EB-1 wrote:
Placement of the EVF is not so critical as it the image viewed. In bright light they just look harsh/grainy and highlights/shadows are not what the eye sees with a good pentaprism. If the Sony sensor has such a great DR, why is the EVF so much worse? Is it a cost issue?
EBH
lightskyland wrote:
Did you set contrast to -3, turn DRO on, and reduce sharpening? Assuming you were shooting RAW.
That opens up the EVF shadows a ton.
EB-1 wrote:
Thanks. No, I was not familiar with it. Is there any chance they will improve the graininess/waviness? It's somewhat like the old fluorescent lights flickering.
EBH
mogul wrote:
Lucky when using the EVF, you will be able to see and get the picture, unlike the OVF which will be completely dark and you won't even try.
I'm not sure what that means. Of course an EVF may be good for astrophotography or caves or some extremely dark area, but not anything I'd normally encounter. A camera built for the <1% exceptions is not for me. Ideally there should be a hybrid of some kind that would only suck power when needed and still allow extended optical viewing without power.
EB-1 wrote:
I'm not sure what that means. Of course an EVF may be good for astrophotography or caves or some extremely dark area, but not anything I'd normally encounter. A camera built for the <1% exceptions is not for me. Ideally there should be a hybrid of some kind that would only suck power when needed and still allow extended optical viewing without power.
EBH
Nikon has a patent on such a system - optical TTL when the mirror is down, EVF using the same viewfinder optics when the mirror is up. That would be very very cool if that is workable.
ClausC wrote:
I have decoded Canon CR2 file at one time, the raw-part of the file was encoded according to ITU T.81 Loss Less Compression and does not involve JPEG. It compresses the raw-image a factor of 2 (approx).
kwalsh wrote:
Um, you do realize that ITU T.81 is in fact the specification for lossless JPEG compression. Which is what was being referred to. You basically proved yourself wrong with your own post and reference to the spec.
Actually ClausC was replying to my earlier statement that lossless JPEG worked by performing a regular lossy JPEG compression and then entropy encoding the difference between that and the original data. It actually works by doing an entirely different lossy operation and then entropy encoding the difference, that's what he meant by it "doesn't involve JPEG", i.e. it doesn't do a DCT transform and quantise the result at any point.
kwalsh wrote:
And those thinking the Sony file is equivalent to 11 bit also are fairly clueless. But that said clearly in some circumstances it leaves artifacts which is frustrating to those it impacts.
Indeed, there's a huge difference between a linear 11-bit encoding and a the nonlinear 11-bit encoding used by Sony, which is much closer to a linear 14-bit encoding.
Anyone with an A7 II know if the back part of it is also magnesium alloy? Most of the images I've seen only show the mount, front, and top being metal and the back is apparently plastic.
This image from DPREVIEW, however, shows the back of the A7r II is also magnesium alloy:
It does take the A7II vertical grip. So I'm assuming it's the same exact dimensions but would not be surprised if it's better materials in some areas over the A7II. The EVF is also bigger and they did put a lock on the mode dial. That's all we know at the moment
Matt Grum wrote:
Actually ClausC was replying to my earlier statement that lossless JPEG worked by performing a regular lossy JPEG compression and then entropy encoding the difference between that and the original data. It actually works by doing an entirely different lossy operation and then entropy encoding the difference, that's what he meant by it "doesn't involve JPEG", i.e. it doesn't do a DCT transform and quantise the result at any point.
Ah, thank you for clarifying! And apologies to ClausC that I wasn't able to follow who was referring to what in the big long thread!
Yes, that whole confusion of what exactly someone means by "JPEG". And indeed lossless JPEG doesn't use DCT as part of its algorithm. I see now in the context ClausC was using JPEG as short hand for DCT.
with pre-orders starting Wednesday, anyone want to make an "over / under" on when they actually start shipping?
i think SAR (I don't even know why i mention this b/c i think he just pulls "stuff" out of his arse) a few wks back stated ("guessed") that inventory for a new camera was building up and would be ready to go shortly after the announcement.
I'm certainly hoping the a7r2 is available within a month, not the 2 months that was indicated. And hopefully both Batis start shipping in 2 wks as well! (has anyone heard anything about how that release is coming?)
Against popular beliefs, the aperture inside Nikon lenses is still mechanical and it is activated by a lever in the body. I am not sure whether or how adapter manufacturers can design an adapter that includes both a mechanical and electronic portions in the same unit. I am sure it could be done but at what cost? And how much of battery power will be used to achieve this? The battery charge is one of the many negative factors of the A7 series cameras...
sjms wrote:
in the end it still puts out an 11bit lossy raw chunk of data per shot. so all that up front stuff is essentially nulled out. with all that overwhelming technology why do the put the results out in a crushed form?
Matt Grum wrote:
Indeed, there's a huge difference between a linear 11-bit encoding and a the nonlinear 11-bit encoding used by Sony, which is much closer to a linear 14-bit encoding.
When you look objectively at it, a camera with a 14Ev dynamic range really only needs 10 bits of encoding to utilize all of the information inherent in the base material. DR is not tone resolution.
But normally you have some shortcuts in the application of the gamma-encoding, so you get some quantization losses from lowered precision. Hence the doubling of precision up to 11 bits.
IMO 11 bits of gamma-compressed data should cover all sensor needs up to 100% efficient FF sensors of 20MP+ pixel density. You MIGHT need 12 bits for a 10MP FF sensor. This depends on the full well capacity of each pixel - anything up to 100k e- per pixel is well within a 10-bit gamma coding step. I think both Sony and Nikon's variations on the gamma-encoding schemes are totally ok.
Sony does seem to have slipped up somewhat on the OTHER part of the compression scheme though. But even though I often maul my files to quite some extremes, I've yet to find the quantization accuracy loss around very high contrast edges to really have an effect on what I do.
On startrails and other types of scenarios when you can have a 6-8Ev contrast that's bright-detail-on-dark-background with extreme sharpness, it is indeed visible if you push the files hard. That is of course a bad thing for those who need to do that on some occasions, and that shouldn't be belittled as I've seen on some occasions.
I do however doubt that any normal photographer shooting PJ-type or even HQ landscape photography will ever be able to A-B-ref test the results from compressed/noncompressed reliably on natural subjects, even with the lossy compression Sony uses. Some architectural work might be affected in a very slight manner, but that's about it.
The back part of the camera appears to be the same material as the rest of the body, at least to my eye and touch.
johntruong wrote:
Anyone with an A7 II know if the back part of it is also magnesium alloy? Most of the images I've seen only show the mount, front, and top being metal and the back is apparently plastic.
This image from DPREVIEW, however, shows the back of the A7r II is also magnesium alloy:
