p.16 #1 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
tsdevine wrote:
I can't really given guidance on the GM 35 comparison, I've never shot one. I find my 28-45 does well when I compare it to my CV 35 APO, I'm not sure I have spent the time and effort to test at various distances, etc. But my 28-45 is sharp wide open, but like most lenses, it does improve when stopping down.
Bryan's TDP comparison tool seems to imply it's in the ballpark of the GM 35 wide open, note you have to copy and paste the whole link, if you click on it, FM chops off part of the URL.
In terms of batches, I sort of doubt it. These lenses are complex enough, I don't think variation would come from a whole bunch of them having the same component out of spec, or assembled slightly improperly in the same way for a whole batch.
And sample variation isn't limited to a single lens model, you might have a better than average GM 35....Show more →
Thank you, I looked at this comparison before and I also read the lenstip reviews of the 35mm GM, 35mm DG DN F2 and 28-45mm. I had the 35mm DG DN F2 and like on the lenstip reviews, my 35mm GM was much sharper at f1.4 than my Sigma at f2 (at close to mid distances). At 2.8 the Sigma was as sharp as my 35mm GM, exactly like on the lens tip image resolution chart. It is why I was surprised to see my 35mm GM much sharper than my 28-45mm at 28mm wide open because lenstip shows a similar center performance wide open for both the 35mm GM and the 28-45mm at 28mm.
p.16 #2 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
philip_pj wrote:
Many factors are in play. It's the reason to be specific about methodology, focus distances, rationale for conclusions, etc. Then, copy variation, etc. Most believe the GM 35/1.4 and CV 35/2 are close to identical in performance. Some believe Sony cooks the books (RAWs) for at least some lenses, so RAW converters need to be detailed. It's messy. Below are the LensTip wide apertures for the GM and CV 35s.
I agree, from my experience the focus distance plays a big role on the performances. By exemple my 35mm DG DN F2 was much sharper at long distance than at short distance, exactly like my 28-45mm wide open.
Also, I use the 28-45mm L-mount version on my S5II/S1 and I use the 35mm GM on my A7C/A7III. I maybe wrong but I always found the "sharpness" of my 35mm GM and 135mm GM a bit unnatural, like if the amount of details were not better than my best Panasonic/leica lenses but more like if a bit of extra digital sharpening was added to renforce the sense of resolution.
I only shoot in raw and I almost never change the default sharpening in LR. So I wouldn't be surprised if Sony adds some sharpening inside their raw files. And I can add that when watching some images samples from the 28-45mm and the S5II at 28mm at short distance or at long distance, I don't see a real difference with my copy, like on the Cameralabs samples on Flickr (the short distance flowers pic looks a bit soft).
p.16 #3 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Keunish wrote:
[...]
I only shoot in raw and I almost never change the default sharpening in LR. So I wouldn't be surprised if Sony adds some sharpening inside their raw files.[...]
Here i don't follow really: Let's see if I get this right.
You look at your images "developed" using LR with sharpening at default value and then you suspect Sony adds sharpening to the raw files. Correct?
I'm surprised as I haven't noticed this. Now I checked again and took a few ARW files from a "sharpness" test (high contrast, small details, performed the last time I bought a new lens), loaded the images in RawTheerapy, checked that the sharpening was off and enlarged the view to a whopping 500% size. There was no signs of sharpening. This was with the A7CR.
Then of course I'm not that educated about sharpening of raw files so I may be wrong.
p.16 #4 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Jonas B wrote:
Here i don't follow really: Let's see if I get this right.
You look at your images "developed" using LR with sharpening at default value and then you suspect Sony adds sharpening to the raw files. Correct?
I'm surprised as I haven't noticed this. Now I checked again and took a few ARW files from a "sharpness" test (high contrast, small details, performed the last time I bought a new lens), loaded the images in RawTheerapy, checked that the sharpening was off and enlarged the view to a whopping 500% size. There was no signs of sharpening. This was with the A7CR.
Then of course I'm not that educated about sharpening of raw files so I may be wrong.
Yes, if Sony applied any RAW sharpening it would have been picked up by review sites already (e.g. optyczne.pl tests for this).
What we might be seeing here is the effect of low-pass filter maybe?
It may be my eyes, but I would consider all of these to be sharp. It seems to track reasonably well to the Lenstip resolution test. Again it's possible your copy is a little off, or maybe you have a unicorn copy of the GM 35.
@Keunish, do you have any samples you can post? Both showing the dramatic change when stopping down a bit, and maybe some comparisons with your GM 35?
p.16 #6 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Keunish wrote:
I just received my copy, my sample looks good at 35mm and 45mm, however at 28mm it looks slightly soft wide open at close distance, it is very sharp at long distance though.
And I can see the top center is also (very) slightly sharper than the middle center even at 35mm and 45mm. Is decentering can be visible at the top or bottom center of the image ?
To me it sounds as the lens is faulty. You clearly describe the problem and the lens "sharpness" behaviour doesn't reflect neither the Lenstip charts nor the experience other users in this thread have.
p.16 #7 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
j4nu wrote:
Yes, if Sony applied any RAW sharpening it would have been picked up by review sites already (e.g. optyczne.pl tests for this).
What we might be seeing here is the effect of low-pass filter maybe?
If you are thinking about the problematic lens Keunish describes I don't think it is about any low-pass filter. He gets a sharper border than center for example.
p.16 #8 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Jonas B wrote:
If you are thinking about the problematic lens Keunish describes I don't think it is about any low-pass filter. He gets a sharper border than center for example.
I'm thinking about the differences he sees between the cameras ...
p.16 #9 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
j4nu wrote:
I'm thinking about the differences he sees between the cameras ...
Ah, the "sharpening of RAW file" misunderstanding. I think my brain already dismissed that. ;-)
I presume you have checked that the other camera he is talking about has a low-pass filter. You are probably right.
p.16 #12 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
tsdevine wrote:
I'm trying to wrap my head around how Sony could sharpen a RAW file prior to the demosaic step. I can understand how it could sharpen the JPEG.
Well, they can remove stars and fix vignetting... so surely there are some believing in sharpened RAW-files. Personally I heard Canon does that...
p.16 #13 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Jonas B wrote:
Well, they can remove stars and fix vignetting... so surely there are some believing in sharpened RAW-files. Personally I heard Canon does that...
;-)
Well, fixing vignetting is scaling a boost to exposure for each pixel based on distance from center. You can do that at a pixel at a time, doesn't matter what color that pixel was sensitive to.
Removing hot pixels and/or noise filtering, I still sort of see how that could be done in a localized fashion based on maybe departure of intensity in a super small set of pixels (i.e. star eater.)
Not saying it can't be done. Just that I'm trying to wrap my head around it. There are many things that are possible that I can't wrap my head around, especially without someone explaining how it is done. And even then there are no guarantees.
p.16 #14 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
tsdevine wrote:
[...] There are many things that are possible that I can't wrap my head around, especially without someone explaining how it is done. And even then there are no guarantees.
p.16 #15 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
'always found the "sharpness" of my 35mm GM and 135mm GM a bit unnatural'
Fred has mentioned this in a 35GM review:
'The Sony FE 35mm f/1.4 GM is very well corrected for lateral CA which is great news for landscape shooters. For this comparison below, CA correction was turned off in-camera and in-post but *keep in mind that Lightroom has an automatic correction profile applied to Sony lenses that can't be turned off. (Unless covering the lens' contacts)*.'
Imagine having to cover lens-camera interface contact points to obtain a correct assessment of lens performance. Regarding the rendering: being fair, Sony uses a lot of asph surfacing which is reflected in that painted on look all the GM's have, for in-focus content shown over the top of smooth background blur. It's a signature look of theirs.
And you may wish to see another source, regarding the Batis and Sony FE 85/1.8 lenses' MTF:
'Stopping down doesn’t improve the quality anymore (nor is this needed anyway). It is somewhat odd that the resolution figures stay exceedingly high at and beyond f/8. Diffraction effects should have a more deteriorating effect here. We suspect that f/11, as displayed by the camera, is really more like f/8 in the real world (interestingly we have seen this in another Sony 85mm lens as well).'
'It may come as a surprise to some but the Sony FE 85mm f/1.8 is tack sharp and it rivals the Zeiss Batis in this respect. The goodness starts straight at f/1.8. *The quality is already so high that the image quality doesn’t really increase anymore when stopping down. Interestingly the characteristic also mirrors the Zeiss lens at smaller aperture settings. Diffraction kicks in “too late”*'
And you can see their otherworldly f1.8 results at the above links. Unlike the GMs, neither Batis 85mm nor FE 85/1.8 has any asph surfaces so maybe it was decided to give them a leg up, for their intended user groups. The results OL obtained do not in any way match the actual MTF produced by Zeiss using their own optical bench - the industry standard. These OL test results are more like what Leica's SL primes deliver. It's not the end of the world, and their processing is clearly very sophisticated.
p.16 #16 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Was going to take photos of Asheville when I went up to check on my house and clean up the yard after power and water (not potable) had been restored. However, after driving around seeing all the places my wife and I frequented in the River Arts District and Biltmore Village were destroyed, I felt it was disrespectful to take photos. Thus, here are a few more from my trip to Chicago. I really like the 28-45.
p.16 #17 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Keunish wrote:
I agree, from my experience the focus distance plays a big role on the performances. By exemple my 35mm DG DN F2 was much sharper at long distance than at short distance, exactly like my 28-45mm wide open.
Also, I use the 28-45mm L-mount version on my S5II/S1 and I use the 35mm GM on my A7C/A7III. I maybe wrong but I always found the "sharpness" of my 35mm GM and 135mm GM a bit unnatural, like if the amount of details were not better than my best Panasonic/leica lenses but more like if a bit of extra digital sharpening was added to renforce the sense of resolution.
I only shoot in raw and I almost never change the default sharpening in LR. So I wouldn't be surprised if Sony adds some sharpening inside their raw files. And I can add that when watching some images samples from the 28-45mm and the S5II at 28mm at short distance or at long distance, I don't see a real difference with my copy, like on the Cameralabs samples on Flickr (the short distance flowers pic looks a bit soft)....Show more →
I think most zooms exhibit some performance degradation at MFD, of those I've tried only Sony ones didn't.
I don't remember if I have any Sigma shots @28mm and MFD but I'll check...
I don't think the lens is so bad but I have some focus inconsistency at f1.8 at short distance and the DOF looks very thin with this lens wide open, almost like the aperture was bigger.
p.16 #19 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Did you you autofocus or manual focus for these? The closer the focus point is to the lens, the less depth of field there is. So even with a small focus point, it's possible there are slight variations on "where" the camera focused. It may seem trivial on paper, but the closer the subject, the more this can affect what is in the plane of focus.
I have fence post shots taken at 28mm as well, and I don't notice a huge difference between f/1.8, f/2 and f/2.8. I did manually focus those test shots though.
Also, I believe the lens is a bit sharper at 35 than 28, so when reviewers said the zoom was as sharp as the GM 35 wide open, they might have been speaking about the zoom at 35mm.
In the end I don't know what to tell you. Most of my shooting is mid distance and further. In the cases where I have shot close focus, I wasn't disappointed with the results. But I also wasn't comparing it to my other lenses in the range. (Other than the fence posts shots earlier in the thread.)
p.16 #20 · Sigma 28-45mm f/1.8 DG DN Art Prime Comparison / Image Thread
Keunish wrote:
[...]Here is some comparisons processed with ACR sharpness at default (40) :
(images)
[...]I don't think the lens is so bad but I have some focus inconsistency at f1.8 at short distance and the DOF looks very thin with this lens wide open, almost like the aperture was bigger.
I know I expressed my opinion on the lens in post #6 above thinking there is a problem with the lens.
After having seen images I'm not that sure:
pair 1 (sponge) and 3 (doll with parrot) shows the problem you described. But, how close up were you when taking the images?
pair 2 (part of tree): can't really be used as the focus point has changed together with the aperture.
pair 4 (doll with parrot): The 35mm lens is probably not at all as close as the zoom.
pair 5 (landscape): f1.8 and f2 seem to be very similar
the last 3 images: to my eyes there is nothing strange going on there
It seems as you safely can use f1.8 unless you are very very close to the subject. Then everything is fine using f2. So, it is again proved that an image says more than a thousand words... I change my mind and think you can safely keep the lens.
PS: Looking at images at 150% in LR with default sharpening makes the water muddy. There also is a lack of fine details in the images of the doll. I've tried to make my best looking at these.
...
