A77: why long lenses seem to be so out resolved?

Started 6 months ago | Discussions thread
JimmyMelbourne
Senior MemberPosts: 1,732Gear list
Like?
Re: A77: why long lenses seem to be so out resolved?
In reply to VirtualMirage, 6 months ago

VirtualMirage wrote:

JimmyMelbourne wrote:

Hi Paul

I certainly understand and appreciate what you have written, and also share this view. The only reason I raise this is we see a lot of longer zoom lenses that have been tested on the a77 rating low. If it was a case of one or two bad lens design, i can appreciate that, but when we see across the board low scores specific to the a77, it makes me wonder why, and if there is something specific to the a77 that may lead to this trend. Hence my original post.

But it isn't the A77. Look at the same lenses on other camera models, Sony including. You don't see any of the other models scoring higher than the A77, do you? See here:

http://www.dxomark.com/Lenses/Compare/Side-by-side/Sony-70-200mm-F2-8-G-on-Sony-SLT-Alpha-77-versus-Sony-70-200mm-F2-8-G-on-Sony-Alpha-550-versus-Sony-70-200mm-F2-8-G-on-Sony-Alpha-700___246_734_246_624_246_562

For some reason this link does not work for me.

The A77 is 24MP, the A550 is 14.2MP, and the A700 is 12MP.

Now let's compare the same lens with the A77, A900, and A99:

http://www.dxomark.com/Lenses/Compare/Side-by-side/Sony-70-200mm-F2-8-G-on-Sony-SLT-Alpha-77-versus-Sony-70-200mm-F2-8-G-on-Sony-Alpha-900-versus-Sony-70-200mm-F2-8-G-on-Sony-SLT-Alpha-99___246_734_246_371_246_831

This link is OK.

The conclusion to come to that is it is the lens is more of a limiting factor than the sensor.

I know this is the primary principle, and this is the underlying principle that DXO is trying to prove with their numbers.

So basically what it all means to me is that we have 24 meg APS-C sensors, but will never realise their full potential. We may as well be shooting with zoom lenses and 10 Megapixel sensors that work better together. Or alternatively shoot primes to get the closest match as possible.

Here's the Sigma 35mm F/1.4 A HSM on the A77, D7100 (no OLPF), and D5200 (with OLPF) (all 24MP):

http://www.dxomark.com/Lenses/Compare/Side-by-side/Sigma-35mm-F14-DG-HSM-A-Sony-on-Sony-SLT-Alpha-77-versus-Sigma-35mm-F14-DG-HSM-A-Nikon-on-Nikon-D7100-versus-Sigma-35mm-F14-DG-HSM-A-Nikon-on-Nikon-D5200___1060_734_1057_865_1057_850

This link does not work either.

Remember that the strength of the aliasing filter can affect perceived sharpness.

The only thing that I keep noticing on DxO's site is the tests tend to consistently show the Sony's as being weak in the corners. And that isn't with just the A77, it is with all the Sony APS-C sensors. I can see that being more problematic with the E-mount cameras and having such a short flange distance, requiring some optimized micro lenses on the sensor, but not so much the A-mount. It makes me question how and what DxO is testing and if they do sample variation. Until I can see other sites or testers verify this, I cannot conclude that DxO's results are 100% correct. They do give us a good start and give us an idea of some expectation, but that is about it. Error tolerances, testing methods, equipment used, etc. are all factors that can sway their results where the same lens on the multiple cameras could score close to each other.

So again, you are looking at it from the wrong perspective and assume the problem is the A77.

I am not assuming something is wrong with the a77. I am testing whether the 24 Megapixel sensor and camera implementation is contributing to the lack of resolution a lens can achieve with this sensor, particularly what was implemented in the a77. When we place a lens on a camera, the sensor is not the only hard variable. There is other hardware, firmware that come into play.

As mentioned, I understand the reason for the numbers, what DXO is trying to achieve by comparing lenses and their resolvability on certain sensors. What this process does not factor in is the variability on the camera side. Then again as you have mentioned it does not account for variation in lens quality control either.

What you really need to be asking are the following:

  • Why do the DxO tests consistently show the Sony APS-C sensors testing worse in the corners when comparing results of the same lens across multiple 1.5x APS-C sensors?
  • Why do the DxO tests consistently show the Sony APS-C sensors testing better in light transmission when comparing results of the same lens across multiple 1.5x APS-C sensors?
  • Why do the DxO tests consistently show the Sony APS-C sensors testing better in vignetting when comparing results of the same lens across multiple 1.5x APS-C sensors?
  • If the above holds true, does Sony do something different with their sensor microlenses that hurts corner sharpness in exchange for improved vignetting and light transmission?
-- hide signature --

Paul

 JimmyMelbourne's gear list:JimmyMelbourne's gear list
Sony Alpha DSLR-A300 Sony SLT-A77 Sony SLT-A37 NEX5R Sony 70-200mm F2.8 G +18 more
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow