The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

E-5 Noise = E-30/ball park E-3

raist3d

Well-known member
http://translate.google.com/transla...43.1-Test_aparatu-Olympus_E-5_Wst%C4%99p.html

After DXo came forward with the Olympus E-5 measurements, many could not believe the results and some others blamed Dxo as unreliable/bad/whatever.
Turns out yet another review (linked above) validates what DXo found checking RAWS and most importantly- comparing them with the E-30 and E-5.

RAW shooters beware: you are getting pretty much the same E-30 DR/noise performance with a little more detail at high iso (and noticeably more detail at low iso). If you are a high ISO shooter and heard that the E-5 has like one more stop DR, that probably came from some JPEG to JPEG engine comparisons, which doesn't show the potential of the camera for you.

For a jpeg shooter of course this is a nice advantage. the E-5 JPEG engine is really nice, but certainly there's only so much you can do with the original data. I think Olympus does superb with it though (JPEG).

The review pretty much validates DXo's results.

- Raist
 

raist3d

Well-known member
i dont agree
Not sure how that proves your point. The link I gave is to the E-30/E-3 not the E-PL1. DXo does mark the E-PL1 a bit noisier and with less DR than e30 or e-5. It's like the e-620 is a tad worse than the e-30, but they are pretty much ball park overall (i.e. no "miracle 1 stop better").

- Raist
 

Riley

New member
lets get something straight here Raist

You proposal depends upon DxO measurement that all the 12Mp sensors are alike in noise yes? That is indeed how they measure at DxO, and the EPL1 and E5 sensors are no different, indeed they track each other almost perfectly, certainly at 1600 ISO. Are you with it that far?

Now explain to me why the EPL1 looks noisier than the E5
 

Riley

New member
ok 1 more
this is the K5 review being plundered and used for no good reason but to display EPL1 at ISO1600 and E5 at both ISO1600 & 3200. This time its the colour chart that is in the shade between two bottles

 

raist3d

Well-known member
"
lets get something straight here Raist

You proposal depends upon DxO measurement that all the 12Mp sensors are alike in noise yes? That is indeed how they measure at DxO, and the EPL1 and E5 sensors are no different, indeed they track each other almost perfectly, certainly at 1600 ISO. Are you with it that far?

Now explain to me why the EPL1 looks noisier than the E5
Actually DXo put the EPL1 with more noise than the E-30 and E-5.

As for noisier there can certainly be many reasons including the processing, the circuitry around it. But the E-30 according to DXo does do better and very close to the E-5. Just like when I mentioned the E-30 did a hair better than the 620 but nobody believed me.

Moreover, if you are going to show this, then you can't deny the link I had put in about the other review where it shows the E-30 and E-5 and E-3 side to side right?

Dpreview's shot for the E_PL1 also looks a bit under exposed.

Or maybe Olympus changed the tone curve for one camera vs the other. Or both.

- Raist
 

raist3d

Well-known member
ISO1600 for E5 and EPL1 are just about the same in DxO
But DXo did rate the E_PL1 lower in DR and in high ISO. There's more to it than just that graph stand alone.

Anyway I already explained. Olympus on the lower end cameras doesn't provide the same level of the higher quality of the expensive models. The E-30 does better than the E_PL1, and I contended a while back, a tiny bit better than the e-620.

DXo shows the E-5 does do better than all of the current Olympus 4/3rds by a hair (over E-30). E-3 doing a hair better in some other areas.

- Raist
 

Riley

New member
But DXo did rate the E_PL1 lower in DR and in high ISO. There's more to it than just that graph stand alone.

Anyway I already explained. Olympus on the lower end cameras doesn't provide the same level of the higher quality of the expensive models. The E-30 does better than the E_PL1, and I contended a while back, a tiny bit better than the e-620.

DXo shows the E-5 does do better than all of the current Olympus 4/3rds by a hair (over E-30). E-3 doing a hair better in some other areas.

- Raist
I think my point is, that the difference is in the visible noise in images (why else would we be interested?), that difference is greater than the differences recorded in DxO, and that therefore describing the noise characteristics as close to the same, in reality just doesnt stack up.

 
Last edited:

raist3d

Well-known member
But you are making the assumption that nothing went wrong on dpreview or methodology. The E_PL1 shot is clearly under exposed.

Given DXo makes software for RAW conversion, given their ratings have consistently matched my experience with the cameras I have, I certainly give DXo the benefit.

But, Dxo did point out that the E_PL1 performs worse than the E-30 or E-5. This is important to note, because it goes in line with what Dpreview is showing also.

(i..e I don't think they are really contradicting each other per se, and it's possible with dpreview's methodology the problem will show more in the way you are linking that it shows).
 

raist3d

Well-known member
Moreover I want to add, DXo does point that the E-5 does a hair better than the E-3. That's what I would expect about "tweaking" the same sensor. Many at the other forum are claiming they are seeing "at least 1 stop" or as ridiculous as 2 stops better DR or noise, and most of the data they have been saying this from in many cases has come from comparing JPEGS.

It is this kind of faulty reasoning that I put on critique- not the camera itself (but most think I am just putting the E-5 down). For a sensor tweak to give a full one stop of DR more or noise is quite simply, science fiction-delusional. It's the same sensor, there aren't going to be any super miracles on that end.

The resolution gain with AA is far more real because as we know AA filters eat resolution and the E-620/E-30 had a relatively strong one.
 

Riley

New member
Moreover I want to add, DXo does point that the E-5 does a hair better than the E-3. That's what I would expect about "tweaking" the same sensor. Many at the other forum are claiming they are seeing "at least 1 stop" or as ridiculous as 2 stops better DR or noise, and most of the data they have been saying this from in many cases has come from comparing JPEGS.
theres easily something like a stop on some occasions at sub 640 ISO, that difference becomes less apparent at high ISO where the image seems to look like its falling apart. Look at skies etc (its also much nicer to shoot skies)

It is this kind of faulty reasoning that I put on critique- not the camera itself (but most think I am just putting the E-5 down). For a sensor tweak to give a full one stop of DR more or noise is quite simply, science fiction-delusional. It's the same sensor, there aren't going to be any super miracles on that end.

if you are operating from the basis that the sensor is identical, i can see why you might think that, you are bound to see 'faulty reasoning'. I have an E5 and Im thinking a number of things are different, including the colour gamut, and perhaps the transparency of the bayer layer. No reviews seem to look at this

http://forums.dpreview.com/forums/readflat.asp?forum=1018&message=36969959

This is known to give better high ISO performance, and is one of the reasons that other marques suck at jpegs. Yet Im thinking the electrical timings on the sensor are going to convince DxO principles that the sensors are the same, in as far as that is concerned they very well might be.

I have no idea how DxO gather data, and im quite sceptical of how others interpret it (with good reason), so Im not entirely convinced by their charts or the worth of them and in part my questioning of your view was about that dependence on DxO data. Theres so much FUD involving DxO I start to switch off right away, just as the moire aliasing scare went around.

Images to me are another matter, as we can see that for ourselves, and what I see is less noise. Ive tried to relay that as best I can with both lit and shadow crops, yet you seem unconvinced. Im not too disturbed if you remain so, but Im likely to cast my own view. The 'hair better' you see in DxO Im afraid I dont, and it is not adding up to the differences I see in images of the same scenes.

The resolution gain with AA is far more real because as we know AA filters eat resolution and the E-620/E-30 had a relatively strong one.
you do realise that the additional resolution will give you more tolerance for NR. On that basis it has better noise performance
 
Top