a580 beats a55, matches D7000 DXO reports

[Dusty]

Just checked SonyAlphaRumors for the latest rumors, and they lead off today with the DXO Labs report that the 580 beats out the 55 and essentially matches Nikon's D7000 in IQ.

So it seems that the quality hit of the SLT is measurable.

Dusty

[Scooterman]

I have had a look at those figures as well and it just proves how good the 580 is and a fraction of the price of the Nikon.
Also I have seen reviews of the Nikon 7000 and it can tend to over expose where as I have found the 580 is just about right also much more head room in RAW over the A-550, I have both.
The AF is superb on the 580 perfect, for my moving wild life shots using a Sigma 100-300mm f4 or the Tamron 200-500mm.
Regards
Richard

[bfitzgerald]

Another day another DxO score to work folks up

I just feel sorry for the guy who bought a Canon 60d as that scored not that well on the site

[Scooterman]

Another day another DxO score to work folks up

I just feel sorry for the guy who bought a Canon 60d as that scored not that well on the site

You don’t really Barry you are just having a snigger.
Mind many Canon users are disappointed in the 60D, I believe they were expecting better.
Richard

[alphaomega]

Lots of people take good pictures with the A55 and David has it as his main camera now. These new cameras are now so good that the fact that the A580 may be marginally better than the A55 in the laboratory is of no real consequence. You want a smaller camera, 10 shots per second, better video etc. go for the A55 and you will get good pictures regardless of what the A580 or competition can do.

[Greg Beetham]

Yeah well the DxO mark as someone previously pointed out rates sensors by noise mainly, and the A55 looses about a third of a stop against the A580 and that seems to be about the only major difference as far as image quality goes..and the 60D is slightly noisier than either but looking at the images in real life it seems to be fine, very very fine actually.
Greg

[bakubo]

Mind many Canon users are disappointed in the 60D, I believe they were expecting better.

Got to say that over on dpreview the 60D gets very few complaints by people who have it. First time I have ever seen a new Canon or any other brand not get quite a few posts about problems after it has been in people's hands awhile. I have had quite a few DSLRs over the last 8 years and without a doubt this is the best of the bunch with no QC problems so far -- knock on wood. So far, I have not found a single thing about it that annoys me and that's a first for a digital camera. Of course, then there are always the gear weenies who work themselves up into a lather worrying themselves and everyone else to death about something on forums.

[bfitzgerald]

Greg nailed it that's the problem with DxO the numbers could lead you to draw conclusions say high ISO performance that simply don't matter very much in the real world. You could say wow xyz camera beats xyz camera by hundreds of points for low light but in reality the difference could be fairly small.

[Neonsquare]

The simple thing is, that the manufacturers all got their things together very well. Practical Image Quality already was a minor distinction in the last generation and it is now so marginally that even benchmarks like DxO really lose their very last bit of value when the question is what camera one should buy.

A little (!) surprise to me is, how close the DxO values of the A580 really are against the D7000. The often mentioned point, that Nikon may make a better use of Sony Sensors by having better software and perhaps signal processing is now not justified anymore. Were will be further innovation here? Will there be more investments in lens engineering again?

[Greg Beetham]

Not a bad little question, they will have to dream up 'something' to make buying a new camera attractive, maybe if they make em to begin failing after a couple of years...nah they wouldn't do that....
Greg

ps. There was a suggestion in a discussion at dpr about whether Sony should maybe start using 14 bit like Nikon do and squeeze a few more drops of DR out of the sensors.....

[Neonsquare]

@ 14 bit
This is a really old thing. It was always handled as the reason of Nikons technical edge over Sony when using the same sensors. The current results show, that this is maybe not necessarily the case. One might argue, that Sony would do much better now when putting 14 bit on top, but I actually start to doubt this. My guess is, that Sonys sensors up to now are built to being processed with 12 bit and they just learned enough tricks in software signal processing to get to similar results like Nikon. AFAIK DxOs computation of DR is very loose in regard to the noise level and one SWAG of me always was, that Nikons noise processing works by having a lower noise level, getting more noise from the sensor into their signal processing. This worked perfectly well with the older sensors (e.g. D90) but doesn't seem to work that well with this new one.

[alphaomega]

neonsquare wrote

AFAIK DxOs computation of DR is very loose in regard to the noise level and one SWAG of me always was, that Nikons noise processing works by having a lower noise level, getting more noise from the sensor into their signal processing. This worked perfectly well with the older sensors (e.g. D90) but doesn't seem to work that well with this new one.

I think this is David Kilpatrick territory. I remember him writing something about the difference between Sony and Nikon sensor noise output. I can't remember where, but I think it went along the lines of Nikon set their black point higher than Sony and got better high ISO output, whereas Sony's ISO100-400 or so output was richer and more pleasing. In other words Sony scored at low ISO and Nikon as you progressed up the ISO scale. I predict that there will be a DK clarification in next to no time as I have probably got it wrong.

[alphaomega]

Reference my views just above, that was perhaps a couple of years ago with A350 and A900 comparisons with Nikons. It is old hat now with the high ISO performance of A580, NEX-5 and A55 to mention the current "top" cameras.

[Edgars L]

I'm confused about DxO scores.
If they testing sensor (as they clamed), they should move mirror away and test only sensor. How the same sensor could get different score? How sensor could lose DR just with translucent mirror? It is corrupted by different processing?

With A55 mirror in the way, sensor reaching 2/3 stop less light or overall 7 DXO points, then how much points I will lost, if I stop down my lens from f2.8 to f8???

[douglasf13]

I'm confused about DxO scores.
If they testing sensor (as they clamed), they should move mirror away and test only sensor. How the same sensor could get different score? How sensor could lose DR just with translucent mirror? It is corrupted by different processing?

With A55 mirror in the way, sensor reaching 2/3 stop less light or overall 7 DXO points, then how much points I will lost, if I stop down my lens from f2.8 to f8???

There are a lot of things on top of a sensor, and, since the mirror on the A55 isn't supposed to be moved for shooting, it makes sense to test with the mirror in place.

As for DR, flare and internal reflections negatively affect DR, and anytime you place another piece of glass in the equation, more flare is introduced. Some only shoot prime lenses with 3-5 elements, because more complicated lens designs with many elements negatively affects DR.