Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
A77 sensor on dxomark
#1
Take off the 1/2 light lost to mirror, and Nex-7 might even score higher than D7000/K-5



[Image: 6175505931_bb5efccdc6_o.jpg]
#2
[quote name='Sathe Wild' timestamp='1316806937' post='11796']

Take off the 1/2 light lost to mirror, and Nex-7 might even score higher than D7000/K-5

[/quote]

You only lose 1/3rd of a stop.



Concerning the data (and scores)... unaltered RAW always shows a curve. The A77 results show a straight like, except at the highest ISO settings. This means we do not see the sensor performance, but rather doctored data where noise has been filtered (at every ISO, probably except base ISO).



The D7000 curve shows relatively low doctoring of the data (however a slight loss of detail at higher ISOs does point to some doctoring), the K5 heavy doctoring only from a certain value.



Not sure why DXO only sometimes "detects" this. But they do not take it into account with their scoring. So, don't place too much importance in the DXO sensor scores, especially from the A77.



Anyway the camera seems to perform fine... no need to look at what DXO says it does.
#3
[quote name='Brightcolours' timestamp='1316809032' post='11799']

You only lose 1/3rd of a stop.

[/quote]



It's 1/3 light and that makes it 1/2 stop. See A55 vs A580 score difference.



The sensor without pellicle mirror is likely to score higher than D7000/K-5
#4
[quote name='Sathe Wild' timestamp='1316810469' post='11801']

It's 1/3 light and that makes it 1/2 stop. See A55 vs A580 score difference.



The sensor without pellicle mirror is likely to score higher than D7000/K-5

[/quote]



Nikon users will still not migrate. Sony has surely a competitive edge on the camera side now but I wouldn't sign this with respect to their lens lineup. Frankly I'd say that they've the weakest offerings among the classic 4.



This is also the biggest constraint for the NEX7 in Sony E land.
#5
[quote name='Brightcolours' timestamp='1316809032' post='11799']Concerning the data (and scores)... unaltered RAW always shows a curve. The A77 results show a straight like[/quote]



Nothing strange, that's due to A/D converter structure on Sony sensors. They have patent rights on that and I doubt we will see that technology on Canikon sensors.





[quote name='Sathe Wild' timestamp='1316810469' post='11801']

The sensor without pellicle mirror is likely to score higher than D7000/K-5

[/quote]



Probably it can score higher than D7000/K-5 but if you look at graphs this 24MP sensor and 16MP Sony sensor have almost identical performance.
#6
[quote name='Bare' timestamp='1316811626' post='11804']

Nothing strange, that's due to A/D converter structure on Sony sensors. They have patent rights on that and I doubt we will see that technology on Canikon sensors.



[/quote]

Nothing strange? Very strange. The 16mp Sony shows the normal curve in the D7000 (not on the K5 as it does heavy NR from a certain ISO and up). Just like the Canon 18 mp sensor, really. And that Sony sensor has that quiet ADC already. This A77 shows a straight line, not the usual curve, and also a tell tale angle with the highest ISO setting.



We will most probably see that normal curve again when it is being used in the Nikon D400...
#7
[quote name='Brightcolours' timestamp='1316818577' post='11806']

Nothing strange? Very strange. The 16mp Sony shows the normal curve in the D7000 (not on the K5 as it does heavy NR from a certain ISO and up). Just like the Canon 18 mp sensor, really. And that Sony sensor has that quiet ADC already. This A77 shows a straight line, not the usual curve, and also a tell tale angle with the highest ISO setting.



We will most probably see that normal curve again when it is being used in the Nikon D400...

[/quote]





Dxomark score is calculated at the base ISO and lowlight score (for A77) is calculated at ISO 800



Any curve at ISO 3200 has no affect on dxomark score.
#8
[quote name='Sathe Wild' timestamp='1316849010' post='11807']

Dxomark score is calculated at the base ISO and lowlight score (for A77) is calculated at ISO 800



Any curve at ISO 3200 has no affect on dxomark score.

[/quote]

Funny thing is that when you look at real (jpg) images of the alpha 77 at imaging resource or dpreview they seem considerably noiser with blocked up shadows compared to the 7d, D7000, not to mention the 5D Mark II. That leaves only two explanations: Either sony have messed up their jpg processing completly or dxo just continues to prove the irrellvance of their measurments for practical applications. (I presume the latter is the case). In particular the statement that the alpha comes extremly close to the 5D mark II seems rather ridiculous in view of the real life images floating around on the web.
#9
Oh my god they have done it again: Heres another reason why its best to ignore DXO:



http://www.dxomark.com/index.php/News/Dx...and-review

I believe I dont need to comment their finding that the older 70-200 L IS has better resolutuion than the newer Mark II version)
#10
[quote name='jenbenn' timestamp='1316959907' post='11842']

Oh my god they have done it again: Heres another reason why its best to ignore DXO:



http://www.dxomark.com/index.php/News/Dx...and-review

I believe I dont need to comment their finding that the older 70-200 L IS has better resolutuion than the newer Mark II version)

[/quote]

DXOmark's lens tests are either highly unreliable, or just plain nonsensical. If just highly unreliable, lets put them together with SLRgear.
  


Forum Jump:


Users browsing this thread:
1 Guest(s)