Olympus mZD 9-18mm f4-5.6 Experience Diary - Page 4

Amin Sabet · #31 July 26th, 2010, 07:09 AM

Quote:

Originally Posted by Brian Mosley

No distortion correction applied to either raw file, CA correction applied to the mZD 9-18, clicked on Auto levels on both, adjusted sharpening to bring more detail out of the mZD 9-18.

My point is that if you're interested enough in pixel level detail, using the right tools - the difference between the two lenses can be much less than it seems over there.

Distortion correction has been automatically applied to both RAW files, and like Lightroom, C1 doesn't give you the option of disabling that distortion correction, which I think is a shame.

If you are going to selectively give the Olympus image more sharpening, might as well not even show the Panasonic crops. It invalidates the comparison. Using selective processing, one can make just about any two decent lenses look like similar performers.

Brian Mosley · #32 July 26th, 2010, 07:19 AM

Amin, Capture One 5 PRO has a Distortion Correction slider with enable/disable option... I had this option disabled for both raw files.

Why would you do a pixel level comparison without CA correction, when it's easily available?

Cheers

Brian

Amin Sabet · #33 July 26th, 2010, 08:57 AM

Quote:

Originally Posted by Brian Mosley

Amin, Capture One 5 PRO has a Distortion Correction slider with enable/disable option.

Lightroom 3 has the same thing, and it is equally nonfunctional. With the LX3, Panasonic made Adobe disable the ability to turn off distortion correction in ACR/LR (1,2). I suspect a similar thing is going on here. The disable option is nonfunctional. If you process the same RAW files using dcraw or Raw Developer, you'll see the actual distortion, which is marked. I showed this in the review at S.C.

Quote:

Originally Posted by Brian Mosley

Why would you do a pixel level comparison without CA correction, when it's easily available?

The same reason I show RAW conversions and not in-camera JPEGs. At S.C., the main emphasis has always been on the raw potential of the files coming off the sensor. There are many sites doing the opposite, focusing on the corrected files. I assume that people coming to S.C. for the tests know how to address CA in their own workflow. For some, this is not done in the RAW conversion but afterwards, and those folks may want to know that the CA is there needing that extra step.

The day will come soon when these CA and barrel distortion corrections are baked irreversibly into the RAW files. At that time, this discussion will be irrelevant. Until then, I think that those looking to get the most out of their gear will want to know exactly what is getting captured.

Quote:

Originally Posted by Brian Mosley

My point is that if you're interested enough in pixel level detail, using the right tools - the difference between the two lenses can be much less than it seems over there.

If the goal is to show that image quality is good enough with the Olympus - and I believe that it is and said so in my review - no comparison crops are needed are needed. However, if you are trying to give insight into the difference between the two lenses, why sharpen one of their files more than the other? Why omit the two corners which show the greatest difference?

Using the right tools, I can show that a lens is much different than itself. Sharpening makes all the difference.

Djarum · #34 July 26th, 2010, 10:28 AM

Since this this discussion is more developed here, maybe I'll ask this. Distortion correction aside, looking at the original RAW files, I notice a few things. First is lateral CA with the Olympus. Now I understand that in camera jpeg Panny cameras will fix purple fringing, but this is different than latteral CA. Does panny fix lateral CA too in camera? Secondly, I also notice that there is a slight exposure difference even though that the shutter and aperature are the same. Is this due to processing or contrast differences in the actual lens?

I think review sites in general are going to have to go to two different approaches on judging lenses. The first is judging the standard jpeg, since many shooters shoot jpeg, and all they care about is jpeg out put. Then I think they are going to have to go to a raw conversion that does not apply any sort of corrections what so ever. Its a shame that certain raw converters are forced to apply correction settings by the manufacturer to convert the raw files. Why can't there be a universal standard for converting these files?(one reason I don't do RAW normally). Geesh, and here I thought the professional video industry had issues with SMPTE.

On edit, I do have a question. Does the lateral CA go away in the Olympus standard jpeg output?

Amin Sabet · #35 July 26th, 2010, 11:44 AM

Dj, I virtually never shoot JPEG and so don't know the answer to your questions about in-camera correction of color fringing. I believe (but am not sure) that the Panasonic cameras fix lateral CA on in-camera JPEGs.

The differences in apparent exposure may be due to: 1) less light falloff with the Olympus lens, a nice characteristic of this lens; 2) changing light conditions (modestly changing cloud cover). If you look at the EXIF on the RAW files, the exposures were not totally constant. Some f/8 images were 1/160s, others 1/200s.

The reason I didn't compare sharpness on crops without distortion correction is that few people will use these lenses without distortion correction. I can process the files in Raw Developer without distortion correction, and the level of detail goes up, but as soon as you correct the distortion, there is a price to pay in resolution. I still think it's a shame that Lightroom, C1, and others don't let you shut off the distortion correction, because sometimes one wants that little bit of extra wide angle effect and maximum resolution.

Djarum · #36 July 26th, 2010, 12:50 PM

Thanks Amin for the candid response.

I'm not a wideangle person mainly due to the distortion issues with wideangle lenses. I was suprised to see the level of lateral CA though in the Olympus lens.

Brian Mosley · #37 July 29th, 2010, 05:20 PM

Taken this evening, scouting for our upcoming Safari Group event on the 21st August...

Giant ferris wheel, Sheffield city centre.

E-P2 + mZD 9-18mm f4-5.6
1/800s f/4.5 at 9.0mm iso200

Cheers

Brian

Amin Sabet · #38 July 29th, 2010, 05:52 PM

Very nice. Makes me a little dizzy like a Ferris wheel should.

PeterB666 · #39 July 29th, 2010, 10:40 PM

I have no problems with image quality from the 9-18mm lens. It certainly is a compact and convenient lens.

I have struggled to get anywhere near the detail out of MFT compared to the Nikon cropped sensor cameras and full-frame Canon cameras that most of the people I shoot with manage. Part of that is my processing of images. I have gone part way there by changing to PhotoShop as my main editor and being more careful about destructive processing. Ultimately, the MFT and FT sensor is smaller than your average dSLR and Olympus and Panasonic need to do more work to bridge the gap.

Then again, it is all about compromise and I do appreciate the small size.

Djarum · #40 July 30th, 2010, 11:42 AM

Maybe in a different topic, but I'm not sure how detail is related to the actual size of sensor. Detail relates to the lens and number of pixels in the sensor.