Hello Cyber. First of all thank you for your amazing work in trying to make retro gaming feel as authentic as possible.
I’ve been trying around and playing with shaders for a while now and decided to stick with yours.
To further enhance my experience and gather some knowledge in the process I have a set of questions which you’re hopefully able to answer. I hope they’re not too many.
Info: I play docked on the Steamdeck with 4K HDR output through Retroarch to my Sony A8 OLED. I mostly stick to your Mini LED packs.
It says BFI in the shader names and I know it stands for Black Frame insertion. Is BFI actually necessary for these to look as intended? Retroarch through Emudeck doesn’t support BFI yet AFAIK and my TV isn’t capable of 120Hz anyway. I thought they mostly look good without.
I sometimes have a hard time to decide between these very distinct names. Fine, Sharp, Ultimate, Ultimate Fine Advanced. You get the gist. When I get up close I CAN tell the difference but I’m not always sure WHAT it is. Sometimes the mask density is higher. Is it emulating different quality TVs?
Often the shader names already have the name of the mask in them. When I then go into parameters I can change the mask though. When I go into the shader parameters of a shader with “slot mask” in the title, is it advised to also change the setting to slot mask or can I change it to Aperture grill if I prefer?
I realized most of these have 4K in the title but out of box without adjusting parameters they always seem to be at 8K by default in the screen resolution parameters, so the masks look way too large and I can see the individual RGB pixels from afar, as each color is made of 16 pixels instead of 4 then. I manually have to set them to 4K.
I’m not quite sure what to set my peak paperwhite luminance as. I found out peak luminance for my TV (Sony A8) is 650nits, but there’s no info on the internet about peak paperwhite luminance. I saw a retrocrisis video where he set his to 50, but that image is just way too dark then.
Why are they actually divided by consoles? I get why they’re divided by cable signals for example, but some of the SNES titled shaders give me the see through Sonic waterfall just as the Genesis titled shaders. What’s the reason behind that?
Welcome @IlMonco, and thanks for your interest and these great questions.
I know everyone calls them shaders but I mainly make Shader Presets which are settings for shaders made by other awesome people. So feel free to thank the shader devs as well because without them, there probably wouldn’t be any CyberLab Shader Preset Packs.
No, that’s really a note to help me remember the settings used when creating the presets. Compensating for BFI when making a preset would make for a very different if BFI was off in the process but you don’t have to use it.
I make so many presets and am always evolving them that I probably couldn’t tell you the difference either but sometimes when I come back to improve or adjust something I go into experimental mode. That does not mean that there was something wrong with what I did before so the previous good preset is usually kept and the new preset is simply a variation so it needs a new filename in order to differentiate. Hmmm…how do you differentiate such minor or evolutionary changes? I tried I, II III e.t.c but got tired of it so now we have this hyperbolic nomenclature.
Don’t take it too seriously. Most of these presets are self contained so you can go through them one by one and delete the ones you don’t like or make note of the ones you do. I just made a new preset pack, the name of it is Epic. I’m going to try my best to only include my best of the best presets and not clutter it too much but I’m not making any promises…lol
I don’t think I’ll ever stop adding and making presets once I am able to.
In general the longer and more extreme the name the more refined the preset. In some cases you can look through the thread and see where I posted screenshots of photos and the names of presets and you can try those.
You can also try to stick to the newer presets as newer ones tend to be more refined. I’m always noticing stuff that can be improved and learning as I go along but not all the time a new preset or pack means that the old ones are obsolete or irrelevant. Sometimes, I perfect then I start over or try something different or new and I come up with a new naming scheme or new preset pack to reflect that fresh start.
Some of the terms have meaning though.
Fine means a higher TVL or smaller pitched mask features (aka higher mask density).
Sharp usually means, no (edge) smoothing is applied.
Advanced is a new generation of presets which have additional features like noise, rounded corners and the ability to crop overscan or garbage from within the shader parameters.
Ultimate presets are supposed to be more refined than vanilla presets but sometimes I refine an already Ultimate preset. Hmmm…Ultra Ultimate? Ultimate Ultra? Mega for Megadrive/Genesis? Super for SNES? Turbo for Turbo Duo/Grafx16/PC-Engine?
Do you see where this is going now?
You shouldn’t have to change a shader with “Slot Mask” in the title to Slot Mask in the Shader Parameters. That’s a bug or typo if you see something like that and of course you can change it to whatever Mask type you prefer, although I kinda already do that for most presets and I compensate for the differences in certain characteristics which happen when you switch Mask types. This is also one of the reasons why there are so many presets, for every preset there can be at least 3 versions, 1 for each Mask Type to start with.
This is not a bug, this is a feature. 8K and 4K just allow for different Masks to be selected which would give the approximate TVL thats listed in the Resolution parameter provided that you run the preset at the selected display resolution specified.
However, you can find that a mask with a different subpixel layout that is only available using the 8K setting looks better than a mask of a slightly different or similar TVL that’s available using the 4K Display’s Resolution. Sometimes the Mask height might be more appropriate for the 8K mask Vs a similar but not identical mask in the 4K set. I first started using 8K when I noticed that the centre of the Slot in one of my Slot Mask presets wouldn’t align exactly with the center of the scanline but the 8K mask lined up exactly!
In more recent times I noticed that many of the 8K masks that I liked used RYCBX instead of RGB and although they generally looked almost exactly like what I wanted, that additional subpixel bothered me so in the latest addition of my Epic preset pack, I modified all of them to use RGBX/RBGX/BGRX masks instead.
It’s still under the same 8K setting.
You can use the TVL settings while using any of the Display Resolution settings to adjust the size of the Mask features. You don’t have to switch to 4K to do that. You can’t easily adjust the height of the Mask in Sony Megatron Colour Video Monitor though. That seems to be fixed. Which is why I choose to use 8K over 4K in some cases. You complained about the Mask looking way too large but did you compare the height of the 4K and 8K masks? The 4K slot masks are taller than the 8K ones.
So you can have mask that’s just as chunky using the 4K setting if you lower the TVL.
These things are just a guide and a starting point to setting the shader up. It varies according to your Picture Mode, Tonemapping settings e.t.c.
You really have to use your eyes to dial these in perfectly. Too much brightness leads to harsh looking highlights and oversaturated, clipped colours.
Too little and it leads to darkness and very inaccurate looking colours.
The trick is to use the reported Peak Brightness/Luminance to get a close enough figure for your Peak Luminance, then crank up the Paper White Luminance until it looks bright and good to you without going overboard and clipping all the details. The number doesn’t matter. How it looks matters.
I don’t know how he arrived at a value of 50 for his Paper White Luminance for any TV but the vast majority of his videos are excellent while a few might sometimes be a bit oversimplified or its possible that he may not have fully understood what I might have been trying to convey in explaining how to go about setting up the shader or presets.
As you see, I’m the opposite in that, I try to provide as much details as possible in my explanations but this doesn’t always work for the masses. What RetroCrisis does is very appreciated at the end of the day. Maybe that settings of 50 is just what he needs for his display but definitely not anyone who owns an OLED TV of any kind.
On my OLED I used 630 Peak and 630 Paper White Luminance. For my "Near Field presets, I used 630 Peak and 450 Paper White.
If you have a similar WOLED based OLED TV, I don’t see your settings being vastly different.
Because different consoles have different output chips and circuitry even among the same generation and family. There are several variations of Sega Megadrive/Genesis hardware and each would look different on the same composite cable.
How would I capture these differences if I used one preset for all variants within the family?
Well there’s dithering in SNES games too as well as Turbo Duo family games and if blended properly can provide more colours and transparency effects so why not exploit what the software can do to bring forth such elements?
In my presets I try to achieve those things by default most of the time so it’s not necessarily something you’ll read much about or hear me talk about but you’ll definitely see in action when you use my presets.
You’re most welcome. Thanks for the impromptu Q & A and feel free to ask more and also to share more!
