Discussion in 'Olympus Cameras' started by ntblowz, Sep 10, 2013.
Except Fuji X-trans
Compare to E-M5/GX7
The pics don't work...
Pretty good comparison. Actually I was expecting gx7 to be better
Sent from my M040 using Tapatalk 4
Lets talk when we have raw files available to compare... I'm not doubting that its very close, maybe as good, but I'm not buying any broad conclusions from jpegs either...
I supposed these were RAW comparisons, with the E-M1 files being converted using Olympus Viewer or similar software. Am I wrong?
I think they usually use SOOC jpegs and then eventually add raw comparisons once raw files are supported by Adobe and/or Apple. I could be wrong, but I doubt the folks at DPR are wrestling with every manufacturer's included raw processing software...
If you check here:
Olympus E-M1 First Impressions Review: Digital Photography Review
and switch between JPEG and RAW on the drop down list, you'll see major differences (JPEGs are generally softer, certainly heavy NR added, etc). This is why I supposed the "RAW" option represents direct conversions from in camera RAW files.
so tempting...should I go with E-M1 instead of GX7?
Wondering the same thing...
As a JPEG shooter the images from ISO 200-6400 look soft in comparison to the E-M5. The RAW files look pretty good, though.
The DPR jpegs are showing lots of softness too in JPEG. I'm not sure why Oly would apply more NR than the E-M5 at similar settings. I'm not interested in the camera per-se anyways, I just hope the technology trickles down.
Same, One big feature I wanted was focus peaking on manual lenses and according to TheCameraStoreTV it isnt available on manual lenses only on lenses with electrical contacts.
Edit: Nvm, according to a comment on 43rumors you have to assign the function to a button for it to work with manual lenses
Eh, I wouldn't say it kicks APS-C butt. D7100 clearly looks better, which is not surprising considering the bigger sensor. Looking at JPEG results isn't meaningful due to differences in in-camera NR. The RAW comparisons show the E-M1 as being similar to the E-M5, but a bit sharper due to the AA filter removal (and a little less purple fringing, maybe due to lens used or improved in-camera correction?)
D7100 is also 50% more megapixels, so keep in mind that if it was downsized to the same 16MP as the E-M1, it would look even better in comparison. The same holds true of the D800...at a pixel level it may not be super great at high ISO, but you have so many megapixels to work with in terms of downsizing.
How's it compare to the Fuji X-Trans sensors (which are APS-C)?
Robin's part I review has high ISO testing and raw images:
Robin Wong: Olympus OM-D E-M1 Review: Introduction and High ISO Shooting
Actually, the X-PRO1 kicks the E-M1's butt in RAW at ISO 3200 as I suspected. It also kicks the others (I won't count the E-M5).
The X-PRO1 image appears slightly less sharp, but it's the "grain effect" of the noise in the others. When you denoise them to the level the X-PRO1 is already, they will look awful by comparison. Whereas you can apply sharpening to the X-PRO1 and it's lack of noise will give you a great image.
Pay particular attention to the shadow noise. Very telling.
Idle question.. why do these tests always keep the same iso levels, and not seem to test at optimum native iso levels? Especially if, we consider looking for optimum noise reduction at pixel levels, we should be smart enough to be using native iso's. Right?
If we are looking for HIGH ISO performance comparatively between systems, why would we not do it this way? I don't understand your "native ISO" term.
Probably that's because Fuji is employing noise-reduction in the RAW files. The other question is what conversion software the E-M1 is using. Traditionally, Olympus has employed different levels of NR in their own software on different models (including when NR is disabled).
In short - it's hard to compare RAW noise at high ISOs without knowing the conversion settings.
Native iso. where your camera operates at it's best and natively. EG: 5d mkii iso around 160 are pushed and pulled from that iso until you then hit 320. The further from that 160 point to your next native iso will have more noise. Push and pulled from there till you get to 640. then 1260. because of that, iso 1260 has less noise than iso200.
Canon 5d Native ISO Test - Empty Bucket Studios
a decent article on the 5dii natives as an example.
Separate names with a comma.