In addition to support for the A7 / A7r, three FE lens profiles were added:
Sony Sony FE 28-70mm F3.5-5.6 OSS
Sony Sony FE 35mm F2.8 ZA
Sony Sony FE 55mm F1.8 ZA
Actually profiles for a whole bunch of Sony lenses. But the biggest deal to me is that it's crazy fast. Its like I have a new computer. I'm not imagining this, am I?
p.127 #3 · Official: Sony A7 and A7R Fullframe Mirrorless
Taylor Sherman wrote:
Bleh. Well, Sony were also responsible for lossy music compression with the MiniDisc. . .
Yeah, I immediately thought of MP3/ACC (variable bit rate compression) vs Flac/ Apple Lossless when I read about the Sony Raw compression. I don't know how much a difference, if any, it makes but would like to have the old option that Sony once offered with regard to raw options (lossy cRaw as today vs a larger lossless Raw) for piece of mind.
p.127 #4 · Official: Sony A7 and A7R Fullframe Mirrorless
I also think of double-blind ABX testing when I see all this crap about "losses in compression". So many "users" claim lots of things, but when confronted with real comparison data they tend to guess wrong more often than a random generator. Which then would basically mean that a dice (the most simple random generator except for the coin-flip, which requires more physical effort) knows more about audio/imaging than they do.
Nikon uses a very similar value-compression scheme, and even after taking requests about worst-case targets and worst case post processing for really stressing the differences I got 0% detectability with 100% statistical significance in ABX testing with reasonably discerning user groups.
They were very loud in complaining about it, but in the end they couldn't find even one visible difference. There are measurable differences that takes some numerical analysis, but so far I've had none, zero, null persons able to tell the difference by looking at identical image pairs taken with different settings.
We even went as far as allowing people to work the raw files for themselves at location, just so that I could make sure they didn't go into EXIF and check recording settings. That resulted in a zero detectability too.
The only thing that did make a difference (with Nikon then) was switching from 12 to 14 bit, since that has had some sensor readout setting effect in some Nikon cameras. 14-bit settings lowers electronic read noise in most Exmor sensors.
Most gradation and posterizing problems users have are related to bad camera and screen profiles, and raw converters not handling color profiles correctly in communication with the OS, this goes for both Win and OSX.
p.127 #5 · Official: Sony A7 and A7R Fullframe Mirrorless
Samuli Vahonen wrote:
Wow, first time enjoyed shooting A7. I was able to improve most of the handhold ergonomics issues. Solution was rather simple, but seems to work: remove FE35, insert Metabones adapter and 1.4/35. After this operation handhold become much much easier.
With the small lens I hold camera 95% with right hand; trying to squeeze desperately from "grip" (that minimalistic grip is just too small for anyone older than 10 years...) and same time pressing all those of buttons spread to random unergonomic places.
With proper size lens left hand is naturally below the lens and if needed I can release right hand and camera doesn't drop to ground. Button pushing become much easier. Also with proper size lens I was able to handhold 1/30s with 35mm and have keeper rate >75%. So 1.4/35 highly recommend with A7 by me. Otus should be good fit as well.
p.127 #6 · Official: Sony A7 and A7R Fullframe Mirrorless
arbitrage wrote:
Pointless....just use a D800E if you are going to attach giant lenses...IMHO
A7R is to the D800E for Nikon Live View shooters as A7R is to the 5DM3 Dynamic Range/Resolution for Canon shooters In other words, Nikon's implementation of LV on the D800/E is horrible and is not an insignificant reason to cross-grade.
p.127 #7 · Official: Sony A7 and A7R Fullframe Mirrorless
philip_pj wrote:
That one looks poor, sebboh. Here are couple of 24Mp a99 images, nothing fancy done. I seriously like the way the sensors/processing handle highlights in general.
Do you know if those were shot with single frame or continuous with the a99 as some have said different raw compression is used depending upon the frame rate on the a99. I guess a look at the raw file size would be a clue. If it's near 36MB, it would be the older method but if it's close to 25MB or so it would be the higher compressed cRaw type method.
I would be surprised if a visible difference would show up with a fairly straight forward raw conversion. What I'm more curious about is what occurs with gradations when one starts really pushing and pulling on the shadows and highlights in raw conversion - or pushing channels to extremes like one often does with a conversion to B&W. In those situations, I wonder if the difference in raw compression (lossy vs lossless) might show up (if bits are being thrown away in areas that do not appear to have subtle gradations in the original scene, what happens when one attempts to recover those areas during raw conversion?).
p.127 #8 · Official: Sony A7 and A7R Fullframe Mirrorless
theSuede wrote:
Nikon uses a very similar value-compression scheme, and even after taking requests about worst-case targets and worst case post processing for really stressing the differences I got 0% detectability with 100% statistical significance in ABX testing with reasonably discerning user groups.
That's great info to know and would seem to address my question about what happens with extreme post processing (at least with Nikon). I do notice that Nikon seems to compress a little less than Sony though I suppose Sony could just be using a more efficient compression method.
p.127 #9 · Official: Sony A7 and A7R Fullframe Mirrorless
snapsy wrote:
A7R is to the D800E for Nikon Live View shooters as A7R is to the 5DM3 Dynamic Range/Resolution for Canon shooters In other words, Nikon's implementation of LV on the D800/E is horrible and is not an insignificant reason to cross-grade.
Nikon's LV implementation was the single main reason I did not keep my D800E (or replace it with another as my copy also suffered AF issues.)
p.127 #12 · Official: Sony A7 and A7R Fullframe Mirrorless
Tariq Gibran wrote:
I would be surprised if a visible difference would show up with a fairly straight forward raw conversion. What I'm more curious about is what occurs with gradations when one starts really pushing and pulling on the shadows and highlights in raw conversion - or pushing channels to extremes like one often does with a conversion to B&W. In those situations, I wonder if the difference in raw compression (lossy vs lossless) might show up (if bits are being thrown away in areas that do not appear to have subtle gradations in the original scene, what happens when one attempts to recover those areas during raw conversion?). ...Show more →
Pushing the shadows normally reduces the appearance of posterization because the added noise dithers the gradations.
Both Tim and I Pre-ordered the whole L Bracket last evening. But it does seem odd that the Bottom Plate that has all of the work is listed for $75 and the whole L Bracket is $175.
p.127 #15 · Official: Sony A7 and A7R Fullframe Mirrorless
snapsy wrote:
Pushing the shadows normally reduces the appearance of posterization because the added noise dithers the gradations.
At base ISO with a Sony sensor, there is a lot of headroom (many stops) before noise even becomes visible in the shadows so I don't know how dithered noise would make a big difference. At higher ISO's/ really extreme pushes, that makes sense.
p.127 #17 · Official: Sony A7 and A7R Fullframe Mirrorless
Tariq Gibran wrote:
At base ISO with a Sony sensor, there is a lot of headroom (many stops) before noise even becomes visible in the shadows so I don't know how dithered noise would make a big difference. At higher ISO's/ really extreme pushes, that makes sense.
The Exmor has exceptionally low read noise but it takes very little noise to dither shadow gradations, esp when you consider how few raw bits are needed to represent shadow tonations in the first place (ie, noise precludes the need for extra tonal representation).
p.127 #18 · Official: Sony A7 and A7R Fullframe Mirrorless
snapsy wrote:
The Exmor has exceptionally low read noise but it takes very little noise to dither shadow gradations, esp when you consider how few raw bits are needed to represent shadow tonations in the first place (ie, noise precludes the need for extra tonal representation).
p.127 #19 · Official: Sony A7 and A7R Fullframe Mirrorless
arbitrage wrote:
Pointless....just use a D800E if you are going to attach giant lenses...IMHO
How? First; I don't own one. Second; it's live view implementation sucks and I don't use optical viewfinders? Third; I have all my lenses on EF mount or shorter lens mount to sensor distance. Just going out to shoot with few Canon FL/FD lenses...try to do that with D800E...
Also since I shoot with 2-3 bodies, it would be damn expensive to get 3x D800E and absolute no change of flying (weight regulations) with such camera backbag... A7 is really cheap, basically just price of one medium quality lens (new).
Everybody isn't looking for the same thing. Personally I see people who try to get symmetrical wide angle lenses (=high ray angle) to work with A7(r) pointless, but that just my opinion.
p.127 #20 · Official: Sony A7 and A7R Fullframe Mirrorless
arbitrage wrote:
Pointless....just use a D800E if you are going to attach giant lenses...IMHO
I don't see why this is pointless. Relatively speaking how is a relatively large lens on an A7/R that much different than a relatively large lens (most telephotos) on a D800E?
In other words, Nikon's implementation of LV on the D800/E is horrible and is not an insignificant reason to cross-grade.