way to f/8 and f/11.

 

Sheldon

 

BTW - Yakim! How could you go and buy a 28mm f/1.8 after all your staunch defense of the f/2.8? (Actually, I totally understand why) I hope you get a sharper copy!

[mike_meloy]

Sheldon,

Did you AF these pics? I really can't say that I see anything that is clearly in focus with any

of the lens' samples. It may be that you were shooting into a shaded area.

When you AF with WA lens at wide apertures on a 1.6 camera, turn the FTM focus to the

right just past the point of focus and then press the shutter half way to AF the pic. I know

the 28 2.8 doesn't have FTM, so you get what you pay for I guess.

Your examples of the 28 1.8 do not represent the quality of the 28 1.8 that I own. I can

give some of my examples here:

 

http://www.pbase.com/bosphorousman/inbox

 

These pics were shot for testing of the lens, mostly.

I use this lens mostly for indoor, availible light photos. Outside at wide aps, it has a lot of

CA.

[grant g]

"I have recently decided to buy the 28/1.8..." -Yakim

 

The sky is falling! LOL. Yakim, I'm glad you are open enough to have changed

your mind about this one. :-)

 

I have both and having to choose one, I'd keep the 1.8 hands down. The 2.8

is only preferred for hiking and daytime landscapes.

[sheldonnalos]

Mike -

 

The pics were autofocused, but the full frame shots has elements in it that go from 15 feet away to infinity, so you can see if things are front/back focused. Unfortunately, that's not the case. The lens shows similar performance indoors at closer distances when wide open. I think it's just a poor performer/poor copy.

 

Sheldon

[mike_meloy]

Sheldon,

I agree, it appears that it is a poor performer. My first 28 1.8 had to be sent back to B&H

because of inconsistant wide ap performance below f3.5. It was the only lens I have ever

had to return but its replacement has been just fine. Hope you find a good one to use for

comparison.

[benb]

I've had both, I think my f/1.8 copy is one of the good ones, it has always subjectively

seemed better then the 28/2.8 I had. Especially color & contrast as noted.

 

But I use it for available light stuff too, and for that the f/1.8 is obviously superior. But I

have a 5D and I've shot quite a few pictures which are very sharp on it already. It is

noticeably better then my 17-40 f/4L USM. I had a 10D up till a month ago and the f/1.8

lens always seemed better on that one too.

 

Keeping with the pets, 5D, EF 28mm f/1.8, 1/100th @ f/1.8 @ ISO 100.

 

http://bsquared.net/2005/november/IMG_1849_web.jpg

[benb]

Oops, the shoot I linked to in my previous post was ISO 800, not ISO 100.

 

In any case I think it is a great lens. It has just a teeny bit of light fall off wide open, not

much worse then the 50/1.4, great focus, always nice and sharp.

 

I used it as a normal on my 10D but it is really coming into it's own on the 5D, I anticipate

getting lots of very nice shots with it.

 

Actually here are a few more shot in NYC a few weeks ago, I really like this lens.

 

http://www.bsquared.net/2005/november/11162005/images/IMG_1607_800.jpg

 

http://www.bsquared.net/2005/november/11162005/images/IMG_1615_800.jpg

 

Here's a landscape shot, stopped down with a polarizer, again nice color, this has

absolutely minimal post processing:

 

http://www.bsquared.net/2005/october/10302005/images/IMG_0046.jpg

[andrew robertson]

Yes, it shoots pictures. Yes, they look nice when reduced severely. How does the exact

frame look when shot with both lenses?

[mike dixon]

Sheldon, you've got a <i>very</i> bad sample of the 28/1.8 lens. It's my most-used lens on digiRebel, and I'm consistently getting much sharper shots than that. Here's a full-frame image taken at f4, shot in RAW (parameter 2, which is the low sharpness/saturation setting).

 

(see the followup post for a 100% crop)<div></div>

[mike dixon]

Here's a 100% crop from that image with <b>no</b> sharpening applied.<div></div>

[mike dixon]

Another set shot at f2.8.

 

(Neither set are great photos, but they were the first I found to illustrate my point.)

[mike dixon]

Here's the 100% crop.<div></div>

[benb]

BTW I am not going to stoop to posting pixel peeping 100% crops but all the photos I posted

are nice and sharp at 100% viewing just like the one Mike posted. (Assuming they don't have

blur due to DoF)

 

Point is unless you have a broken or defective 28/1.8, if you're not getting good pictures out

of it, it's you, not the lens.

[benb]

Sorry I got angry. I am really sick of all the pixel peeping though.

 

Here are some crops, these are from a 5D so when I say corner it means the corner.

 

All of these are handheld except for the f/22 shot. I would actually say mine perhaps gets

worse at f/16 and f/22, although I haven't really had any ruined shots, and most of my

shots at those apertures are slow shutter speeds so it could just be my technique.

 

I think the original posters shot may have flared. The f/1.8 is more vulnerable to flare. It

doesn't come with a hood and it can be quite deceptive on a cropped body. When I was

using it on my 10D I could barely tell when it was flaring through the viewfinder. The

issue is there may be a bright light which is in the lenses field of view but is being cropped

out by the camera. You could even have the sun in the field of view and not realize it.

The hood helps but you can still screw it up as you're not seeing everything that the lens is

seeing.

 

A lot of my f/1.8 shots are indeed not sharp, but that has more to do with what I'm

shooting then the lens. I don't know if I would say it is sharp as the 50/1.4 wide open. I

only recently got a 50/1.4 and I've already gotten a few handheld f/1.4 shots that are tack

sharp in the center. The crop I am including was actually a mistake, I should not have

taken that shot at f/1.8 but it is pretty sharp.