2021 Sony Master Series A90J (55", 65" and 83") [Owners' & Discussion Thread]

[ray0414]

Originally Posted by samlop10

Very close to what Dnice got (~750 nits calibrated and accurate settings seems correct for these sets).

I wonder if the CX John tested had DTM turned on or if it was using the static tone mapping. I prefer static since it is more accurate to the source (and if he had DTM turned on that might explain why the image on the CX was brightening more around bright objects compared to the Sony).

Overall though, the improvements still seem small imo. It definitely improved on previous Sony OLEDs but in other ways it seems to be catching up more to previous LG OLEDs instead of surpassing them in everything. I wonder how it’ll compare to this year’s LGs. I wouldn’t be surprised if the G1s turn out to be brighter than the A90Js.

Hate to the bearer of bad news, but when Dnice was called into the livestream yesterday, his exact qoute was that the A90J was "trouncing the CX in all brightness categories" now after the panel got brighter after a week's usuage. There's also multiple owners at avs now saying the same thing who also have LG oleds.

Also. See the attached picture.

[apriori]

Man, Samlop10, you really seem to have it in for this TV. The impressions I’m reading are all at odds with your assumption that the A90J is playing a catch up game to the CX. We should be excited that this is a significant leap forward in OLEDs.

[samlop10]

Originally Posted by ray0414

Hate to the bearer of bad news, but when Dnice was called into the livestream yesterday, his exact qoute was that the A90J was "trouncing the CX in all brightness categories" now after the panel got brighter after a week's usuage. There's also multiple owners at avs now saying the same thing who also have LG oleds.

Also. See the attached picture.

What’s he measuring now though? Another user calibrated and measured his own set (see tama’s quote above) and he also said he got just above 750 nits peak. So did Dnice mean it was brighter with APL and/or non accurate modes? or did he actually get a brighter peak calibrated and with settings favoring accuracy?

Because like mentioned before, it’s definitely brighter in non-accurate picture modes. But if you prefer the accurate one(s), the peak brightness is comparable to previous LG OLEDs.

[NeilZ]

Originally Posted by samlop10

Why would the processor do that? Like, what would be the purpose of it doing that and limiting the brightness of test patterns (which includes white boxes)? If that were the case it would mess up the calibration since calibration requires test patterns. Doesn’t make sense.

And if that were the case (despite the fact it does not make sense) what happens when you turn the processor off because you want 4K/HDR content to be accurate to the source? Does the peak brightness drop because it’s not doing any recognition of objects for bright highlights?

It's not the reality creation or light/color reading type of processing we're talking about, from my understanding we're talking about the built in processing. The tv recognizes that the test pattern is not actual content and as such is not pushing the panel. Is it weird for the test patterns to betray the true capabilities of the panel? Sure. Still this helps explain the readings and viewing experience mismatch that has been going around.

I think Vincent will probably get more in details over this. Some other users are now also reporting that actual content with most accurate out of box settings are reaching close to 1000 nits in specular highlights.

[tama]

Originally Posted by samlop10

What’s he measuring now though? Another user calibrated and measured his own set (see tama’s quote above) and he also said he got just above 750 nits peak. So did Dnice mean it was brighter with APL and/or non accurate modes? or did he actually get a brighter peak calibrated and with settings favoring accuracy?

Because like mentioned before, it’s definitely brighter in non-accurate picture modes. But if you prefer the accurate one(s), the peak brightness is comparable to previous LG OLEDs.

They're jumping between 2 threads over at AVSFourm so I'm trying to keep up when I can. But D-nice stated when he made his comparisons his Tvs where in their correct calibrated modes. And he feels there's good improvement but also he broke in his set with 100 hours before calibration but he said 10-15 hours later after calibration it seemed his New Sony really brightened up. He remeasured 780 nits a day or two after his initial reading but before he made the statement about the TV really taking off. Not sure if anything has changed since.

Still waiting to see if other well know trusted calibrator obtain sets and chime in. Chad B. Or even Vincent Teoh

[Kingchin]

Originally Posted by ray0414

It had been pointed out that the term Evo panel is actually used by LG Electronics, it LG Display, therefore Sony could get into copyright trouble using the term or saying they use it.

And also, the evo panels are not exclusive to LG. Both Panasonic and Philips have confirmed they will also be using them this year. I'm not sure if they used the actual term evo or not but the green Emmitter will be in their top oled tvs this year.

LG Display is a subsidiary of LG Corp so the decision would lay with them to either sell the EVO panel to Sony and Panasonic or keep it exclusive to themselves for this year.

Obviously the Sony engineer who told Robert Zohn that the Sony A90J doesn't use the EVO panel would know if it was the LG G1 panel the question was about. If there was some legal agreement then why didn't he just say no comment.

Instead of saying that the Sony A90J doesn't use the New LG EVO panel and that the high brightness is actually only done from Sony's new processor together with Sony's new custom heatsink.

There has been speculation gossip that Sony, Panasonic and Philips will use the EVO panel but no actual statements from Sony, Panasonic and Philips that they definitely use the EVO panel. If you have links to prove otherwise then post them by all means.

LG Display other new panel for 2021 could also have a different green layer and also blue layer. If that's the case then obviously the spectral response of the Sony A90J will look different from last year's tv's.

I'm not bashing the Sony A90J as I'm almost set to buy the 65inch version here in the UK next month. Looks like a cracking tv, and it could well turn out to use the EVO panel. I just like clarification on details that should be told to consumers before I pay £3,500 for a product.

[samlop10]

Originally Posted by apriori

Man, Samlop10, you really seem to have it in for this TV. The impressions I’m reading are all at odds with your assumption that the A90J is playing a catch up game to the CX. We should be excited that this is a significant leap forward in OLEDs.

Look, I’ll explain my, I guess, little background with this model.

When it was first announced I got pretty excited. The last big innovative TV Sony had released was the Z9D. Since then, it kind of seemed like they had gone backwards, and later they had started providing meager improvements from year to year, but nothing innovative. With this model they said they would have a new (and better) panel along with the heat sink. Then the very early impressions had measured 1200+ nits. Obviously, I knew once the sets were calibrated and the settings adjusted, that would drop. But even then I thought Sony had finally done something to innovate TVs instead of just provide meager improvements with those two new changes. I was actually very seriously considering buying the 83” model (see one of my early posts on this thread).

Thing is, now that reviews and measurements have started, the nuances of how those improvements play out are quite important in my opinion. And if you prefer accuracy, those improvements are barely present, imo anyway; I’ll explain:

One thing that is both a pro and a con of OLEDs now (from LG, Sony, and Panasonic, at least) is they are quite accurate in how they render picture up to their panels’ capabilities. So now, the only substantial improvements to make are brightness across all colors (i.e. peak white as well as color volume). Based on the measurements we have gotten so far, if you prefer accuracy, the drop from that 1200+ nits is substantial (dropping around 450 nits, close to ~750 for final peak brightness after calibration and accurate settings), to the point that you will only get small improvements in peak brightness. So after those reports, it really makes it hard, at least for me, to get excited and justify getting it given the bumps in picture quality from my current set are relatively minor. I know a lot of people are excited about this model and I come off as a party pooper, but I also think it is important for people to realize what they are actually getting from this model and decide for themselves whether this model is worth it, based on measurements, their own preferences and priorities, and the jump in quality based on their current TVs.

Having said that, if you don’t mind altering the picture from its source to sacrifice accuracy (or if you have previous Sony model, or a 2+ years older LG OLED model), then yes, the improvements are higher (although I still would not consider them substantial, but that is my personal opinion). You get picture modes with higher brightness (at the expense of clipping and a brighter image than it should be), better motion (with motion interpolation), and better post-processing with the new XR processor (changing how objects look from their source). Some people might be ok with that, and if they think that makes it worth buying, then by all means. Each person can decide what to do with their own money.

However, if you value accuracy (and/or you have a recent LG model), the only improvement here is really APL, which very little real-world content (i.e. series and movies) would actually take advantage of. Color is supposed to be a little better than previous LG OLEDs (although they are no slouch in that anyway), near-black is also a little improved (only important if you watch content with heavy compression, and even then, the last previous models were already improved in that), and that is pretty much it. Other than APL, the improvements when using settings favoring accuracy are pretty minor given how well previous models do. And that is not even mentioning the lower number of HDMI 2.1 inputs (4 on LGs, 2 on Sonys). I think this matters, even if it is just a matter of principle (comparing models and prices between models from different brands). This is a premium model, LG has included 4 HDMI 2.1 inputs from over two years ago (while still costing less, even on their non-premium OLED models), and Sony still cannot offer more of them to even match them. They have yet to release a receiver (or HDMI switcher) that can handle more than one HDMI 2.1 input without bugs, so they become more important, especially if you rely on eARC.

So again, if you’re ok with sacrificing some accuracy, it seems the improvements are a little higher, but if you value accuracy more, then the improvements are pretty minor imo. Hence why I am hesitant to just get carried away by what people are saying and get excited about this model. If those measurements improve after calibration while still in settings favoring accuracy, then I will get excited if those numbers are actually substantial (i.e., 900+ nits, and I would be willing to overlook the lack of HDMI 2.1 inputs. But, given how human eyes work, any lower than that will not be very noticeable and tone mapping is still going to be required for like 99% of HDR content).

And, even after all that, that is still not comparing it against this year’s LG OLEDs, which are supposed to also provide improvements, especially the G1 models. So I would at least wait for results from these as well before making a decision.

So, I think those nuances (improvements of non-accurate modes vs accurate ones) are coming into play here, with most people seemingly judging it in non-accurate modes given the current results of the measurements. So after all that, it makes me skeptical given how the numbers do not really match with what people are supposedly seeing (at least not with picture-accurate settings), and thus it makes it hard for me to get excited about it since I value accuracy more, despite other various people getting excited.

[samlop10]

Originally Posted by NeilZ

It's not the reality creation or light/color reading type of processing we're talking about, from my understanding we're talking about the built in processing. The tv recognizes that the test pattern is not actual content and as such is not pushing the panel. Is it weird for the test patterns to betray the true capabilities of the panel? Sure. Still this helps explain the readings and viewing experience mismatch that has been going around.

I think Vincent will probably get more in details over this. Some other users are now also reporting that actual content with most accurate out of box settings are reaching close to 1000 nits in specular highlights.

It was just posted above using 75% to measure peak brightness is standard for calibrations, hence the headroom:

Originally Posted by tama

[Show spoiler]

This was posted: Post #1497

https://www.avsforum.com/threads/202...184217/page-75

Those guidances seem universal for HDR televisions, not just Sony. I guess Sony could still be leaving even more headroom by trying to detect patterns and lowering their brightness when displayed. But have they actually tried measuring those peaks with real world content after calibration and accuracy-favoring settings? The numbers of those measurements would still matter. But if they are substantial (900+nits), then that would be pretty cool. Seems like an important thing to do if that is the case. Otherwise it’s just speculation.

[Kingchin]

Originally Posted by samlop10

Look, I’ll explain my, I guess, little background with this model.

When it was first announced I got pretty excited. The last big innovative TV Sony had released was the Z9D. Since then, it kind of seemed like they had gone backwards, and later they had started providing meager improvements from year to year, but nothing innovative. With this model they said they would have a new (and better) panel along with the heat sink. Then the very early impressions had measured 1200+ nits. Obviously, I knew once the sets were calibrated and the settings adjusted, that would drop. But even then I thought Sony had finally done something to innovate TVs instead of just provide meager improvements with those two new changes. I was actually very seriously considering buying the 83” model (see one of my early posts on this thread).

Thing is, now that reviews and measurements have started, the nuances of how those improvements play out are quite important in my opinion. And if you prefer accuracy, those improvements are barely present, imo anyway; I’ll explain:

One thing that is both a pro and a con of OLEDs now (from LG, Sony, and Panasonic, at least) is they are quite accurate in how they render picture up to their panels’ capabilities. So now, the only substantial improvements to make are brightness across all colors (i.e. peak white as well as color volume). Based on the measurements we have gotten so far, if you prefer accuracy, the drop from that 1200+ nits is substantial (dropping around 450 nits, close to ~750 for final peak brightness after calibration and accurate settings), to the point that you will only get small improvements in peak brightness. So after those reports, it really makes it hard, at least for me, to get excited and justify getting it given the bumps in picture quality from my current set are relatively minor. I know a lot of people are excited about this model and I come off as a party pooper, but I also think it is important for people to realize what they are actually getting from this model and decide for themselves whether this model is worth it, based on measurements, their own preferences and priorities, and the jump in quality based on their current TVs.

Having said that, if you don’t mind altering the picture from its source to sacrifice accuracy (or if you have previous Sony model, or a 2+ years older LG OLED model), then yes, the improvements are higher (although I still would not consider them substantial, but that is my personal opinion). You get picture modes with higher brightness (at the expense of clipping and a brighter image than it should be), better motion (with motion interpolation), and better post-processing with the new XR processor (changing how objects look from their source). Some people might be ok with that, and if they think that makes it worth buying, then by all means. Each person can decide what to do with their own money.

However, if you value accuracy (and/or you have a recent LG model), the only improvement here is really APL, which very little real-world content (i.e. series and movies) would actually take advantage of. Color is supposed to be a little better than previous LG OLEDs (although they are no slouch in that anyway), near-black is also a little improved (only important if you watch content with heavy compression, and even then, the last previous models were already improved in that), and that is pretty much it. Other than APL, the improvements when using settings favoring accuracy are pretty minor given how well previous models do. And that is not even mentioning the lower number of HDMI 2.1 inputs (4 on LGs, 2 on Sonys). I think this matters, even if it is just a matter of principle (comparing models and prices between models from different brands). This is a premium model, LG has included 4 HDMI 2.1 inputs from over two years ago (while still costing less, even on their non-premium OLED models), and Sony still cannot offer more of them to even match them. They have yet to release a receiver (or HDMI switcher) that can handle more than one HDMI 2.1 input without bugs, so they become more important, especially if you rely on eARC.

So again, if you’re ok with sacrificing some accuracy, it seems the improvements are a little higher, but if you value accuracy more, then the improvements are pretty minor imo. Hence why I am hesitant to just get carried away by what people are saying and get excited about this model. If those measurements improve after calibration while still in settings favoring accuracy, then I will get excited if those numbers are actually substantial (i.e., 900+ nits, and I would be willing to overlook the lack of HDMI 2.1 inputs. But, given how human eyes work, any lower than that will not be very noticeable and tone mapping is still going to be required for like 99% of HDR content).

And, even after all that, that is still not comparing it against this year’s LG OLEDs, which are supposed to also provide improvements, especially the G1 models. So I would at least wait for results from these as well before making a decision.

So, I think those nuances (improvements of non-accurate modes vs accurate ones) are coming into play here, with most people seemingly judging it in non-accurate modes given the current results of the measurements. So after all that, it makes me skeptical given how the numbers do not really match with what people are supposedly seeing (at least not with picture-accurate settings), and thus it makes it hard for me to get excited about it since I value accuracy more, despite other various people getting excited.

I do think some of the improvements from the Sony A90J are actually a decent improvement from last years OLEDs. Like a cleaner panel, less vertical banding, no ABL in SDR and virtually none in HDR apart from very difficult scenes. Brightness in specular highlights in HDR seem to also be improved compared to last years OLEDs.

I myself like a accurate picture faithful to the film directors intent. Good guys but why on earth D-nice, John Reformato, Robert Zohn said on the YouTube Livestream that they preferred having Reality Creation turned on even on 4k sources. I don't agree with doing that for blu ray, 4k blu ray and I'm probably not the only one who will turn Reality Creation off.


They even kept repeating how magnificent real looking Reality Creation is. When it actually uses noise reduction and detail enhancement to do the opposite by artificially changing the original image source.

They should have at least tested a few 4k scenes with the artificial junk like Reality Creation turned off and commented what the picture quality was like done that way compared to last years OLEDs that had been calibrated correctly for film accuracy.

[samlop10]

Originally Posted by tama

They're jumping between 2 threads over at AVSFourm so I'm trying to keep up when I can. But D-nice stated when he made his comparisons his Tvs where in their correct calibrated modes. And he feels there's good improvement but also he broke in his set with 100 hours before calibration but he said 10-15 hours later after calibration it seemed his New Sony really brightened up. He remeasured 780 nits a day or two after his initial reading but before he made the statement about the TV really taking off. Not sure if anything has changed since.

Still waiting to see if other well know trusted calibrator obtain sets and chime in. Chad B. Or even Vincent Teoh

Yeah, if that is the case then it would be important to see the new measurements from Dnice.

If other calibrators/reviewers share their measurements, it would also be important for them to share their own results of previous OLED models (like Dnice has done so far) since, going by what calibrators have said on AVSForums and other places, measurement readings can vary depending on the specific calibration equipment used, even when measuring the exact same set. I.e., two calibrators could end up getting different numbers from the exact same TV if they use different models of colorimeters, hence why it would be important for them to also share their results from measurements of previous OLED models to have points of reference for each of their calibration equipments and fairly see how they compare to this one. Complicated, but important if one wants more reliable results that fairly compare previous OLED models.

[samlop10]

Originally Posted by Kingchin

I do think some of the improvements from the Sony A90J are actually a decent improvement from last years OLEDs. Like a cleaner panel, less vertical banding, no ABL in SDR and virtually none in HDR apart from very difficult scenes.

I myself like a accurate picture faithful to the film directors intent. Why on earth D-nice, John Reformato, Robert Zohn said on the YouTube Livestream that they preferred having Reality Creation on even on 4k sources I don't agree with doing that.


They even kept repeating how magnificent real looking Reality Creation is. When it actually uses noise reduction and detail enhancement to do the opposite by artificially changing the original image source.

They should have at least tested a few 4k scenes with all the artificial junk turned off and commented what the picture quality was like done that way compared to last years OLEDs that were calibrated correctly for film accuracy.

I think to an extent they feel the need to show improvements to help the industry keep going business-wise since to an extent their jobs depend on it. Even when some people would not benefit from those improvements much. That is all my from my own thoughts, but it does make sense I think.

If they had turned all that post-processing off affecting source-accuracy, then the improvements would be relatively marginal, and I do not think that helps the industry on the long-run. After all, like I said, to an extent, they need people to keep buying new TVs since their jobs (calibrators, reviewers, TV sellers) usually do better when people are more interested in new TVs and they buy them. I do not blame them since after all most of us need work to make a living, but it does make things a little murky for the consumers. I appreciate at least some users on AVSForums are relatively honest with their results. That way we can analyze things such as these and have more objective results to help us make better buying decisions.

[Kingchin]

Originally Posted by tama

From D-Nice:
I’ll talk to Robert about this tomorrow. It is the new panel design that LGE calls EVO. The Spectral Response proves that. Blue is totally different compared to previous years. I’ll post the individual RGB Spectral Responses tomorrow

Even experts get stuff wrong, D-Nice has previous history of stating facts that Vincent Teoh proved wrong

Here's the latest from a guy who has now received his LG G1 -
PicsArt_03-12-12.30.06.jpg
PicsArt_03-12-12.28.01.jpg

The Sony A90J very well might use the EVO panel but we will find out very soon. As some people now have the LG G1 with the embargo ending next week. There will be way's to definitely find out if the LG G1 panel is different from the Sony A90J panel.

[NeilZ]

The A90J does use the new panel. It's obvious by looking at the energy ratings alone. Panasonic had a heat sink and used nothing close to those numbers. You don't get 20% more efficient out of thin air. Also no past LG oled could ever reach 1400 nits in vivid. Sony is not known to drive panels that hard either. There's no way this is last year's oled panel.

[NeilZ]

Originally Posted by samlop10

It was just posted above using 75% to measure peak brightness is standard for calibrations, hence the headroom:



Those guidances seem universal for HDR televisions, not just Sony. I guess Sony could still be leaving even more headroom by trying to detect patterns and lowering their brightness when displayed. But have they actually tried measuring those peaks with real world content after calibration and accuracy-favoring settings? The numbers of those measurements would still matter. But if they are substantial (900+nits), then that would be pretty cool. Seems like an important thing to do if that is the case. Otherwise it’s just speculation.

That makes a lot of sense, but it's not the way most tv's work. On Samsung brand you can definitely get the full readings on test patterns, and according to Fomo on LG panels, when they see you display test patterns they drive the panel even harder (if true, he might have been joking). I can't speak for the last two years of Sony (OLED's or LCD's) but prior to that I believe test patterns also used to reflect full performance. I never noticed the picture seeming much brighter in actual content. With the A90J practically every person who got their hands on a set is reporting that actual content is noticeably brighter than the test pattern measurements. It's the first time I hear of such a thing. So until told otherwise I'm inclined to believe this is the first year Sony actually follows the guideline above and displays modest test results while leaving headroom for specular highlights.

[NeilZ]

Also interestingly if 75% of max brightness by the avs forums link above holds true, then the A90J potentially reaches 1000 nits give or take some. This is in line with initial expectations for the panel + heatsink combo. It's really stunning that Sony would let their set be modest in test patterns, but it bodes well for actual performance, with possibly little to no tone mapping required on the A90J.

[samlop10]

Originally Posted by NeilZ

The A90J does use the new panel. It's obvious by looking at the energy ratings alone. Panasonic had a heat sink and used nothing close to those numbers. You don't get 20% more efficient out of thin air. Also no past LG oled could ever reach 1400 nits in vivid. Sony is not known to drive panels that hard either. There's no way this is last year's oled panel.

I believe one of the theories was that Sony is indeed using a new panel for the A90J, but not necessarily the evo one, as is also theorized that even the C1 models are using a new panel, but not the evo one. There’s no confirmation of either one yet.

I think what matters are the readings after calibration in different settings. The Sony can go beyond 1200 nits but hardly anyone that cares about PQ is going to use that mode given how it screws up other aspects of the PQ (colors more than anything else).

[samlop10]

Originally Posted by NeilZ

Also interestingly if 75% of max brightness by the avs forums link above holds true, then the A90J potentially reaches 1000 nits give or take some. This is in line with initial expectations for the panel + heatsink combo. It's really stunning that Sony would let their set be modest in test patterns, but it bodes well for actual performance, with possibly little to no tone mapping required on the A90J.

Wouldn’t that mean, given the 75%, that other OLEDs can also technically go beyond their measured peak brightness when measured using test patterns for calibration? And if so, how does that compare against this Sony?

And even if that’s the case for the Sony, we’ve yet to get those measurements using real-world content. For all we know it might barely pass 800 nits after calibration in accuracy-favoring modes with that real-world content, nowhere near 900+ or even 1000 nits. We just don’t know yet. It’d definitely be nice, but there’s no confirmation yet. Also important would be consistent results between different people measuring this with the same settings after calibration, while also having them provide their individual measurements of previous OLED models to have good points of reference to compare against given the reading variation from one calibration equipment to the next.

[Kingchin]

Originally Posted by NeilZ

The A90J does use the new panel. It's obvious by looking at the energy ratings alone. Panasonic had a heat sink and used nothing close to those numbers. You don't get 20% more efficient out of thin air. Also no past LG oled could ever reach 1400 nits in vivid. Sony is not known to drive panels that hard either. There's no way this is last year's oled panel.

You don't know for sure if the Sony A90J uses the EVO panel. You, me, D-Nice, Vincent Teoh etc can only speculate at this point.

All we know is the main Sony engineer of the Sony A90J definitely told Robert Zohn that it doesn't use the EVO panel.

Do some research and you will find LG Display have two new OLED panels for 2021. So who's to say it isn't the other 2021 OLED panel in the Sony A90J.

Eric Beeckmans who has both the Sony A90J and LG G1 will confirm for sure after the LG G1 embargo ends on March 17th.
PicsArt_03-12-02.14.48.jpg
PicsArt_03-12-12.28.01.jpg

[ray0414]

Originally Posted by Kingchin

Even experts get stuff wrong, D-Nice has previous history of stating facts that Vincent Teoh proved wrong

Here's the latest from a guy who has now received his LG G1 -
[ATTACH]]

The Sony A90J very well might use the EVO panel but we will find out very soon. As some people now have the LG G1 with the embargo ending next week. There will be way's to definitely find out if the LG G1 panel is different from the Sony A90J panel.

He since clarified. Sony didn't say anything.

[Kingchin]

Originally Posted by ray0414

He since clarified. Sony didn't say anything.

WTF

You obviously never watched Stop The FOMO livestream Sony A90J calibration video with Robert Zohn and John Reformato Robert Zohn confirmed the main Sony engineer of the Sony A90J told him that it doesn't use the EVO panel. He said it uses Sony's new processor and Sony's new custom heatsink to achieve the high brightness.

This is fact. But whether the main Sony engineer told him the truth we don't know.