[Winder]

Is that what's causing these lines? Because I had been told that having PDAF points on the sensor had no down side and they didn't show up in the finished image. At the same time, in the images shown in the first post, I wouldn't have said that was "strong back lighting," just typical lighting for a portrait shot.

It is the PDAF array. DPR intentionally setup the shot to show the effect. All mirrorless cameras that have on sensor PDAF will show the effect under the right circumstances. It has been discussed before on mirrorless forums. Cameras like the A9 or A7III which have over 600 PDAF points will be more prone to it. You also need fast glass. If you are working in studio you can stop down to F/2.8 and probably not have the problem. The image used to show the effect was shot wide open specifically to show the reflection of the PDAF lines. It should have been shot at F/5.6 or F/8. There wouldn't have been a problem.

There is a reason you don't see anyone complaining about the problem with the A9 in the real world. For most of us its not an actual problem. I can see certain artistic situations where you want to shoot wide open into a bright light source, but that type of work is not what you buy an A9 for. I would buy a Fuji GFX50 if I was doing that kind of work. It doesn't have PDAF on sensor.

[photoptimist]

It is the PDAF array. DPR intentionally setup the shot to show the effect. All mirrorless cameras that have on sensor PDAF will show the effect under the right circumstances. It has been discussed before on mirrorless forums. Cameras like the A9 or A7III which have over 600 PDAF points will be more prone to it. You also need fast glass. If you are working in studio you can stop down to F/2.8 and probably not have the problem. The image used to show the effect was shot wide open specifically to show the reflection of the PDAF lines. It should have been shot at F/5.6 or F/8. There wouldn't have been a problem.

There is a reason you don't see anyone complaining about the problem with the A9 in the real world. For most of us its not an actual problem. I can see certain artistic situations where you want to shoot wide open into a bright light source, but that type of work is not what you buy an A9 for. I would buy a Fuji GFX50 if I was doing that kind of work. It doesn't have PDAF on sensor.

LOL! So mirrorless buyers should not waste money on fast glass such as 8 of the 9 Sigma Art lenses because they have unusably large apertures?

That said, I agree with you that "for most of us its not an actual problem" and suspect that most of supposed failings of Pentax fall into the same "storm in a teacup" pattern of click-bait internet rantings.

(P.S I'd be curious to see if tilt/shift lenses also suffer from this artifact. Tilting and shifting also creates the same kinds of extreme off-axis, unusual-angled light as a large aperture lens or a lens flare.)

[surfar]

Pretty annoying if you see it regularly in your photos, though

Oh Yes, however one photographer recently shot over 4100 images and didnt see this effect in any?

Pick on the the new kid at school/Tall poppy syndrome!....A7iii is top of a few selling charts for a reason.

On the other hand,$$$ony menus and WR are the weak points.

[BigMackCam]

Oh Yes, however one photographer recently shot over 4100 images and didnt see this effect in any?

Just because that photographer didn't see the effect doesn't mean that it can't happen in certain circumstances, though. Plenty of photographers have shot thousands of compressed .ARW photos and never noticed compression noise artefacts on high contrast edges. Some still argue that the problem doesn't even exist in normal shooting. I have numerous photos that prove otherwise.

Pick on the the new kid at school/Tall poppy sysndrome!....A7iii is top of a few selling charts for a reason.

Well, I'm certainly not picking on the new kid. I like Sony cameras... you know I do. But I see no harm in acknowledging that this is a potential weakness in certain shooting conditions. It's a minor one for most people, I suspect, but a weakness nonetheless. Better to know it's there, then the photographer can make appropriate choices to avoid it happening. After all, you wouldn't want to start processing your photos after a day's shoot, and find that some of your favourites have this artefact. If that happened, it would be cold comfort to be told that someone else took 4,100 shots without it appearing, right?

All things considered, the A7 MkIII looks like a fine camera to me, and one I'd be very happy to own. Like every other, we can't expect it to be perfect. But early indications suggest it's very good

Although this is a non-Pentax thread, @photoptimist makes a very good point... some of the complaints levelled at Pentax equipment are for equally minor things. But goodness me, it seems like an impossible task to convince people of that...

[surfar]

Well, I'm certainly not picking on the new kid

Not aimed at you...ha ha

@photoptimist makes a very good point...

Yes,always...

Bottom line is,the pefect camera exists to the buyer only...or Never!

[adriandavidpayne]

Oh Yes, however one photographer recently shot over 4100 images and didnt see this effect in any?

Pick on the the new kid at school/Tall poppy syndrome!....A7iii is top of a few selling charts for a reason.

On the other hand,$$$ony menus and WR are the weak points.

I'm here to tell you the Problem is Real! I got the camera yesterday. Went to the skate park today and took about 100 test shots. I would say there were 15-20 affected images where backlight was involved. Serious Problem! Check out the image.
This is a definite deal breaker! Any natural light photographer would have nightmares with this camera. As a strobe user, I'm beginning to wonder if the same thing is going to happen with the rim lights I use. Calling B&H tomorrow to discuss options.

[BigMackCam]

I'm here to tell you the Problem is Real! I got the camera yesterday. Went to the skate park today and took about 100 test shots. I would say there were 15-20 affected images where backlight was involved. Serious Problem! Check out the image.
This is a definite deal breaker! Any natural light photographer would have nightmares with this camera. As a strobe user, I'm beginning to wonder if the same thing is going to happen with the rim lights I use. Calling B&H tomorrow to discuss options.

That's very noticeable indeed Have you shot any of the earlier A7 and A7II models? I have the A7II and I've never noticed this issue, but then in all of my shooting I rarely take shots with so much intentional flare. I do wonder how many people will actually find this to be a real problem in normal use. But for those shooting into very strong light sources and eliciting flare as a result, it's going to be frustrating...

---

Last edited by BigMackCam; 04-13-2018 at 03:09 AM.

[Rondec]

Oh Yes, however one photographer recently shot over 4100 images and didnt see this effect in any?

Pick on the the new kid at school/Tall poppy syndrome!....A7iii is top of a few selling charts for a reason.

On the other hand,$$$ony menus and WR are the weak points.

What difference does it make that he shot 4100 photos or 10000? Clearly if you don't shoot a style where you shoot back lit photos with wide apertures it won't be a problem -- folks who shoot landscape, or architecture, or wildlife probably won't see this much. But folks who do wedding photography or portraits shoot those sort of images all of the time and they will see it.

I'm not a portrait photographer, but I do shoot photos of my kids and I just think this would show up in images like the one I attached.

I guess the price you pay for phenomenal auto focus performance in a mirrorless camera...

[beholder3]

I'm here to tell you the Problem is Real! I got the camera yesterday. Went to the skate park today and took about 100 test shots. I would say there were 15-20 affected images where backlight was involved. Serious Problem! Check out the image.
This is a definite deal breaker! Any natural light photographer would have nightmares with this camera. As a strobe user, I'm beginning to wonder if the same thing is going to happen with the rim lights I use. Calling B&H tomorrow to discuss options.

Did you consider posting that in the Amazon/DPR Sony section, where people have been analysing the effect?

[adriandavidpayne]

Did you consider posting that in the Amazon/DPR Sony section, where people have been analysing the effect?

I did a search for Amazon/DPR Sony but I can't find anything but a link for CDs Where can I find it?

[surfar]

Check out the image.

Yes, thats disappointing with your ratio for sure.The problems existed for quite a while though, same as the negative aspects of lots of other gear.

[clackers]

Yes, thats disappointing with your ratio for sure.The problems existed for quite a while though, same as the negative aspects of lots of other gear.

Yeah, but they're trying to convert more and more of the sensor pixels to AF ones, Surfar, and right across the sensor, not just in a central area, so this is increasingly a problem.

Having a mirror to do AF fixes this.

[BigMackCam]

I wonder if someone will come up with a software solution to address this - at least partially - in post processing. Given that it's a predictable repeating pattern, it should be readily identifiable through an algorithm. Once found, the software could use neighbouring pixels above and below the affected lines to calculate new HSL values. I'm sure there's more to it than that, but it seems like this approach might at least improve affected images.

Sony has quietly addressed previous IQ issues in software. An "Edge noise reduction" function was added to its own Image Data Converter software to deal with high contrast edge "noise" artefacts caused by RAW compression. Maybe they'll do something similar here...

[Rondec]

Yes, thats disappointing with your ratio for sure.The problems existed for quite a while though, same as the negative aspects of lots of other gear.

I thought you said it wasn't an issue. "Over 4100 photos taken without showing this effect..." not?

(Clearly it is an issue for some photographers and not for others. Sony just needs to be honest with those purchasing their gear that if they shoot this style of photography they need to look elsewhere).

[photoptimist]

I wonder if someone will come up with a software solution to address this - at least partially - in post processing. Given that it's a predictable repeating pattern, it should be readily identifiable through an algorithm. Once found, the software could use neighbouring pixels above and below the affected lines to calculate new HSL values. I'm sure there's more to it than that, but it seems like this approach might at least improve affected images.

Sony has quietly addressed previous IQ issues in software. An "Edge noise reduction" function was added to its own Image Data Converter software to deal with high contrast edge "noise" artefacts caused by RAW compression. Maybe they'll do something similar here...

Maybe. But it's likely to create a "Sony line-eater" problem to go with the "Sony star-eater" problem.

The central problem with any kind of artifact removal software is in what it does to image features that just happen to look similar to the artifact. Hair, spider's webs, star trails, thin edges on buildings, fire works, etc. all can create thin lines that may be horizontal or vertical in some part of the image. Based on the strip-artifact images being presented, an astounding 17% of all the pixels of the A7Riii sensor have the potential to show this effect. That means the chance of some linear image feature getting eaten by the post processor would be quite high.