Olympus-OM
[Top] [All Lists]

Re: [OM] E-M5 II announced

Subject: Re: [OM] E-M5 II announced
From: Moose <olymoose@xxxxxxxxx>
Date: Mon, 09 Feb 2015 13:33:08 -0800
On 2/8/2015 12:56 PM, Mike Gordon via olympus wrote:
Yes, seems very weird that a Bayer raw is recreated.  Just because it is weird, 
doesn't mean it isn't true.  Yet another dissection of the High Res Raw format 
confirms this.
There are no Bayer artifacts discernible however, so the strategy works well and the 
results described as "so far look stunning."

Weeeelllll ... That depends on what you look at, and perhaps how you look at it.

Viewer 3 will open one, but the view is highly pixelated, with ugly artifacts.

RawTherapee will open and convert. Using their default demosaicing, the output TIFF is big, and has a lot of detail. With a little post, some parts of it look as good as the HD JPEG. The fiddler on the beer label, for example, is pretty amazing. But the fabric doesn't come close to the JPEG in detail in the weave.

DPR offers HD RAW in their comparator, from a Beta version of ACR. And again the converted RAW is soft, and in some areas, slightly smeary.

Whatever the camera is doing in processing the slightly downsized JPEG is pretty Da*n tricky and sophisticated. Whether the result of somehow backing into a Bayer array or that Raw converters don't as yet handle them properly, the Raw converter results so far are poor in comparison.

The comparisons of HD JPEGs to a D810 are interesting. The yellow fabric conversion on IR clearly favors the Oly, as does the solid green screen printing on the S Smith label. But the apron of the standing woman on the left of the pencil drawing in the DPR sample has a diagonal pattern that's clearly shown in the Nikon and hardly even hinted at by the Oly. There are also slight, but distinct, halos around high contrast detail, as text, with the Oly, there but almost invisible with the Nikon JPEG. The Nikon has strong moiré in the small, central circular target, the Oly, none.

There is also the issue of the lens used and where in its field a particular part of the test subject is placed. For the E-M5 Mark II, IR used the 4/3 50/2, DPR the µ4/3 45/1.8. For the Nikon, IR = 70/2.8 DPR = 85/1.4G. Might the results for a piece in one of the targets be different with a different lens, if placed in a different part of the frame?

So yes, the HD JPEGs are spectacular from a 4/3 format, 16 MP camera. No, they don't, overall, beat a top FF, 36 MP camera, but can beat it, where moiré rears its head. And, of course, they only work on static subjects and are very slow to produce.

In Progress 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