Discussion in 'Olympus Cameras' started by With_Eyes_Unclouded, Oct 28, 2013.
Looks like the E-M1 sensor essentially offers a modest increase in DR, and a miniscule decrease in high ISO noise. I can believe that.
One thing they don't measure is long-exposure noise - I think there they'd find the E-M1 differs more noticeably from the older cameras.
Looks like, for landscape purposes, in any case, it doesn't really offer anything statistically significant over the GH3 other than better weather sealing and that nifty long exposure trick it has where you can see how the exposure is going (and stop it if necessary) in the middle of the exposure.
With prices on the GH3 falling, this will be a tough call this Winter.
Actually... it's a bit perplexing.
I've seen a number of high ISO RAW comparisons between the E-M1 and the E-M5, and the former looks way better in every respect. Which makes me wonder if the newer processor "cooks" the RAW data in a different manner, resulting in a final image with less noise and better color retention, overall. Not to mention lens corrections (done in software) in camera, that may also contribute to a better image, esp. at high ISO.
As I expected, it is very similar to the dxo score from nex-5n to nex-6 change, eg a little better dynamic range w/ high iso loss due to the new sensor pixels....
Such small differences compared to the other (E-M5, GX7, GH3) that its not visible to the human eye.
The trend is clear: its not the image quality that seperate the new cameras - its the handling, GUI, feeling and lens park.
Really? Any particular examples that you found notable?
One thing to remember is that Olympus Viewer does not apply the same settings to different models, so Noise Filter = Low means very different things on different cameras... Third-party RAW converters of course don't have this issue.
Interesting that they found an issue with a much lower ISO result.. I wonder if what ginobonetti found here is proof of an ISO performance issue with the EM-1. https://www.mu-43.com/showthread.php?t=53621
I wonder if we'll see more posts about ISO and the EM-1 and see if what Gino found gets resolved as to why it happened to him.
Also dpreview review is out:
Like gx7, they published after dxo score, maybe just a coincidence...
The dip in low-light ISO score is very strange and rather worrisome. I wonder why the score dropped like that.
Is it possible that a firmware update / improvements to the RAW engine in LR will improve things further? (I'm new here) but similarly when I first got my hands on the XE1, software made a nice improvement on IQ and RAW processing improved in LR from LR4.2 to 4.4 etc...
According to Olympus its a new/tweaked sensor (phase detect and lack of AA filter) so it stands to reason that early RAW support will probably only be so-so... ?
If not, is this something that firmware could fix ?
Maybe not strange - maybe a manifestation of the sensor being tuned to give closer to TRUE ISO 100 performance at baseline. As the DR seems to be best in class, this indicates that maybe the ISO 100 baseline is real. Frankly, this is a good tradeoff, IMHO.
No, the RAW engine/processing output is a JPG. There may be some processing done to the RAW file, but what DxO measures is based on the camera's RAW file output. They don't generally test JPGs.
The dynamic range increase goes from about 1/2 stop to 2/3 stop, which can a visible difference in a scene with a large dynamic range.
I one were to look at the graphs of S/N ratio, one would see that there is about 1/4 - 1/3 stop difference up to about ISO 200 and goes to 0 above that.
Looking at DxO numbers, without looking at the graphs can be very misleading.
A quick explanation: DxO's ISO scores are a bit iffy, particularly with modern sensors. That's not to say they aren't a good overall gauge (and probably the best numerical measurement we have), but they don't give the full picture.
Let's use the most obvious example. The E-M1 scores a 757 for ISO. The GH1 scores a 772.
Yes, that's the GH1. Not 2, not 3, but the very first camera in the GH series. It doesn't matter what the numbers say; any amount of anecdotal final-product evidence will confirm that the the E-M1 obviously performs better than the GH1 for high ISO.
Furthermore, from DxO's own description of their scoring process:
"When shooting a moving scene such as a sports event, action photographers’ primary objective is to freeze the motion, giving priority to short exposure time. To compensate for the lack of exposure, they have to increase the ISO setting, which means the SNR will decrease. How far can they go while keeping decent quality? Our low-light ISO metric will tell them.
The SNR indicates how much noise is present in an image compared to the actual information (signal). The higher the SNR value, the better the image looks, because details aren't drowned by noise. SNR strength is given in dB, which is a logarithmic scale: an increase of 6 dB corresponds to doubling the SNR, which equates to half the noise for the same signal.
An SNR value of 30dB means excellent image quality. Thus low-light ISO is the highest ISO setting for a camera that allows it to achieve an SNR of 30dB while keeping a good dynamic range of 9 EVs and a color depth of 18bits.
A difference in low-light ISO of 25% represents 1/3 EV and is only slightly noticeable.
As cameras improve, low-light ISO will continuously increase, making this scale open."
Thanks for the explanation!
The firmware question that I was asking was around ginobonetti's iso noise issues that he was talking about when comparing the em5 and the em1.
Finally, as the sensor is quite new, is it possible that improvements to the RAW interpreter that DXOmark use would yield better results over time , similarly to the way that Adobe are able to rework their RAW engine in LR over time. Or at least the way things improved on my XE1 over time... ??
Anyway, most of this is just theoretical testing and has little relation to real world use in the field where the ability to capture the decisive moment quickly is the most important!
Those figures of merit aren't very useful, because they are based on a single ISO performance only. If you every shoot at any other ISO you have to look at the Big Boy graphs.
Come on everyone. There's no problem with the ISO score! there's nothing worrisome here. There's no troubling issue.
A difference of 757 vs. 826 isn't anything. Don't you remember how ISO works? 100, 200, 400, 800, 1600, etc.? The difference between 757 and 826 is minuscule. This is nothing more than sample variation. The E-P5 is about 1/6th of a stop better than the E-M5 and the E-M1 is about a 1/6th of a stop worse than the E-M5. DxO themselves said that the difference is small enough that these are all likely the same sensor model...which means in all probability, any given instance of this particular sensor might be more like the E-P5 or more like the E-M1. The same variation exists for all cameras that share the same sensor tech.
All this tells you is that for any given camera with the Sony-made 16MP μ43 sensor, you can expect that your own will likely fall somewhere in the middle of these.
Mountains out of mole hills...
As the happy owner of an E-M5, I'm glad to see this...the E-M1 is certainly a great camera, but doesn't look like it's worth it for me to ditch my E-M5, especially because I'm one of the few people who prefer the smaller camera's form factor/ergonomics!
That's what I though until I used one. It's actually only slightly bigger, but the camera's form factor/ergonomics are much better. It's much more comfortable to grip it with your right hand to the extent that it feels lighter in the hand than the 5 (which I've had for a year and a half).
Personally, I'm kinda disappointed with the results. Was hoping for better performance specially in low light from a "PRO" camera.
Separate names with a comma.