I'm referring to the Sony A700 review at dpreview.
They were saying that noise reduction was applied to the RAW files of the A700, and therefore making the use of any other noise reduction software difficult.
I'm just repeating what they were writing, but can't tell if they are right or wrong.
Camera tests should always be taken with a decent amount of salt when it comes to esthetics of image quality, but this would be a technical detail that can be measured.
Obviously, douglasf13 thinks it is a false statement.
I'm just trying to say it's very unclear. That dpreview has stimulated quite a discussion about this topic. It's relevant to me, because the "a900" chip appears to be the same general design, and I plan on buying it.
A few things about the dpreview. Phil and Simon used ACR for that review, which has been blasted for it's handling of high ISO A700 files for months and months. Many Sony users have abandoned Adobe until recently, and it looks like Adobe has finally straightened things out in the latest software updates. Whew, I can use lightroom again
The dpreview assumption that NR is applied to the RAW is more or less based on two things: the poor output of of the older ACR, and this illustration (image also attached:) [a href=\"http://a.img-dpreview.com/reviews/SonyDSLRA700/images/features/processing2.jpg]http://a.img-dpreview.com/reviews/SonyDSLR...processing2.jpg[/url]
In this illustration, you'll notice THREE stages of RAW NR!! An analog, a digital, and a final BIONZ NR. The final two stages are unique to Sony AFAIK. Here is where things get tricky. The A700 has a low, normal, and high NR setting that kicks in at around ISO 1600, and DOES affect RAW. However, we don't know where in that signal path the adjustments are made.
At PMA, Sony reps were grilled about this quite a bit, and the conclusion that was drawn was basically that, when set to low (with firmware v 3.0) there is no more NR that Sony can turn off with another firmware update, and that should lead us to believe that the BIONZ, which is firmware adjustable, is OFF at this setting, but there is still a certain amount of non-defeatable on-chip NR going on, which is common with all CMOS cameras...albeit in the Sony it's probably the unique, digital, adjustable, on-chip NR! It seems that having an adjustable, on-chip digital NR is new, and it is highly touted by Sony for better or worse, but it shouldn't be considered "cooked" any more than the analog NR applied right before it. Sony only really knows what's going on here, but if it can be proven that the BIONZ processor (3rd stage) adds RAW NR even in the lowest setting, I'll certainly concede.
I say that it's semantics because had Sony labeled the NR options as off, normal, and high in the software, there would be a lot less confusion. Kinda like my amp goes to ELEVEN! lol.
If dpreview were to re-do the high ISO A700 test with the new ACR, it would be a bit different story. Like I said earlier, the D300 has a finer grain, A700 has better color. Take your pick. I still prefer the D300 a bit, but it's really close. Oh yeah, also, the D300's ISO sensitivity is a bit off from the A700. In identical shooting scenarios, with the same ISO and aperture, the D300's shutter speed tends to be about a half to full stop slower than the A700 to get proper exposure. ie. D300's ISO 3200 is equivalent to the A700's ISO 1600 or so. It's not appropriate to compare their ISOs straight across.
Wow, sorry for the long post. I just wanted to point out a few things, since it looks like our little Sony cams are gonna be brought up a bit more in this forum. We all know Canon and Nikon make top flight gear, and I'm hoping Sony will at least be in the conversation over the next couple years. Cheers! -d