The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Fun with Olympus EM5 Mark ii 40mp only

scho

Well-known member
Sandy's AccuRaw Monochrome will also now process the HR ORF files from the E-M5II. Default output file size is 9274x6932 pixels. Nice option if you want high resolution monochrome files. Here is a sample image taken awhile ago and processed through AccuRaw Monochrome. Click for larger file that is a 25% linear reduction of the original.



Two more from AccuRaw Monochrome with some toning.



 
Last edited:

scho

Well-known member
LR CC now supports E-M5II ORF and HR ORF files. Also, some nice new features for merging files to stacks, HDR, and panos. Here I used 3 HR ORF files from the EM5II + 75mm in LR CC to create a focus stack in Helicon Focus without having to use intermediary TIFS. Click for full size 64 MP image.

 

scho

Well-known member
Two HR ORF files merged in Helicon focus. EM5-II+12-40. Evening shot of rock garden. Click for full size 64 MP image.

 

Knorp

Well-known member
Hi Carl,

first, thanks for posting these examples :thumbup:
Now that shot with the piece of wood is pretty clean, whereas the rock garden is very noisy.
Also artefacts are found all over the image.
Any thought what may have caused the difference between these two shots ?
Low light ? Some wind ? Other processing ?

Kind regards.
 

scho

Well-known member
Hi Carl,

first, thanks for posting these examples :thumbup:
Now that shot with the piece of wood is pretty clean, whereas the rock garden is very noisy.
Also artefacts are found all over the image.
Any thought what may have caused the difference between these two shots ?
Low light ? Some wind ? Other processing ?

Kind regards.
Hi Bart,

Thanks for the comments. The differences between the two images are lens (75 vs 12-40), aperture (f/5.6 vs f/8), and processing of the rock garden image included some contrast adjustment with Topaz Adjust. It is probably the latter that is causing what you refer to as noise and artifacts. There are no motion artifacts (diagonal line patterns) that I could see.
 

scott kirkpatrick

Well-known member
Two more 64 MPx example files to play with

My first tries, medium weight tripod, 4 sec delay before each shot:



and a long shot -- a Moire tester



The originals and various full size (30 MB) jpegs are all here.

Other files: ...i.jpg means that I rendered the 16 MB first frame of the series. And ...OLY.jpg is the in-camera processed 40 MPx jpg that Olympus provides. The picture of the building has Moire on the window sun-grilles when viewed at 100%. It's faint but visible in the jpegs which I developed in Capture One or in the AccuRaw results from Sandy McGuffog, but it is strong and accompanied by a "mosaic" or "stairstep" artifact in the in-camera 40 MPx jpeg.

Think about that for a moment. In the course of taking the 8 exposures, each pixel location has had all four of the Bayer filters in front of it for a single exposure (and the same applies to the second group of four shots, displaced by half a pixel-spacing in the x and y directions. So this has become in principle a Foveon-like image, with all three colors measured at every pixel. So why is there any Moire at all?

Maybe I've got the wrong impression about where the chip moves for its 8 shots. Or maybe this is because, in order to allow fairly standard rendering programs, which expect to see a Bayer array image, Olympus has had to translate their lovely Foveon style image back into a virtual Bayer array with finer spacing, which can then indeed have Moire! Sandy tells me that the big 100 MB .ORF files are coded to represent a 64 MPx array of pixels with standard RGGB Bayer color filters.

scott

Both pictures with 75/1.8@f/8 ISO 200
 

scho

Well-known member
Re: Two more 64 MPx example files to play with

My first tries, medium weight tripod, 4 sec delay before each shot:


Other files: ...i.jpg means that I rendered the 16 MB first frame of the series. And ...OLY.jpg is the in-camera processed 40 MPx jpg that Olympus provides. The picture of the building has Moire on the window sun-grilles when viewed at 100%. It's faint but visible in the jpegs which I developed in Capture One or in the AccuRaw results from Sandy McGuffog, but it is strong and accompanied by a "mosaic" or "stairstep" artifact in the in-camera 40 MPx jpeg.

Think about that for a moment. In the course of taking the 8 exposures, each pixel location has had all four of the Bayer filters in front of it for a single exposure (and the same applies to the second group of four shots, displaced by half a pixel-spacing in the x and y directions. So this has become in principle a Foveon-like image, with all three colors measured at every pixel. So why is there any Moire at all?

Maybe I've got the wrong impression about where the chip moves for its 8 shots. Or maybe this is because, in order to allow fairly standard rendering programs, which expect to see a Bayer array image, Olympus has had to translate their lovely Foveon style image back into a virtual Bayer array with finer spacing, which can then indeed have Moire! Sandy tells me that the big 100 MB .ORF files are coded to represent a 64 MPx array of pixels with standard RGGB Bayer color filters.

scott

Both pictures with 75/1.8@f/8 ISO 200
Hi Scott,

Results are looking good. I downloaded your distant buildings ORF file and processed through both LR and Irident. Similar results to what you obtained with C1 and AccuRaw. Screen grab comparison below with Irident left and Lightroom right. Looks like a bit less moire in the Irident file, but some strange demosaicing artifacts that look similar to what I've seen before from Fuji X files.
Click for full size file.
 

sandymc

New member
Re: Two more 64 MPx example files to play with

Hi Scott,
Results are looking good. I downloaded your distant buildings ORF file and processed through both LR and Irident. Similar results to what you obtained with C1 and AccuRaw. Screen grab comparison below with Irident left and Lightroom right. Looks like a bit less moire in the Irident file, but some strange demosaicing artifacts that look similar to what I've seen before from Fuji X files.
Scott, Carl,

FYI, this is what the same section looks like processed via AccuRaw (at default settings)

Sandy
 

scott kirkpatrick

Well-known member
This is just a test. Bookshelf at 2 m, 75/1.8 @ f/5.6. The full image:



and an excerpt at full size (1200 pixels wide)



The .ORF, and jpegs from the big ORF, the first frame of the 8 and one courtesy of Olympus are here.

(The bookshelf slopes, not the camera.)

scott

P.S. I shot first at f/8, but there was a loss of contrast due to diffraction. From 2.8 to 5.6 was equally sharp. At 1.8, there was some loss of resolution, probably due to depth of field -- the books aren't lined up very well.
 
Last edited:

scott kirkpatrick

Well-known member
Here's another stress test:



It's a huge cemetery on the western approaches to Jerusalem. I may need to get an even sturdier tripod. This 8-shot HR composite was only slightly sharper then the first frame taken by itself (and of course the moving branches were not a problem in the individual frame).

OM-D E-M5.2 12/2.0 @ f/8 processed in C1 8.2.2
The original files will be found here


scott
 

scho

Well-known member
Boathouse window. A relatively safe subject for the E-M5II in HR mode:rolleyes:. E-M5II+12-40.



Another window HR shot with embedded link to full size (52 MP) version.

 
Last edited:

scott kirkpatrick

Well-known member
Here's one with enough detail to make 64 MPixels work a bit. First the easy-downloadin' version:



and then a link to the RealThing.
Note the topless runner in orange shorts appearing in two positions on the field.

The full .ORF, the jpg created by Olympus, and jpgs rendered by CaptureOne 8.2.2 from the first frame and from the 64MPx composite can all be found here.

scott

Oh, 40-150/2.8 @f/5.6. Distance about 500-600 m.
 

Knorp

Well-known member
Hi there Scott,

Here's one with enough detail to make 64 MPixels work a bit. First the easy-downloadin' version:
This "easy-downloading" version (still 2,400×1,800 - 3.0 megabytes) is pretty good when viewing full-size.

and then a link to the RealThing.
Note the topless runner in orange shorts appearing in two positions on the field.
The "RealThing" however and looking at the scaffolding appears to be less crisp.

Kind regards.
 

Annna T

Active member
Here's one with enough detail to make 64 MPixels work a bit. First the easy-downloadin' version:


and then a link to the RealThing.
Note the topless runner in orange shorts appearing in two positions on the field.

The full .ORF, the jpg created by Olympus, and jpgs rendered by CaptureOne 8.2.2 from the first frame and from the 64MPx composite can all be found here.

scott

Oh, 40-150/2.8 @f/5.6. Distance about 500-600 m.
Please, refrain from posting pictures larger than the max authorized of 1200 pix in the largest dimension. It is useless because everything is downsampled to 1200 pixels anyway. But this is a nuisance for those if us using ipads : the downsizing occurs on our individual machines and this freezes the ipads. Once there are several of those pictures in the same thread, the ipads become irresponsive : it takes several minutes to download and resample. The like buttons don't work as intended and the comment editor crashes or freezes.
This isn't fair for other forumers and the benefit you hope to get isn't there anyway. I too measured 2400 dpi on the longer side of the picture you included directly in your post.
 

scott kirkpatrick

Well-known member
The "RealThing" however and looking at the scaffolding appears to be less crisp.
I sharpened the full 64 MP jpeg as much as I dared. Viewed at 50% (down to 16 MP) it is crisper and clearer than the first frame seen at 100%, which doesn't always happen, if your tripod is a little old, or the wind is blowing. So I sharpened it twice as much (see P5240423 1.jpg) and now I think it looks pretty good at 100%, but I'll have to check it at home on a big screen to be sure there are no artifacts.

scott

P.S. Driving the sharpening slider in COne 8.2.2 all the way to the right works pretty well on this file.
 
Last edited:

scott kirkpatrick

Well-known member
Here's a Moire-challenging shot, cloth flowers. First the whole frame, reduced beyond recognition to 1200 pixels width:



and then a 100% crop:



The full size versions (64 MP, 40 MP, and the first frame at 16 MP) are all here


using the 40-150/2.8 at 150/8. Sharpened in Capture One 8.2.2 to 80 on a scale of 100.

scott
 

scott kirkpatrick

Well-known member
Oh c'mon Scott, it isn't that bad, is it ?
:ROTFL:

Kind regards.
Well, with the site's new flaky software, I can often see my pictures at 2400 pixels wide. I was looking at stuff on an Android tablet the other night, and everything came down wider than 1200, because I had squeezed the frame width before connecting to the site.

scott
 
Top