A7RII posterization?

[Iliah Borg]

Iliah Borg wrote:
the profiles I use at the raw conversion stage can't cause posterization as they are simple matrix transforms

so you never ever use LUT profiles in your raw conversions (I mean the regular commercial work) ?

Right. I do it in 2 stages, raw conversion using a matrix transform, and after the raw conversion applying a correction transform with gamut mapping control. Correction transform is LUT-based.

 

[timparkin]

Iliah Borg wrote:
the profiles I use at the raw conversion stage can't cause posterization as they are simple matrix transforms

so you never ever use LUT profiles in your raw conversions (I mean the regular commercial work) ?

Right. I do it in 2 stages, raw conversion using a matrix transform, and after the raw conversion applying a correction transform with gamut mapping control. Correction transform is LUT-based.

 

[Iliah Borg]

Red channel looks very weak

Not just weak. SONY are in need of a bit more transparency in their manuals to avoid this when a photographer needs a cleaner image:

If I read that correctly, you're just showing that the Sony files are 13 bit

No, that's not what the screenshot shows.

Care to clarify this a bit some more? Otherwize people will be left a bit in dark about the meaning of the screenshot.

Very few values in the red channel over a largish selection, and missing pixels at one of the values in the middle. This is posterization, and it will become more and more visible down the road..

--
http://www.libraw.org/

 

[Brad]

I am so far very happy with my A7RII but this discussion and Lloyd's picture did give me the incentive to review images I made with large bodies of water. This particular image was taken with the A7RII and an old but excellent condition Takumar 28mm 3.5 . I opened the raw file in PS and applied a small amount of distortion correction...nothing else...the OOC Jpg looks nearly identical...I'm not sure how to describe the effect in the lower left corner, is this posterization or something else?




100% crop

Water often looks posterized, even from film.



Ignoring the coarse granularity, note the areas of apparently flat tone.

Appreciate the comments

 

[Dave Sanders]

His business model still represents a clear conflict of interest (as does anyone who's source of income is an add supported website, but to a lesser extent).

He also has a certain narrative (in this case it's 'Sony doesn't understand photography') and sticks to it. He has the opposite opinion of Sigma's DP Merrill cameras, I sent him some evidence of their artifacts and RAW file cooking but he wasn't interested as it didn't fit the narrative he'd established.

His business model is much more subscription rather than ad based, though obviously clicks don't hurt. His main sponsor is Zeiss and, personally, that is a bias I do see. He rags on Nikon and Canon for 'not getting it' just as much as he does Sony for poor execution. His reviews are mostly photos with not a lot of words. I respect the fact the he has a passion first, hiking and climbing, and uses photography in its service.

Most people who dismiss him have never spent the money to actually see what kind of content he provides. There is some misinformation in this thread, I am merely countering it. I find that, due to the sheer volume of photos he provides, you can really make up your own mind, much more so than with other reviewers like Rockwell or Huff who truly do just trade in clickbait.

 

[mdcromer]

Red channel looks very weak

Not just weak. SONY are in need of a bit more transparency in their manuals to avoid this when a photographer needs a cleaner image:

If I read that correctly, you're just showing that the Sony files are 13 bit

No, that's not what the screenshot shows.

--
http://www.libraw.org/

I'm sorry, I can't guess what this screenshot is supposed to tell me, other than that alternate values are missing due to 13 bits (and missing values for one single expected value, which we've also seen before from Jim Kasson's research). Oh, and the red channel seems very low as well. What else is relevant to this discussion?

 

[timparkin]

Iliah Borg wrote:
the profiles I use at the raw conversion stage can't cause posterization as they are simple matrix transforms

so you never ever use LUT profiles in your raw conversions (I mean the regular commercial work) ?

Right. I do it in 2 stages, raw conversion using a matrix transform, and after the raw conversion applying a correction transform with gamut mapping control. Correction transform is LUT-based.

 

[mdcromer]

Sony A7RII + sony 70-300 SSM + LA-EA3 @300mm f5,6

Is that the same problem i'm experiencing here?

Jpg X.Fine ooc. It is the same with steady shot on or off. I think it related to the particular lens because the Sony-Zeiss 135 f1,8 does not have this kind of artifacts.

Heat shimmer I think, combined with JPEG noise reduction.

 

[mdcromer]

Thank you for your attention by the way...

It is happening only with focal lengths over the 135mm but i can see it with every kind of weather so i don't think it is connected with haze...and my Nikon in the same conditions does not show anything similar!

It's not haze, it's atmospheric disturbance / refraction. It happens with any long-distance view through the atmosphere, magnified by longer lenses.

It's a huge issue for astronomy and why telescopes are located as high up as possible (to minimize it by reducing the amount of atmosphere light rays have to travel through).

 

[mdcromer]

Potential posterization issue with A7RII documented here: http://diglloyd.com/blog/2015/20150819_1136-SonyA7R_II-posterization-BlueLake.html

Lloyd shared the RAW with a reputable getdpi forum member and professional photographer, and he confirmed the issue, at least with this particular exposure, in this thread: http://www.getdpi.com/forum/sony/55836-seeing-any-posterization-issues-a7r-ll.html

Was a thread about this already, but it seems to have been lost. Perhaps we can all discuss this like adults, without the usual accusations of fanboy, troll, shill, willfully blind to any criticism of one's chosen brand, etc.

I would hate to think dpreview is attempting to quash discussion of a potential problem with a new camera. This is, after all, how products improve.

This is not posterization issue it is a color management issue ...

It can be seen in RawDigger is that he drove with semi-mechanical shutter time which halves the bit resolution in the readout. Well documented, should be avoided at base ISO and ISO200. ISO400 and up it makes no difference, the noise is much higher than the lowest quantization level.

I think by "semi-mechanical shutter" you mean EFCS.

That doesn't reduce bit depth, only the fully electronic shutter does that.

The missing values in the histogram are because the Sony cameras to date are only 13 bit, not 14, so every other value is missing. Not sure what causes the single missing expected value - but we've seen that in Jim Kasson's tests for years. Perhaps some kind of A/D artifact?

 

[billzilla2000]

He did post them. Most people don't subscribe to his site or don't care to visit. Lloyd's testing is always transparent and full of high resolution images. One can make up their own mind. His other shots are outstanding, but he saw something he didn't like and posted. He's a reviewer, not a cheer leader.
--
Dave Sanders

He's a clickbaiter shock jock looking to stir up sensational attention. Has been doing the 'OMG SONY SUCKS' thing for months, years maybe.

Lloyd's 'testing' has mostly been other people's samples and he emphasizes the bad and dumps (again) the analyses of others out in order to lend his grave assertions some credence without full context. I take anything he says with a huge grain of salt, at minimum.

I'm not a huge fan of his photography but your assertion that he uses what others do is patently false. Not a subscriber, I take it. He provides multiple scenes for every lens or camera, and within that scene, every aperture. He generally leaves the photos to speak for themselves and provides little analysis.
--
Dave Sanders

His cRAW initial 'findings' were based on the research and samples of others. His data is lifted to balance his sweeping, pulpit-pounding conclusions. Nothing FALSE about it, that's what he did.

He's a hack writer and photographer baiting clicks, IMO and I don't have to pay the guy to see his style at work in the public access articles.

 

[Dave Sanders]

He did post them. Most people don't subscribe to his site or don't care to visit. Lloyd's testing is always transparent and full of high resolution images. One can make up their own mind. His other shots are outstanding, but he saw something he didn't like and posted. He's a reviewer, not a cheer leader.
--
Dave Sanders

He's a clickbaiter shock jock looking to stir up sensational attention. Has been doing the 'OMG SONY SUCKS' thing for months, years maybe.

Lloyd's 'testing' has mostly been other people's samples and he emphasizes the bad and dumps (again) the analyses of others out in order to lend his grave assertions some credence without full context. I take anything he says with a huge grain of salt, at minimum.

I'm not a huge fan of his photography but your assertion that he uses what others do is patently false. Not a subscriber, I take it. He provides multiple scenes for every lens or camera, and within that scene, every aperture. He generally leaves the photos to speak for themselves and provides little analysis.
--
Dave Sanders

His cRAW initial 'findings' were based on the research and samples of others. His data is lifted to balance his sweeping, pulpit-pounding conclusions. Nothing FALSE about it, that's what he did.

He's a hack writer and photographer baiting clicks, IMO and I don't have to pay the guy to see his style at work in the public access articles.

He still provides far better photos for every camera and lens he reviews than any other reviewer. As an experienced photographer, all I need is a photo. I can make up my own mind about what I see.
--
Dave Sanders

 

[billzilla2000]

You analysis here, as usual, is full of unsubstantiated hyperbole.

Ironic, seeing your defense of someone whose writings rely on that very thing, to a great extent.

His reviews, far from being useful only for neophytes, are quite the opposite: useful for a photographer with enough experience and skills to analyze a photo and an aperture series and make their own conclusions. His test if the Batis 85 has a large number of portraits. Again, not a subscriber, I take it. Though your preference for Huff and Rockwell say all I need to know.

You can like the guy all you want, but his writing is clearly heavily biased and laden with needlessly dramatic exaggerations and wild conclusions based on loose or exceedingly elaborate data that requires a great deal of context to understand and 'decode'. Add to that a clear dislike for Sony products and a strong case of confirmation bias and you have the antithesis of what I'd call reliable reference.

If you agree with his opinions and bias, I can see why you'd be so quick to stick up for him.

 

[JimKasson]

but also a lot of damage is done in conversion from ProPhoto RGB to output space.

Iliah, can you describe how this could be avoided?

By using soft-proofing in Photoshop when doing the conversion.

Right you are. Lr soft proofing is broken.

Jim

 

[billzilla2000]

He did post them. Most people don't subscribe to his site or don't care to visit. Lloyd's testing is always transparent and full of high resolution images. One can make up their own mind. His other shots are outstanding, but he saw something he didn't like and posted. He's a reviewer, not a cheer leader.
--
Dave Sanders

He's a clickbaiter shock jock looking to stir up sensational attention. Has been doing the 'OMG SONY SUCKS' thing for months, years maybe.

Lloyd's 'testing' has mostly been other people's samples and he emphasizes the bad and dumps (again) the analyses of others out in order to lend his grave assertions some credence without full context. I take anything he says with a huge grain of salt, at minimum.

I'm not a huge fan of his photography but your assertion that he uses what others do is patently false. Not a subscriber, I take it. He provides multiple scenes for every lens or camera, and within that scene, every aperture. He generally leaves the photos to speak for themselves and provides little analysis.
--
Dave Sanders

His cRAW initial 'findings' were based on the research and samples of others. His data is lifted to balance his sweeping, pulpit-pounding conclusions. Nothing FALSE about it, that's what he did.

He's a hack writer and photographer baiting clicks, IMO and I don't have to pay the guy to see his style at work in the public access articles.

He still provides far better photos for every camera and lens he reviews than any other reviewer. As an experienced photographer, all I need is a photo. I can make up my own mind about what I see.
--
Dave Sanders

Comparative analysis requires more than just 'a photo'. You need a clear, controlled process and an unbiased approach. Lloyd lacks any of that. I thought you weren't 'a fan of his photography'?

You don't have to defend your subscription here. Some of us think he's a Sony bashing clown and hack, but you choose to purchase his writing and that's fine if you feel it adds something to your decision making process.

 

[Dave Sanders]

He did post them. Most people don't subscribe to his site or don't care to visit. Lloyd's testing is always transparent and full of high resolution images. One can make up their own mind. His other shots are outstanding, but he saw something he didn't like and posted. He's a reviewer, not a cheer leader.
--
Dave Sanders

He's a clickbaiter shock jock looking to stir up sensational attention. Has been doing the 'OMG SONY SUCKS' thing for months, years maybe.

Lloyd's 'testing' has mostly been other people's samples and he emphasizes the bad and dumps (again) the analyses of others out in order to lend his grave assertions some credence without full context. I take anything he says with a huge grain of salt, at minimum.

I'm not a huge fan of his photography but your assertion that he uses what others do is patently false. Not a subscriber, I take it. He provides multiple scenes for every lens or camera, and within that scene, every aperture. He generally leaves the photos to speak for themselves and provides little analysis.
--
Dave Sanders

His cRAW initial 'findings' were based on the research and samples of others. His data is lifted to balance his sweeping, pulpit-pounding conclusions. Nothing FALSE about it, that's what he did.

He's a hack writer and photographer baiting clicks, IMO and I don't have to pay the guy to see his style at work in the public access articles.

He still provides far better photos for every camera and lens he reviews than any other reviewer. As an experienced photographer, all I need is a photo. I can make up my own mind about what I see.
--
Dave Sanders

Comparative analysis requires more than just 'a photo'. You need a clear, controlled process and an unbiased approach. Lloyd lacks any of that. I thought you weren't 'a fan of his photography'?

You don't have to defend your subscription here. Some of us think he's a Sony bashing clown and hack, but you choose to purchase his writing and that's fine if you feel it adds something to your decision making process.

I'm not a big fan of Lloyd's photography, but he does provide a consistent and controlled testing process. I have yet to see anyone on this board, yourself or myself included, provide anything remotely equivalent. I'd love to see some photos shot in the real world at every aperture with the same composition. And I'm not being facetious, I really would. It would be highly instructive. Those are precisely the type of posts that I come to this forum for. Perhaps I'll get off my butt and do some when I'm back from vacation.

Anyways, you see an obvious bias in what he does, fair enough. I find his photos instructive, especially when he reviews lenses and provides an aperture series...it is because of this the B25 is the top of my acquisitions list. But we can agree to disagree here.

Dave Sanders

 

[Dave Sanders]

You analysis here, as usual, is full of unsubstantiated hyperbole.

Ironic, seeing your defense of someone whose writings rely on that very thing, to a great extent.

His reviews, far from being useful only for neophytes, are quite the opposite: useful for a photographer with enough experience and skills to analyze a photo and an aperture series and make their own conclusions. His test if the Batis 85 has a large number of portraits. Again, not a subscriber, I take it. Though your preference for Huff and Rockwell say all I need to know.

You can like the guy all you want, but his writing is clearly heavily biased and laden with needlessly dramatic exaggerations and wild conclusions based on loose or exceedingly elaborate data that requires a great deal of context to understand and 'decode'. Add to that a clear dislike for Sony products and a strong case of confirmation bias and you have the antithesis of what I'd call reliable reference.

If you agree with his opinions and bias, I can see why you'd be so quick to stick up for him.

If you disagree with his opinions, I can see why you're so quick to denigrate him. Feel free to post photos of your own testing to refute.
--
Dave Sanders

 

[billzilla2000]

You analysis here, as usual, is full of unsubstantiated hyperbole.

Ironic, seeing your defense of someone whose writings rely on that very thing, to a great extent.

His reviews, far from being useful only for neophytes, are quite the opposite: useful for a photographer with enough experience and skills to analyze a photo and an aperture series and make their own conclusions. His test if the Batis 85 has a large number of portraits. Again, not a subscriber, I take it. Though your preference for Huff and Rockwell say all I need to know.

You can like the guy all you want, but his writing is clearly heavily biased and laden with needlessly dramatic exaggerations and wild conclusions based on loose or exceedingly elaborate data that requires a great deal of context to understand and 'decode'. Add to that a clear dislike for Sony products and a strong case of confirmation bias and you have the antithesis of what I'd call reliable reference.

If you agree with his opinions and bias, I can see why you'd be so quick to stick up for him.

If you disagree with his opinions, I can see why you're so quick to denigrate him. Feel free to post photos of your own testing to refute.
--
Dave Sanders

That's the problem. He doesn't post photos that prove the extreme language several people have quoted verbatim from his articles. His posterized sample is clearly not indicative of average real world shooting, there's more there that needs to be addressed.

Should I post a scope pic of a dirty A7 sensor and then go on about what dust magnets the cameras are? I could post a dozen horrible dirty sensor pics and go on a tirade like Chambers does, but the contextual truth many already know is, it's in how careful you are with changing lenses, your upkeep process and how often you clean, which tools, etc. Same basic thing.

He makes clickbait out of minutiae taken out of context and blown out of proportion, then tops it off with heavy doses of 'Sony doesn't know how to make cameras' and so forth. And you're saying I need to post some kind of series of photos to prove that point? No, sorry.

 

[Ari Aikomus]

Iliah Borg wrote:
the profiles I use at the raw conversion stage can't cause posterization as they are simple matrix transforms

so you never ever use LUT profiles in your raw conversions (I mean the regular commercial work) ?

Right. I do it in 2 stages, raw conversion using a matrix transform, and after the raw conversion applying a correction transform with gamut mapping control. Correction transform is LUT-based.

 

[mdcromer]

It appears that the only thing that needs to be done is for Lloyd to change his color management workflow to use a consistent color space.