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!

OM-D E-M5 Mark II Pre-Order

k-hawinkler

Well-known member
The Old Hill Road


Olympus OM-D E-M5 Mark II + Panasonic LUMIX G Leica DG Nocticron 42.5mm f/1.2 ASPH Power OIS Lens.
Hi res raw file processed with Olympus High Res Shot Raw File Photoshop Plug-in, CS6, and Nik.

Image size 1152x864: above
Image size 2304x1728: http://winklers.smugmug.com/2015-03.../0/O/_3110098_CS6_2304x1728_Old_Hill_Road.jpg
Image size 4608x3456: http://winklers.smugmug.com/2015-03.../0/O/_3110098_CS6_4608x3456_Old_Hill_Road.jpg
Image size 9216x6912: http://winklers.smugmug.com/2015-03.../0/O/_3110098_CS6_9216x6912_Old_Hill_Road.jpg
 
Last edited:

k-hawinkler

Well-known member

Olympus OM-D E-M5 Mark II + OLYMPUS M.75-300mm F4.8-6.7 II Lens.
Hi res raw file processed with Olympus High Res Shot Raw File Photoshop Plug-in, CS6, and Nik.

Image size 1152x864: above
Image size 2304x1728: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-gvsgsGf/0/O/_3110025_CS6_2304x1728.jpg
Image size 4608x3456: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-zC5FhvD/0/O/_3110025_CS6_4608x3456.jpg
Image size 9216x6912: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-XdcVZ8V/0/O/_3110025_CS6_9216x6912.jpg
 
Last edited:

scho

Well-known member

Olympus OM-D E-M5 Mark II + OLYMPUS M.75-300mm F4.8-6.7 II Lens.
Hi res raw file processed with Olympus High Res Shot Raw File Photoshop Plug-in, CS6, and Nik.

Image size 1152x864: above
Image size 2304x1728: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-gvsgsGf/0/O/_3110025_CS6_2304x1728.jpg
Image size 4608x3456: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-zC5FhvD/0/O/_3110025_CS6_4608x3456.jpg
Image size 9216x6912: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-XdcVZ8V/0/O/_3110025_CS6_9216x6912.jpg
Very nice K-H. What are your sharpening settings in NIK? I've been using about 50-70 % adaptive raw pre-sharpening in the LR plugin which works well for most of my HR images.
 

k-hawinkler

Well-known member

Olympus OM-D E-M5 Mark II + OLYMPUS M.75-300mm F4.8-6.7 II Lens.
Hi res raw file processed with Olympus High Res Shot Raw File Photoshop Plug-in, CS6, and Nik.

Image size 1152x864: above
Image size 2304x1728: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-jsdQpqP/0/O/_3110003_CS6_2304x1728.jpg
Image size 4608x3456: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-LsZq7RK/0/O/_3110003_CS6_4608x3456.jpg
Image size 9216x6912: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-hZHMD9p/0/O/_3110003_CS6_9216x6912.jpg

Crop


Image size 1168x657: above
Image size 2336x1314: http://winklers.smugmug.com/2015-03...i-Qp5SkzS/0/O/_3110003_CS6_2336x1314_Crop.jpg
Image size 4672x2628: http://winklers.smugmug.com/2015-03...i-FwT4jmZ/0/O/_3110003_CS6_4672x2628_Crop.jpg
 

k-hawinkler

Well-known member
Very nice K-H. What are your sharpening settings in NIK? I've been using about 50-70 % adaptive raw pre-sharpening in the LR plugin which works well for most of my HR images.
Many thanks Carl.
I typically use the default value (50%, never more than that though) of the Nik adaptive raw pre-sharpening, but exclude the sky from sharpening. If I use output sharpening at all, it would be typically at 25% or at most 30%. Higher can very easily produce jaggies. Also before any sharpening I use Define to rid the image of noise caused by high ISO or by increased clarity.
 

k-hawinkler

Well-known member

Olympus OM-D E-M5 Mark II + Panasonic LUMIX G Leica DG Nocticron 42.5mm f/1.2 ASPH Power OIS Lens.
Hi res raw file processed with Olympus High Res Shot Raw File Photoshop Plug-in, CS6, and Nik.

Image size 1152x648: above
Image size 2304x1296: http://winklers.smugmug.com/2015-03-11-E-M5-II-Nocticron/i-8sz9SKV/0/O/_3110065_CS6_2304x1296.jpg
Image size 4608x2592: http://winklers.smugmug.com/2015-03-11-E-M5-II-Nocticron/i-6dtVw53/0/O/_3110065_CS6_4608x2592.jpg
Image size 9216x5184: http://winklers.smugmug.com/2015-03-11-E-M5-II-Nocticron/i-JLNGLhs/0/O/_3110065_CS6_9216x5184.jpg
 

k-hawinkler

Well-known member

Olympus OM-D E-M5 Mark II + OLYMPUS M.75-300mm F4.8-6.7 II Lens.
Hi res raw file processed with Olympus High Res Shot Raw File Photoshop Plug-in, CS6, and Nik.

Image size 1152x864: above
Image size 2304x1728: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-gvsgsGf/0/O/_3110025_CS6_2304x1728.jpg
Image size 4608x3456: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-zC5FhvD/0/O/_3110025_CS6_4608x3456.jpg
Image size 9216x6912: http://winklers.smugmug.com/2015-03-11-E-M5-II-M75-300mm/i-XdcVZ8V/0/O/_3110025_CS6_9216x6912.jpg

A quick note. If one carefully looks at this hi res image then one can detect artifacts/jaggies around sharp edges. I also compared the images derived from the hi res ORF file with either CS6 or Iridient and found that CS6 seems to generate significantly stronger artifacts than Iridient. In order to avoid these kinds of issues it's probably best to display or print these hi res images at 50% resolution.
 

ptomsu

Workshop Member
K-H,

many thanks for posting these and taking the effort!

Having said that, one can see significant artifacts in the high res images , so I hope this will improve with further development of conversion SW.

Peter
 

scott kirkpatrick

Well-known member
K-H, I have the same reaction to the 64MPx images that you are getting. They seem to have been taken a bit too far, perhaps losing something. Olympus seems to think that the real power of their multishot HR is 40 MPx, and they have designed an in-camera transform to jpg to achieve that. Maybe if you output 48 MPx jpeg files from Iridient's processing rather than 64 MPx, you would also catch that sweet spot.

scott
 

scho

Well-known member
A quick note. If one carefully looks at this hi res image then one can detect artifacts/jaggies around sharp edges. I also compared the images derived from the hi res ORF file with either CS6 or Iridient and found that CS6 seems to generate significantly stronger artifacts than Iridient. In order to avoid these kinds of issues it's probably best to display or print these hi res images at 50% resolution.
I see some cross hatching artifacts in a few areas of the rocks that are indicative of movement. I assume you were not experiencing an earthquake so could it also be some camera movement?

Artifacts in these HR files tend to be subject (contrast, edge types), environmental (light, wind etc.), and post processing affected and I wouldn't generalize based on a few samples. I've made several excellent images and large prints using the 64 MP tif output from the plugin as well as Irident 40 MP output. It takes awhile to learn what will and will not work, but for me I like what is possible now and this little camera is giving my Sony A7r a run for the money.
 

jlm

Workshop Member
they are all different due to changes in the base shape and hole locations, etc... the Em1 does not need the grip portion but both the EM-5 versions can use one.

the rotating LCD on the Mark II makes that end not so useable for the L-plate, so for that model, i am putting it on the other end (with card access, of course)
 

k-hawinkler

Well-known member
A quick note. If one carefully looks at this hi res image then one can detect artifacts/jaggies around sharp edges. I also compared the images derived from the hi res ORF file with either CS6 or Iridient and found that CS6 seems to generate significantly stronger artifacts than Iridient. In order to avoid these kinds of issues it's probably best to display or print these hi res images at 50% resolution.
K-H, I have the same reaction to the 64MPx images that you are getting. They seem to have been taken a bit too far, perhaps losing something. Olympus seems to think that the real power of their multishot HR is 40 MPx, and they have designed an in-camera transform to jpg to achieve that. Maybe if you output 48 MPx jpeg files from Iridient's processing rather than 64 MPx, you would also catch that sweet spot.

scott
I see some cross hatching artifacts in a few areas of the rocks that are indicative of movement. I assume you were not experiencing an earthquake so could it also be some camera movement?

Artifacts in these HR files tend to be subject (contrast, edge types), environmental (light, wind etc.), and post processing affected and I wouldn't generalize based on a few samples. I've made several excellent images and large prints using the 64 MP tif output from the plugin as well as Irident 40 MP output. It takes awhile to learn what will and will not work, but for me I like what is possible now and this little camera is giving my Sony A7r a run for the money.

Scott and Carl, Many thanks for your feedback. Much appreciated.

After checking into this image a bit more I would like to come back to the apparent issue in a bit more detail. As my snow image above has demonstrated at least to me, the E-M5 II is capable of high resolution, pretty much artifact free quality images, provided one uses sufficient care in shooting and post-processing an image. So if something doesn't look quite right, the question arises: WHY?

The snow scene I shot from within my house, wind not causing camera motion. The tree branches may have been minimally affected by wind, but certainly not the fence wires. Also the grey stripes, in the first version of the image derived from the raw file, could be traced back to post-processing inadequacies and eventually avoided.

The current image



was shot about 15 yards away from a road with heavy car traffic. There might also have been an occasional minimal puff of wind. This could have possibly had a slight impact on camera motion with negative consequences. So I would like to examine the central part of the image a little closer.

When shooting in high res mode the E-M5 Mark II, one has to use a tripod and one gets three files per shot if the camera settings are set to SL F + RAW and High Res Shot, namely an .ORF, .ORI, and a 7296x5472 .JPG image. Of course, PS CS6 can read the .JPG image. With the help of the Olympus High Res Shot Raw File Photoshop Plug-in Software Software Download | Olympus High Res Shot Raw File Photoshop Plug-in | OLYMPUS IMAGING CS6 can also read in the .ORF file, resulting in a 9216x6912 image. However, so far I am not aware how to directly read in a .ORI file with CS6.

In contrast Iridient can read in all 3 files, the .ORF, .ORI, and the 7296x5472 .JPG image.

Iridient can also save the .ORF file as a 16 bit .tif image in these pixel sizes 4608x3456, 7296x5472, 9216x6912, and 12288x9216. Other pixel sizes seem possible, but I have not tried them.

Likewise Iridient handles .ORI files.

Whereas for the .ORF file the E-M5 Mark II has combined several shots into a single file, with advantages and apparent disadvantages, the .ORI file contains the data just from a single shot. So it's very convenient to retreat to a single shot image if the multiple shot composite doesn't give the desired outcome, for whatever reason.

In the following I have used CS6 (with Olympus Plugin) and Iridient to generate images and crop out the central part.

Processed with CS6 (with Olympus Plugin), default settings. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.

I clearly can see the diagonal stripes in many parts of this crop, going from the upper left to the lower right.

Iridient has the following panels: Expo Color Mono Curves Detail Lens.
Processed with Iridient, default settings except all boxes checked in the Lens panel. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.

Here the diagonal stripes seem less pronounced and are barely visible to my eyes.
Clearly some sort of pattern is visible as an artifact in the image.
Can one make that pattern more visible or preferentially make it disappear?

Processed with Iridient, default settings except all boxes checked in the Lens panel. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.
Here I also set in the Detail panel, Edge Detail from 100 to 500 = max, Texture/Micro Detail from 5 to 25 = max.

Now the stripes have become more visible to me eyes.

So in conclusion I agree with Carl that some unintended camera motion must have affected this composite image.


The next question on mind is what does the image look like that can be derived from the one shot .ORI file?
Here is the crop from it, processed the same way as the last image and mapped to the same size.
Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.

Of course no sign of camera motion. To my eyes this image also has coarser features.
 
Last edited:

scho

Well-known member
Scott and Carl, Many thanks for your feedback. Much appreciated.

After checking into this image a bit more I would like to come back to the apparent issue in a bit more detail. As my snow image above has demonstrated at least to me, the E-M5 II is capable of high resolution, pretty much artifact free quality images, provided one uses sufficient care in shooting and post-processing an image. So if something doesn't look quite right, the question arises: WHY?

The snow scene I shot from within my house, wind not causing camera motion. The tree branches may have been minimally affected by wind, but certainly not the fence wires. Also the grey stripes, in the first version of the image derived from the raw file, could be traced back to post-processing inadequacies and eventually avoided.

The current image



was shot about 15 yards away from a road with heavy car traffic. There might also have been an occasional minimal puff of wind. This could have possibly had a slight impact on camera motion with negative consequences. So I would like to examine the central part of the image a little closer.

When shooting in high res mode the E-M5 Mark II, one has to use a tripod and one gets three files per shot if the camera settings are set to SL F + RAW and High Res Shot, namely an .ORF, .ORI, and a 7296x5472 .JPG image. Of course, PS CS6 can read the .JPG image. With the help of the Olympus High Res Shot Raw File Photoshop Plug-in Software Software Download | Olympus High Res Shot Raw File Photoshop Plug-in | OLYMPUS IMAGING CS6 can also read in the .ORF file, resulting in a 9216x6912 image. However, so far I am not aware how to directly read in a .ORI file with CS6.

In contrast Iridient can read in all 3 files, the .ORF, .ORI, and the 7296x5472 .JPG image.

Iridient can also save the .ORF file as a 16 bit .tif image in these pixel sizes 4608x3456, 7296x5472, 9216x6912, and 12288x9216. Other pixel sizes seem possible, but I have not tried them.

Likewise Iridient handles .ORI files.

Whereas for the .ORF file the E-M5 Mark II has combined several shots into a single file, with advantages and apparent disadvantages, the .ORI file contains the data just from a single shot. So it's very convenient to retreat to a single shot image if the multiple shot composite doesn't give the desired outcome, for whatever reason.

In the following I have used CS6 (with Olympus Plugin) and Iridient to generate images and crop out the central part.

Processed with CS6 (with Olympus Plugin), default settings. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.

I clearly can see the diagonal stripes in many parts of this crop, going from the upper left to the lower right.

Iridient has the following panels: Expo Color Mono Curves Detail Lens.
Processed with Iridient, default settings except all boxes checked in the Lens panel. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.

Here the diagonal stripes seem less pronounced and are barely visible to my eyes.
Clearly some sort of pattern is visible as an artifact in the image.
Can one make that pattern more visible or preferentially make it disappear?

Processed with Iridient, default settings except all boxes checked in the Lens panel. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.
Here I also set in the Detail panel, Edge Detail from 100 to 500 = max, Texture/Micro Detail from 5 to 25 = max.

Now the stripes have become more visible to me eyes.

So in conclusion I agree with Carl that some unintended camera motion must have affected this composite image.


The next question on mind is what does the image look like that can be derived from the one shot .ORI file?
Here is the crop from it, processed the same way as the last image and mapped to the same size.
Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.

Of course no sign of camera motion. To my eyes this image also has coarser features.
K-H, Thanks for your tests and comparisons. Have you tried using the ORI file as temporary background layer and then brushing out the diagonal line artifacts? Also, I find that simply downsizing will remove most of these motion induced diagonal line artifacts you are being in 64 MP files at 100%.
 
Last edited:

k-hawinkler

Well-known member
K-H, Thanks for your tests and comparisons. Have you tried using the ORI file as temporary background layer and then brushing out the diagonal line artifacts?

Thanks Carl. No, I won't do that as I have a better shot.
But first let me finish my train of thought.

So next I take the multiple shot images and compare them to the single shot image but at 4608x3456, so at 50% size for the larger images.

.ORF processed with CS6 (with Olympus Plugin), default settings. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.


.ORF processed with Iridient and in the Detail panel, Edge Detail = 500 = max, Texture/Micro Detail = 25 = max. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.


.ORI processed with Iridient and in the Detail panel, Edge Detail = 500 = max, Texture/Micro Detail = 25 = max. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.

This was a surprise as it looks completely over processed.

So I decided to set the two parameters back to its default values.
.ORI processed with Iridient and in the Detail panel, Edge Detail = 100, Texture/Micro Detail = 5. Image shot at Date Time: Mar 11, 2015, 6:21:19 PM.


It appears that through this mapping process the artifacts seem to have been removed from the image.
However, my eyes are not good enough to determine whether the down-mapped multiple shot images are any better than the single shot image.

Luckily at Date Time: Mar 11, 2015, 6:21:30 PM, i.e. 11 seconds after the previous shot at Date Time: Mar 11, 2015, 6:21:19 PM I took another image that doesn't seem as badly affected by camera motion.

Here is a crop of the image derived with CS6 from the .ORF file and mapped to 12288x9216.


This is the corresponding image derived with Iridient.


And this is the downsized complete image developed with Iridient from the .ORF file.
 

k-hawinkler

Well-known member
Now a fun shot. I am really amazed by what this camera, the E-M5 II, can do.



I was far from any vehicular traffic and there was no noticeable wind.
Please, let me know if you can detect any obvious or not so obvious artifacts in the high resolution versions below. TIA.
The original image from Iridient was 12288x9216 in size.
I then cropped it to 12288x6912. Then stepwise cut the linear size in half. Here we go.

1536x864: image above
3072x1728: http://winklers.smugmug.com/2015-03...072x1728_200_10_Define_Sharpen_Crop_16by9.jpg
6144x3456: http://winklers.smugmug.com/2015-03...144x3456_200_10_Define_Sharpen_Crop_16by9.jpg
12288x6912: http://winklers.smugmug.com/2015-03...288x6912_200_10_Define_Sharpen_Crop_16by9.jpg <--- patience please!!!
 

scho

Well-known member
Now a fun shot. I am really amazed by what this camera, the E-M5 II, can do.



I was far from any vehicular traffic and there was no noticeable wind.
Please, let me know if you can detect any obvious or not so obvious artifacts in the high resolution versions below. TIA.
The original image from Iridient was 12288x9216 in size.
I then cropped it to 12288x6912. Then stepwise cut the linear size in half. Here we go.

1536x864: image above
3072x1728: http://winklers.smugmug.com/2015-03...072x1728_200_10_Define_Sharpen_Crop_16by9.jpg
6144x3456: http://winklers.smugmug.com/2015-03...144x3456_200_10_Define_Sharpen_Crop_16by9.jpg
12288x6912: http://winklers.smugmug.com/2015-03...288x6912_200_10_Define_Sharpen_Crop_16by9.jpg <--- patience please!!!
Very nice detail. The sweet spot seems to be between 20-30 MP based on the 50% downsize of the largest file. I didn't see any artifacts.
 

k-hawinkler

Well-known member
A couple shots from Bandelier National Monument.

• Image derived from the high res .ORF composite file.


• Image derived from the low res .ORI single shot file.



Now a very lucky shot, image derived from the high res .ORF composite file. :cool:
 

scho

Well-known member
A couple shots from Bandelier National Monument.

• Image derived from the high res .ORF composite file.


• Image derived from the low res .ORI single shot file.



Now a very lucky shot, image derived from the high res .ORF composite file. :cool:
They all look good. Was that "lucky" shot a 100% crop detail? What happened to the wooden ladders they used to have leaning against the cliff to give tourists access to the 2nd story dwellings?
 
Top