I MA my 28mm 1.8 and want your opinion guys:

This was shot at ISO 100, 1/250 (430EX II), 1.8

Mine's sharper than that. I just used it this morning at a botanical conservatory, shot a lot at 1.8 and 2.0 and they came out fantastic.

MA = micro adjust?

Also did you focus on the bottle? Because at this range and at that aperture, I believe the DOF is very shallow, so even if you focused on the Snitch, it may off just because he isn't flat like the bottle.

Looks OOF, I would try again with Manual Focus in Live View. Your MA may not be correct

Dudewithoutape wrote:
MA = micro adjust?

Also did you focus on the bottle? Because at this range and at that aperture, I believe the DOF is very shallow, so even if you focused on the Snitch, it may off just because he isn't flat like the bottle.

Yep. MA=Micro adjust.
I focused on the bottle but not using the center point. I used the nearest focus point.

You can't use autofocus to test for sharpness as it introduces an uncontrolled variable into the mix

Dawei Ye wrote:
You can't use autofocus to test for sharpness as it introduces an uncontrolled variable into the mix

But, how could I check if the adjust I did is ok if I don't use an autofocus ?

Dawei Ye wrote:
You can't use autofocus to test for sharpness as it introduces an uncontrolled variable into the mix


Yea you said the same thing with my tests. WE GET IT, you don't like to use AF.

Fact is, when you use your lens, it's being used in AF mode, not MF. So why not test what it does in AF mode?

Alright... so I redid my tests and see what I've got.

Using the left focus point:

I bought my 28 in here used about a year ago. I compared it to my 85 1.8, and it failed wide open, no contest. Contrast and sharpness was not usable for me. I sent it into Canon for checking, spent around $125, got it back and it's still unusable to me at 1.8. The lens did seem better at 2.0 and up after I got it back. I chalked it up as basically, it's a 2.0 lens now. No problem. If they made it any better, it would probably take away rom their 24 and 35 L sales.

The bokeh is actually pretty nice. Build and focus feel fine to me. Not ready to ditch a mid-range zoom for it though.

Is this FF or crop? Perhaps your copy just isn't sharp in that area? It could also be something wrong with that AF point. One way to test if its the lens of the camera AF is to make sure its on the same plane (like the first pic) AF center point on it (half shutter press) and then (I forget the technical point) but move/shift the camera to the right, so that its in the same area as the first shot. I feel stupid, I cannot think of the term right now, OMG.

Edit - Pan? Frame? ugh.
Edit 2 - I realize I have no idea what MA actually does, I mean I have an idea from the term, but my Rebel doesn't have that so, I'm trying to help with very limited knowledge, sorry.

Dudewithoutape wrote:
Is this FF or crop? Perhaps your copy just isn't sharp in that area? It could also be something wrong with that AF point. One way to test if its the lens of the camera AF is to make sure its on the same plane (like the first pic) AF center point on it (half shutter press) and then (I forget the technical point) but move/shift the camera to the right, so that its in the same area as the first shot. I feel stupid, I cannot think of the term right now, OMG.

Edit - Pan? Frame? ugh.
Edit 2 - I realize I have no idea what MA actually does, I mean I have an idea from the term, but my Rebel doesn't have that so, I'm trying to help with very limited knowledge, sorry.

Don't worry about all those technical terms...
I thought that if I shoot and recompose, it would be worst because I would move more than choosing the nearest focus where I want to focus and using 1.8, that would affect my focus point (I think... )
I think it's something about the limits of this lens. I've heard that it's not (that) sharp at 1.8 and I wanted to know if my copy falls into the acceptable range. I guess not

I tried several adjustments with micro-adjust and -5 was the one that I had the best results... maybe I'll try it again.

Thanks !

Which camera is this? Most of the XXD models have "high precision" AF at F2.8 or larger with the center point and normal precision with outer points. The Rebels only have single axis AF sensors on outer points. So basically the center point should be much better than outer points.

Gochugogi wrote:
Which camera is this? Most of the XXD models have "high precision" AF at F2.8 or larger with the center point and normal precision with outer points. The Rebels only have single axis AF sensors on outer points. So basically the center point should be much better than outer points.

It's a 50D. The test I redid using the center focus is much better than using the outer focus.

fmikio wrote:

But, how could I check if the adjust I did is ok if I don't use an autofocus ?

Ok but your thread title clearly says "28 f/1.8 Sharpness" not "AF performance" I'm just saying they are two separate issues and if you wanted to confirm whether your lens was sharp or not - use MF first. THEN do your stuff with MA and whatever. It's important to get the temporal sequencing correct and not fudge about with 2 uncontrolled variables in the same test

Who knows whether the softness is caused by the lens being soft, or being out of focus, or both. At least once you have confirmed sharpness with MF, you can be sure that any softness in AF results is due to AF inaccuracy, rather than lens optical issues

Will Patterson wrote:
Fact is, when you use your lens, it's being used in AF mode, not MF. So why not test what it does in AF mode?

This has nothing to do about how the lens is used in practice. It's about identifying the causal reason why performance is not up to scratch. To do this, you have to eliminate conflicting variables. Otherwise, how can you possibly know whether it is because the lens is not optically up to scratch, or whether an AF intolerance is the reason for the softness?

But...But...you say, "I use my lenses as AF lenses." This is besides the point. Being a wedding photographer like yourself, if you take your lens to canon because you think it's not working well - do you expect Canon to send its technicians to a random Wedding on the weekend to do some test shots using AF in order to check the optical specifications of your lens? hey...is that not how the lens is being used?

The reason MF must be used when testing for lens sharpness is the exact reason why a Tripod is used (to eliminate another factor as a cause (in the tripod's case - motion blur). If you want to test for AF accuracy, then fine, but the topic of the thread is "28mm 1.8 sharpness"

The center AF shot looks pretty decent to me.

Dawei Ye wrote:
fmikio wrote:

But, how could I check if the adjust I did is ok if I don't use an autofocus ?

Ok but your thread title clearly says "28 f/1.8 Sharpness" not "AF performance" I'm just saying they are two separate issues and if you wanted to confirm whether your lens was sharp or not - use MF first. THEN do your stuff with MA and whatever. It's important to get the temporal sequencing correct and not fudge about with 2 uncontrolled variables in the same test

Who knows whether the softness is caused by the lens being soft, or being out of focus, or both. At least once you have confirmed sharpness with MF, you can be sure that any softness in AF results is due to AF inaccuracy, rather than lens optical issues

Will Patterson wrote:
Fact is, when you use your lens, it's being used in AF mode, not MF. So why not test what it does in AF mode?

This has nothing to do about how the lens is used in practice. It's about identifying the causal reason why performance is not up to scratch. To do this, you have to eliminate conflicting variables. Otherwise, how can you possibly know whether it is because the lens is not optically up to scratch, or whether an AF intolerance is the reason for the softness?

But...But...you say, "I use my lenses as AF lenses." This is besides the point. Being a wedding photographer like yourself, if you take your lens to canon because you think it's not working well - do you expect Canon to send its technicians to a random Wedding on the weekend to do some test shots using AF in order to check the optical specifications of your lens? hey...is that not how the lens is being used?

The reason MF must be used when testing for lens sharpness is the exact reason why a Tripod is used (to eliminate another factor as a cause (in the tripod's case - motion blur). If you want to test for AF accuracy, then fine, but the topic of the thread is "28mm 1.8 sharpness"

what use is a sharp lens if it isn't focusing accurately?

Will Patterson wrote:what use is a sharp lens if it isn't focusing accurately?

The point is that you'll want to figure out what the problem is in order to take an appropriate action. If the lens isn't sharp at all, then you can just return the lens from where you bought the lens.

But, if the lens is sharp, but doesn't auto focus correctly, then you can micro adjust the lens to your camera or send it to Canon for adjustment.

Each problem requires a different solution.

I treat mine as an f/2.0 lens too. At this aperture it is very sharp in the center of the frame, with good contrast and flare resistance. It does soften up outside of the center much more than the 'L' primes. Still, a good lens for candids and reportage.

Here's an example I've posted before shot at f/2.0, full res on a 5D2:

I have no problem using mine wide open -