![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
I am pretty curious how the SEIKI pulls off 1080p scaling during 4K. If it's done upstream at the motherboard level, then life for 4K 120Hz could potentially become a lot simpler because the motherboard is apparently already outputting the full pixel data rate necessary for 4K 120Hz.
You will have to fill the pixels on the screen so it will be pixel doubling.
As long as at the panel level, it's all merged into a sequential scan. Among some display engineers, there are known motion artifacts that shows up when coarse multi-scanning directly at the panel level (multi-scan artifacts would look like stationary tear lines between independently-scanned sections of the panel, multiscanning would make www.testufo.com/#test=blurtrail&thickness=8 problematic with tearing that shows up between the independently scanned sections, because on a time-basis, the last pixel rows of a multiscanned section of display is severely time-offset from the first pixel-row of an adjacent multiscanned section, creating an effect that looks exactly like tearing, but the tearlines are stationary!)My mistake everyone, you indeed don't need to store 1 billion pixels, only 16.6 million (two frames), or 497664000 bits... thanks for pointing that out!
Still... I think my last paragraph has the more ideal solution.
Thanks this should be the one we need:
)It skipped half the frames(according to the photographs) so I would say 60 or close to 60 but that shall be fixed with a firmware update later on.
Well the Asus 39 inch screens board will be pin to pin comptaible as they use the same panel and if you look into it you can pretty much get newer non eDP contorllers hooked up to the panel (once HDMI 2.0 + new timing controllers come along)
Actually, the math is incorrect; for 88%, that's closer to 76% definitely telling them apart and 24% (12%+12%) doing a random choice because they couldn't tell them apart.
1 Gbit/s+ a pair would still get us 20 Gbps+ meaning a 100Hz refresh rate
Reduced timings allows you to make it as little as 5% overhead or so.
So the 1080p scaling is done at the panel level rather than at the motherboard level? There must be a motherboard level chip, and then a panel level electronics (sometimes a small chip embedded on a thin board in the edge of the panel, etc). Can anything be done further upstream (i.e. closer to the panel), to bypass this bottleneck? This would certainly raise the difficulty of achieving 120Hz, but since clearly all the pixels are refreshable at 120Hz on the panel glass, I'm curious which stage of the display internal pipeline this is occuring... (aka "how is 1080p@120Hz is actually possible on the 4K panel?" and "where is the scaling done?") Basically we'd want to find the stage of the 4K pipleine where the scaler is, and then figure out how to do 4K 120Hz at that point. It might be happening later in the pipleine than we expected, for example.
That stands as a house, I will PM Moddermike to see if he can explain if the 20 LVDS pairs can get 120Hz trough in any way.
Yes that is weird my first question was how does motion interpolation occur @ 120Hz since that would mean 30 real fps getting duplicated/mixed to make 120fps then those will be displayed @120Hz still meaning 120 frames every second going in the panel and being shown on the panel even though they are not real this means it should be possible unless the panel itself holds the motion interpolation electronics.
It is due to the crappy scaler used in the Seiki mine can do 1440P up to 80 or so Hz.Can this monitor do 2560x1440? If so, what's the max HZ?
I'm hoping it can at least do 2560x1440 at 60hz seeing how my ZR30W can do 2560x1600 over HDMI to DVI (Which is probably single link DVI overclocked?).
EDIT:
Doesn't look like it supports that resolution
Does SEIKI support motion interpolation of 4K to 120Hz? I thought it did not.
I don't know if the Seiki has motion interpolation the skyworth has and also at 4K (judging by the input lag)
Not on the Seiki but Chinese domestic monitors like the Skyworth have a different scaler even though the panel is the same.
I figured that they were the same but now you point me to it they have different screw hole locations.are you sure the panel is the same? I am starting to have my doubts:
- true the 8th minute of the following video shows that the Skyworth 50E780U 4K HDTV has Chimei Innolux panel
- but if you take a look what is really inside, then the panel looks different from behind:
- compare to the panel from seiki SE50UY04 (Chimenei Innolux V500DK1_LS1)
as shown here or on the last page here
SE50UY04 (Chimenei Innolux V500DK1_LS1)
Skyworth 50E780U ( ??? )
)I figured that they were the same but now you point me to it they have different screw hole locations.
Here you can see the revs
As you can see there are 2 revs and one version without backlight in other words only the LCD sheet and electronics. (maybe something for someone looking for making a lightboost panel
Seems like one has true 10 bit the other has fake 10 bit.
Also it seems that the panels indeed do have a integrated MEMC system.
Not a problem for the 39 inch model though as they only have 1 version of those.
