A7RII, the good, the bad and the ugly

andrewD2 · Aug 9, 2015

DVT80111 said:
DVT80111 said:

Rather than posting crops, I'll just say that the girl on the left has 3 large star tattoos on her arm:

Click to expand...

I would be very impressed. Thank you.

Easy to get pulled in searching images for tiny details. Little bit of USM on the crop.

Andrew

DVT80111 · Aug 9, 2015

Thank you. I like what I see. Amazing resolution.

SurfFish · Aug 9, 2015

Very nice review Andrew! Do see any compression artifacts by any chance on 8mm sample you provided? Also I wander if the artifact sample would still be the same if shot at ISO 800 at similar shutter speed.

Thank you for your time!

Adam-T · Aug 9, 2015

andrewD2 said:
Yes, very nice church. Here is the crop:

A bit more rez than a 13yr old canon D60

I had to use the 300 F4L (X the 1.6 APS-C crop factor) to get that back then . now all you need is the 24-70 and an R2 .

I hope the compression issue gets resolved, I plan to replace my ageing Mill-stone wedding camera (1DS3) with the A7R2 ..

Canon 10D this time @ 300mm X 1.6 12 years ago - times have changed all you need is 70mm and an R2. the 1D Mk1 I also had at the time didn`t cut it for this , 4Mp , 1.3X Crop factor

--
** Please ignore the Typos, I'm the world's worst Typist **

andrewD2 · Aug 9, 2015

I used 1DsIII + 1DIV, then 5DIII + 1DIV, then 5DIII + 6D. 1DsIII is a great camera.

No way with a flip out LCD that is near the camera base you can throw it in and out of the bag in the same way as a 1DsIII. 5DIII is awesome, bang on all day. Hope the 5DIV has less low ISO banding, which is why the pairing with a 6D helps.

Andrew

andrewD2 · Aug 9, 2015

SurfFish said:
Very nice review Andrew! Do see any compression artifacts by any chance on 8mm sample you provided? Also I wander if the artifact sample would still be the same if shot at ISO 800 at similar shutter speed.

Thank you for your time!

That's a very good question. I don't see it on the 17.8MP fisheye image. The stonework lattices are smaller in the frame, and yes maybe ISO800 is a factor, I'd thought about that myself. I'll test late next week if I don't get chance today.

Andrew

Esstee · Aug 9, 2015

Confirmed that this is indeed compression artifacts and that I won't buying this system so long as this isn't addressed by Sony.

TBH. I find it rather disappointing that Sony would release a new body with such a problem given that the DR advantages are in the shadow detail. Though after looking at the RAW file for this image it becomes obvious that this is a serious issue and that Sony needs to address it sooner than later. Otherwise, the A7R II won't receive the respect it needs to succeed by those seeking to make the move from DSLR kits.

PS. Until recently I never really got the opportunity to see the effects of Sony's RAW file compression artifacts beyond that of very specified processing methods. And so this particular example marks the first time I've been able to see the effects under what I'd call, normal image processing methods.

Adam-T · Aug 9, 2015

andrewD2 said:
I used 1DsIII + 1DIV, then 5DIII + 1DIV, then 5DIII + 6D. 1DsIII is a great camera.

I tend to use work stuff till it breaks or becomes totally uncompetitive , I`ve had the 1DS3 for 5 years (1DS2 before that and 1DS-1 before that with various seconders and backups) and it`s been a gem but like a London Taxi which has been around the clock 6 times, it`s in need of replacement and I`m not liking what I`m seeing from the 5DS (Which I was planning as replacement) , especially noise floor ..

the sensors in the 6D and 5D3 are basically warmed over 1DS3 sensors so not ploughing money in there (may grab a used 6D to tide me over if the A7 doesn`t work out for weddings) and hope the 5D4 gets a 32/36Mp sensor with competitive DR and noise levels than what they have now..

--
** Please ignore the Typos, I'm the world's worst Typist **

SQLGuy · Aug 9, 2015

50/1.8?

Or FE55?

maiaibing · Aug 9, 2015

andrewD2 said:
mdcromer said:

I didn't see any RAW compression artifacts, just lens chromatic aberration in the area you highlighted.

Click to expand...

CA does not have a barcode pattern. End of. It looks nothing like CA.
If you are struggling (my girlfriend and I can see it at 50% zoom) this is at 200% from the more central window.
Click for full view.

Thank you for your review.

I'm not sure this is a compression issue although I understand why this springs to mind when seeing the pattern. I remember the issues Canon's 5DII had with bloated highlights until it was fixed in firmware.

However, its of course not acceptable and must be fixed one way or the other. Especially since its a scenery that people may want to use a camera like the A7RII for.

SONY would do good to come out fast and say they are looking at a fix.At this rate and size it can and I dare say I trust will be fixed - even if its a compression issue.

Canon remained silent at first, then they were seemingly in denial - and then came the fix. Not recommended.

andrewD2 · Aug 9, 2015

SQLGuy said:
50/1.8?

Or FE55?

Robgo2 · Aug 9, 2015

Andrew,

With regard to the supposed compression artifact that you circled in red, please convince me that this is not simply an irregular area in the rough wooden window frame, because that's what it looks like to me.

Rob

lemming2008 · Aug 9, 2015

Robgo2 said:
Andrew,

With regard to the supposed compression artifact that you circled in red, please convince me that this is not simply an irregular area in the rough wooden window frame, because that's what it looks like to me.

Rob

What would convince you? The shots are already there, the rest is hard to convey. The red circled zone is hard is indeed debatable in isolation. But check out post #6 in this thread, Andrew's zoomed-in view of the central windows. The horizontal compression artifacts in dark areas are obvious there. The circled area in the other shot does look very similar to that.

billzilla2000 · Aug 9, 2015

andrewD2 said:
mdcromer said:

I didn't see any RAW compression artifacts, just lens chromatic aberration in the area you highlighted.

Click to expand...

andrewD2 said:
CA does not have a barcode pattern. End of. It looks nothing like CA.
If you are struggling (my girlfriend and I can see it at 50% zoom) this is at 200% from the more central window.
Click for full view.

No, it's definitely cRAW artifacts... You just need 200% crops at +4 EV in this case (nuking the image for any practical use but nit-picking in forums).

In short, you have to be looking and you almost always have to have an underexposed image bumped by +2 EV or more with adjusted contrast and shadows to accentuate the effect. Or Rawdigger's colorized compression tiles trick, which of course doesn't exactly prove anything.

billzilla2000 · Aug 9, 2015

Esstee said:
Confirmed that this is indeed compression artifacts and that I won't buying this system so long as this isn't addressed by Sony.

If only you knew how rare and hard to show these artifacts are.

What OP isn't saying is how much the image had to be jacked to produce such obvious results - at 200%. Look at the ridiculous amount of noise visible from the EV adjustment alone.

Esstee said:
TBH. I find it rather disappointing that Sony would release a new body with such a problem given that the DR advantages are in the shadow detail.

People have been happy with RX1/RX1R and A99 and A-series SLT IQ, along with NEX-7 and A6000 for years, now. All the exact same Sony ARW rev v2.3.

Esstee said:
Though after looking at the RAW file for this image it becomes obvious that this is a serious issue and that Sony needs to address it sooner than later.

Not really. Just to appease the extremists, maybe.

Esstee said:
Otherwise, the A7R II won't receive the respect it needs to succeed by those seeking to make the move from DSLR kits.

Who gives a fug about respect? Sony's never had it - never will, by the Canikon old guard.

I've been a Sony owner since SLT days and no longer do I care what Canikoners think.

Esstee said:
PS. Until recently I never really got the opportunity to see the effects of Sony's RAW file compression artifacts beyond that of very specified processing methods. And so this particular example marks the first time I've been able to see the effects under what I'd call, normal image processing methods.

Not normal whatsoever. This image of the stained glass window is not usable for anything other than forum 'finger pointing'.

andrewD2 · Aug 10, 2015

Robgo2 said:
Andrew,

With regard to the supposed compression artifact that you circled in red, please convince me that this is not simply an irregular area in the rough wooden window frame, because that's what it looks like to me.

Rob

Hi Rob,
Not unless its made of those thinner lego blocks.
Seriously 1) its not wood, its stone. 2) the texture is the same on the stone all the way down.
3) it happened to be aligned perfectly to my camera that was tilted ever so slightly.

Andrew

Mfcrisis · Aug 10, 2015

Pedro

Please don't quote an entire email with multiple photos - it's really annoying to have to scroll through the same photos multiple times.

Thank you

Nic

pegelli · Aug 10, 2015

Hi Andrew, thanks for this post. It indeed doesn't look so good on the columns and there's indeed strange artifacts there. I happened to be in a cathedral in Amiens 2 weeks ago with my NEX6 which (if I'm not mistaken) uses the same lossy compression algorithm as the newer Sony Cameras. I especially looked at the windows in my underexposed shots but couldn't find anything that even closely resembles this. Can you share which raw converter you used and what the develop settings in there were? I'd also be interested to test your raw file if you're willing to send it to me by We Transfer or another method. Let me know and I'll send you my email address in a private message.

I'm seriously considering this camera, but this issue is nagging me and is still stopping me from pressing the "buy now" button.

Here's the best (or worst) result I could come up with. NEX6 + E10-18, iso100, 1/80, f5
Developed in Lightroom, + 1.0 EV, +100 shadows, -100 highlights, + 100 darks, 50/0.5/50 sharpening, 200% size:

Looking for barcodes (but did not find them

Thanks in advance for any help.

andrewD2 said:
mdcromer said:

I didn't see any RAW compression artifacts, just lens chromatic aberration in the area you highlighted.

Click to expand...

CA does not have a barcode pattern. End of. It looks nothing like CA.
If you are struggling (my girlfriend and I can see it at 50% zoom) this is at 200% from the more central window.
Click for full view.

Camera ILCE-7RM2Focal Length 55mmAperture f/3.5Exposure 1/160sISO 100

lemming2008 · Aug 10, 2015

User7079951100 said:
Hi Andrew, thanks for this post. It indeed doesn't look so good on the columns and there's indeed strange artifacts there. I happened to be in a cathedral in Amiens 2 weeks ago with my NEX6 which (if I'm not mistaken) uses the same lossy compression algorithm as the newer Sony Cameras. I especially looked at the windows in my underexposed shots but couldn't find anything that even closely resembles this. Can you share which raw converter you used and what the develop settings in there were? I'd also be interested to test your raw file if you're willing to send it to me by We Transfer or another method. Let me know and I'll send you my email address in a private message.

I'm seriously considering this camera, but this issue is nagging me and is still stopping me from pressing the "buy now" button.

Here's the best (or worst) result I could come up with. NEX6 + E10-18, iso100, 1/80, f5
Developed in Lightroom, + 1.0 EV, +100 shadows, -100 highlights, + 100 darks, 50/0.5/50 sharpening, 200% size:

Looking for barcodes (but did not find them

Thanks in advance for any help.

andrewD2 said:

mdcromer said:

I didn't see any RAW compression artifacts, just lens chromatic aberration in the area you highlighted.

Click to expand...

CA does not have a barcode pattern. End of. It looks nothing like CA.
If you are struggling (my girlfriend and I can see it at 50% zoom) this is at 200% from the more central window.
Click for full view.

Camera ILCE-7RM2Focal Length 55mmAperture f/3.5Exposure 1/160sISO 100

Click to expand...

... So you took a picture with a different sensor tech, a different resolution, a different processor, and an entirely different location, and you're wondering why you didn't see the same effects? Even if the cameras were running exactly the same algorithm (which actually seems unlikely), such different inputs would produce entirely different results.

pegelli · Aug 11, 2015

Hi lemming2008, thanks for your input. Since Sony is getting a lot of bad rap for the raw compression on all their cameras I am actually trying to see the effect in the camera I have but so far failed to do that. No big deal and the actual objective of my post is to better understand what the conditions were under which the effect was observed in the new A7rII and see if that puts me further on the trail of reproducing it with my camera. Actually I'm hoping to never be able to reproduce it but can't be sure until I've investigated all possibilities.

lemming2008 said:
User7079951100 said:

Hi Andrew, thanks for this post. It indeed doesn't look so good on the columns and there's indeed strange artifacts there. I happened to be in a cathedral in Amiens 2 weeks ago with my NEX6 which (if I'm not mistaken) uses the same lossy compression algorithm as the newer Sony Cameras. I especially looked at the windows in my underexposed shots but couldn't find anything that even closely resembles this. Can you share which raw converter you used and what the develop settings in there were? I'd also be interested to test your raw file if you're willing to send it to me by We Transfer or another method. Let me know and I'll send you my email address in a private message.

I'm seriously considering this camera, but this issue is nagging me and is still stopping me from pressing the "buy now" button.

Here's the best (or worst) result I could come up with. NEX6 + E10-18, iso100, 1/80, f5
Developed in Lightroom, + 1.0 EV, +100 shadows, -100 highlights, + 100 darks, 50/0.5/50 sharpening, 200% size:

Looking for barcodes (but did not find them

Thanks in advance for any help.

andrewD2 said:

mdcromer said:

I didn't see any RAW compression artifacts, just lens chromatic aberration in the area you highlighted.

Click to expand...

CA does not have a barcode pattern. End of. It looks nothing like CA.
If you are struggling (my girlfriend and I can see it at 50% zoom) this is at 200% from the more central window.
Click for full view.

Camera ILCE-7RM2Focal Length 55mmAperture f/3.5Exposure 1/160sISO 100

Click to expand...

Click to expand...

... So you took a picture with a different sensor tech, a different resolution, a different processor, and an entirely different location, and you're wondering why you didn't see the same effects? Even if the cameras were running exactly the same algorithm (which actually seems unlikely), such different inputs would produce entirely different results.

A7RII, the good, the bad and the ugly

Veteran Member

Senior Member

Leading Member

Forum Pro

Veteran Member

Veteran Member

Leading Member

Forum Pro

Forum Pro

Veteran Member

Veteran Member

Veteran Member

Forum Enthusiast

Veteran Member

Veteran Member

Veteran Member

Forum Enthusiast

Forum Enthusiast

Forum Enthusiast

Forum Enthusiast

Keyboard shortcuts