Sony 16-50mm image sharpness comparison on the Sony a6000

Raiatea Arcuri Photography · Jun 11, 2014

I've had the Sony a6000 for a few weeks now and it's met all me expectations. The autofocus is fast and it tracks very well. The high ISO performance is almost a full stop better than the NEX 6. It processes very fast and the viewfinder has no lag. It has lots of customizable buttons and the grip is a big upgrade over the NEX 5r.The only let down is the kit lens. It isn't the best and it's okay for general purpose use. I did a nice group portrait with 24 people at 32mm, f/5 and the sharpness let me down. I printed it on an 8x10 and the people on the edges weren't sharp. It was more than enough resolution to print, it's just that the edges weren't sharp.

I wanted to see just how sharp the lens was, so I went out and tested it at 16mm, 24mm, 35mm, and 50mm, with a range of apertures. I used a tripod with the timer in aperture priority mode @ISO 100 with manual focus to make sure the images were clean and that there was no shake. These were all shot in RAW and the only thing done to them was distortion and vignette correction.

Base image at 16mm.

Center 16mm.

Center edge 16mm.

Corner 16mm.

Base image at 24mm.

Center 24mm.

Center edge 24mm.

Corner 24mm.

Base image at 35mm.

Center 35mm.

Center edge 35mm.

Corner 35mm.

Base image at 50mm.

Center 50mm.

Center edge 50mm.

Corner 50mm.

Overall, this lens is a mixed bag. It might be sharper at a bigger aperture, or it might be sharper at a smaller aperture. I found that its sharpest point is at 35mm at f/8. When I use this lens I usually try to avoid 16mm because it has heavy distortion, vignetting, and it's very soft in the corners. Zooming to 18mm improves it a good amount. The 18-55mm kit lens has better sharpness and less vignetting.

Should I pay $150 more for this lens with the camera?

If you already have another lens you're happy with, than don't buy it. If you want a better mid-range zoom lens, than buy the 18-55mm, or the Zeiss 16-70mm, or the Sony FE 28-70mm.

If you’re looking at primes, I have the 50mm f/1.8 which is excellent for portraits. The 35mm f/1.8 is a good general use lens. The 20mm f/2.8 is much better than the 16mm f/2.8.

If you’re not going to print large prints or you’re using it for the web, than the 16-50mm is a good option. It's lightweight, compact, and is an okay all-purpose lens for travel.

If you have any questions or you would like me to do another comparison, please comment.

[My equipment: Sony a6000, Sony NEX 5r, Sony 18-55mm, Sony 16-50mm, Sony 50mm f/1.8, Sony 16mm f/2.8.]

WillMc · Jun 11, 2014

Thanks! Very useful.

joe6pack · Jun 11, 2014

Any movement in the vegetation will ruin your test.

blue_skies · Jun 11, 2014

Raiatea Arcuri Photography said:
I've had the Sony a6000 for a few weeks now and it's met all me expectations. The autofocus is fast and it tracks very well. The high ISO performance is almost a full stop better than the NEX 6. It processes very fast and the viewfinder has no lag. It has lots of customizable buttons and the grip is a big upgrade over the NEX 5r.The only let down is the kit lens. It isn't the best and it's okay for general purpose use.

You mention E1855 as better, but based on what?

Raiatea Arcuri Photography said:
I did a nice group portrait with 24 people at 32mm, f/5 and the sharpness let me down. I printed it on an 8x10 and the people on the edges weren't sharp. It was more than enough resolution to print, it's just that the edges weren't sharp.

Over-frame and crop is the better technique for such a shot, if using a 'compromised' lens.

Raiatea Arcuri Photography said:
I wanted to see just how sharp the lens was, so I went out and tested it at 16mm, 24mm, 35mm, and 50mm, with a range of apertures. I used a tripod with the timer in aperture priority mode @ISO 100 with manual focus to make sure the images were clean and that there was no shake. These were all shot in RAW and the only thing done to them was distortion and vignette correction.Overall, this lens is a mixed bag. It might be sharper at a bigger aperture, or it might be sharper at a smaller aperture. I found that its sharpest point is at 35mm at f/8. When I use this lens I usually try to avoid 16mm because it has heavy distortion, vignetting, and it's very soft in the corners. Zooming to 18mm improves it a good amount. The 18-55mm kit lens has better sharpness and less vignetting.

Not sure how these images tell you this: the corner performance is measured on an OOF area? If the AF point is where I think it is, then the corners are not even expected to be in focus.

A better approach for such a shot is a layered (multi-focus) approach.

If you want to check corner performance, at least chose a subject which has a reasonable change of getting its corners in the DOF and focal plane range.

Also, shoot multiple shots - I have seen that AF may vary, causing the corners to change dramatically.

Raiatea Arcuri Photography said:
Should I pay $150 more for this lens with the camera?

I think so - if you use the lens correctly, it is rather stellar.

Raiatea Arcuri Photography said:
If you already have another lens you're happy with, than don't buy it. If you want a better mid-range zoom lens, than buy the 18-55mm, or the Zeiss 16-70mm, or the Sony FE 28-70mm.

I don't see how/where the E1855, or the other zoom lenses, would have done better in your test - DOF and OFF is aperture/distance related, not lens related, so all these lenses would have failed the same test.

Raiatea Arcuri Photography said:
If you’re looking at primes, I have the 50mm f/1.8 which is excellent for portraits. The 35mm f/1.8 is a good general use lens. The 20mm f/2.8 is much better than the 16mm f/2.8.

And even the primes would have failed this test

Raiatea Arcuri Photography said:
If you’re not going to print large prints or you’re using it for the web, than the 16-50mm is a good option. It's lightweight, compact, and is an okay all-purpose lens for travel.

I would say that if you use it in daytime, stopped down, the lens is as good as higher IQ zoom or prime lenses.

If you use it wide open, relying on OSS and long shutter times, then I would say 'forget it' - you will get much better results with a fast prime - higher shutter speeds and lower ISO yield much better images.

Raiatea Arcuri Photography said:
If you have any questions or you would like me to do another comparison, please comment.

Sure, find a more reasonable subject for corner performance measurements, or calculate based on DOF and verify with a high IQ lens what your expected results are.

Raiatea Arcuri Photography said:
[My equipment: Sony a6000, Sony NEX 5r, Sony 18-55mm, Sony 16-50mm, Sony 50mm f/1.8, Sony 16mm f/2.8.]

Loan/borrow the E24Z or the Sigma 30mm and use that as a reference point. Then shoot the zoom lens at that same FOV. Use MF, and not AF, to assure that the AF point is the same in all shots. And use a tripod.

120 to 35 · Jun 11, 2014

This scene is not a good one for checking lens sharpness. The subject-to-camera distance varies a lot between the corner and the center, so the differences you see are due mainly to depth of field rather than lens sharpness. A picture of the front of a building that covers the top half of the frame and is taken with the camera pointing horizontally is a more suitable picture to check lens sharpness (using the top corners).

viking79 · Jun 11, 2014

Thanks, I think your samples do a good job depicting the lens's performance. Depth of field should be fine: at 16mm it has very deep depth of field, and at 50mm your subjects are further away.

I do notice on my test charts that the lens has substantial field curvature, so you might try focusing at the edge too, but overall I think your assessment is accurate.

Eric

tomtom50 · Jun 11, 2014

viking79 said:
Thanks, I think your samples do a good job depicting the lens's performance. Depth of field should be fine: at 16mm it has very deep depth of field, and at 50mm your subjects are further away.

I do notice on my test charts that the lens has substantial field curvature, so you might try focusing at the edge too, but overall I think your assessment is accurate.

Eric

Your test is very careful and I agree with Viking that the scene is appropriate to the depth of field. In answer to Viking's question about curvature of field, I have tried focusing nearer the edges when they count, and it doesn't work that well. The lens just doesn't have good edges and corners.

I personally like vegetation tests because it shows the level of detail. Building fronts are artificial and do not always have a smooth range of variation of detail, vegetation usually does.

I have the 16-50, I liked it until I started making bigger prints and found the Sigma 19mm & 30mm pictures just plain looked better. I also have an EOS-M, and its kit lens is a lot better than the 16-50, really very close to the Sigmas.

I got a great buy on an a6000, but the lack of a good kit lens is a real disadvantage and if the a6000 price were not so good I'd have got out of Nex. M43 has two (12-32 and 14-45 Panasonic). Fuji has at least one as does Samsung and Canon. Sony doesn't. I have borrowed the 18-55 and it also is not very good. For now if I want kit range (which is often, 28-80 35mm equiv is most of what I use) I will grab the EOS-M unless I need the speed of the Sony.

There is a reason why DPR's review of the a6000 put a relatively weak lens line as a disadvantage for the camera.

Raiatea Arcuri Photography · Jun 11, 2014

Thanks Henry, I will keep your information in mind.

cxsparc · Jun 11, 2014

I disagree, in my opinion this is an elaborate, but flawed test in many ways:

1. For making any verdict based on this arbitrary scene, showing the results of another lens to compare is fundamental, yet missing here. This makes the statement regarding 1855 unsubstantiated.

2. No information on whether oss was turned off.

3. A uniform green scene is a challenge for the Bayer sensor and pp with low contrast to start with. Its also unrealistic because you usually couldn't care less for the details in the grass. Also, the light is so diffuse you can't see any shadow, making this a very low contrast scene.

4. The other major flaw is taking the crops at different depths of the scene. This is ridiculous because we usually desire a more or less flat focus plane. And the lens designers at least attempt to make it like that. Taking the crops at these different depths shows the effect of the aperture and focal length on the dof and not off the lens quality.

In contrast, my tests of the 1650 were done on a large subject parallel to the focus plane, with crops taken at approximately similar depth and most important, with other lenses in the same setup to compare. You can find those on my blog, download and check for yourself:
http://www.dpreview.com/forums/thread/3671891

Not surprisingly, I have come to different conclusions. Not precluding sample variation, my 1855 was retired immediately

.
My 1650 yields IQ similar to the 1670 when stopped down. In the blind quiz I posted recently in this forum, the majority even mistook the 1650 shots for the Zeiss:
http://www.dpreview.com/forums/thread/3674590

Yes, the 1650 has its shortcomings:
The extreme corners never get perfect at lower focal lengths. Wide open, the IQ is not great. And its powerzoom. And it lacks a hood.

But in practice, I enjoy this lens very much for its sharpness and clarity in the central part of the image, the 16 mm option and its super compactness. This lens shows the progress in lens design at Sony in the last years.
--
Nex-Blog: http://luxorphotoart.blogspot.de/

GaryW · Jun 12, 2014

cxsparc said:
I disagree, in my opinion this is an elaborate, but flawed test in many ways:

What if I also did a similarly-flawed test, but came to the same results as you did?

cxsparc said:
1. For making any verdict based on this arbitrary scene, showing the results of another lens to compare is fundamental, yet missing here. This makes the statement regarding 1855 unsubstantiated.

Yeah, you really need to take two (or more!) lenses and swap 'em at the same scene, under the same conditions.

cxsparc said:
2. No information on whether oss was turned off.

I forgot to do that in my tests. Hmm.

cxsparc said:
3. A uniform green scene is a challenge for the Bayer sensor and pp with low contrast to start with. Its also unrealistic because you usually couldn't care less for the details in the grass. Also, the light is so diffuse you can't see any shadow, making this a very low contrast scene.

Since there are more green pixels, you'd think that green scenes would do better?!

cxsparc said:
4. The other major flaw is taking the crops at different depths of the scene. This is ridiculous because we usually desire a more or less flat focus plane. And the lens designers at least attempt to make it like that. Taking the crops at these different depths shows the effect of the aperture and focal length on the dof and not off the lens quality.

Yeah, having a scene with depth is problematic, but sometimes you have to go with what you have.

cxsparc said:
In contrast, my tests of the 1650 were done on a large subject parallel to the focus plane.....

Not surprisingly, I have come to different conclusions. Not precluding sample variation, my 1855 was retired immediately .
My 1650 yields IQ similar to the 1670 when stopped down. In the blind quiz I posted recently in this forum, the majority even mistook the 1650 shots for the Zeiss:
http://www.dpreview.com/forums/thread/3674590

I have had similar results, with the 16-50 often looking better than the 18-55. I've been posting about this for a while now, with a lot of people saying it can't be so. I know people really want an expect the 18-55 to be better -- I expected it as well -- but it's not, at least not at all focal lengths.

cxsparc said:
Yes, the 1650 has its shortcomings:
The extreme corners never get perfect at lower focal lengths.

True. And I couldn't get really good results at 20mm. Which is weird, as I've had decent results at 16mm (at least in the center, and with close subjects).

But at 30mm, it was as good as the 18-55, I think, and much better at 50mm. This seems to be in complete contrast with the review on DxOMark, so I don't understand.

cxsparc said:
Wide open, the IQ is not great. And its powerzoom. And it lacks a hood.

Both kit lenses do better at f5.6 and beyond.

cxsparc said:
But in practice, I enjoy this lens very much for its sharpness and clarity in the central part of the image, the 16 mm option and its super compactness. This lens shows the progress in lens design at Sony in the last years.

It seems to me that the designers know how to make something work for the real world, with center sharpness being more important than corner sharpness, compact size, fast focusing, etc. All of this in a lens you pay just $150 more for when you buy the camera. Seems like a deal to me.

And if the lens really is so bad, you'd think there'd be more comparisons to show it.

cxsparc said:
--
Nex-Blog: http://luxorphotoart.blogspot.de/

123Mike · Jun 12, 2014

What a totally awesome post you just made! It will last forever. I'm going to do a similar test, following your model of presenting, and pick the 18-55, 18-105, and then some unusual choices like Rokinon 8mm, and SAM 35/1.8 using a focal reducer and also Minolta 28-85 and 24-85 RS using a focal reducer and also without focal reducers! I'll be back!

As for spending $150 on this kit lens..... *NO* ! Apply this money to the 18-105 lens instead! But... time for some tests !

123Mike · Jun 12, 2014

blue_skies said:
You mention E1855 as better, but based on what?

Isn't it well known by now though, that the 18-55 is higher quality than the 16-50 at any focal length? Check out the reviews on Kurt Munger for one. The 16-50 needs all kinds of stopping down, whereas the 18-55 looks a lot better without having to stop down!

mailman88 · Jun 12, 2014

.....A6000 Brain Surgery is not needed, right? Crop a A6000 pic at 200% and you want better results....Won't happen.

mailman88 · Jun 12, 2014

I grant you, you need a better lens. But... just look at the following threads from the A6000 with better glass...stunning results!

http://www.dpreview.com/forums/post/53630702

http://www.dpreview.com/forums/post/53642759

Raiatea Arcuri Photography · Jun 12, 2014

I forgot to mention, these are 100% crops.

hector69 · Jun 12, 2014

Excellent review. I did similar tests of the 16-50mm vs. 18-55mm and the latter is noticeably sharper. Unfortunately, I bought the SELPZ 16-50mm for my A6000 based on other reviews in this forum. Now I am disappointed because I think maybe the AF speed is a bit better than the 18-55mm on the A6000. I am now using again the sharp Sigma 30mm f2.8 as a compact altertnative to the kit lens. Altough it is a prine lens, it is priced similarly as compared with the 16-50mm and with f2.8 it is a better lens for the A6000 except for video (no OSS). I will use also the old 18-55mm sometimes because it seems sharper thn the PZ 16-50. The Zeiss 28-70mm is too expensive foe most amateur photographers like me.

123Mike · Jun 13, 2014

hector69 said:
Excellent review. I did similar tests of the 16-50mm vs. 18-55mm and the latter is noticeably sharper. Unfortunately, I bought the SELPZ 16-50mm for my A6000 based on other reviews in this forum. Now I am disappointed because I think maybe the AF speed is a bit better than the 18-55mm on the A6000. I am now using again the sharp Sigma 30mm f2.8 as a compact altertnative to the kit lens. Altough it is a prine lens, it is priced similarly as compared with the 16-50mm and with f2.8 it is a better lens for the A6000 except for video (no OSS). I will use also the old 18-55mm sometimes because it seems sharper thn the PZ 16-50. The Zeiss 28-70mm is too expensive foe most amateur photographers like me.

Actually, for video, the A6000 uses digital stabilization, which, for video, is MUCH more effective than OSS is. It' s possible that the A6000 does both at the same time, I don't know. But what I do know is that digital stabilization which is the scrolling around and lining up of frames, works very well.

The A3000 does not do digital stabilization for instance, and it's up to the lens' OSS to stabilize it. It's not very effective at all, not for video at least. Way too easy to get all shaky. My previous A57 did digital stabilization and was WAY more stable. Before that the A33 tried to stabilize using IBIS, and that was also awful. But the A6000 when you start the video, you'll see a crop happening. It zooms in a bit, to create room around the edges so that it can start scrolling around the frames.

So for video, the 30mm Sigma lens should be fine. Although focusing won't be PDAF. But that's another whole matter: does video even use PDAF? With video it's hard to discuss things, but usually when you start talking about the nitty gritty details, not a whole lot of people have anything to say about it. I've tried it multiple times, and the feedback was mostly crickets....

cxsparc · Jun 13, 2014

Seems there is sample variation, because my 1855 is significantly worse then my 1650.

Comparing to the 18105 is debatable, because the thing with the 1650 is that it is inexpensive, small, and adds 16mm to the table - and boy, that 2mm is a big difference in photography. And my copy is very good at f5.6-f8.

123Mike · Jun 13, 2014

cxsparc said:
Seems there is sample variation, because my 1855 is significantly worse then my 1650.

Comparing to the 18105 is debatable, because the thing with the 1650 is that it is inexpensive, small, and adds 16mm to the table - and boy, that 2mm is a big difference in photography. And my copy is very good at f5.6-f8.

You can't compare the 18-105 to either of them.

According to tests, the 16-50 needs stopping down for it to become sharp, pretty at least some, at all focus lengths, whereas the 18-55 can be sharp wide open.

But it's not the first time I've heard of copy variations... Apparently, there is also a rumor that the 18-55 was/is made in two places. Taiwan and somewhere else? China probably. If that is true, then perhaps that's what those 18-55 differences are all about...

For wide angle, the Rokinon (Samyang) 8mm f/2.8 is totally cool! I just got a used one, and holy crap it's sharp! It actually offers a really weird proposition. In theory you could stop bothering zooming in at all, and just crop and defish after the fact... Well that can't work always of course, but walking around on the street for instance, it could! It's WAY sharper than I expected.

cxsparc · Jun 13, 2014

123Mike said:
cxsparc said:

Seems there is sample variation, because my 1855 is significantly worse then my 1650.

Comparing to the 18105 is debatable, because the thing with the 1650 is that it is inexpensive, small, and adds 16mm to the table - and boy, that 2mm is a big difference in photography. And my copy is very good at f5.6-f8.

Click to expand...

You can't compare the 18-105 to either of them.

According to tests, the 16-50 needs stopping down for it to become sharp, pretty at least some, at all focus lengths, whereas the 18-55 can be sharp wide open.

But it's not the first time I've heard of copy variations... Apparently, there is also a rumor that the 18-55 was/is made in two places. Taiwan and somewhere else? China probably. If that is true, then perhaps that's what those 18-55 differences are all about...

For wide angle, the Rokinon (Samyang) 8mm f/2.8 is totally cool! I just got a used one, and holy crap it's sharp! It actually offers a really weird proposition. In theory you could stop bothering zooming in at all, and just crop and defish after the fact... Well that can't work always of course, but walking around on the street for instance, it could! It's WAY sharper than I expected.

I really wonder what 1855 lens you are talking about, sharp wide open? No way

At least not mine.

Sony 16-50mm image sharpness comparison on the Sony a6000

Member

Attachments

Well-known member

Senior Member

Forum Pro

Senior Member

Forum Pro

Senior Member

Member

Senior Member

Forum Pro

Senior Member

Senior Member

Veteran Member

Veteran Member

Member

Well-known member

Senior Member

Senior Member

Senior Member

Senior Member

Similar threads

Keyboard shortcuts