Does DXO Optics turn Sony's average kit and cheap zooms into superb performers on par with Zeiss?

Started 5 months ago | Discussions thread
blue_skies
Senior MemberPosts: 7,507Gear list
Like?
Re: Use lenses for what they are good at, postprocessing makes everything better
In reply to ProfHankD, 5 months ago

ProfHankD wrote:

40daystogo wrote:

From DXO's marketing literature, it claims that DXO Optics software can compensate even for corner softness, and other lens aberrations. Now, on these forums, the Sony A6000's 16-50 kit lens, and the reasonably priced 55-210 zoom get a lot of criticism. So my question is, when DXO's software tools work their magic on the RAW files from these so-so lenses, what level does it bring it up to? To the level of the Zeiss lenses? What I'm asking is, how much of a boost can be achieved by software alone, rather than spending massive bucks on expensive Zeiss lenses.

More sophisticated digital processing will get more and more out of imperfect (i.e., all real) lenses and DxO has proven that they are very serious about camera calibration and computational use of that data; I think they are very good and will only get better.

As for the Zeiss lenses, well, there are a few really amazingly good lenses for E-mount now, and even they will do still better using postprocessing like DxO provides. However, the truth is that most lenses are pretty good if used at their best aperture and zoom focal length -- including the much-maligned E-mount kit zooms. If we ignore noise, it is theoretically possible, but software is unlikely to ever bring IQ of a kit zoom wide open at it's worst focal length up to that of the best lenses wide open (with similar postprocessing). The question should be: do you need to shoot at settings that will make the flaws in the kit zooms unacceptably visible?

I shoot mostly with old manual focus lenses and rarely do calibrated corrections when shooting for non-technical reasons (as opposed to my computational photography research). In fact, I usually use the OOC JPEGs although I usually shoot with JPEG+raw so I have the option of fixing things if needed. The catch is that I have a lot of lenses, so I pick carefully and avoid using lenses for what they are not good at. In sum, take a look at what the kit zooms can do well, which is quite a lot when stopped down a couple of stops. If that covers your needs, don't worry about better glass... if not, that will tell you what other glass to consider buying.

One last note: new postprocessing will improve old captures too. Shots I took with a Canon G1 more than a decade ago look much better when I postprocess the raws using modern software.

I agree, each lens has its 'sweet' spot, and if used at those settings, such lenses perform admirably.

A lot of the discussion is related to wide-open and wide-angle lenses, with both effects being visible in the final IQ to the naked eye (full image view) without post-processing. But both wide-open and wide-angle bring other issues (ISO, SS, composure, subject, angle/framing, and so on).

In general, the IQ of the images post post-processing is much closer than the IQ of the images prior to post-processing.

If you only shoot JPG, you rely on the camera's choices (or menu settings), whereas in post-processing you can optimize for the format of choice (e.g. monitor use only, increased anti-aliasing, contrast, acutance, color tone). If printing, different optimizations should be chosen.

Once you make such adjustments, the IQ of a lens is mostly relegated to pixel peepers (my corner is sharper than your corner) - but the fact is, when displayed on a monitor, without zooming in, this hardly matters.

DxO software 'smartly' combines several options that do yield to great results (e.g. their PRIME NR algorithm), but there is nothing that you cannot achieve with sitting behind your favorite post-processing tool and going step-by-step.

A higher IQ lens gives you a better starting point, imho, and with a less IQ lens you may have to do more work (or even crop) to get a high IQ result. But the way in which we use most images, the IQ difference is not a deal breaker.

If you are creating stock footage, paid services, or large print, the difference is (can be) substantial. For most of us it is not such a big deal (I think).

One example that comes to mind now I am comparing the E16+UWA and the Touit 12 side by side: the Touit is much sharper lens, but when using it at f/2.8, the DOF limits do show up, and foreground/background softening is visible. Stop down to a better f-stop, e.g. f/5.6, and the image is a lot better. But so is the E16+UWA image. Push both images through post, and this magical 'large' IQ difference becomes rather non-meaningful.

But, we do shoot at f/2.8 sometimes, and then it does matter, so better IQ always wins.

If you use a tripod, static subjects, and use lenses at their 'optimum', the IQ difference are far less, and post-processing can bring them very close, I think.

-- hide signature --

Cheers,
Henry

 blue_skies's gear list:blue_skies's gear list
Canon PowerShot S95 Sony Alpha NEX-7 Sony Alpha NEX-6 Sony Alpha 7 Sony a6000 +34 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