Go Back   Steve's Digicams Forums > Digital SLR and Interchangeable Lens Cameras > Olympus dSLR

Reply
 
Thread Tools Search this Thread
Old Dec 4, 2010, 2:10 PM   #11
Administrator
 
Join Date: Jun 2003
Location: Savannah, GA (USA)
Posts: 22,378
Default

P.S.

I also tried a conversion without the -w (lower case) parameter (which tells dcraw.c to use the as shot white balance, which are basically rgb multipliers stored in the metadata) like this:

dcrawMS -W -T -4 E5hSLI3200_NR_OFF.ORF E3hSL_FILTEROFF.ORF

Basically, the -W (case sensitive) tells it not to brighten the image, the -T tells it to use a TIFF versus PPM output, and the -4 tells it to use a linear conversion (no curve applied).

I get the same thing without applying any multipliers for the White Balance settings (by omitting the -w parameter)...

The E-3 shot is brighter using the same subject, lighting, ISO speed, aperture and shutter speed, as compared to the conversion from the E-5 shot.

So, on the surface, it does appear that DxOMark is correct... that is, the E-5 is not as sensitive as set compared to the E-3 when looking at the files converted from RAW using a linear conversion (at least with the ISO 3200 shots from imaging-resource.com taken in simulated daylight lighting used in Studio shots, as I haven't converted any tungsten lighting images yet, as in the Mannequin shots they have available). I'm seeing the same thing regardless of whether or not I apply RGB multipliers for White Balance correction with the -w option. The E-3 shots are brighter using the same subjects, camera settings, lighting, etc. (as imaging-resource uses controlled lighting conditions for their tests).

Note that you may also want to try it without the -T, which tells it to output to TIFF (as the default .ppm output is probably going to be more suitable for many editors and image viewers that don't properly interpret the TIFF output from dcraw.c).
JimC is offline   Reply With Quote
Old Dec 4, 2010, 3:09 PM   #12
Administrator
 
Join Date: Jun 2003
Location: Savannah, GA (USA)
Posts: 22,378
Default

Another P.S.

But, again, take those types of tests with a "grain of salt", as a linear conversion is one thing, and a more usable image is another thing (after applying curves to get the end result looking like you want via in camera processing or using popular raw converters).

So, even though the E-5 appears to be less sensitive than set for ISO speed at the raw level, they probably did it that way for a good reason (perhaps to protect highlights more, and/or because of sensor response curve differences which can impact the processing needed for more usable results in more lighting conditions after applying tone curves and multipliers for WB).
JimC is offline   Reply With Quote
Old Dec 4, 2010, 5:21 PM   #13
Senior Member
 
tkurkowski's Avatar
 
Join Date: Apr 2007
Location: Maryland, USA
Posts: 3,625
Default

Quote:
Originally Posted by fldspringer View Post
Adobe doesn't support the E-5 files and while it can be fooled into processing them, it must be done on a command line hex editor and is a hassle I've done only once and hope I never need to again. Hence I did use JPEG for the E-5.
Did you do this hack?

http://forum.fourthirdsphoto.com/showthread.php?t=62217

Ted
tkurkowski is offline   Reply With Quote
Old Dec 4, 2010, 7:24 PM   #14
Senior Member
 
Join Date: Feb 2007
Location: Wisconsin
Posts: 1,241
Default

Quote:
Originally Posted by tkurkowski View Post
Let me tell you, I'm not very good with the command line anymore. Its almost impossible for me to get a whole line in there without screwing it up.

I did it once, and that is enough. I'm not doin' another raw converter either. I'll survive with Raw/Sfine Jpegs and using the jpegs until Adobe comes through.

I'm gett'n old and set in my ways. I'm working enough these days that when I'm playing, it better be fun.

Greg
__________________
Greg

https://dogsportphoto.smugmug.com/
fldspringer is offline   Reply With Quote
Old Dec 4, 2010, 7:37 PM   #15
Senior Member
 
Join Date: Feb 2007
Location: Wisconsin
Posts: 1,241
Default

Quote:
Originally Posted by JimC View Post
Another P.S.

But, again, take those types of tests with a "grain of salt", as a linear conversion is one thing, and a more usable image is another thing (after applying curves to get the end result looking like you want via in camera processing or using popular raw converters).

So, even though the E-5 appears to be less sensitive than set for ISO speed at the raw level, they probably did it that way for a good reason (perhaps to protect highlights more, and/or because of sensor response curve differences which can impact the processing needed for more usable results in more lighting conditions after applying tone curves and multipliers for WB).
Its interesting, but I'm much more concerned with what the camera does.

I have a feeling your correct about the highlight protection, but if they are bringing up an under-exposed photo and are keeping the noise in check that well, I'm also impressed. There is no free lunch here. Bringing up an underexposed photo is about the same as shooting at the higher ISO as far as the finished product is concerned.

This camera isn't a full frame 35mm and it doesn't perform like one, but it is allowing me to gain a little shutter speed in a pinch. That's nice. If its overstating the ISO or not is a side note as long as I get the results in the end.
__________________
Greg

https://dogsportphoto.smugmug.com/
fldspringer is offline   Reply With Quote
Old Dec 4, 2010, 7:38 PM   #16
Administrator
 
Join Date: Jun 2003
Location: Savannah, GA (USA)
Posts: 22,378
Default

FWIW, out of curiosity, I spent a bit more time playing with the raw files from the E-3 and E-5, and I don't see much difference in them if you convert them so that the output brightness is similar. IOW, any difference is mostly negligible.

Noise levels, grain pattern, detail retention, etc., look very close between the images produced by these cameras using the same subjects, lighting, and camera settings at ISO 3200 (using the imaging-resource.com studio samples), using dcraw.c to convert them (which isn't going to do as much to the images as popular raw converters like Adobe's, etc.).

But, the JPEG output from the cameras is another story entirely (again, using the ISO 3200 samples at imaging-resource.com).

It appears that the noise reduction algorithms being used with the E-5 are a lot better (far less noise with good detail retention). The only place I see where the E-5 doesn't do well is on some of the red fabric in the studio, where most of the fine detail is smeared away from the E-5 produced jpeg, whereas the E-3 leaves plenty of detail in the same fabric from it's camera produced jpeg.

But, for whatever reason, a number of other cameras have the same issue with that particular red fabric.

If you use the comparometer there, you'll see what I mean (E-5 files look *much* cleaner).

http://www.imaging-resource.com/IMCOMP/COMPS01.HTM

Now, without spending more time with more samples in a greater variety of conditions, it's hard to say for sure.... But, my impression from playing with the raw files is that there is really not much difference between these cameras' output at the raw level if you use the same camera settings and convert so that you end up with the same image brightness in the output images. But, the E-5's JPEG output appears to be much better, apparently using far more sophisticated noise reduction algorithms.
JimC is offline   Reply With Quote
Old Dec 4, 2010, 8:58 PM   #17
Administrator
 
Join Date: Jun 2003
Location: Savannah, GA (USA)
Posts: 22,378
Default

Quote:
Originally Posted by JimC View Post
...But, for whatever reason, a number of other cameras have the same issue with that particular red fabric.
BTW, if I had to guess... it's because the same Noise Reduction engines are being used by the cameras having issues with it.

For example, Apical makes no secret that Sony, Olympus and Nikon all license pieces of it's iridix engine for use in their cameras' image processing. You can see more about iridix here:

http://www.apical-imaging.com/iridix

So, it wouldn't surprise me to find out that companies like Olympus and Sony are using some of their other technology, too (for example, the sinter engine for noise reduction).

http://www.apical-imaging.com/sinter

From my perspective, it's just too much of a coincidence in how some cameras from different brands have the same issues with certain colors and textures when using newer noise reduction algorithms that appear to give far cleaner output while preserving a lot of detail (and one of the red fabric swatches in the imaging-resource.com examples is a good indicator that something is in common with the way some newer cameras are suppressing noise).
JimC is offline   Reply With Quote
Old Dec 4, 2010, 9:29 PM   #18
Senior Member
 
dlpin's Avatar
 
Join Date: Jul 2008
Posts: 143
Default

Quote:
Originally Posted by JimC View Post
FWIW, out of curiosity, I spent a bit more time playing with the raw files from the E-3 and E-5, and I don't see much difference in them if you convert them so that the output brightness is similar. IOW, any difference is mostly negligible.

Noise levels, grain pattern, detail retention, etc., look very close between the images produced by these cameras using the same subjects, lighting, and camera settings at ISO 3200 (using the imaging-resource.com studio samples), using dcraw.c to convert them (which isn't going to do as much to the images as popular raw converters like Adobe's, etc.).

But, the JPEG output from the cameras is another story entirely (again, using the ISO 3200 samples at imaging-resource.com).

It appears that the noise reduction algorithms being used with the E-5 are a lot better (far less noise with good detail retention). The only place I see where the E-5 doesn't do well is on some of the red fabric in the studio, where most of the fine detail is smeared away from the E-5 produced jpeg, whereas the E-3 leaves plenty of detail in the same fabric from it's camera produced jpeg.

But, for whatever reason, a number of other cameras have the same issue with that particular red fabric.

If you use the comparometer there, you'll see what I mean (E-5 files look *much* cleaner).

http://www.imaging-resource.com/IMCOMP/COMPS01.HTM

Now, without spending more time with more samples in a greater variety of conditions, it's hard to say for sure.... But, my impression from playing with the raw files is that there is really not much difference between these cameras' output at the raw level if you use the same camera settings and convert so that you end up with the same image brightness in the output images. But, the E-5's JPEG output appears to be much better, apparently using far more sophisticated noise reduction algorithms.
This lines up with what I saw when I downloaded a set of images from another review site.

I've always expected the e3 and e5 to perform within in the same ball park if I just did the most basic raw conversion. Much like the d5000 and d90, or the d7000 and sony a55.

This isn't just restricted to olympus, as most manufacturers seem to be doing this forced obsolescence nowadays, where the output of older cameras could be much improved if they allowed them to be processed with current technology, but that is blocked by software. One example with regards to olympus is that if you process an e-3 raw image through olympus software you don't have access to the art filters, but if you hack the file to say it is from an e-5 suddenly you can use it.
dlpin is offline   Reply With Quote
Old Dec 4, 2010, 9:40 PM   #19
Administrator
 
Join Date: Jun 2003
Location: Savannah, GA (USA)
Posts: 22,378
Default

Well... there's nothing stopping you from using third party raw conversion tools, noise reduction tools, etc. ;-)

IOW, even if Olympus restricts features you can use in it's raw converter with a given camera model, you've got lots of alternatives (Lightroom, Bibble, DxO Optics Pro, etc.). The E-5 is relatively new. So, it may take a while before more products support it. But, chances are, that won't take very long.

That is one nice thing about shooting raw. As technology improves (demosaic algorithms, noise reduction algorithms, highlight recovery algorithms, lens correction algorithms, etc.), you can always go back and reprocess cherished images later for even better results.

It's amazing how sophisticated some of the newer algorithms are becoming (for example, where you can see a *lot* of noise with a basic raw conversion using something like dcraw.c, yet end up with a relatively clean and detailed output using the very sophisticated noise reduction algorithms showing up in newer software).
JimC is offline   Reply With Quote
Old Dec 4, 2010, 10:03 PM   #20
Administrator
 
Join Date: Jun 2003
Location: Savannah, GA (USA)
Posts: 22,378
Default

But, the market is very competitive, and newer generations of cameras do tend to have improvements in many areas.

For example, I see Olympus added a number of features in the E-5 that you don't get with the E-3 (higher resolution sensor, video recording, live view using Contrast Detect AF with face detection, improved shutter durability, higher resolution LCD and more).

So, even though the E-3 may be able to produce great results using newer software to convert from raw, many buyers are still going to want the features found in competing models when shopping for a new camera, and not everyone wants to bother with raw conversion and post processing of images (so the improved in camera processing can be a plus for many users).
JimC is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 12:07 PM.