Review - Sigma 24-105mm f/4 DG OS

neuroanatomist said:
JVLphoto said:
iMagic said:
Ugg. Focus shift.
Yeah, there's a term for that right? "Non-parfocal" but neither is the Canon... for what that's worth.

No, they are two different things.

Parfocal means a lens maintains focus while being zoomed, it's a very useful feature for shooting video (whereas with stills, it's generally easy to refocus after zooming). The Canon 24-105L is not parfocal (the 17-40, 16-35, and 70-200/2.8 non-IS are parfocal). Not sure on the Sigma, but I'd guess it's not parfocal.

Focus shift means the focus changes when the lens is stopped down. No problem shooting wide open, but if you stop the lens down to f/5.6 or f/8 with a close subject, that subject will likely not be in crisp focus with the Sigma 24-105. With more distant subjects, the effect is masked by the deeper DoF. The 50/1.2L is notorious for focus shift (people call it a 'backfocus problem' usually because of a lack of understanding the real issue).

Ah! Okay, yes, that actually makes sense. Thanks for the clarification!
 
Upvote 0
photo212 said:
Some of us want a lens we can know will perform in the rain and snow. Is this as weather resistant as Canon's L-series lenses? Can this lens go to the windy beach and not get clogged with sand? I've had Canon's 24-105mm f/4L IS in all sorts of conditions, and I have not thought twice about it. Dumb luck or specs, I do not know.

Is there a reliable rating system for such a question?

Hey, no weather sealing, not to mention the double-barrel zoom makes an extra area "exposed" for potential element influx. I don't think Sigma has an "All weather" rating, and neither does Canon (not even all the white L lenses are weather sealed).
 
Upvote 0
neuroanatomist said:
JVLphoto said:
iMagic said:
Ugg. Focus shift.
Yeah, there's a term for that right? "Non-parfocal" but neither is the Canon... for what that's worth.

No, they are two different things.

Parfocal means a lens maintains focus while being zoomed, it's a very useful feature for shooting video (whereas with stills, it's generally easy to refocus after zooming). The Canon 24-105L is not parfocal (the 17-40, 16-35, and 70-200/2.8 non-IS are parfocal). Not sure on the Sigma, but I'd guess it's not parfocal.

Focus shift means the focus changes when the lens is stopped down. No problem shooting wide open, but if you stop the lens down to f/5.6 or f/8 with a close subject, that subject will likely not be in crisp focus with the Sigma 24-105. With more distant subjects, the effect is masked by the deeper DoF. The 50/1.2L is notorious for focus shift (people call it a 'backfocus problem' usually because of a lack of understanding the real issue).

I experienced some Sigma focus shift today in fact. I was a bit perplexed when I was using LV to manually focus the 50mm at f/4, and when I took the shot it was slightly off. The 24-105L didn't have this issue. Is this due to the fact that when focusing the aperture is wide open at f/1.4 on the Siggy and f/4 on the Canon? That is kinda useless then if using LV. Makes AF accuracy even more important.
 
Upvote 0
Re: Review - Sigma 24-105mm f/4 OS

JVLphoto said:
traveller said:
Why do people insist on posting images of bookshelves to demonstrate how "sharp" a lens is? There was not one detail in those images that was resolved differently by either lens, mainly due to the dearth of any high frequency detail that might show up resolution differences. If it wasn't for the distortion differences, I might have actually thought that they were samples from the same lens.

I think I'll wait for the results from Roger Cicala to decide whether the Sigma is a worthwhile upgrade. I would recommend that in future, you either do reviews properly or stop publishing this rubbish as CR's "official" review.

Out of 23 "reviews" this is the first and *only* time we posted photos of my dusty bookshelves. I certainly don't insist on it, or like it, but it does show difference of detail in the corners, vignetting and other real-world variables.

Roger is an incredible technical reviewer with all the right tools to measure and analyze lenses and I too look forward to his reviews and articles. I just pay less attention to the charts and more to how it works for me professionally. The final image is what matters to me and my clients. I'm also open to hearing your take on what entails a "proper" review for my future rubbish.

Please accept my apologies for my last comment, it was harsh and rude; I appreciate anyone who takes the time to post a review and take the flak for it!

I wouldn't expect anyone who isn't properly set up for it to attempt quantitative testing, nor given the number of technical review sites available, would I feel the need for it. I would prefer to see you use large crops from photos taken in your professional style, like the photos that you used to illustrate the review, to demonstrate your points.

I won't mention the name of the website that uses the "bookshelf-of-doom" to "show" lens "sharpness", suffice to say that I don't put any credence in their reviews!
 
Upvote 0
Zv said:
I experienced some Sigma focus shift today in fact. I was a bit perplexed when I was using LV to manually focus the 50mm at f/4, and when I took the shot it was slightly off. The 24-105L didn't have this issue. Is this due to the fact that when focusing the aperture is wide open at f/1.4 on the Siggy and f/4 on the Canon? That is kinda useless then if using LV. Makes AF accuracy even more important.

It has to do with the lens design, not the max aperture (although when focus shift affects a lens, it's usually more evident with a faster lens). AF is always done wide open - the thing about focus shift is that the lens is the shift occurs when the lens is stopped down, which happens after AF is complete. It doesn't matter if focus is live view or phase detect, the problem is there. In live view, you can hold the DoF preview button and focus manually to avoid the issue.
 
Upvote 0
tomscott said:
Is it just me that thinks the crops look very similar?

Nope, they are very close, distortion is a bit different at 24mm and the real difference are seen at the 100% crops. Even then it's not quite the res you need. See below (hope they help).

https://dl.dropboxusercontent.com/u/12373943/Canon%2024-105%20-%2024mm%20f4.jpg Canon 24-105 f/4 IS (at f/4) 9.8MB
https://dl.dropboxusercontent.com/u/12373943/Sigma%2024-105%20-%2024mm%20f4.jpg Sigma 24-105 f/4 OS (at f/4)
 
Upvote 0
Re: Review - Sigma 24-105mm f/4 OS

traveller said:
JVLphoto said:
traveller said:
Why do people insist on posting images of bookshelves to demonstrate how "sharp" a lens is? There was not one detail in those images that was resolved differently by either lens, mainly due to the dearth of any high frequency detail that might show up resolution differences. If it wasn't for the distortion differences, I might have actually thought that they were samples from the same lens.

I think I'll wait for the results from Roger Cicala to decide whether the Sigma is a worthwhile upgrade. I would recommend that in future, you either do reviews properly or stop publishing this rubbish as CR's "official" review.

Out of 23 "reviews" this is the first and *only* time we posted photos of my dusty bookshelves. I certainly don't insist on it, or like it, but it does show difference of detail in the corners, vignetting and other real-world variables.

Roger is an incredible technical reviewer with all the right tools to measure and analyze lenses and I too look forward to his reviews and articles. I just pay less attention to the charts and more to how it works for me professionally. The final image is what matters to me and my clients. I'm also open to hearing your take on what entails a "proper" review for my future rubbish.

Please accept my apologies for my last comment, it was harsh and rude; I appreciate anyone who takes the time to post a review and take the flak for it!

I wouldn't expect anyone who isn't properly set up for it to attempt quantitative testing, nor given the number of technical review sites available, would I feel the need for it. I would prefer to see you use large crops from photos taken in your professional style, like the photos that you used to illustrate the review, to demonstrate your points.

I won't mention the name of the website that uses the "bookshelf-of-doom" to "show" lens "sharpness", suffice to say that I don't put any credence in their reviews!

Accepted. Funny, I get flak when I don't do tests and when I do too (though the later I make no excuses for: I am not set up to do that *at all*).

And I'll definitely post more images in higher res as the day goes on. In most cases my images come processed to some point, I mean, most do right? My bookshelves are unedited though :-P
 
Upvote 0
JVLphoto said:
tomscott said:
Is it just me that thinks the crops look very similar?

Nope, they are very close, distortion is a bit different at 24mm and the real difference are seen at the 100% crops. Even then it's not quite the res you need. See below (hope they help).

https://dl.dropboxusercontent.com/u/12373943/Canon%2024-105%20-%2024mm%20f4.jpg Canon 24-105 f/4 IS (at f/4) 9.8MB
https://dl.dropboxusercontent.com/u/12373943/Sigma%2024-105%20-%2024mm%20f4.jpg Sigma 24-105 f/4 OS (at f/4)

Thanks for the links. Its interesting the top of the image of the Canon is much sharper than the Sigma (RunDMC) and the bottom of the sigma is slightly sharper. Centres are almost identicle.
 
Upvote 0
Gah! Charts!

Not to be rude, but the review doesn't answer some of the real world questions that I, as a Canon 24-105/4 user, have.

For example, how does the Sigma perform when using the OS to stabilize the video when following a moving subject while walking and pulling focus? This is a place where the Canon does a great job, and one of the reasons so many video guys use it.

Or, when in studio doing product photography, or other industrial/catalog work, how does the Sigma perform when shooting against lit white seamless at f/8? Do the blacks wash out, or does it stay sharp? Does kick-back from a backdrop lit a stop brighter than the subject cause the Sigma to become less contrasty? Lots of product guys use the 24-105 in this situation because it performs well under these conditions. I recall both Zarias and Ukandu lady saying the Canon 24-105 is the go-to lens in these conditions in their creative Live workshops.
 
Upvote 0
Rey said:
Gah! Charts!

Not to be rude, but the review doesn't answer some of the real world questions that I, as a Canon 24-105/4 user, have.

For example, how does the Sigma perform when using the OS to stabilize the video when following a moving subject while walking and pulling focus? This is a place where the Canon does a great job, and one of the reasons so many video guys use it.

Or, when in studio doing product photography, or other industrial/catalog work, how does the Sigma perform when shooting against lit white seamless at f/8? Do the blacks wash out, or does it stay sharp? Does kick-back from a backdrop lit a stop brighter than the subject cause the Sigma to become less contrasty? Lots of product guys use the 24-105 in this situation because it performs well under these conditions. I recall both Zarias and Ukandu lady saying the Canon 24-105 is the go-to lens in these conditions in their creative Live workshops.

Ugh, and you even specifically asked me for these. I'm the rude one for not addressing them.

To answer the video, I'm not a video shooter and any "test" I could do with it would lack any real insight into what a video shooter needs. I haven't talked about video in my other reviews because of this and held true here. I will assume that the "always on" OS being as silent as it was would suit this function incredibly well. It worked well enough for the stills.

I wanted to try the blown out backdrop, but ran out of time with the lens and had to send it away. Without testing this *specitifically* I do have a few good shots with direct light and there's still good contrast and detail. Sorry for missing this :(
 
Upvote 0
I believe the EF 24-105 is parfocal in one direction; from long to short, ie you can focus on a subject at 105 mm and then zoom out to 24 holding focus, but not the other way round. It is specifically designed this way with a compensating cam inside to keep it right. ( Another fact gleaned from CR ;) )

Interesting on the superior corners of the Sigma, but is it perceivable at normal magnifications ? The corners are (one) of the EF 24-105's weak areas and with the 24-70 IS the difference is perceivable at realistic magnifications.
 
Upvote 0
JVLphoto said:
Rey said:
Gah! Charts!

Not to be rude, but the review doesn't answer some of the real world questions that I, as a Canon 24-105/4 user, have.

For example, how does the Sigma perform when using the OS to stabilize the video when following a moving subject while walking and pulling focus? This is a place where the Canon does a great job, and one of the reasons so many video guys use it.

Or, when in studio doing product photography, or other industrial/catalog work, how does the Sigma perform when shooting against lit white seamless at f/8? Do the blacks wash out, or does it stay sharp? Does kick-back from a backdrop lit a stop brighter than the subject cause the Sigma to become less contrasty? Lots of product guys use the 24-105 in this situation because it performs well under these conditions. I recall both Zarias and Ukandu lady saying the Canon 24-105 is the go-to lens in these conditions in their creative Live workshops.

Ugh, and you even specifically asked me for these. I'm the rude one for not addressing them.

To answer the video, I'm not a video shooter and any "test" I could do with it would lack any real insight into what a video shooter needs. I haven't talked about video in my other reviews because of this and held true here. I will assume that the "always on" OS being as silent as it was would suit this function incredibly well. It worked well enough for the stills.

I wanted to try the blown out backdrop, but ran out of time with the lens and had to send it away. Without testing this *specitifically* I do have a few good shots with direct light and there's still good contrast and detail. Sorry for missing this :(
No problem, its not like I'm going to ditch my 24-105L if the Sigma is better in these conditions (I'd rather milk every ounce of ROI from my current gear), but if I were building a kit today, I would like to know the answers to those questions.
 
Upvote 0
Sporgon said:
I believe the EF 24-105 is parfocal in one direction; from long to short, ie you can focus on a subject at 105 mm and then zoom out to 24 holding focus, but not the other way round. It is specifically designed this way with a compensating cam inside to keep it right. ( Another fact gleaned from CR ;) )

Chuck Westfall stated, "There's a cam inside the 24-105mm lens that is designed to maintain an accurate focus when the lens is zoomed from tele towards wide." So, I suppose you could say it's parfocal…but I wouldn't. A truly parfocal lens remains in focus regardless of the zoom direction. Also, zooming from 105mm to 24mm, assuming you're not moving the camera, results in a dramatic increase in DoF as you zoom out. A subject at 5 m at 105mm f/4 has a DoF of approximately 0.25 m on either side of the subject…at 5 m, 24mm f/4 the DoF runs from 3 m in front of the subject to infinity behind - so, yeah, focus is maintained…no cam required. ::)
 
Upvote 0
neuroanatomist said:
Sporgon said:
I believe the EF 24-105 is parfocal in one direction; from long to short, ie you can focus on a subject at 105 mm and then zoom out to 24 holding focus, but not the other way round. It is specifically designed this way with a compensating cam inside to keep it right. ( Another fact gleaned from CR ;) )

Chuck Westfall stated, "There's a cam inside the 24-105mm lens that is designed to maintain an accurate focus when the lens is zoomed from tele towards wide." So, I suppose you could say it's parfocal…but I wouldn't. A truly parfocal lens remains in focus regardless of the zoom direction. Also, zooming from 105mm to 24mm, assuming you're not moving the camera, results in a dramatic increase in DoF as you zoom out. A subject at 5 m at 105mm f/4 has a DoF of approximately 0.25 m on either side of the subject…at 5 m, 24mm f/4 the DoF runs from 3 m in front of the subject to infinity behind - so, yeah, focus is maintained…no cam required. ::)

Yes I'd thought about the greater DoF as you zoom out from medium tele to quite an extreme wide angle. Maybe the cam still holds focus from 105 to 70 - or maybe there's no cam at all ;D Seems strange that they would go to the trouble of engineering it to hold focus one way ( the easy way ) but not the other, but perhaps in video work zooming from close up to wide is a much more frequently used technique than visa versa.
 
Upvote 0
Hi,

Thanks for taking the time and effort to review this lens.
If people don't like your review maybe they can do one themselves....


I've had sigma lenses before and maybe these new "Art" designated lenses are higher quality than the last generation.
However, if my Canon 24-105 lens needs to be replaced, I think I'll stick to a Canon. My previous lens was a 17-70mm and I had to replace it twice to get one that was on the money.

Sigma's new lenses do look great but the ultimate question in my head is can they back it up with robust product that continually performs in its lifetime.
 
Upvote 0