[Top] [All Lists]

Re: [OM] Olympus 12mm is not 12mm?

Subject: Re: [OM] Olympus 12mm is not 12mm?
From: Moose <olymoose@xxxxxxxxx>
Date: Fri, 26 Jun 2015 17:25:17 -0700
On 6/19/2015 11:44 PM, C.H.Ling wrote:
Just found an interesting issue with the lens angle of view on the newer vs older M4/3 bodies and also Viewer 3 vs RawTherapee.


The angle of view is further widened on both images.

I previously posted some distortion correction images with the 12-50 and the Panny 12-32 lenses. I didn't point out the difference in AOV at the time, and it's not as obvious as it might be, as I sized everything the same. <http://www.moosemystic.net/Gallery/tech/DXO/LensCorrect.htm>

Here's a more real world sort of image (i.e. one I already had at hand.) All the versions retain their size relationship to the original Raw file, so you can see where there is expansion vs. compression. <http://www.moosemystic.net/Gallery/tech/DXO/_A001643_dist_corr.htm>

I made no effort to deal with differences in other aspects. May be unfair to RawTherapee, for example, with its blah rendition, 'cause I'm unfamiliar with its defaults.

As many of you know I prefer Olympus RAW converter but since E-P5 I found the image edges are very chunky, looks over sharpened with USM at high radius (say 1.5-2 pixels) even I set the sharpen to minimum. This forced me to look for other RAW converters. widened.

Further checking the images after this discovery, I believe Olympus has made the following process to its image on both JPEG and its RAW engine:

- linear distortion correction
- chroma and fringe adjustment
- crop the image heavy
- Resize the image to fit the pixel size, in that case enlarge the image to fit 
4608x3456 on the E-P5

4608x3456 is the JPG size. The Raw file size is 4640x3472. Both DxO and PtLens enlarge the image dimensions, not overall, but mostly in the corners as a result of barrel distortion correction.

The last process creates a big issue on the natural look of the image, which looks realy bad for pixel peeping. Now I understand why the images taken with my OM Zuikos looked much better than the 4/3 native lenses as no correction/cropping is done with the non native lenses.

I partially chose this image for the fine leaf detail in the corners. I have previously noticed some significant deterioration in the far corner of a SWA image corrected with PTL. In that case there was a beer/soda can on the ground which was distorted and softer after correction. In this sample, at 100%, the corner detail looks fine.

You are certainly correct that Oly is cropping the post correction image (and ACR is obviously using the same exact algorithms). However, I don't see where they would be up-sampling. I suspect the whole image is enlarged in the correction process (why?), then simply cropped to size.

Comparing the angle of view of my old 11-22 at 11mm and the 12-24 at 12mm, I believe the 12mm is much narrower than quoted on a E-P5 with Olympus Viewer. As you know even half a millimeter counted at wide angle, it is too bad to see the result like that.

This is a tricky business. There is sample variation, and as you point out, even small variations have a big effect on AOV. Modern Photo allowed FL variation of ±5% in their tests. Here are the results for some OM lenses:

       FL Spec.    FL Meas.  Var.
Zuiko    18/3.5    18.76     4.2%
Zuiko    21/2      20.15    -4.0%
Zuiko    40/2      41.33     3.3%
Zuiko    50/1.2    52.48     5.0%
Zuiko    50.1.4    50.94     1.9%
Zuiko    50/1.8    51.87     3.7%
Zuiko    90/2      90.44     0.5%
Zuiko    100/2    100.65     0.6%
Zuiko    180/2    180.57     0.3%

As you can see, the particular samples of the 18 and 21 mm lenses varied a lot, 
and in opposite directions.

       FL Spec.     AOV   FL Meas.   AOV
Zuiko    18/3.5    100.5    18.76    98.1
Zuiko    21/2       91.7    20.15    94.1

The assumed 9° difference in AOV is actually 4°. As a result, it might hardly 
be worth having both lenses!

The results with another pair might go the other way.

Comparing the spec. 11 mm of one zoom to the 12 mm of another can't be generalized; it only applies to those particular samples.

View Point Moose

What if the Hokey Pokey *IS* what it's all about?
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz