Is my lens soft or DxO great?

[C2020]

I thought my Nikon 18-200mm VR took pretty sharp images until I started using Dxo Optics. Here is an example I took yesterday with my D80 and 18-200. The EXIF data is included. Post processing is so much sharper I'm concerned my lens is not as sharp as it could be. I would appreciate comments on whether it's my lens' softness or Dx0's greatness. This discrepancy seems to present frequently.

I've just ordered the new Sigma 18-50mm f/2.8 to help me shoot indoors with out a flash. If it turns out the 18-200mm is soft I'm may sell it and get a Nikon 80-200mm f/2.8.

Subscribe to see EXIF info for this image (if available)

 

[blw777]

I don't get it. DxO processed the images, and they look great compared to an unprocessed copy. What's wrong with that? Why would you conclude that the lens is "soft" from this experiment? Seems to me that if the results are there, the lens can't be at fault, right?

The 80-200/f2.8 is sharper than the 18-200, especially at 200mm, and at high shutter speeds where VR can't make a difference. If you processed shots from the 80-200, they would likely be sharper too.

The 18-200 should be the "softest" off all of the lenses in your .sig - but if this is how they come out, what's to complain about?

 

[C2020]

The difference is so marked in many cases it seems like the 18-200mm is not up to the task. I guess when the Sigma 18-50 f/2.8 arrives I'll have a better comparison with DxO.

We all want sharp photos. The less post processing, the more time to take the shots.

 

[Logan]

with an 11x zoom factor its not going to be the sharpest of the bunch, but when it comes to convenience, it cant be beat.

 

[Gale]

The image looks fine here
All images need some procesing

 

[davidwcramer]

Your exif data is limited. I take it this was shot using jpg, not raw? Most data can stand some post sharpening, unless you have it set to very high in camera. As Brian said above, if you had a soft lens, you wouldn't be able to get a sharp image with post processing.

 

[haze2]

I think your lens is fine. I was reading one of Scott Kelby's Photoshop books last night and he said flat-out, all digital pictures need sharpening.

 

[blw777]

The difference is so marked in many cases it seems like the 18-200mm is not up to the task. I guess when the Sigma 18-50 f/2.8 arrives I'll have a better comparison with DxO.

I think you're missing the point entirely. The only thing that matters is what you hang on the wall. You're getting what look to me like great results. How can you possibly think that "the lens is soft," ie it's not doing its job? If it's not up to the task, where did that image come from? All that information is in the image.

The state of the art in digital (and that means the world, not just Nikon) is that processing is part of the process. You will find that all of your other lenses also do better with postprocessing. That includes the legends, like the 70-200VR, the 28-70, the 200 Micro - all of them.

 

[weiran]

DxO adds the optimum amount of sharpening depending on the lens you have, the difference looks just like any unsharpened picture with a decent amount of unsharp mask applied to it.

Your Sigma 18-50 is likely to be a lot sharper than the 18-200 in the same range.

 

[C2020]

Thanks for the responses everyone, a lot of things now make more sense.

 

[SP77]

When I run auto-optimize in iView Pro I can get similar results vs unprocessed originals no matter the lens I'm using. Sometimes it looks better and other times it doesn't.

 

[jcovert]

Plus DxO is flat-out awesome. I'll be forever grateful for getting turned on to that secret weapon. I've been with PS since version 2.5 but I don't miss the work personally. I keep waiting for DxO to mess one up so I'll have to go fix it in Photoshop...and I'm still waiting.

 

[Harry Lavo]

It works wonders with my Sigma 18-200 as well....

DxO corrects for lens distortions as well as doing its own best guess at post processing. It can make quite a difference on raw files, as can be seen in the following comparison I have up on pbase:

http://www.pbase.com/hflphotos/dxocomparison

Just use the next/previous links to go back and forth. The roof is telling, as is are the fire excapes.

Moreover, the software lends itself to batch processing, which can save a lot of pp time.

Too bad their ordering system online is flawed and makes ordering more difficult than it should be.

 

[blw777]

Yeah, it looks like it'd be terrific, but I have a D100 and a D2h. So far they support a whopping total of eight lenses between the two cameras, of which I own a big fat zero. You'd think that as I have something north of 15 lenses, chances would be decent that one or another of them would be supported... *sigh*

Oh, and they only do JPEG on these cameras.

I guess I'll wait until I upgrade cameras to go back to their (rather interesting) web site... next year, perhaps?

 

[DABO]

I'm in the same boat. I have a D200, which is supported, but NONE of my most used lenses are supported (70-180/4-5.6, 35-70/2.8, 105/1.8, 50/1.2, 16/3.5, Sigma 100-300/4).

DAB

 

[SP77]

I just downloaded and tried out the DxO trial last night and HOLY SMOKES! I'm going to order this ASAP. I was considering upgrading to the 18-70 from my 18-55, but why bother? DxO fixes all my qualms with the 18-55 with its post-processing, and the 18-70 would probably only perform marginally better than the 18-55 anyways.

DxO is far more useful to me and only $160? A new 18-70 is $300, or $200 used on ebay. That's a no brainer. Plus I'm in the middle of reloading almost all of my albums on Fotki (putting up full-sized images) since we're decorating our house (just bought) and I'll be doing a lot of printing. Might as well run them all through DxO before uploading.

Looks like I'll need some more RAM too. :biggrin:

 

[weiran]

DxO is unlikely to support the older lenses you have DABO, especially the MF ones. Although if you poke them enough they may support the 100-300mm f/4, I put in a request for the Tamron 17-50 and Sigma 70-200 with the D80 and now they've both appeared on the latest release.

 

[SP77]

It won't address the specific optical flaws that track with the lens, but as long as you're using a supported camera there's still a ton that it'll do.

http://www.dxo.com/intl/photo/dxo_optics_pro/product_comparison

Check out the bottom chart.

 

[ultimind]

I got extremely ill when I attempted to use DxO the first time. It's UI experience on my brand spanking new Macbookpro was like trying to work with raw files on a 500mhz G3 iBook.

 

[Pa]

This is turning into a DxO thread, so I thought I'd chime in.

I had not heard of DxO until I read this thread and another one last night. I looked them up on the web and was impressed by what I saw, so I invested the time (and disk space) in the download for a 21-day trial. I have been trying it out some today, and so far I am underwhelmed. It does support the cameras and lenses I use most, but the automatically processed results fall far short of what I can do with Capture NX and/or Photoshop CS3.

Are you folks getting good results with the more manual approach with DxO?