New german OM-D review

DonParrot said:

Louis_Dobson said:

Another futile JPG test?

Click to expand...

Hi Louis,

as I see it, these tests are futile only to those who've got the knowledge and time for PPing all their pics or enjoy to do so. And I'd bet a high amount of money that this doesn't apply to the majority of the camera users - what may be different when it comes to those active in forums like this one but they defintely represent just a small group of all the camera users. The vast majority, however, won't pp their pictures on a regular basis and those are exactly the ones who will check tests in magazines and websites before purchasing a new camera.

So, I'd say that these JPEG tests are anything but futile as they are helpful (more or less) for a high majority of those interested in buying a new camera.

For camera nerds like us - and even this group comprises quite a lot of JPEG shooters - there are other sources such as this website/forum for obtaining the information they need.

--



Why not - if there's enough space on the sofa...

I'm a HOlygan

Click to expand...

Okay, I keep reading this. I understand the perspective. But it seems a point is being glazed over here. If something must be done to improve the RAW file there will be a penalty or cost in the .JPG. If you're going to smear the details to fix the noise you will see that in the .JPG. The quality and the DR of the sensor ultimately affects everyone. Will you care about the differences? Maybe not. There are definite quality/technology differences here and many of us want to see the wolf under the sheep's clothing because we care about what we're REALLY getting - not just what can be cooked up in a .JPG.

Anyone who's dealt with the shadow noise that you just cannot fix without totally smearing what's there knows and dislikes it. The pens are notorious for the shadow (read noise) and we really are wanting a technology improvement here not just tweaks that appear to solve this ongoing irritation. I don't want to be fooled. If the package says organic then it pretty danged well be!

Marketing hype has occasionally driven all of us into the night screaming.

Dan

I think you are fools if you think that you can have a higher density sensor, without additional noise reduction.

Oly counteracts it with thinning the AA filter and some magic detail enhancement, which might just be a better denoising algorithm.

Whether this is cooked into RAWs or just implemented in Jpeg does mean little.

Unfortunately buit in denoising started a generation ago, and you can't disable it.

Because of the laws of read noise, the camera will have a better SNR than before at higher ISO. There is no treason, just a tradeoff.

So my impression is that you get a better overall camera, but with possible artefacts here and there dependiing on how strongly the firmware acts.

Am.

--
Photostream: http://www.flickr.com/photos/amalric

That is correct - more wells in the same area mean less light per well and therefore more noise. That is a physical truth that cannot be denied. Furthermore there are architecture considerations with the ADCs which have been beat to death so I won't go there again.

Louis_Dobson said:

rrr_hhh wrote:

Click to expand...

.....

Louis_Dobson said:

Louis_Dobson said:

1) The true ISO they measured was 120 at base, not 200, but they also wrote that the camera was under exposing by 1 and 1/3 EV, which I find confusing. This allows them to preserve the highlights very well.

Click to expand...

What they mean is the camera produces JPGs about 1/3 EV under-exposed (acceptable margin of error) but to do so takes a RAW about a stop under what most people would. Note that you cannot under or over expose a RAW technically, you choose what you want! Oly produce a dark RAW to preserve highlights. To me this is clear, but I find it hard to explain, and I can see how some people get fixated on the ides that it is "cheating". It isn't cheating of course, it is Oly making you the best JPG they can in the way they think best, and it seems to work.

Click to expand...

Here is exactly what they say :

Die OM-D E-M5 bietet eine hohe Empfindlichkeit von bis zu ISO 25.600. Allerdings zeigte die Messung im Labor, dass die Empfindlichkeiten, die bei ISO 200 beginnen, durchweg 1 2/3 Blendenstufen niedriger ausfallen als eingestellt. Die Kamera fängt laut Messung bei etwa ISO 120 an und endet bei ISO 14.200. Das bedeutet aber auch, dass die Kamera insgesamt eher vorsichtig belichtet und ausgefressene Lichter nicht zur Tagesordnung gehören.

Translation (as good as I can) :

"The OM-D E-M5 offers a high sensivity topping at ISO 25600. However measurements in the labs show that the sensitivities which start at ISO 200 fall 1 2/3 F stops under the setup ISO all the way up. Second our measurements, the camera starts at base ISO 120 and end at ISO14200. This means that the camera exposes rather carefully and that blown out highlights aren't part of the pack."

May be it is just a spelling error on their part, but for me going from ISO 120 up to ISO 200 there is not 1 and 2/3 stop, but rather only 2/3 stops. That is what got me confused.

Louis_Dobson said:

Getting a lot of good reviews. Mine arrives next week.

Click to expand...

Yes, reviewers seem to like it. You are luckier than me. Here none are expected before middle to end May !

--
rrr_hhh

eques said:

rrr_hhh wrote:

Click to expand...

....

eques said:

eques said:

3) they find the camera is well usable up to ISO 1600, both for noise and details.

Click to expand...

They say the signal noise ratio is just good at ISO 200 and mediocre at higher ISOs (sinking from just under 40 dB to 35 dB); however details are well preserved up to ISO 3200, at higher ISOs the NR is rather agressive.

Click to expand...

It is not exactly what they are saying. You make things appear worse than what they say :

Den Signal-Rauschabstand kann man nur als befriedigend bezeichnen. Allenfalls bei ISO 200 kratzt die E-M5 am guten Bereich, der bei 40 dB beginnt. Bis ISO 1.600 dagegen ist sie auf ausreichendem Niveau von über 35 dB, darüber unterscheiden sich Bild- und Rauschsignal nicht mehr deutlich genug. Auch die Rauschunterdrückung ist nur bis einschließlich ISO 1.600 in der Lage, das Bild einigermaßen rauschfrei zu halten, darüber steigt das Helligkeitsrauschen immer stärker an und wird sichtbar, über ISO 12.800 sogar stark. Das Farbrauschen hat Olympus besser im Griff, es taucht erst bei über ISO 12.800 auf, wo man von einer Kamera mit einem Sensor im Four-Thirds-Format aber auch keine Rauschfreiheit erwarten kann, selbst APS-C-Kameras können dies nicht leisten.

Translation :

"The signal to noise ratio can only be described as satisfying. Anyway, at 200ISO, the E-M5 is touching the good score which begin with 40 dB. Untill ISO 1600 it stays in the satisfying level above 35dB. Even the noise filter is able to reduce the noise up to and including 1600 ISO. Above 1600 the luminance noise becomes stronger and stronger and becomes visible with 12800 ISO. However Olympus hold the color noise well in hand : it begun to appear only above 12.800, where one can't expect a noise free picture from am MFT sensor. Not even the APSC sensor are able to reach it."

Then they go on prizing the high dynamical range : above 11 up to 1600 ISO included (1600 ISO seems like a critical level for a lot of IQ parameters), 10 f stops at 3200, aka surprisingly good results up to 3200 ISO.

Only with 6400 could they confirm a loss of details. And then Colorfoto concludes this paragraph on IQ, by stating that the E-M5 sensor beats (locker=loose ?) the Samsung NX200 in spite of its smaller sensor.

--
rrr_hhh

Uncle Albert said:

"The slaughter of the piggy bank to buy the OM-D E-M5 you will regret it hardly..."

Thank you, Google Translate! Made my day.

Click to expand...

That is the equivalent of to "break the bank", aka :



--
rrr_hhh