A7RII initial thoughts and images

[adrian tyler]

If you look at that horrible image I posted, with the inverted channels, you can see how well my junkworthy 1Ds3 discriminates the  *color shades* inside the leaves. I have no particular love for this camera, but you are doubtless aware of its reputation, uses and limitations: In its day it was Canon's studio and MF replacement product, a slow camera whose images degrade around ISO 400, but intended to convey nuance.

Here is a similarly inspired treatment of Quentin's image above, and you can see there is simply not that much color information in that image. My take on this is that Sony decided to capture and encode as much luminance information as they could in the available 14 bits (?) of DR, with the intent of allowing people to shoot for movies in available light, and they set the design priority on capturing and encoding luminance information rather than color differentiation. A different set of priorities would have meant 2 stops less DR and more precise color, which is basically the tradeoff on those old CCD backs and early studio oriented CMOS cameras. Contrarywise to the way everyone else here expects the laws of physics to be waived, I still believe in conservation laws in physics, and the in the no free lunch theorem.

In the full afternoon shadow in Paris, my shot was soft as with difficulty I eked out 1/80@F8@ISO400 on my ancient 1DsIII, precisely because the color filters on this camera are very orthogonal so it is not very sensitive, and to shoot it higher means destroying the color.
 
Do some more testing, by all means, but I think I've now seen enough images here and elsewhere to reach a conclusion; however every camera has advantages and disadvantages, and this one can certainly see deep into both shadow and sky, in a unique way. If it is at the price of some shades of yellow-green and green and blue green which most men would not percieve, who cares?  In exchange, one will get well defined and textured images of the clouds above a landscape. The A7IIR is certainly the camera of the year. And if I may be allowed to say so, I think you have a very nice sample of an extraordinary 50mm lens that is almost symbiotic with that body.

I'm sure everybody here will tell you that I'm an idiot who understands nothing about digital cameras and even less about color. You can come back in a few months and tell us what you have learnt from experience

Edmund

thanks edmund, very interesting post.
adrian

[Quentin]

Edmund

Very interesting.  More a sage, certainly not an idiot!

If you are right, then it's is the exact reverse of a camera like the foveon-equipped Merrill cameras I have long been a fan of.

[NancyP]

There's a choice: Foveon Sigma Merrill camera or old-style CCD MF camera, top usable ISO in color of 400, but fabulous fine color differentiation
vs. Sony, top usable ISO in color of ?6400  or ?12800 but less color differentiation

Speaking of which, the Pentax 645Ds are now selling for under USD 4,000.00

[ErikKaffehr]

Hi Ronald,

I think you are mixed up on this. DR has nothing to do with colour. DR is just about (gray) levels above some noise threshold. It may be that the Sony sensor gives up some color separation for better high ISO performance, but you may also ask if there are some colours to separate. Excessive separation of colour is also just another artefact.

Best regards
Erik

If you look at that horrible image I posted, with the inverted channels, you can see how well my junkworthy 1Ds3 discriminates the  *color shades* inside the leaves. I have no particular love for this camera, but you are doubtless aware of its reputation, uses and limitations: In its day it was Canon's studio and MF replacement product, a slow camera whose images degrade around ISO 400, but intended to convey nuance.

Here is a similarly inspired treatment of Quentin's image above, and you can see there is simply not that much color information in that image. My take on this is that Sony decided to capture and encode as much luminance information as they could in the available 14 bits (?) of DR, with the intent of allowing people to shoot for movies in available light, and they set the design priority on capturing and encoding luminance information rather than color differentiation. A different set of priorities would have meant 2 stops less DR and more precise color, which is basically the tradeoff on those old CCD backs and early studio oriented CMOS cameras. Contrarywise to the way everyone else here expects the laws of physics to be waived, I still believe in conservation laws in physics, and the in the no free lunch theorem.

In the full afternoon shadow in Paris, my shot was soft as with difficulty I eked out 1/80@F8@ISO400 on my ancient 1DsIII, precisely because the color filters on this camera are very orthogonal so it is not very sensitive, and to shoot it higher means destroying the color.
 
Do some more testing, by all means, but I think I've now seen enough images here and elsewhere to reach a conclusion; however every camera has advantages and disadvantages, and this one can certainly see deep into both shadow and sky, in a unique way. If it is at the price of some evanescent shades of yellow-green and green and blue green which most men would not percieve, who cares?  In exchange, one will get well defined and textured images of the clouds above a landscape. The A7IIR is certainly the camera of the year. And if I may be allowed to say so, I think you have a very nice sample of an extraordinary 50mm lens that is almost symbiotic with that body.

I'm sure everybody here will tell you that I'm an idiot who understands nothing about digital cameras and even less about color. You can come back in a few months and tell us what you have learnt from experience

Edmund

[Kevin Gallagher]

Don't forget the US prices are usually exclusive of tax which is added at checkout.
Plus if history is anything to go by, the launch price often falls a chunk within 3 months. One of the few exceptions was the E-M5 and that was because of its near-cult status creating a back-order, though it seems as though the A7RII is heading the same way from all the chatter and hype at the moment.

For reference, the 5DIII is £2,300. In US the prices are 2,600 (5D3) and 3,200 (A7RII) so the differential is less (BH prices) in UK than US which is interesting.

Hi Mike, there is one saving grace for some of us in that we can order from an online dealer who is out of State and not pay sales tax (at least not yet!). If you coupe that with the ofter times free shipping its a pretty good deal for we in the US.

[eronald]

Erik,

 I am often confused, but maybe not today. Think about it a bit. The GedankenExperiment that is appropriate is to compare the  behavior of ONE cell of a sensor with some CFA filter to that of a cell from a sensor with no filter. Or if you prefer, you could look up the comparison data between the normal M and the M no chrome

 And anyway, whether I'm right or wrong, does it matter to *you*? Can *most men* who buy dSLRs (not young women) actually see the difference between a VERY good camera and a decent camera such as the A7RII? Do you have such good color vision yourself? Is your computer system good at showing you colors?

I suggest the following test

http://www.xrite.com/online-color-test-challenge

We can go to the usual discussion about the Luther Ives condition etc, but I don't think it would help as I don't understand all that stuff anyway.

Edmund

Hi Ronald,

I think you are mixed up on this. DR has nothing to do with colour. DR is just about (gray) levels above some noise threshold. It may be that the Sony sensor gives up some color separation for better high ISO performance, but you may also ask if there are some colours to separate. Excessive separation of colour is also just another artefact.

Best regards
Erik

[NancyP]

Well, that (X-rite) was entertaining. Rubik cube-ish color puzzle. Reminds me a lot of trying to fine-read the better quality pH strips, something I used to do a lot. I am glad something is perfect about my high-myopia, esotropic, floater-bedevilled eyes. (Floaters only somewhat irritating, mostly when looking through microscope, waiting for annoying floater to cross the field of vision).

[eronald]

Well, that (X-rite) was entertaining. Rubik cube-ish color puzzle. Reminds me a lot of trying to fine-read the better quality pH strips, something I used to do a lot. I am glad something is perfect about my high-myopia, esotropic, floater-bedevilled eyes. (Floaters only somewhat irritating, mostly when looking through microscope, waiting for annoying floater to cross the field of vision).

Here's another just for you  http://igame.com/eye-test/

Edmund

By going here: http://www.fredmiranda.com/forum/topic/1379628
I found this video (last in series) https://youtu.be/0o6olPXUcNU

I think the color is good, not incredible but really good IMHO. Texture is also good.

Another plus point for the A7R2 is probably complete color compatibility with Sony's  FS7 documentary Super 35 camera
 
Brands N and C are not going to be happy.

I suspect the next shoe to drop will be some sort of still/video hybrid that is marketed as a prosumer video camera. The gap is now too large between the FS7 and the AX100, and the A7RII is not really saleable to the institutional video crowd who want to put up a camera to film a speech or a marriage or a football match.  In fact, there are already rumors that there will be no dSLR A7s replacement, but rather some sort of video body.

Edmund

[adrian tyler]

... and how much of that information is lost in an inkjet print? let alone 4 colour offset litho, surely what we capture has some other purpose than being viewed on a screen. it's interesting and useful information now that all of us photographers handle our own repro, but the really ineresting thing for me (because i dont hammer nails with the thing all day long) with this new sony system is it's size, flexibility and the freedom it affords. i'm sure that working with raw files from most of the current pro cameras we can adjust them to our (or our clients) requirements ...

[eronald]

I'm male, 66, with aging eyes that are dependent on specs for close work and distance. I also have UV damage and the beginnings of macular degeneration. I've been working with colour on a critical and daily basis since the age of 16.

I've taken a number of these colour tests over recent years and have achieved perfect scores on all, but, and it's a big but, only when using a high quality and calibrated monitor. The scores when using laptops have on occasion been less than perfect.

No surprise. I get mediocre results when I use my calibrated "retina" laptop, which is pretty much always. The Apple laptops seem to be getting worse and more unpleasant colorwise in each generation, for much the same -or opposite- reason as the cameras: Apple wants to crank the subjective brightness as high as possible while conserving backlight power. I think there even was a class action lawsuit at some point when some users discovered that the bit depth advertised was a joke.

My take on the A7RII -after this digression- is that it is really a prosumer video camera at heart. Sony is historically a pro video company, before it ever was a still camera company,  videos are what the kids want these days,   The A7RII will chew up little memory bandwidth when using the encoding tricks (eg. slog) and codecs which the video guys need to make their exposures work in the cheap real world, where Raw storage of video is not an option. And its images will then grade easily to produce video colors which people in the various markets will accept.

But the need for high ISO, DR and the wish to create a non-noisy compressable datastream means that some color discrimination has been sacrificed, compared with a single use stills camera. Also, we all know by now that even the Raw data is subject to a lossy encoding on chip, I conjecture to save chip communication bandwidth, also gaining some thermal dissipation advantage.

Our of the box, my impression is that The A7IIR will supply nice TV-conventional color, with skin tones that will offend no one in the target markets. The clips will be gradable rather than brittle, and the pro user is expected to paint in his own color when required. It'll take superb high-resolution images; it'll make movies by moonlight, it'll do anything decently. Has anybody complained that a Swiss Army knife is not a chainsaw? I don't see Adrian complaining about his test images in the woods. Those are superb images.

Would I try to use this thing for art reproduction? No way But this is just one of my polemic posts - reality may prove me a sceptic.

Edmund

[adrian tyler]

^

Adrian, not sure if your post above was a response to mine, but just to be clear mine was intended as a commentary on colour tests and eyesight rather than on reproduction.

sorry, i though it bore some  kind of relation to what the sony a7r II can or cannot capture,  and the mediums though which we can perceive that, my mistake?!

[eronald]

sorry, i though it bore some  kind of relation to what the sony a7r II can or cannot capture,  and the mediums though which we can perceive that, my mistake?!

I think he was stating that the displays are so bad anyway that there is no reason to ask for perfect camera color.
That is also my opinion. In a way.
Where the sh*t hits the fan in my experience is when you give a customer the files, and she expects the dress to be the right shade, or a painter expects his colors in the repro.

Edmund

[adrian tyler]

edmund, another great post, thanks
interesting the art-repro analogy as my wife is the principal of a fine art (old masters) restoration studio. she was telling me a dealer was complaining about what i have suspected for a while, these pesky digital captures are TOO GOOD for what many dealers want, you just get TOOOO MUCH, warts and all, and it can't be percieved with the naked eye. i've seen (an photographed) many many many exquisite premier league works of art (zurbarán, ribera, goya...) that have been really adversley affected when viewed on screen, microfibres, "craqualet" re-lining marks, not to mention varnish, all the dirty secrets of the last say 400 years and in many cases i've had to back off and down res...
ok, some dealers want all these mega-pixels, lord know why, and for a curator it may be interesting, but beware it can make a beautiful old master look badly damaged.
give me an 8x10" chrome please !!

i'm probably gonna get flamed for that comment!

[adrian tyler]

i'm gonna try out the new sony on easel painting when i get to town, it's not my speciality but i am curoius about this... i'll see if i can find something i can share... could be an interesting post when the dust settles.

[eronald]

i'm gonna try out the new sony on easel painting when i get to town, it's not my speciality but i am curoius about this... i'll see if i can find something i can share... could be an interesting post when the dust settles.

Try it! If the customer is happy - everything is great.

The issues start once the requirements are tightly formulated and the equipment doesn't cooperate - but that is not going to happen in your case.

I used to meet with guys doing art repro; obsessional people who would go round with a spectro measuring the paint on the canvas ...some of that crowd sometimes post on the colorsync list, they work to a degree of precision different from the rest of us.  

I once had a meeting at the Louvre where I was shown a 30m long linear accelerator that is buried under the glass pyramid, and is used to determine whether objects are real or not. I guess the reproductions fakes are becoming very very good They also made up some sort of repro camera using a cryogenically cooled multispectral satellite imager to create files for conservation in case the Mona Lisa gets vaporized by North Korea - or sold off to Saudi Arabia mysteriously stolen yet again

Edmund

[eronald]

I haven't a clue what the Sony A7RII can or cannot capture. I'll leave that judgment to those for whom it matters.

You know, I think I'll do the same
Are you going to buy one? I think I'll defer the problem until next year, at least.

Edmund

[NancyP]

That Louvre conservation department stuff is pretty cool, eronald. Are you a materials scientist or an artist? I can't think of a lot of people who visit the high-end instrumentation involved in authentication of art materials. This (chemistry and materials science as applied to art) was the career path not taken for me, I was a little intimidated by the thought of dealing with irate multimillionaires and lots of time testifying in court, which is where you end up if you determine if something expensive is a forgery.

[eronald]

That Louvre conservation department stuff is pretty cool, eronald. Are you a materials scientist or an artist? I can't think of a lot of people who visit the high-end instrumentation involved in authentication of art materials. This (chemistry and materials science as applied to art) was the career path not taken for me, I was a little intimidated by the thought of dealing with irate multimillionaires and lots of time testifying in court, which is where you end up if you determine if something expensive is a forgery.

I think Leger wrote that any living painter who is asked to authenticate a *good* forgery should immediately authenticate it and sign it, because the embarassment caused to dealers, buyers and experts by a forgery making it past them, and the doubts created by the knowledge of forgeries entering the market are worse than the effects of just owning the damn thing.

I was a color consultant for some time.

What do you do?

[Jack Hogan]

I was a color consultant for some time.

Interesting posts Edmund, though  I am not sure I understand what you mean by 'green discrimination' and I don't have anything to compare your negatives to.  Can you explain a bit more, a bit more quantitatively?  Do you mean that the color filters are less selective than in other sensors in order to gain a little extra sensitivity?  If so that should be easy enough to quantify for people like Jim Kasson.

Jack

[adrian tyler]

That Louvre conservation department stuff is pretty cool, eronald. Are you a materials scientist or an artist? I can't think of a lot of people who visit the high-end instrumentation involved in authentication of art materials. This (chemistry and materials science as applied to art) was the career path not taken for me, I was a little intimidated by the thought of dealing with irate multimillionaires and lots of time testifying in court, which is where you end up if you determine if something expensive is a forgery.

a bit off topic but you may find this of interest nancy, it's a bit clunky but it has an engish option and a technical analisis section, some pretty incredible gear:
http://www.illanyromerorestauracion.com