At last we have it! This is the very first full size RAW(!) Sony A7rII image samples you can download and analyze. You can download it here: http://we.tl/X6ACE5UQem
Thanks to the anonymous source who sent me this!
A7rII and RX camera preorder list:
In USA:
A7rII at Amazon, BHphoto, Adorama, SonyStore US, FocusCamera, Uniquephoto.
RX10m2 at Amazon, BHphoto, Adorama, FocusCamera, Uniquephoto.
RX100m4 at Amazon, BHphoto, Adorama, FocusCamera, Uniquephoto.
In Europe at:
A7rII at Sony DE, UK, FR, IT, ES, NL, BE, CH, AT, SE, FI, NO, PT. Wex UK.
RX100m4 at DE, UK, FR, IT, ES, NL, BE, CH, AT, SE, FI, NO, PT. Wex UK.
RX10m2 at DE, UK, FR, IT, ES, NL, BE, CH, AT, SE, FI, NO, PT. Wex UK.
In Asia at:
A7rII at Digitalrev, CameraPro.
RX10m2 at Digitalrev, CameraPro.
RX100m4 at Digitalrev, CameraPro.
via Xitek Forum.
143 Responses
And we can open it with Capture One!
First try wouldn’t import into capture one so I assigned capture one as the default program for ARW files and it worked. This shot is just awful! My question is why wast your time shooting it. Where did this person get the camera?
Also it can be opened quite nicely through free Raw Therapee
And Adobe seems to be on vacation…
I was just shooting Roman Abramovich’s Yacht ‘Eclipse’ for real. 😀
I have proof, all the raw’s 😀 And my name.
Sensor needs a clean already.
Raining and using vintage tele with adapter too. Not cleanest example.
Fairly sharp then given the vintage glass, adapter & precipitation.
Pardon me for being curious, but what is the significance of a very ordinary shot of a boat in an unrecognisable and mundane location?
I can’t for the life of me see why anyone should ask you to provide proof of taking these and I, for one, am quite happy to take your word for it. They’re not pictures I would bother to retain on my hard drive, but if they please you then that’s fine.
All of you will leave my life alone and my work or be shown to be who you are.
You can also prove you are a real identity and photography, like you made me before.
Yes! Everyone! Leave Illuminated/ enlightened/ unlimited/ whatever observer alone!
Hey, “The real photographer”, go try to sell those noisy, non-sharp, with chromatic abberation pictures. I’d like to see how long it will take you to sell that trash.
Real photographers work sells itself. You on the other hand are a wannabe. You lie so much about being a photographer, you believe yourself….
This is all you are. you don’t even see that anymore, NEVER parallel you with me again. Or even that idea through me upon my significant other.
I can’t tell them when they are years and thousands of miles away. You lose and any sign I lose more you are going up that list for sure. As you know, I know all of it too, my survival doesnt allow me to lose that, not yet, and it was right, but whats more, I know when its been driven into reverse, Ive know a long time now, 2 years is a long time in my life. didnt I learn photography in about 3 years or something even from Canon rebel…You know each of others, I dont, whats the problem? so noone can torture me for it? But I would tell, don’t worry about that, not now.
It is very difficult to understand what you want to say.
nice boat, what is the price tag on one of those?
Can’t open with Lightroom! Come on, Adobe!
Has Adobe ever supported a camera before it’s out with firmware 1.0?
Not too bad for a cinny lens at 6400. You can jpeg it in Sony RAW (straight -no sharpening) then run it through PS unsharp mask. The file is pretty hardy.
Real Nice low light camera, Sony is doing a great job.
here is a processed picture for the people that use only lightroom
http://tinyurl.com/qb9mtt3
Thank you – I really appreciate not having to install capture one just to get a look!
Depth of field in the image looks very slim, only tiny area around back of window looks in critical focus. In good focus file looks great for 6400, but in focus area has little fine detail so hard to assess loss of detail. There is subject detail (e.g. feather taped to windshield) in OOF areas, but so blurred it is hard to know how much is due to OOF and how much to NR.
That said, compared to my Nex 7 (older sensor, even smaller pixel pitch) this looks fantastic – on Nex I wince at ISO 800 and rarely tolerate iso 400. Don’t think I’d even bother trying 6400 on Nex7. Big question on detail though – will the absence of low-pass filter on A7Rii afford sharpness/detail very close to Nex 7 (on per sensor area basis – A7rii 18MP crop vs. Nex 7 24 MP) or will I give up detail per area and have to get a little closer to my subjects (which there is plenty of room to do w/ FF sensor….). High iso impact on detail is on top of any per-sensor-area difference. Will take some careful shooting and comparing to know just how much detail A7rii can resolve under various conditions…..
Again thank you – very encouraging result. Can’t wait for my preorder to arrive so I can run my own tests with release firmware.
http://i.imgur.com/h2ycpaG.jpg
What are those circular lines?! Click to see full size.
Damn, did Sony mess up another potentially perfect camera?
After applying some noise reduction it still looks great for such level of sensitive
https://drive.google.com/file/d/0B3B_X7txibJTRUU3dDBEemU5XzA/view?usp=sharing
Where did you apply the NR? In the RAW image file processor, or in other third party software..
Thx!
not that bad for iso6400; will have to process it at home with DxO Optics and see what happens… just tried it with C1 Express and have the feeling, that the noise reduction can’t compete with DxO’s “Prime” mechanism…but have to try…
Anyway, combined with the IBIS that camera probably makes me selling my a7s, which I usually avoid using for stills above iso12800, in worst case maybe iso25600.
In my experience, C1 was worse than Lightroom. I’ve read DxO Prime is the best at noise handling, and based on comparison(i think it was on DPReview), DXO came out on top strictly for noise handling. Color processing/retention title went to someone else.
Well… tried to open it with current DxO Pro 10, but didn’t worked… so far, DxO does not support the a7r II… hopefully they will in a few days, I already contacted the customer support bothering them with “when can I have the a7rII profile” 😉
Found them on flickr: https://www.flickr.com/photos/58521449@N00/sets/72157655011661809
one more full size raw ISO1600
https://www.dropbox.com/s/ub2jagdhl9blg5x/DSC09049.ARW?dl=0
How did you open it? No go with Lightroom or C1 Express.
Upgrade needed for C1 full version.
no, not needed. download lates Express version for Sony, just as I did
Express didn’t seem to work for me but maybe there was some other issue.
Use Raw Therapee (it’s free)
Thanks! That worked!
ISO 6400 and underexposed. Very little can be learned about high IQ performance with this file. However, ETTR post shows typical Sony 11+7 artifacts in the form of lines in the shadow areas.
Maybe, just maybe its a pre-production sample.
It’s a pre-production Software V0.01
Can’t really see what your arrows are pointing at – there’s nothing there that looks anything like Sony CRAW artifacts…
Exactly.
Yes there are. Maybe lines is not quite the correct description but there are digital artefacts. Like clumps of darker pixels than lighter pixels and so on. Clearly visible when you click on the image and view it 100%.
That’s called noise, CRAW artifacts have a very specific appearance and can be identified directly from the RAW file itself using freely available tools.
I was referring to the digital noise of course when I referred to the artefacts! I was referring to how you said you couldn’t really see what the arrows are pointing.
Right, and I was replying to the original assertion that “ETTR post shows typical Sony 11+7 artifacts”, as there’s nothing anywhere near the arrows that looks like typical Sony lossy artifacts.
I’m running it through RAWSHACK now…
That’s just to say most are clueless about the lossy artifact and trying to find fault even it’s not there. I wonder since when digital noise becomes artifacts when “all” digital cameras would have them. Matt, it’s nice that you have an overlay over the problematic area.
You have eyes, don’t you, click to see:
http://i.imgur.com/h2ycpaG.jpg
Yes I have eyes, and I also have access to software to analyse RAW files. Those lines are NOT present in the RAW date.
http://i.imgur.com/h2ycpaG.jpg
Here, click to see full size.
WTF Sony?!
All I did is pull the shadows and turn down noise reduction.
you really want to see sony artifact? this is not from sony a7r ii but from a7ii… check it out… there are more out there if you are willing to search…
I’m not denying compression artifacts exist, they just don’t exist in that image, despite claims that they do….
Sorry about my misunderstanding… However, glad that you aware of that, unlike many others who are denying…
I’ve created an overlay by reading the compression levels directly from the RAW file. The brighter red areas are most compressed. As you can see the compressed areas surround edges with large brightness variations. There is actually next to no lossy compression going on, at all, in the areas pointed to by your yellow arrows:
http://mattgrum.com/ar7m2_compression_levels.png
click to enlarge
So to answer your question of what are the lines from, it’s definitely NOT “11+7 artifacts”.
So it’s trying to compress the more important part of the picture. I.e. the stuff that’s more exposed that is probably the subject of the picture.
you need to know what compress mean
The compression works by taking each line of 16 pixels and storing each pixel as an offset from the darkest pixel in the line. If there is a large brightness variation within the line then these offsets have to be coarser to span the possible values.
What I’ve overlaid in red in that example image is the step size of the offsets in each 16 pixel block. So the parts with lots of contrast are most compressed, but that’s also where you’re less likely to see banding, which is the principal behind the compression scheme.
Anywhere you have a smooth gradient where banding would be very noticeable is stored using the finest offsets. The only problem occurs when you have a smooth area that just touches a much brighter region, resulting in a coarse step size and (highly localised) posterisation. There are very few such areas in that image as a whole.
A bunch of horseshit excuses for Sony.
Why not just give users the 14 bit file they paid for and Sony marketing has been lying about for 3 years+. It wasn’t until a month ago that a Sony exec admitted that their files are a problem. Sony great on innovation but REALLY pathetic after the sale responsiveness.
They could have fixed 11+7 lossy compression a long time ago just by a firmware upgrade that was at the expense of FPS.
I’m not excusing anything, just explaining how things actually work and what the impacts are in the real world, so people understand the issue and can make up their own mind.
Contrast this with creating an entire website dedicated to a cause that in reality affects very very few people, pointing at anything and everything as an example of a compression artifact even when they are not (see the erroneous yellow arrows, above), and posting the same messages in every thread, but under different usernames (trying to get round a ban?)
You’re the equivalent of the guy in the street preaching about the end of the world that everyone does their best to ignore.
Thanks for the explanation.
Nothing you’ve ever said here has been of any worthwhile value. Anyone who reads these forums understands your *insane* Nikon vs. Sony bias (which You even created a website for).
Dude, you should shut up and start listening.
Look at this, click to see it full size:
http://i.imgur.com/h2ycpaG.jpg
What in the world are those lines?!?!
I just pulled the shadows a bit as image was underexposed!!!
WTF Sony?!
Same old Sony Song. Less than 10% of the discrete color values delivered. Grossly underexposed as noted earlier (even at ISO 6400).
Btw. your average monitor can only display 0.0004% of the possible discrete colour values contained in a true 14-bit RAW file.
Don’t buy or use AVERAGE monitors.
By average I mean 8 bits per colour channel, even some very high end monitors are 8 bit. But if you’re lucky enough to have a 10 bit monitor then you can display a whopping 0.02% of the possible discrete colour values contained in a true 14-bit RAW file…
And then you’d have to be lucky enough to have the hardware and software to support 10 bit output (spoiler alert: none of the current Apple machines can for example).
hey dude, can you show some examples of canon and nikon raws in that program ? be interesting to see how they fair
All over my blog – google sonyvnikon and there are several comparisons
whats your blog called ? i found one by some guy named max with sony v nikon blue sheet with this type of data on it going on about the 11+7bits raw thing.. but doesn’t show the actual image he shot to compare properly.. I’m a old school d3s user still, i haven’t updated yet as it still does exactly what i needed at the time (weddings) apart from video which is now 98% of my work now. i spent a month with a borrowed A7r for a food photography gig for a high end restaurant here in china and it worked perfectly for that, just doing posters now with those images at 1470mm wide and back lit and look great still. plus full colour menu photos etc i didn’t notice any 11+7bit issues in my images.. not sure exactly how this 11+7 thing should affect the image unless your shooting blank blue skys ?
ok maxberlin88 is you right ?
How could DXO ever test the A7R on par with the D810 in color depth. BTW: Nice copyright infringement in your avatar.
Can anyone transform it to a .dng please? (for old software users)
I think it looks pretty good. Also my system running C1 Pro Sony processes the files with no noticeably slow down.
If you are on a Mac, Buy the $17.99″Intensify” on the app store. It will open them and clean them up quick.
Or just download the free Capture One for Sony.
Would love to see a well exposed 50 (100) Iso Raw in dng. Could not open it in the Latest Lightroom:-(
For ISO 6400, I think this looks great. Wish we had a properly exposed low ISO shot of the same thing for comparison. There was certainly no need for them to push the shutter to 1/1250 for a simple shot of a stationary object.
What is the lens used?
the 28-135 f4 lens.
So much for the much touted BSI sensor – in Capture One Pro 8.3.1, all I see is a very noisy (heavy luminance and colour noise) ISO6400 file that I’d never dream of using or sending images made with to a client. Hope the lower sensitivities are a lot, lot, lot better than this file…….
Does capture one have the same import defaults like LR (low sharpness gain and low color noise removal)? Have you taken into account that this picture was shot in artificial light? To which high MP Sensor do you compare the noise levels? At what resolution do you usually hand out the product of your work?
Please shed some light on your assessment.
Capture One applies heavy luminance and colour noise reduction and low sharpening to this file by default so no, not the same as LR. Adjusting any of the settings does’t make a miracle happen. Comparing this to any other high MP sensor does not get around the fact that based on this sole example, the A7RII is not that great at ISO6400.
Do you expect it to be miraculously as clean as ISO 1600 or ISO 800 on your current cam at pixel level? That’s why one should compare it to other high MP sensor to see how much better (or worse) it is. To say that it’s not great without a point of reference just doesn’t make much sense.
I still think your opinion is baseless. May be best in class is not enough for your needs, but do not put it as if 5dsr or other high MP Sensor Cameras would outperform this sample. I have posted a small comparison in a seperate comment.
As was proven earlier today using other import software show the issue is with capture one itself.
The 24mp of my Sony Alpha 77 are way enough for me. I really like the A7rII, but i think, waiting for the A7sII would be the better option for me. If the camera comes with about 16-24Mp ~ it will be enough for me. I would prefer better iso-capabilitys 😀 But if i would have the money, i just would go for both offcourse ^^
I’m trying to keep mine for 2 more years, until A77mIII comes out. Hopefully it will come out and Sony will not suddenly decide to drop A-Mount.
A99II might be a worthy alternative, if it will have a mirror lockup design with A7rII sensor AF capabilities. Mirrorless AF feels too slow for me. I got spoiled by the speed of SAL1650.
yeah this probably isn’t the best raw image to put forward for this camera. Regardless of being underexposed, it exhibits quite a bit of noise. I don’t see any raw compression artifacts, but based on this i would not be shooting at ISO 6400 if i could avoid it.
This isn’t an official image, so I can only hope it’s early software we are seeing.
I know the following is far from a scientific comparison:
I took the A7R2 raw file, and tried to skip almost all processing in rawTherapee. Neutral exposure module, no sharpening, no noise reduction, white balance as chosen by camera and only demosaic with amaze. I tried to find a 1:1 crop with highlights and shadows
http://WDMyCloud.device2088953.wd2go.com:80/api/1.0/rest/file_contents/Public/Shared%20Pictures/A7R2AtISO6400.png?device_user_id=15291585&request_auth_code=a3ed27d6487ea94334f1eda5ff5e09f107c312b6a3544831d6f03c3273df6648
I compare it to other high MP Sensors at ISO 6400 under artificial light taken from DPReview Studio test.
http://WDMyCloud.device2088953.wd2go.com:80/api/1.0/rest/file_contents/Public/Shared%20Pictures/highMPSensorsAtISO6400.png?device_user_id=15291585&request_auth_code=9ac9af4983e9af4521f7b87e96e70944777b3eeab90a904094b42a88bb8a4206
Imho the A7R2 has the edge but not by much.
Meh, the comments cannot handle long URIs
A7R2:
http://www.qpic.ws/images/a7r2atiso6400.png
other high MP Sensors:
http://www.qpic.ws/images/highmpsensorsatiso6400.png
If this is without noise reduction, this A7R II is about 2 stops better than the D810. I’m not sure you would call that “not much.” I compared the blue-violet noise patch and the well-lit black ground against the the sample above.
And this is the competition at ISO 3200
http://www.qpic.ws/images/imagextx.jpg
If this was the only client approved shot, what would your Capture One settings be in regards to sharpening, noise reduction, clarity, exposure, etc? Thoughts?
This would not be the only client approved shot as I wouldn’t make the mistake to shoot this camera at that ISO unless I was going for a ‘look’ however, to play the game and after only a very quick play:
Most important sharpening all the way to zero.
NR L=60
NR Details=40
NR C=40
NR Single Pixel=28
Film Grain=Fine Grain
Film Grain impact=60
Exposure=0.75
Contrast=10
Brightness=17
HDR H=52
HDR S=19
WB Kelvin=4200
WB Tint=1.9
Clarity Structure=40
Probably missed something but even with these settings you’d have a lost client unless you were shooting for this gritty look.
Haha, I understand and completely agree. Just curious if this was recoverable by any means. Perhaps good for black & white at these high isos where some grain might be expected?
For me your statement makes no direct sense. For what output size did you investigate the image? At 100%? Downsized to 16 or even 12MP? Makes a big difference. Did you really expect a 42MP file at ISO6400 to provide better noise performance at pixel level than an A7s?
The noise is pretty intrusive what ever size. What baffles me is why you consider downsizing the images to 12MP to improve noise when one of the biggest selling points of this camera is its 42MP sensor? Too early to tell from this single file but I fail to understand why anyone would be impressed with this especially next to the Canon 5DSR ISO6400 shots I’ve seen (which many people slated).
Then you clearly don’t understand the difference. If you are looking at pixels then you have already lost.
I’m looking at the intrusive nature of the noise pattern and severity of it at ISO6400 at the output sizes I normally use, not 100% pixel peeping on screen. Seems you like to put words in peoples mouths.
We use an D810 professionally, for example. We know, that it fares worse at pixel level at high ISOs compared to our D750 or a D4s. However, this isn’t a problem at all. We use the camera at full resolution at lower ISO to benefit form the immense DR and resolution. For weddings downsized to 12 or 16MP you get excellent images with hardly any noise for publishing in the large wedding photo book. Many are using an A7s or D4 as main cameras, so downsizing to these file sizes makes perfect sense for us. This means we can use the camera for different objectives, which is great. ISO 6400-12800 is s.th. we run into only at extreme circumstances. Usually, we can always use light and flash to reduce ISO down to low values at night or during events to print larger if desired. When we shoot for a client, why would we deliberately go to ISO6400 if we know the output size in advance? IF the A7rii would deliver cleaner files than our D810 at 36MP, it would be a win for us.
I don’t shoot weddings and my images can be anything from web or half A4 up to 6ft exhibition stand graphics so downsampling to 12MP is not an option. Regardless, this is not about how to improve the look of these files but my early observation that I expected better. Looking at the 5DSR files at ISO6400 as a comparison, I’d want to be downsampling those to 12MP rather than the Sony if I was a wedding shooter regular shooting at these high ISO levels.
You need to be aware that the dpreview screenshot is almost 2:1 pixel size due this retina nonsense of apple while the a7r2 sample crop is 1:1.
I find it hard to put a stop number on the difference, so yes 2 stops might be correct. Either way this looks very promising and even if the price is steep you will get a feature set no other camera on the market has to offer:
-adaptability
-good autofocus with non native lenses
-no more need for auto focus calibration
-high MP with good ISO
-IBIS
-good in body 4k video with less rolling shutter than 1st gen 4k still cameras
I really hope the price will drop around 10-15% when it retails.
The fact its not very good needs no point of reference. Better than my current camera at ISO6400 does not make the A7RII acceptable by default……Does it?
All I’m saying is BSI obviously doesn’t produce the miracle sensor I thought it would and I’m not drinking the koolaid until I see some better examples at an ISO more suited to show this cameras capabilities.
Pretty naive to think that.
Naive to think this camera would give a much cleaner result at high ISO…. quite possibly at this stage.
So it’s exactly what I said early, you expect it to be miraculously as clean as ISO 1600 or 800 on your current cam at pixel level. Unless there is a drastic change in how a sensor collecting light, you can pretty much forget about what you might deem as acceptable at ISO 6400. Let’s just be realistic.
Look, it’s fine in a way. The a7rII’s 6400 is about as good or better as other camera’s in its class–but it’s still not good, not usable in most situations. Sony’s BSI hype promised a game changer and this is simply not it. 42mp is probably just too many pixels to get next gen low light quality. A lot of people don’t need high ISO anyway and this is a great camera for them–but a lot of us do. So for us, this is a big disappointment.
Sony didn’t promise game changing high ISO performance. Rather good high ISO performance and specifically fast sensor read out.
Cameras like the A7S are close (within a stop) to the theoretical limit (both in efficiency and read noise at high ISO) with Bayer color filter already. There’s not that much to gain from Bayer sensors and a single exposure.
The BSI hype was about readout speeds not sensitivity. Sensitivity gains were always going to be modest with a full frame BSI sensor (compared to mobile) because the wiring takes up a small fraction of the pixel area.
You are aware that 2 stops mean 4 times more light and BSI can give around 30% if you’re lucky?
Fanboyism at its best…
I’m well aware of what I am seeing. 2-stop here doesn’t mean 4 more times more light in this case, but the noise grain and pattern of the A7RII at ISO 6400 looks similar to those of others at ISO 1600. Assuming the ISO value is the same across the brand, for the same scene, aperture and exposure, by using ISO 6400, you effectively use 4 times less light than needed and simply amplify the signal by 2 stops. As Ben pointed out, this is not a direct comparison since it’s not the same scene. But from what I see on dpreview and what I’m seeing here, I see 2-stop advantage. And yes I have been a Sony shooter for a while now, but a fanboy, far from it. Maybe you have an agenda of your own LOL.
Meh… Just like Samsung’s BSI APS-C sensor for the NX1, all marketing hype and not really much of an improvement.
It’s the camera displayed at Sony Showroom in Hong Kong (Should be the Causeway Bay Store)
There’s another full res A7RII jpg on flickr now, ISO 100.
https://www.flickr.com/photos/72264019@N00/19500107862/in/dateposted-public/
Looks very nice. Would love to see some RAW images available that were shot like this.
Variable toned gradient backgrounds like this are easier for an 11+7 bit camera. Let’s see the same with a gradient background of a single color.
You really don’t know what you’re talking about despite turning up to every thread to post the comments over and over again.
Gradient backgrounds of a single colour actually represent the best case for the compression algorithm. The compression works by taking each line of 16 pixels and storing each pixel as an offset from the darkest pixel in the line. If there is a
large brightness variation within the line then these offsets have to be coarser to span the possible values. However a smooth gradient of a single colour allows the finest gradients to be used for negligible loss in quality.
It’s funny – trolls like Rokibow think posting the same false info 400 times make them true. Sad really.
What’s going on with the moare on the left?!
http://i.imgur.com/h2ycpaG.jpg
Where does it come from? Damn!
I had similar effects with my rx100II when high Iso shots had distortion correction applied in body at 24mm. Seems like the lens profile Sony saves in body needs some tweaking.
This is terrible for 3200 USD camera.
It’s the RAW converter not the camera, just use a better (or just different) RAW converter.
Capture 1 pro is as good as it gets.
How can you say “Capture 1 pro is as good as it gets” if other raw converters do not reproduce this issue? Or do you think that we did not try recovering the shadows ?
It seems it really is C1Pro problem. This is good to know!
I will pull down my posts.
I have capture One pro 8. I pulled the hell out of the shadows and I didn’t get these affects! I’m getting noise yes but not these moire artefacts?
Well it’s causing artifact in the image, so it’s not the best across the board, you might just need to a different converter until they fix it…
Here’s what I get when I convert the image with dcraw and lift the shadows:
http://www.mattgrum.com/van_crop_dcraw.png
really bad noise MG
That’s a completely vanilla conversion from dcraw. When you turn noise reduction off in dcraw it actually does no noise reduction, unlike the vast majority of converters that still perform some noise reduction when all the sliders are set to zero. People don’t realise their RAW converters are doing a heck of a lot processing even all the sliders are left alone so a straight up conversion like this will look quite different.
Uh C1 Pro is not optimized for the A7r2 yet as is evidenced by import with other converters.
Did you apply sharpening ?
Zero sharpening I just pulled out the shadows.
I tried to twist the raw file as much as I could but I was not able to see this kind of moire. I used rawTherapee.
You don’t get this with dcraw either, so it’s something to do with the RAW converter, possibly applying some sort of lens correction.
I use capture one. I pulled the shadows up to 100% and I increased the exposure by 4 stops and I still was not able to get this affect.
This is also visible in other parts of the picture. WTF is going on?!
It’s definitely CaptureOne doing this, here’s what it looks like with dcraw:
http://www.mattgrum.com/van_crop_dcraw.png
This is good to know. Damn.
Of course it will be noisy on pixel level, what did you people expect? There has been bsi-sensors out there for quite a while, none of them was the holy grail. Compared to other cameras at any given output size, this BSI-sensor is a small improvement, but still an improvement, just as expected.
I said nothing of the kind, just I hoped the BSI technology used in this sensor would give better results that the mediocre one shown here. I’m never going to get a unbiased discussion on a Sony fanbase forum so best we rest my comments here.
Seriously? You are playing the fanboy card?
During this whole discussion you refused to give any comparison, example or meaningful information about why you are so disappointed.
1.You speak of mediocre results, when everything suggests that this sensor outperforms all currently existing full frame sensors above or equal to 36MP.
2.You speak of intrusive noise patterns without showing a single image crop.
3.You are speaking of unacceptable noise when you won’t even tell what output size you are using.
If you demand an unbiased discussion you had better do your own part and give some proof for your opinions.
http://i.imgur.com/h2ycpaG.jpg
Not there?! Dude do you even process your raws?!
Click and see. Nothing has been done to the image, just turned down noise reduction and pulled the shadows to make image to my taste.
What the f. are those lines?!
But now you’re talking about some abstract things like grain patterns.
Obtaining visible grain is not the reason why we raise ISO…
If nice grain is what you’re looking for – just get a Canon 5Dm3. Its is many stops better than this one…
In the real world (with quantitative comparing methods), noise is a loss of information (colour or details). Simple fact is: this sensor can’t be 2 stops better than the previous Sony’s offering and clearly isn’t just looking at the samples. I guess it’s around 1/3 stop (1/2 would be a big achievement).
lol… not sure if you know what you are talking about…
The image is terribly underexposed. Looking at the RAW file and doing a simple conversion using Picasa yields a reasonably good image considering it’s ISO 6400 and grossly underexposed. In the very dark shadow regions, where there is very little data to pull up, it’s as muddled as expected. In the better lit sections of the photo, the noise is far better than my D800 at 6400. If the ISO performance at 12K or 25K matches my D800 at 6400, I’ll be happy with the upgrade, just for stills. The 4K video will be icing on the cake.