I don’t do much dos emulation so I might be wrong but from my tests @1080p to get things to stay “tidy” and having a good screen size is better to use integer overscale:
This is set to VGA in the install and in the core options and VGA double scan in the shaders parameters:
Integer overscale on
Integer scale off
As I said I don’t do much dos emulation lately so I might be missing some settings or knowledge.
Also in the example above I used mask 7 but I think mask 10 can be similar.
Yes, it seems you need at least 6x scale for VGA doublescan to work correctly. Also, for 1080p, arguably mask 0,5, or 7 transformed to shadow mask are better choices than higher pixel masks, unless you’re emulating an untypical high pitch display.
Many pc monitors from the 90s on also had aperture grille masks, so you might as well just leave it at that if you prefer it.
If you use “VGA” in dosbox, you don’t need double-scan in the shader. It’s already double-scanned by dosbox. If you use “SVGA” in dosbox, then you need double-scan in the shader for games that are 240p or 200p. For 640p or 600p games, you never want double scan in the shader regardless of VGA or SVGA.
That’s a reasonable concern, new subpixel layouts are really surreal.
I’m sticking to IPS atm, can’t really comment on what works with new arrangements.
Post some screenshots. If you have a different display with a different layout, you’ll have to have different settings.
Take a read here and see how far things have come with shaders on WRGB panels.
This is also a great opportunity to try out my shader preset packs which were primarily designed on a WRGB OLED and see what all the fuss is about and pickup a few tips along the way as well.
In my experience (27" 2560x1440 RWBG) all this does is make it look wrong, with weird spacing of the mask and slots and incorrect phosphor layout. Really not a solution and shouldn’t be presented as such, maybe just a “try it to see if you hate it less?”
I’m really [very impatiently] waiting for the “Gen 3.5” and it’s killing me how long they drag this incremental shit out. https://tftcentral.co.uk/wp-content/uploads/2024/03/future_panels.jpg I bet they’ll even find a way to justify not including a good BFI or rolling refresh and G-Sync Pulsar, the bastards.
Please elaborate as you’re being a bit vague. Perhaps some photos of what you’re describing might bring a little more light to what you’re referring to.
Do note that different resolutions may require different Mask and TVL settings as well. That is a given so don’t expect Mask 12 to look good at 1440p just because it does at 4K, instead you can try Mask 10 at 1440p if you want to maintain a similar TVL/look. The height of the Slot might also have to be adjusted.
Also, with these things beauty is in the eye of the beholder. You seem to have a pet peeve with anything other than RGB in that exact order, please correct me if I’m wrong? In that case, then many other panel/display types would look very “wrong” as well.
Unless you can prove me wrong, I stand by my advice to @RealNC. No need to come across so aggressively in your criticism as if to check me and stop me in my tracks. That’s the wrong approach in my opinion as not everyone shares your views on the matter and I’m coming from a positive place of trying to assist. So lose a bit of the 'tude and take it down a bit.
It’s not like this is the first time I’ve heard something like this and please correct me if I’m wrong, did it come from you before?
So instead of trying to lambaste my suggestion, why not come up with a solution? Why not pioneer something, make something new, uncover something that wasn’t noticed before. At the end of the day, we use what display technology we have and we try to enjoy it as best as we could…at least some of us do.
As far as I know, @RealNC did not state the resolution of their display.
If it is in fact 1440p then that’s good as we can now try to see what can be gotten out of it. I just thought that they were unhappy with the way their presets looked and probably didn’t know or realize yet that WRGB subpixels align perfectly with Mask Layout 1 and not 0.
Yeah, it is you again. Get over yourself man. This is the thread for CRT-Guest-Advanced. This shader includes a feature called Mask Layout. It’s there for a reason. Why don’t you beg Guest to remove it while you’re at it because it’s not the correct, pure, pristine phosphor order?
Before it was learned that Mask Layout 1 aligned so well with WRGB OLED panels, folks were recommending B&W Masks for those displays and lamenting how horrible OLED was for “accurate” CRT Shader emulation. Now look at where we are? Even in die hard CRT circles, people are extolling the benefits of OLED displays as a possible alternative to their CRTs.
Do you remember how it was before Sony Megatron Color Video Monitor got that RWBG (OLED) layout? So I’m glad people like myself and others got to get things looking a lot better and a lot more like how we wanted things to look as opposed to things remaining how they were. It’s called progress and it’s an improvement. It’s not 100% perfect or identical but not everyone thinks it needs to be.
Please don’t get upset and unload a wall of text on me just because I disagree with your suggestion.
The solution is to just use the regular layout and accept that it’s not quite as good as it would be on a display with a regular structure. Or for the people developing the shaders to get access to multiple displays with which to develop the shaders on.
Presumably for those LCDs that for whatever reason have a rotated panel. Certainly not for RWBG OLEDs, as that would require its own layout given the issues that can be seen on both of the current ones.
There are obviously issues with both, but to my eye and at various distances the regular RGB one ends up looking more normal/usable than the BGR. And I’m not going to go check now but I remember noticing the same thing on my 55" C9 with the same layout at 4K as well. But either way, having the CRT phosphors laid out as RBG throws off the color dynamics on a lot of sprites/textures and makes the presentation inauthentic to the point where you might not want to be using a CRT shader anyway.
Maybe it would require some kind of different subpixel processing to achieve a better spacing? I don’t really know, I’m sure it could be improved if the people who knew what they were doing had these displays and a macro lens to play around with. But I’m really not too concerned, it looks fine as it is for most uses. Plus I’m sure it’s better on the newer Gen 2.5 RGWB panels, and the Gen 3.5 RGB ones should be coming out within a couple of years and then it wouldn’t matter anyway.
But hey, thanks for getting me to waste my afternoon.
Proceeds to unload a wall of text just because I disagreed with the way you disagreed with my suggestion.
This is funny. Hey, I’m all about constructive discussion. I just didn’t appreciate your approach and I said what I had to say and made my point and addressed a number of things in the process.
I don’t know what some people’s issue is with “too many words/walls of text” as if it is some crime or sin. I roll my own way. So it doesn’t matter.
Listen, I’m really glad that I actually got you to finally contribute to finding a solution to this seemingly long standing issue. Posting those subpixel shots is very useful. This is exactly where things started at 4K resolution before the correct mask layout was stumbled upon possibly by accident.
I don’t have a 1440p OLED display to test these with and I probably could have just ran my 4K one at 1440p without scaling if it was working.
All hope is not lost though. I remember when @guest.r made a change to the shader to better align the BGR Mask Layout and it completely broke the subpixel alignment at 4K on my OLED TV and he was kind enough to quickly revert this change.
This makes me think that it’s suitability for WRGB OLED (at least at 4K) was possibly coincidental.
The starting subpixel is also a factor by the way and Guest-CRT-Advanced’s BGR is actually RBG so it’s not equivalent to Sony Megatron Color Video Monitor’s BGR.
Assuming 4K WRGB OLED Panel and a 1440p OLED Panel are identical at the subpixel level, I think it might be possible to have improved alignment.
Subpixel layout looks exactly the same as any other WRGB OLED Panel to me.
One thing I just remembered @DevonCM did you know that if you took a 4K WOLED display and use CRT-Guest-Advanced’s Mask Layout 1 (or Sony Megatron Color Video Monitor’s RWBG (OLED) Subpixel Layout and you rotated your display 180° then used Windows Display settings to correctly orient your desktop, you could have perfect RGB “Phosphor” order on an OLED!
Now I don’t know what’s going on at 1440p but maybe with enough interest the same can probably be done at that resolution.
RTINGS even recommends this method as a way to get improved text clarity.
I’m not sure if anyone has tried to figure out why Guest’s Mask Layout 1 works but we do know that since the White Subpixel is always Off if we’re using Full Strength Masks with no additional bloom or anything, that creates a natural wider gap between the Red Subpixel and the Blue Subpixel. So you’ll have Red Gap (left by white subpixel) Blue Green or depending on which subpixel you start on that can be Blue Green Red Gap (left by white subpixel). Rotate that 180° and it becomes Gap (left by white subpixel) Red Green Blue Gap (left by white subpixel) or depending on where you start Red Green Blue Gap (left by white subpixel).
This is something interesting @hunterk said regarding non RGB layout masks at 1440p. Could be useful for those who wish to improve the WRGB OLED experience at that resolution:
This is factually incorrect and needs to be qualified. If you’re having issues with Mask Layouts on 1440p WOLED displays then say that but CRT-Guest-Advanced’s Mask Layout 1 works perfectly with 4K WOLED Displays and also BGR LCD Displays, while Sony Megatron Color Video Monitor’s RWBG (OLED) - Display Subpixel Layout works perfectly with 4K WOLED Displays.
Your 1440p WOLED issue with not being able to achieve proper subpixel alignment here is different. 4K WOLED has been “solved” for some years now.
My main question was directed mostly towards Guest’s existing Reshade ports that are in need of a update and some presets, mainly now with the Doom Ports.
there is this preset i stumbled upon for retroarch on which i would love to see translated into Retroarch Guest Advanced https://pastebin.com/h6pvMxcK
First time of me seeing the different subpixel combinations with the guest mask size/rgb/rbg switch ups on a OLED so thanks for posting those pics up. I don’t know what it is but something about OLED subpixels I really don’t like, it’s not just the extra “white” that they added to it either.
I’ve been hoping in all this time I’ve held out on buying a OLED that some company would make a standard RGB subpixel layout type but it seems it’s gonna be this new “WRGB” madness here on out. Guess I’ll be sticking with my standard 4k lcd and just make the best of it until it dies out.
The Reshade port is really up-to-date. There are some dynamic resolution change features which can’t be ported and slightly better performance with corner/border functionality. No visual changes are to be gained by updating to the latest guest-advanced version.
@DevilSingh did a splendid port and there are no hurries for the “new corner”.
Those pics @DevonCM posted are not how things look on a 4K OLED at least not mine. That’s from a 1440p monitor.
I’ve been posting photos of OLED Subpixels for years now and all of those masks work properly using Mask Layout 1 on my 4K OLED TV.
Also, in those pics that @DevonCM posted there seem to be additional subpixels being lit for example the Mask 10 BGR one. This could be due to the Shader Parameters allowing colours other than the Phosphor primaries to be lit.
Stuff like this has been discussed extensively in the Sony Megatron Color Video Monitor and even mitigated by the implementation of a Mask Accurate Mode which clamps the phosphor primaries to the RGB subpixel colours so that no extra subpixel are lit.
In my Mega Bezel presets, I haven’t encountered that.
What I’m seeing here is an opportunity to explore and find solutions to these anomalies but that’s just me.
If you’re interested in learning and seeing more you can read up on the Sony Megatron Color Video Monitor thread from a couple years ago and also the OLED Subpixels: How Do They Work? thread.
I’m getting a vibe here that folks here are willing to throw in the towel too easily but that’s just my opinion.
@DevonCM what happens when you use Mask 12, Layout 1 BGR? Do you mind posting pics of that when you get a chance?
Pay attention to the highly zoomed in pics which show the individual subpixels elements as the ones which are zoomed out might be more affected by camera optics.
I see the shots, it’s cool. I just find those little 5 circles that make up each triad subpixel line weird looking instead of it being a solid line. I’m not here to knock anyone using a OLED, that’s their cup of tea was just sharing my opinion with DevonCM. Besides I’ve never personally used a OLED in person so I can’t give too much of my two cents on the matter, just going off screen shots.
Anyway I’m going back to playing in the lab, I hardly have time these days for anything so I gotta make every minute count when I get some free time.
That’s impossible to see with the naked eye. Those are actually the elements that make up the subpixel elements.
Stuff like that is not exclusive to OLED panels though. In my “dissertation”/wall of text, I showed pics from my Toshiba Regza’s VA panel and those subpixels are made up of some vertical strips. So basically what you’re talking about is microscopic stuff. It looks like a straight line to the naked eye. If you zoom in some of the more zoomed out shots you can see this.
What’s common in all of these pics is the use of the RRGGBBX Mask, which is what’s used in CRT-Guest-Advanced Mask 12 but the layout used is RRBBGGX or Layout 1 in CRT-Guest-Advanced and RWBG (OLED) in Sony Megatron Color Video Monitor.
1440p (WOLED) probably just needs some more work to find something that works. 4K is fine.
Man, seeing is believing. It is extremely impactful from a CRT emulation perspective. All of these things we’ve been “wasting” energy and time on discussing is only relevant from up close to the screen literally nose to screen or in photos. You don’t see any issues even with messed up and mismatched layouts. Colours look correct as well. Getting the subpixels aligned is just the icing on the cake.
The micro contrast and lack of blooming is just phenomenal on OLED.
I’m really eyeing MiniLED though because it’s not nice having to worry about burn-in if this is what you’ll be using the display for primarily. My experience is with an ancient OLED and it is said that things have improved but I wouldn’t want to be the guinea pig though.
The thing is, the MiniLED sets that can give OLED some serious competition are basically more expensive making it easy to just buy another OLED.
The next things that is amazing about OLED is the viewing angle. You can watch from anywhere and things are going to look good and saturated with virtually no noticeable colour shift or additional blooming.
From another perspective, CRT Shaders can also look awesome on LCDs but the thing with LCDs in general are the things that can be visible which just shouldn’t be a part of the image, mostly the blooming, flashlighting, low contrast, lack of true blacks. Those things take away from an otherwise more than acceptable image.
OLED has none of those things present in the image.
One more thing about (at least LG and Sony) OLED TVs is that is that they usually come with very accurate calibration out of the box.
On the flip side LCD’s can generally and more easily get brighter than OLED displays and that helps immensely with realistic CRT Emulation when you take into consideration all the stuff that darkens the image like the CRT Mask and scanlines. Then, extra brightness is needed if we want to use things like BFI.