Page 5 of 13

Re: Samsung QD OLED

Posted: 15 Feb 2022, 14:54
by Discorz
Price is surprisingly good.

Here are some testufo predictions of QD OLED.
Alienware AW3423DW, testufo motion test accurate estimate, sample-and-hold.png
Alienware AW3423DW, testufo motion test accurate estimate, sample-and-hold.png (44.93 KiB) Viewed 5057 times

Re: Samsung QD OLED

Posted: 16 Feb 2022, 00:30
by planart
Discorz wrote:
15 Feb 2022, 14:54
Price is surprisingly good.

Here are some testufo predictions of QD OLED.

Alienware AW3423DW, testufo motion test accurate estimate, sample-and-hold.png
What's that simulation based on? Judging by what I've personally seen on 120Hz woled and what's seen here on slow down video from Linus Tech Tips, it should prolly be even a bit more clear with trasistions.

https://www.youtube.com/watch?v=Piteu5gyKq0&t=530s

Re: Samsung QD OLED

Posted: 16 Feb 2022, 01:10
by DrR1pper
planart wrote:
16 Feb 2022, 00:30
Discorz wrote:
15 Feb 2022, 14:54
Price is surprisingly good.

Here are some testufo predictions of QD OLED.

Alienware AW3423DW, testufo motion test accurate estimate, sample-and-hold.png
What's that simulation based on? Judging by what I've personally seen on 120Hz woled and what's seen here on slow down video from Linus Tech Tips, it should prolly be even a bit more clear with trasistions.

https://www.youtube.com/watch?v=Piteu5gyKq0&t=530s
Was that 120hz woled in BFI mode? Because the simulation should be right with the blur coming from sample-and-hold. Slow motion capture of oled (e.g. the LTT video) will appear clear as if BFI is active but it's just because the recording is high framerate and then played back so slow that the moving object is now below the speed that induces motion blur by sample-and-hold on our screens.

https://blurbusters.com/faq/oled-motion-blur

Re: Samsung QD OLED

Posted: 16 Feb 2022, 02:46
by Discorz
planart wrote:
16 Feb 2022, 00:30
What's that simulation based on? Judging by what I've personally seen on 120Hz woled and what's seen here on slow down video from Linus Tech Tips, it should prolly be even a bit more clear with trasistions.

https://www.youtube.com/watch?v=Piteu5gyKq0&t=530s
I've seen LTT do ufo test correctly only couple of times. This oled one and all the other attempts are captured with static camera, which is incorrect. For sample-and-hold display it is important to pursuit the camera to simulate eye tracking. Even in strobed situations there are very few use cases where static capturing would work. So it's not recommended. This is a common issue people do.

No worries, simulation is quite accurate.
DrR1pper wrote:
16 Feb 2022, 01:10
Was that 120hz woled in BFI mode? Because the simulation should be right with the blur coming from sample-and-hold. Slow motion capture of oled (e.g. the LTT video) will appear clear as if BFI is active but it's just because the recording is high framerate and then played back so slow that the moving object is now below the speed that induces motion blur by sample-and-hold.

https://blurbusters.com/faq/oled-motion-blur
It's sample-and-hold. Fixed recording/gaze makes it appear as it's in BFI mode - stroboscopic effect.

We can also simulate BFI mode, but that will still be quite dark on these oleds.

Re: Samsung QD OLED

Posted: 16 Feb 2022, 04:48
by planart
Discorz wrote:
16 Feb 2022, 02:46
planart wrote:
16 Feb 2022, 00:30
What's that simulation based on? Judging by what I've personally seen on 120Hz woled and what's seen here on slow down video from Linus Tech Tips, it should prolly be even a bit more clear with trasistions.

https://www.youtube.com/watch?v=Piteu5gyKq0&t=530s
I've seen LTT do ufo test correctly only couple of times. This oled one and all the other attempts are captured with static camera, which is incorrect. For sample-and-hold display it is important to pursuit the camera to simulate eye tracking. Even in strobed situations there are very few use cases where static capturing would work. So it's not recommended. This is a common issue people do.

No worries, simulation is quite accurate.
Okay, fair enough. It's just that personally I feel that the simulated image looks somewhat similar to what for example I'm actually seeing in terms of blur in something like 160Hz fastIPS panel without VRB. Where I do believe QD-OLED should actually perform quite a lot better in flesh.

With this kind of image I feel it would be interesting to see more reference points on the side of this simulation to get a better idea what's it like in comparison to what we have available today. How it fares against your typical top contenders what's available on the market right now. Like 360Hz IPS variants or 240Hz popular models.

Re: Samsung QD OLED

Posted: 16 Feb 2022, 09:32
by Discorz
planart wrote:
16 Feb 2022, 04:48
Okay, fair enough. It's just that personally I feel that the simulated image looks somewhat similar to what for example I'm actually seeing in terms of blur in something like 160Hz fastIPS panel without VRB. Where I do believe QD-OLED should actually perform quite a lot better in flesh.
In real world its not that far off from fast IPS in terms of response times. On Internet in the other hand people tend to exaggerate things. But sensitivity to it also depends from person to person... Ghosting and blurring are not the same. Ghosting comes from slow
response times, blurring comes from slow refresh rate. In reality you see all these combined plus stroboscopic blurring and other. OLEDs only eliminate rtimes, while BFI reduces blurring. So removing response times is only one part of the journey to 1000+Hz.
planart wrote:
16 Feb 2022, 04:48
With this kind of image I feel it would be interesting to see more reference points on the side of this simulation to get a better idea what's it like in comparison to what we have available today. How it fares against your typical top contenders what's available on the market right now. Like 360Hz IPS variants or 240Hz popular models.
Here you go - UFO Motion Test Collection

Re: Samsung QD OLED

Posted: 16 Feb 2022, 10:04
by jorimt
planart wrote:
16 Feb 2022, 04:48
Okay, fair enough. It's just that personally I feel that the simulated image looks somewhat similar to what for example I'm actually seeing in terms of blur in something like 160Hz fastIPS panel without VRB. Where I do believe QD-OLED should actually perform quite a lot better in flesh.
To reinforce what @Discorz stated, OLED effectively eliminates GtG artifacts (smearing, ghosting, strobe crosstalk), it does not eliminate MPRT (motion picture response time), which is entirely dependent on max refresh + framerate due to the image persistence from sample-and-hold, regardless of pixel response time.

Due to the lack of GtG artifacts, strobed OLED can look closer to static examples, however, especially at slower pps.

Re: Samsung QD OLED

Posted: 17 Feb 2022, 00:12
by planart
Discorz wrote:
16 Feb 2022, 09:32
planart wrote:
16 Feb 2022, 04:48
With this kind of image I feel it would be interesting to see more reference points on the side of this simulation
Here you go - UFO Motion Test Collection
Thanks. Most of these seem to be actual photos or tests by many parties, not synthetic simulations like the image you posted no? This I fell always introduces some amount of human difference from workflows equipment etc.

In the end I'll just have to wait and see. Very often I feel it's hard to actually fully justify the visual effect in real life from these benchmark stopped motion photos / tracking videos. The experience in real life can be quite different than what you'd think from a photo.

I do remember seeing this test on a 120Hz 48" woled on a friends house and it was visually very clean and clear, super nice transistions, even if the movement was somewhat juddier @120Hz mainly even because the pixel response time is so instant. Compared to even 390Hz LCD I have here at my desk. I'm not sure if he had BFI on at that time.

Re: Samsung QD OLED

Posted: 17 Feb 2022, 01:05
by DrR1pper
planart wrote:
17 Feb 2022, 00:12
I'm not sure if he had BFI on at that time.
This would make a huge difference.

Re: Samsung QD OLED

Posted: 17 Feb 2022, 02:37
by Discorz
planart wrote:
17 Feb 2022, 00:12
Thanks. Most of these seem to be actual photos or tests by many parties, not synthetic simulations like the image you posted no? This I fell always introduces some amount of human difference from workflows equipment etc.

In the end I'll just have to wait and see. Very often I feel it's hard to actually fully justify the visual effect in real life from these benchmark stopped motion photos / tracking videos. The experience in real life can be quite different than what you'd think from a photo.

I do remember seeing this test on a 120Hz 48" woled on a friends house and it was visually very clean and clear, super nice transistions, even if the movement was somewhat juddier @120Hz mainly even because the pixel response time is so instant. Compared to even 390Hz LCD I have here at my desk. I'm not sure if he had BFI on at that time.
After you see bunch of them through time you learn to spot camera and human errors.

If u have already tried a few various monitors, panel types and generations you can get some idea of motion performance.

Either testufo was set to slower than 960pps speed, or Motion Pro (BFI) was enabled. If it was very dimm its bfi.