PDA

View Full Version : Dynamic range comparison


snaarman
12th September 2008, 09:15 PM
I thought I would rig up a test to compare the "Cinderella" camera (the E-400) with last year's "Prince" (the E-510).

These were shot at ISO 100, 200, 400, 800 and 1600 with identical settings. No EV compensation. Matrix metering, Aperture priority at f5.6 using the 50mm Zuiko and a tripod. They were all developed from raw with ACR using -1.0 stop compensation and +50 fill light and Tungsten lighting balance. (This replicates the type of rescue strategy I would adopt for problem pictures like this).

The spot metered range on the wall was 1/320th down to 1.3 seconds - which I calculate as something like 9 stops(!) and all the images were well white-clipped in their original form.

These are full width crops reduced to 800 pixels with very light sharpening added.

I think this shows that the E400 does recover from white clipping slightly better than the E510. It also shows the E400 left edge magenta glow at high ISO. Equally the E510 is not vice free either.

The E-400
http://e-group.uk.net/gallery/data/622/e400dr_testx.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/8158)

... and the E510
http://e-group.uk.net/gallery/data/622/e510dr_testx.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/8159)

Any thoughts 400 and 510 owners??

Pete

Zuiko
12th September 2008, 09:35 PM
Compared to film I thought they both performed rather well.

john

Ian
13th September 2008, 08:55 AM
Hi Pete,

From where I'm sitting, the E-510 shots look lighter than the E-400 in the dark end, so that would force out the highlights. But what I can clearly see at ISO 1600 in the E-510 shot is banding, and this is absent with the E-400. In my tests there is more resolution available from the E-510 sensor.

Ian

snaarman
13th September 2008, 09:23 AM
Yes, the banding (E510) and the l.h. purple glow (E400) are visible at ISO 800 and 1600. To be honest, this was such a harsh test that I would normally sacrifice the shadows into black rather than attempt so much "lift" so maybe these issues are not such a big deal.

In general they produce quite similar results up to ISO 400. It may be that the E400 CCD retains something of the old E-1's image quality mystique, equally this may be in the eye of the beholder. (Valve hifi amps anyone?)

As to resolution, in the few direct comparison shots I have tried its very hard to spot a difference. They say the E400 has a stronger optical antialias filter which would take some of the edge off things.

The E-510 is faster than the E400 is most aspects: Downloading smaller RAW files over a faster interface makes it about 3 times quicker. With IS and LV it has the edge for macro and low light.

The E400 is smaller and lighter. Take yer pick :-) For now I plan to keep and use both of them...

Pete

Ian
13th September 2008, 09:41 AM
Yes, the banding (E510) and the l.h. purple glow (E400) are visible at ISO 800 and 1600. To be honest, this was such a harsh test that I would normally sacrifice the shadows into black rather than attempt so much "lift" so maybe these issues are not such a big deal.

In general they produce quite similar results up to ISO 400. It may be that the E400 CCD retains something of the old E-1's image quality mystique, equally this may be in the eye of the beholder. (Valve hifi amps anyone?)

As to resolution, in the few direct comparison shots I have tried its very hard to spot a difference. They say the E400 has a stronger optical antialias filter which would take some of the edge off things.

The E-510 is faster than the E400 is most aspects: Downloading smaller RAW files over a faster interface makes it about 3 times quicker. With IS and LV it has the edge for macro and low light.

The E400 is smaller and lighter. Take yer pick :-) For now I plan to keep and use both of them...

Pete

I find the E-1 'Kodak' colour mystique a bit of a puzzle. The E-400's sensor is not related to the E-1/E-300/E-500 line of FFT sensors, apart from the fact that they are all made by Kodak. In other words the E-400 sensor is not an FFT type. I did a colour chart comparison between the E-1 and E-3 under controlled conditions and there was virtually no difference.

The E-400 appears to have stronger noise filtering and there is no choice of strength, which the E-510 offers. To be honest, I've only compared camera JPEG output in terms of resolution. RAW file comparison would reveal if there is a stronger anti-alias on the E-400.

Ian

snaarman
13th September 2008, 11:44 AM
OK, here's the last comparison: Time to go out and take some pictues (there is a real ale feshtival in town thish afternoon..)

Here 's a test shot to compare the sensor resolutions (as far as I can arrange it). ISO100 for minimum noise. 50mm macro lens - the sharpest lens I have. f5.6 for best results from that lens. Contrasty lighting onto detailed subject. about 1/500th, so that camera shake isn't an issue... Shot in RAW.

In converting both images I added a certain amount of sharpening in ACR: Just enough to improve the image without overdoing it one hopes. (Looks like they didn't quite agree on colour temperature - I used the "As Shot" option in ACR)

What I find is that you can end up with very similar results, but you need to add a little bit more sharpening to the E400 file. Maybe this bears out the stronger antialias filter theory..

Hope all this is useful info :-)

http://e-group.uk.net/gallery/data/622/e510restest.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/8166)

Pete

Ian
13th September 2008, 12:56 PM
OK, here's the last comparison: Time to go out and take some pictues (there is a real ale feshtival in town thish afternoon..)

Here 's a test shot to compare the sensor resolutions (as far as I can arrange it). ISO100 for minimum noise. 50mm macro lens - the sharpest lens I have. f5.6 for best results from that lens. Contrasty lighting onto detailed subject. about 1/500th, so that camera shake isn't an issue... Shot in RAW.

In converting both images I added a certain amount of sharpening in ACR: Just enough to improve the image without overdoing it one hopes. (Looks like they didn't quite agree on colour temperature - I used the "As Shot" option in ACR)

What I find is that you can end up with very similar results, but you need to add a little bit more sharpening to the E400 file. Maybe this bears out the stronger antialias filter theory..

Hope all this is useful info :-)

http://e-group.uk.net/gallery/data/622/e510restest.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/8166)

Pete

Interesting tests - thanks for sharing :)

Ian