re: A99 Fast iso test: raw and jpeg crops
still looks awful.........I have doubts about whether these tests are indicative of the A99 performance.
I will wait to have a look at some decent samples and then decide.
have you got any JPEGS ? or what about RAW converted to TIFF printed.
24,256 kb 24,320 kb and 24,384 kb
Strange that there are different sizes.
It isn't surprising to me. The A77's RAWs will vary in size slightly from one image to the next. The files, despite not being labeled as such, are really a form of compressed RAW. The other thing that can cause a slight variation is any extra information the camera is storing alongside the RAW image information (Lens corrections, GPS coordinates, etc.). Also, don't forget that RAW files usually have a very small sidecar photo that is in a JPEG/JFIF format for quick preview. That can cause the file size to vary a bit as well.
What do you mean it looks good. It looks like crap. This is not even on par with the d800 at 25600 .
There is not just excessive noise in these photos there are major artifacts which distort and corrupt the image. I have noticed the same artifacts in all video I have seen shot on the a99 and yet when I looked at the rx1 video it looked awesome.
I want to get sonys new camera so bad but I fear this is a natural consequence of the SLT.
Absolutely, the SLT design compromises output by at least 3 stops. I'd say closer to 4.
Seems like you'd be better off with those 3 extra stops at Nikon.
Do you read or just bleet? baaahhh.. Im a sony sheep... baaaaahhh.
You just said that the D800 looks better at ISO 25600 than those ISO 3200 crops. That's 3 stops. And then you state that you "fear this is a natural consequence of the SLT".
And then you expect me to take you seriously?
3-4 stops ?? nahh more like a third to one stop but it still does not account for all the noise at that ISO setting and the sensor cannot be that bad if it is the same sensor in the D600.
IDC (cough cough) conversion of highly underexposed bits, pushed a bit as well and that leads to certain conclusions?