Baasha
Limp Gawd
- Joined
- Feb 23, 2014
- Messages
- 251
Nice! So the release is imminent although it looks like Euro/Asia already has it(?). Anyway, definitely will be getting this and perhaps using the 49" Neo G9 as an accessory display! lel
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Samsung Odyssey Neo G9 57" 7680x2160 super ultrawide (mini-LED)
- Thread starter kasakka
- Start date
Its in stock at retailers at least in Sweden. Guessing in part due to the price being a bit hefty for most.
Here in Finland it seems to be in pretty low quantity too. Two retail store chains that have it, one has messed up so that it can't even be ordered but reports "5+ in stock". The other store has already sold their whopping 4 monitors they had in store and reports 1-2 month delivery times.
Samsung's own website is still the best deal because you can buy with 0% interest payment for 12 months.
I'm hoping we will see some discounts in a few months for Black Friday. Hopefully the refresh rate issues etc are resolved by then.
In other news, I asked on Reddit if G95NC owners could try making custom resolutions like 5120x2160. As expected, the option is greyed out on Nvidia since the display uses DSC, but there may be another possibility for doing it via Custom Resolution Utility or registry hacking the Nv_Modes value which seemed to work on the previous G9 model even with DSC.
If those don't work out, the only way to use custom resolutions would be windowed mode and removing window styles from that so it acts like borderless window. This might come with an input lag penalty.
It really sucks that there is no 5120x2160 40" 120+ Hz ultrawide with similar capabilities as the Samsung.
Tried the CRU variant when I had the QN900B with DSX but to no avail. Could be different for the G95NC though.
Blade-Runner
Supreme [H]ardness
- Joined
- Feb 25, 2013
- Messages
- 4,471
Surely it wouldn't have been that hard to include 5120x2160 in the EDID?! Just baffling.
Welcome to monitor developers and their insanity. I mean 3440x1440 has been requested for the previous models but has Samsung bothered to add it? Of course not.
But good news! Someone has got 5120x2160 @ 120 Hz working on their 3080: https://www.reddit.com/r/ultrawidem...o_g9_g95nc_57_and_custom_resolutions/jyg8w7a/
It had to be done by the Nv_mode registry hack described here: https://www.reddit.com/r/ultrawidem...ung_odyssey_neo_g9_custom_resolution_support/
I've also recently discovered a quirk unique to Samsung. Their displays may switch to a completely different EDID depending on refresh rates and whether Adaptive Sync is enabled from the OSD. Why they do this, I don't know. Maybe some EDID size limitation or just because it was easier to manage in the display firmware.
As an example, if Adaptive Sync and 144 Hz is enabled I dump the EDID on my Samsung G70A 4K 144 Hz, I get a pretty barebones EDID that basically supports standard timings (crap resolutions for compatibility that nobody uses) and 4K 144 Hz. If I set the display to 120 Hz, I get a more full featured EDID that describes more stuff.
This sort of stuff may potentially be useful info if some EDID hacks are discovered in the future. I'd be interested to try if I can add a detailed timing for 5120x2160 that way for example.
As an example, if Adaptive Sync and 144 Hz is enabled I dump the EDID on my Samsung G70A 4K 144 Hz, I get a pretty barebones EDID that basically supports standard timings (crap resolutions for compatibility that nobody uses) and 4K 144 Hz. If I set the display to 120 Hz, I get a more full featured EDID that describes more stuff.
This sort of stuff may potentially be useful info if some EDID hacks are discovered in the future. I'd be interested to try if I can add a detailed timing for 5120x2160 that way for example.
Not sure if this monitor has built option to change screen ratios like the QN900B had, but that appeard as different monitors based on what selections you made in the gamebar (which is in itself expected). That also meant that you could have something like VRR activated for one of them but not the other and such.
As an eBay Associate, HardForum may earn from qualifying purchases.
View: https://www.youtube.com/watch?v=tFvhCbpslF8
This fast paced video from I think China shows the PbP and KVM features in action. The KVM actually does seem to work so that you can just switch monitor inputs and the USB devices will follow to the computer assigned to the input. The switching speed also seems a lot faster than many older Samsung monitors where it stays blank for less time.
G95NC so far based on YT videos and Reddit:
- 5120x2160 @ 120 hz is possible via Nv_modes registry hack. HDR seems to work too.
- I've asked a Reddit user to dump their EDID in a couple of ways (separate EDIDs for different modes) so I can open them up in an editor and take a closer look. Maybe just adding the relevant EDID data would make it possible to have 5120x2160 or even 6K x 2K as standard?
- 240 Hz refresh rates seem to be problematic unless running at 5120x1440. For some reason people have had issues activating 3840x2160 @ 240 Hz too even though it should work.
- Neither AMD or Nvidia GPUs can do 8K x 2K @ 240 Hz over HDMI 2.1. We will see if e.g next AMD driver will fix this.
- A few failures and issues have cropped up already. Poor QC or shipping damage? Hard to say.
- One person had a dark patch on the screen where the backlight doesn't work.
- Another had weird electrical noises and flickering on the left side of the screen.
- One person complained about washed out colors, but seems their issue was solved by turning off eco mode and adaptive brightness. Why do those work in HDR mode anyway?
- For Macs, M2 Air was limited to 5120x1440 @ 120 Hz. M2 Pro/Max should have higher capabilities for displays, so maybe they can run the screen better, but so far no reports for these machines. PbP mode does allow for full res at least, so that's an alternative if you can spare two USB-C ports.
- KVM seems to work fine as described in the above video.
PS. The sheer amount of bullshit videos on YT where people are literally talking about this display based on nothing but the product website is ridiculous. It's like one step above those AI voice reading news articles bollocks.
elvn
Supreme [H]ardness
- Joined
- May 5, 2006
- Messages
- 5,377
Just wait for version 2 of this like the odyssey ark 2. /s 
Hopefully the wrinkles get ironed out in this model though. Curious what TCL's version will be like also, especially since it's likely to have dolby vision support.
Hopefully the wrinkles get ironed out in this model though. Curious what TCL's version will be like also, especially since it's likely to have dolby vision support.
Just wait for version 2 of this like the odyssey ark 2. /s
Hopefully the wrinkles get ironed out in this model though. Curious what TCL's version will be like also, especially since it's likely to have dolby vision support.
TCL... ewwww.
elvn
Supreme [H]ardness
- Joined
- May 5, 2006
- Messages
- 5,377
They bought samsung's LCD plant in china a few years ago and I believe they provided some panels to samsung in the past as well. Considering the complaints with samsung they might not be any worse. A flagship TCL from a samsung plant is probably not comparable to a $300 roku tv
In fact, afaik this is the first ever TCL monitor rather than tv so it could be new ground here.
From 2020:
https://www.gsmarena.com/samsung_sells_its_lcd_plant_in_china_to_tcl_for_18_billion-news-45004.php
Really wish she brushed her teeth.
The only review so far I have seen measure overshoot was a South Korean one and those overshoot graphs seem to show the Extreme and Faster modes but even the review states people should use the Default mode which should have next to no overshoot issues.
This is pretty typical of most monitors Samsung releases - the default mode is fine and the Faster/Extreme modes are there to let them advertise some very fast response time but at a lot of overshoot. I wish they had more granularity in this regard.
Both the Odyssey G9 and Neo G9 lock you out of the overdrive settings when VRR is enabled and they all used the Faster mode preset. I expect this one will also be the same. If you look at the website of the Korean review, it states Faster is the default mode which makes sense as the Neo G9 and G9 use Faster when VRR is on
The Samsungs I've had have defaulted to the Standard setting on VRR so it's not necessarily consistent. We will know more when other sites review the display.
Zurginator
Gawd
- Joined
- Nov 16, 2009
- Messages
- 864
I did a rough 3D model based on the dimensions and product images Samsung has put out. This sucker is so much larger than I expected. Desk shown is 30" deep and 6' wide.
I'm no longer sure this is the right monitor for me. The amount of space behind the monitor is insane.
I'm no longer sure this is the right monitor for me. The amount of space behind the monitor is insane.
Yeah I've never liked the way these encroach on your shoulders and completely dominate a desk making all the space under/behind useless. I wish it were curved far less aggressively but then we'd deal with really bad gamma loss due to the poor viewing angle of its VA panel.
Aesthetically it just looks bad on a desk.
Aesthetically it just looks bad on a desk.
Yes, and we need to remember that those poor viewing angles is the real reason for the curve. I believe even Samsung themselves has stopped with the "Curved like the eye" nonsense now. Of course, some people might find the aggressive curve pleasing for various reasons, but for me it is mainly a tradeoff to combat poor viewing angles.
I don't agree with that. I had the 1800R curve CRG9 and always wished it was more curved. I've tried the 49" Neo G9 and it felt nicer to use to me.
So to me it's a definite design feature rather than just to accommodate the VA panel.
It would be cool if they would make it adjustable curve like the LGs when Samsung eventually makes an OLED version of this.
elvn
Supreme [H]ardness
- Joined
- May 5, 2006
- Messages
- 5,377
I think the problem is the curvature is wrong vs the seating distance on practically all uw screens. The radius of the curve is to the center point. For 1000R it is a 1000mm radius. That would be around a 40" view distance screen surface to eyeballs. That wouldn't work for this screen b/c it would shrink the height to your perspective and it would end up looking like a short belt. People would likely keep this screen viewed at a distance like a single 32" 4k would be in the center, but with the longer sides added like wings. A 750R, 750mm curve would be around 30" radius/center point so would be a lot closer to optimal, especially for screens mounted onto and/or sized for a desk.
When sitting at the center point of a curve, all of the pixels are equidistant from your eyes and all of the pixels are on axis, pointing directly at you. When you sit closer than that the center point is pushed well behind you. Then the pixels are off axis more and more the further away from the center of the screen they are, almost like a gradient. This will cause distortion like a fun house mirror. It also exacerbates uniformity issues the farther toward the ends.
The curved black line is what sitting at the 1000R/40inch radius of a screen would be with the solid blue viewing angle being your human central viewing angle of 60 to 50 degrees.
The red line here would be a very long uw screen that provided immersion into the periphery beyond your 60 to 50 degree central viewing angle.
The transparent blue viewing angle is where people sit closer, distorting the curve and pushing the center point behind them.
. . . . . .
The game devs can warp the game to compensate, but really the screens should be designed for viewing at the center point imo. But it is what it is for now. The only screen big enough to view from 1000R, 1000mm, 40 inches without turning into a short belt to your perspective would be the 55inch 4k ark if it was decoupled from the desk and set back that far. It gets around 62ppd at that distance which is ok for 4k but not stellar, and it wouldn't have multiple 4k worth of desktop real estate. Also wouldn't be wider into your periphery for immersion at that view distance.
I really hope that they would eventually design curves to the target view distance and make the screens wider into the periphery as necessary but I don't see that happening any time soon. There was that curvable, varying curve screen that could do 750R though which is about 30 inch center point but I didnt like the rest of the specs. 750R, 30 inch center point would be a lot better layout wise, especially for desktop sized screens imo.
Some games have warping of the game output to compensate, but it's not standard by any means. Nvidia could probably come up with a tool for that if they wanted to as well but afaik there isn't any. Imo it would be better if the screen design didn't have you sitting where the curve would warp(and also exacerbate uniformity issues)in the first place.
What I just wrote in this reply is in regard to appreciably curved screens. I agree with you about 1800R. A 1800R screen is a semicircle of a 71 inch radius circle, (142 inch diameter). That's almost a 6 foot radius or center point, a very small semicircle segment of a much larger ~ 12' diameter circle. A 1800R screen isn't really curved much at all.
. . . . . . . . .
. . . . . . . . .
. . . . . . . . . .
Think of the pixels on the screen like small laser pointers. In a room with a fog machine you'd see the shafts of laser light. When sitting at the center point of the curve, all of the lasers would be on axis to you and pointed directly at you so that for the most part you'd be seeing the points of light. The nearer you sat than that, the more you'd see the shafts of the light beams more sidelong. From your nearer position, the farther the pixels were from center of the screen, the more of the side of the laser beams you'd see. In a graduated fashion the pixels would be more and more off axis the farther they were away from the center and towards the outer ends of the screen.
This will make the screen distorted. Practically all uw and super ultrawides are designed lacking an aggressive enough curve and/or long enough semi-circle segment screen length to be able to sit at the center point of their curve without making the screen look short and belt like. (Outside of maybe the adjustable curve model monitor that could do up to 750R ~> 30" center point but I didn't like the overall specs of that screen). So practically everyone is sitting with the center point way behind them with current curved screens.
The ark is a big 16:9 so it is actually tall enough where you could mount it on a floor tv stand and get enough distance to be near the center point with the screen not being shrunk to a narrow belt to your perspective. It's rez is too low for it's size imo though. Perhaps someyear we'll get an 8k available in that format for higher PPD and quads of something near 4k real-estate (maybe a little less if scaled slightly for clarity/visibility).
Theoretically, they could design an uw or s-uw screen better so that you still get immersion on the sides when sitting at the center point of a curve. For example, the red line being a 120degree arc of a semicircle in the image below. The aggression of the curve/center point would have to appropriate in order to provide enough height to the screen though whatever the screen's actual height dimension might be.
When sitting at the center point of a curve, all of the pixels are equidistant from your eyes and all of the pixels are on axis, pointing directly at you. When you sit closer than that the center point is pushed well behind you. Then the pixels are off axis more and more the further away from the center of the screen they are, almost like a gradient. This will cause distortion like a fun house mirror. It also exacerbates uniformity issues the farther toward the ends.
The curved black line is what sitting at the 1000R/40inch radius of a screen would be with the solid blue viewing angle being your human central viewing angle of 60 to 50 degrees.
The red line here would be a very long uw screen that provided immersion into the periphery beyond your 60 to 50 degree central viewing angle.
The transparent blue viewing angle is where people sit closer, distorting the curve and pushing the center point behind them.
. . . . . .
The game devs can warp the game to compensate, but really the screens should be designed for viewing at the center point imo. But it is what it is for now. The only screen big enough to view from 1000R, 1000mm, 40 inches without turning into a short belt to your perspective would be the 55inch 4k ark if it was decoupled from the desk and set back that far. It gets around 62ppd at that distance which is ok for 4k but not stellar, and it wouldn't have multiple 4k worth of desktop real estate. Also wouldn't be wider into your periphery for immersion at that view distance.
I really hope that they would eventually design curves to the target view distance and make the screens wider into the periphery as necessary but I don't see that happening any time soon. There was that curvable, varying curve screen that could do 750R though which is about 30 inch center point but I didnt like the rest of the specs. 750R, 30 inch center point would be a lot better layout wise, especially for desktop sized screens imo.
Some games have warping of the game output to compensate, but it's not standard by any means. Nvidia could probably come up with a tool for that if they wanted to as well but afaik there isn't any. Imo it would be better if the screen design didn't have you sitting where the curve would warp(and also exacerbate uniformity issues)in the first place.
What I just wrote in this reply is in regard to appreciably curved screens. I agree with you about 1800R. A 1800R screen is a semicircle of a 71 inch radius circle, (142 inch diameter). That's almost a 6 foot radius or center point, a very small semicircle segment of a much larger ~ 12' diameter circle. A 1800R screen isn't really curved much at all.
. . . . . . . . .
. . . . . . . . .
. . . . . . . . . .
Think of the pixels on the screen like small laser pointers. In a room with a fog machine you'd see the shafts of laser light. When sitting at the center point of the curve, all of the lasers would be on axis to you and pointed directly at you so that for the most part you'd be seeing the points of light. The nearer you sat than that, the more you'd see the shafts of the light beams more sidelong. From your nearer position, the farther the pixels were from center of the screen, the more of the side of the laser beams you'd see. In a graduated fashion the pixels would be more and more off axis the farther they were away from the center and towards the outer ends of the screen.
This will make the screen distorted. Practically all uw and super ultrawides are designed lacking an aggressive enough curve and/or long enough semi-circle segment screen length to be able to sit at the center point of their curve without making the screen look short and belt like. (Outside of maybe the adjustable curve model monitor that could do up to 750R ~> 30" center point but I didn't like the overall specs of that screen). So practically everyone is sitting with the center point way behind them with current curved screens.
The ark is a big 16:9 so it is actually tall enough where you could mount it on a floor tv stand and get enough distance to be near the center point with the screen not being shrunk to a narrow belt to your perspective. It's rez is too low for it's size imo though. Perhaps someyear we'll get an 8k available in that format for higher PPD and quads of something near 4k real-estate (maybe a little less if scaled slightly for clarity/visibility).
Theoretically, they could design an uw or s-uw screen better so that you still get immersion on the sides when sitting at the center point of a curve. For example, the red line being a 120degree arc of a semicircle in the image below. The aggression of the curve/center point would have to appropriate in order to provide enough height to the screen though whatever the screen's actual height dimension might be.
Last edited:
elvn
Supreme [H]ardness
- Joined
- May 5, 2006
- Messages
- 5,377
This is 800R .. xenon flex (videocarz.com review)
That equates to a ~ 31.5 inch radius or centerl point. Near the point where all points on the screen will be equidistant from your eyes/head (your head in the center of the circle/curvature ~ at the center point) . . depending where you are sitting.
That equates to a ~ 31.5 inch radius or centerl point. Near the point where all points on the screen will be equidistant from your eyes/head (your head in the center of the circle/curvature ~ at the center point) . . depending where you are sitting.
Last edited:
Baasha
Limp Gawd
- Joined
- Feb 23, 2014
- Messages
- 251
What did you get for your PG32UQX? I'm thinking of using this Samsung 57" Neo G9 (and my 49" Neo G9) on the UberRig while keeping the PG32UQX + Aorus FIU32 on the Z690 Rig. Need to get that Adtec stand tho...
$1100. Hard to sell now days given the age and the fact that Amazon warehouse had tons of them for the same price.
Meanwhile the display is still $3775 at its cheapest here in Finland.
Anyway, back to the G95NC. Someone on Reddit confirmed they can get 8kx2K @ 120 hz over HDMI 2.1 on a M2 Max Macbook Pro 16". So that's nice, makes it a very viable work option for me.
For those of you not aware of Apple's shenanigans, the M2 Macbook Air, M2 Pro Macbook Pro and M2 Max Macbook Pro all have different display support capabilities, with the M2 Max obviously being the best at this. Apple's specs are vague and basically refer only to their own monitors so you literally have to try a specific display to know if it works well on Macs. Rtings does report this in their reviews thankfully.
I'm not here to tell you if you should like the curve or not, but for many years, it seems like Samsungs desire to curve screens has been directly related to it's viewing angels. I would personally much rather have something flat providing that it had good viewing angles like OLED. One big reason for this is that the massive curve for productivity would mean you have to turn your head more to see the edges, unless you sit further away, in which case the curve itself starts to get a bit pointless. For gaming/immersion though, I can absolutely see a value in the curve and being surrounded by the image.
Based on using the LG CX 48" as a monitor for a few years, my primary issue with it was the lack of a curve. Even with near perfect viewing angles it meant that the display needs to be a particular distance away to be comfortable to use because it does not curve towards you so looking at the corners becomes uncomfortable. I felt I had to move my head around too much compared to the Samsung CRG9 (1800R curve but similar width to the LG) I also had at the time. Due to the LG's height I mostly used it closer to a "tall ultrawide", utilizing the bottom 2/3rds of its screen in desktop use.
So I'm totally onboard with devices like the Samsung ARK (though it could be much smaller) or the G95NC. The wider the display the more sense the curve makes.
I believe we need to separate the discussion of "a slight curve for a 16:9 or 21:9 monitor" from the "an aggressive curve for a 32:9 monitor". And also the reason why we like/dislike each. For me, an aggressive curve on a really wide monitor is not preferred as it would both basically take up my entire desk due to it's depth and also that it would force much more head turning to see the edges of the screen compared to a similar flat one.Unless you move further away of course, but then the idea of the aggressive curve is kind of lost. But what's right for me might not be for someone else and vice versa. I should add that I do perhaps 90% work and 10% gaming, so my needs are based on that. For something like a simracing, the G95NC would probably be the ideal monitor as well as the previous G9s.
Let's all, including me, be careful not to turn this thread into yet another LCD vs OLED one
elvn
Supreme [H]ardness
- Joined
- May 5, 2006
- Messages
- 5,377
Curvature vs screen dimensions and how well that applies to different usage scenarios is not off topic to this thread at all though, especially considering the specs and size of this screen.
from
https://www.calculator.net/circle-c...meter=&circumference=&area=&ctype=1&x=57&y=21
Ultrawide and super ultrawides so far have all been a very minor arc of a circle. I don't think any are even 120deg.
The R value is the radius in mm.
The radius is the distance to the center of the circle obviously. So if you are sitting at the center point, all of the pixels would be pointed directly at you and for the most part none would be off-axis horizontally.
When sitting closer, the farther away from the center the pixels are, the more and more off axis they become, almost like a gradient. In that scenario, the pixel's vectors are off-axis/tangential to a growing degree the farther away from center of the screen.. That causes distortion and geometry issues (like a fun house mirror). It will also exacerbate screen uniformity issues (especially on certain panel types), most noticeably on solid fields of bright color - including documents and app backgrounds among other things.
.
Optimal position is the leftmost dot in the animated gif below, where all of the points on the surface are pointed directly at you (on axis). Sitting at the nearer dot will have all of the pixels the farther from the center of the screen becoming more and more off-axis as they will still be pointing at the leftmost dot (the center of the circle in relation to the arc of the curve).
1000R = 1000mm = around 40". Optimally, the closer you sit, the smaller the circle should be. At a desk, that would probably require a radius of 30 inches or less for a screen mounted directly at a desk, (desktop mounted screens which also are relatively short physical height screens besides).
. . . . . .
This is about light reflecting but the angles the pixels are lighting from, when off-axis, can have a similar warping effect to some degree. This is different than the fact that some games also stretch (and can squash) game viewports like a rubber sheet to fit ultrawide aspects. You can tell that it's not just a game stretch/squash problem because the issue also presents itself on regular desktop stuff where for example the geometry (e.g. straight lines, circles, etc) are distorted.
Ultrawide and super-ultrawide curved screens are designed lacking an aggressive enough curve and/or long enough semi-circle segment screen length to be able to sit at the center point of their curve without making the screen look short and belt like. (Outside of maybe the adjustable curve model monitor that could do up to 750R ~> 30" center point or the 800R ~> 31" screens). So practically everyone is sitting with the center point way behind them with current curved screens.
A 1800R screen is a semicircle of a 71 inch radius circle, (142 inch diameter). That's almost a 6 foot radius or center point, a very small semicircle segment / very minor arc of a much larger ~ 12' diameter circle. A 1800R screen isn't really curved much at all. So I feel like people who like 1800R and less curvature are going in the opposite direction from the existing problem and trying to eliminate curvature in the first place for the most part, instead of making an equidistant curve with all of the pixels on axis.
from
https://www.calculator.net/circle-c...meter=&circumference=&area=&ctype=1&x=57&y=21
Ultrawide and super ultrawides so far have all been a very minor arc of a circle. I don't think any are even 120deg.
The R value is the radius in mm.
The radius is the distance to the center of the circle obviously. So if you are sitting at the center point, all of the pixels would be pointed directly at you and for the most part none would be off-axis horizontally.
When sitting closer, the farther away from the center the pixels are, the more and more off axis they become, almost like a gradient. In that scenario, the pixel's vectors are off-axis/tangential to a growing degree the farther away from center of the screen.. That causes distortion and geometry issues (like a fun house mirror). It will also exacerbate screen uniformity issues (especially on certain panel types), most noticeably on solid fields of bright color - including documents and app backgrounds among other things.
.
Optimal position is the leftmost dot in the animated gif below, where all of the points on the surface are pointed directly at you (on axis). Sitting at the nearer dot will have all of the pixels the farther from the center of the screen becoming more and more off-axis as they will still be pointing at the leftmost dot (the center of the circle in relation to the arc of the curve).
1000R = 1000mm = around 40". Optimally, the closer you sit, the smaller the circle should be. At a desk, that would probably require a radius of 30 inches or less for a screen mounted directly at a desk, (desktop mounted screens which also are relatively short physical height screens besides).
. . . . . .
This is about light reflecting but the angles the pixels are lighting from, when off-axis, can have a similar warping effect to some degree. This is different than the fact that some games also stretch (and can squash) game viewports like a rubber sheet to fit ultrawide aspects. You can tell that it's not just a game stretch/squash problem because the issue also presents itself on regular desktop stuff where for example the geometry (e.g. straight lines, circles, etc) are distorted.
. . . . . . .https://www.scientificamerican.com/article/distorted-images/
" If the incident ray strikes the reflective surface at a particular angle, the reflected ray leaves the surface at the same angle—but is located at the other side of the normal. In other words, the incident and reflected ray make a perfectly symmetrical V shape, with the normal as the line of symmetry.". <----- Obviously, pixel lit displays aren't reflecting the image but they will have the vector of your line of sight vs the axis of the pixels on the screen. When the pixels are on axis, that angle is zero. When the pixels are off axis from sitting nearer than the center point of the curve, , that angle grows larger and larger the farther away from the center of the screen the pixels are.
When you are sitting far inside of the center point of a curved screen up close, light coming out of the screen farther from center is as if you bounced light or a tiny ball off of the screen and it reflected off at a tangent, off-axis and farther away from you the farther away from the center of the screen the pixels are. If you were sitting at the center of the circle (semi-circle segment that the screen is), any time you shot a rubber bead at the screen, theoretically in perfect conditions the surface would bounce it right back at you along the same vector in reverse (like in the animated gif above).
I think some games have warping of the game output to compensate but it's not standard by any means. Nvidia could probably come up with a tool for that if they wanted to as well but afaik there isn't any. A lot of people seem to change their FOV in an attempt to compensate for games that have FoV adjustment provided in the game settings or if you can force it with edits/mods - but you should be able to have other FoVs without warping if all of the angles vs.curvature were optimal. Imo it would be better if the screen design didn't have you sitting where the curve would warp(and also exacerbate uniformity issues)in the first place. You can tell that it's not just a game stretch/squash problem because the issue also presents itself on regular desktop stuff where for example the geometry (e.g. straight lines, circles, etc) are compromised.
Ultrawide and super-ultrawide curved screens are designed lacking an aggressive enough curve and/or long enough semi-circle segment screen length to be able to sit at the center point of their curve without making the screen look short and belt like. (Outside of maybe the adjustable curve model monitor that could do up to 750R ~> 30" center point or the 800R ~> 31" screens). So practically everyone is sitting with the center point way behind them with current curved screens.
A 1800R screen is a semicircle of a 71 inch radius circle, (142 inch diameter). That's almost a 6 foot radius or center point, a very small semicircle segment / very minor arc of a much larger ~ 12' diameter circle. A 1800R screen isn't really curved much at all. So I feel like people who like 1800R and less curvature are going in the opposite direction from the existing problem and trying to eliminate curvature in the first place for the most part, instead of making an equidistant curve with all of the pixels on axis.
Theoretically, they could design an uw or s-uw screen better so that you still get immersion on the sides when sitting at the center point of a curve. For example, the red line being a 120degree arc of a semicircle in the image below. The aggression of the curve and the centerpoint would have to appropriate in order to provide enough height to the screen though whatever the screen's actual height dimension might be.
Last edited:
For me it's probably 70-80% work, 20-30% gaming. I really enjoyed using the Samsung CRG9 49" 5120x1440 for work, but just want sharper text/UI because MacOS so heavily favors scaling for text rendering. So the G95NC is the next best thing.
Dual 27-32" monitors often don't work for me because you either put the divide between them right front and center and sit lopsided, or you have one display centered in front of you and another to the side. For window arrangement I prefer horizontally side by side instead of grid type layouts. I am also not a fan of 3+ display setups as it gets kinda overwhelming. For some reason a single wide display is easier for me to manage.
I generally prefer my displays fairly close so for me the curve worked quite nicely for both immersive gaming and work. My current 28" 4K display is at slightly above arm's length away and I have pretty long arms. A 32", or two of them like on the G95NC, would go further away.
With the LG CX 48", the display was at 1m viewing distance and that was ok, but I still often didn't use its full vertical space and for e.g fullscreen videos and gaming felt I wanted to back away a bit. Meanwhile the RGBW pixel structure and low-ish PPI meant that scaling was necessary for decent text rendering while extra viewing distance needed to make it look sharper.
It's always a battle between wanting high quality text rendering for desktop use vs gaming where I don't necessarily need above 4K resolutions.
elvn
Supreme [H]ardness
- Joined
- May 5, 2006
- Messages
- 5,377
I can understand where improvise is coming from, where he said he'd prefer 1800R. I can understand it because of what we have available right now. I've been saying that a more aggressive curve that puts you at the centerpoint of a circle that the minor arc of the screen would be a part of is the optimal way for an ultrawide to be designed, where all of the pixels would be on axis pointer directly at you - but that is not what we have available now in most curved screens. They are practically all designed where the viewer is sitting much closer to the screen than the center point. The 750 - 800R curved screens are the only ones that are anywhere near where your eyes would be sitting at a desk with the centerpoint of their minor arc/circle being at 30" to 31" away from the screen surface, and sitting farther away to sit at the centerpoint on lower curvature, longer R screens is not viable because the screens then turn into short height belts to your perspective.
What we have now imo is not optimal and the way I understand it is I'm complaining that I'd prefer more optimal curvatures vs seating distances, while I believe improvise is saying 1800R is better having very little curvature at all considering what we actually have available overall on most high end uw and s-uw screens rather than a wish list of better designs.
What we have now imo is not optimal and the way I understand it is I'm complaining that I'd prefer more optimal curvatures vs seating distances, while I believe improvise is saying 1800R is better having very little curvature at all considering what we actually have available overall on most high end uw and s-uw screens rather than a wish list of better designs.
Last edited:
elvn
Supreme [H]ardness
- Joined
- May 5, 2006
- Messages
- 5,377
From redd. Perspective/lens seems to be exaggerating things though.
49 g9 vs 57 G95NC
View: https://i.imgur.com/sektRhg.jpeg
49 g9 vs 57 G95NC
View: https://i.imgur.com/sektRhg.jpeg
Someone on Reddit claimed they got 240 Hz over HDMI 2.1 with a Radeon 7900 XTX. So this now becomes purely an Nvidia issue and a problem on older AMD cards it seems. Hopefully it can be fixed on both via driver updates. Otherwise Nvidia deserves a lot of flack for providing an outdated port selection on their flagship GPU.
Well, can't run jedi survivor @ native res.
Lets me chose the 7680x2160 ingame but just crashes before it activates.
Good start.
HDR does not seem as popping as my LG C242 initially.
But i only tested for a few mins so far.
Test more on the weekend.
Lets me chose the 7680x2160 ingame but just crashes before it activates.
Good start.
HDR does not seem as popping as my LG C242 initially.
But i only tested for a few mins so far.
Test more on the weekend.