Please show off what crt shaders can do!

When you consider that most tv is calibrated to 100 nits, 80% OLED achieves that and 100% is 120 nits which is very comfortable. With all of the bells and whistles it is 100 nits. I use mask type 7. It’s completely bright and playable in a completely dark room.

1 Like

It’s not a problem of brightness per se. Sorry to do this to you, but that mask results in banding / posterization at the low end. Masks 5-7 all do this. Needless to say, this is really not good for objective picture quality, but if you’re happy with it then that’s all that matters. I’d check out the other masks I mentioned as you’ll get better/more accurate results with those.

Mask 7:

Mask 3:

1 Like

My calibrator arrived!! I wanted to test a few things brightness wise. You say that a CRT with masks (scanlines or not?) should measure 120nits? is that ok?

I will try to measure a few masks, intensities and compensation for HDTV back light settings. By the way, I should check if there’s a BFI option on my TV (or it’s just set on Retroarch right?) it’s a mid tier Samsung from 4 years ago but I think it would mess with my modelines so…


A higher-end CRT with or without scanlines will have a max of about 120nits. When in double-strike/240p mode, more voltage is applied to each line so the brightness is the same.

For BFI, just set the display to 120Hz and enable BFI in RA and you should be good to go (I might be forgetting something as it’s been a while since I’ve played around with it)

See how your results compare to mine. I’m also taking into account viewing distance; masks must be visible at the correct Lechner distance (2-3x the width of the display).

1 Like

I use the color calibration roms but I find I still have to open games and tweak a lot. I frequently bounce between Street Fighter Alpha 3 and Vagrant Story. One is very bright and the other dark. SFA3 has highly saturated colors close in value that are easy to get lost with shaders. The subtle hues of the clouds and the red’s in the posted image for example. I adjust to much for SFA3 and then Vagrant Story becomes to dark.

One of my shader goals is having a bright, colorful, contrasting image nice scalenline representation. What are you personal shader torture tests? Here are some images of before and after.

1 Like

Yep, no way around it if you want optimum results, really. I know I’m sounding like a broken record, but check out Dogway’s grading shader, it’s a must-have for those who are serious about image quality.

As far as the test roms go, you should use the rom for the system you’re trying to optimize. Settings for SNES using the SNES test suite won’t necessarily translate to other systems perfectly.

Torture tests… I usually fire up something with a lot of greens since I can immediately tell if saturation is off by looking at the greens in certain games. For brightness, anything really dark will do. Super Metroid, Silent Hill, Castlevania… just set brightness to the lowest level where all detail is still visible.

1 Like

@Nesguy I’m using Dogway’s grading shader in one of my venom shader chains. I find it goes off the rails quick for me. Should I put it at the top of my shader chain or the end? Perhaps I should look more into how others are using it.


The CRT mentioned in the displaymate article has measured peak brightness of:

Sony PVM-20L5 176 cd/m2 ( )

I’m reading in LCD monitor testing sites that mostly they calibrate peak brightness to 150 cd/m2 for comparison between different models as the 150 cd/m2 is then mentioned to being the normal viewing luminance in normal daylight conditions (they argue at least). Personally I would see if the 150 can be targeted as it leaves some headroom for tweaking to personal preference.

@Nesguy , the brightness values you mention is that from some source or personal preference?

EDIT: @Dogway maybe you could use the reshade version of guest-venom such that you can actually use the scanlines / mask on the calibration software testscreens that comes with your calibrator.

EDIT 2: here are some recommended luminance levels from the calman site:

Recommended Display Luminance Levels:

  • Bright Room: 200+ cd/m2
  • Dim Room: 140-170 cd/m2
  • Dark Room: 100-140 cd/m2
  • SDR Reference: 100 cd/m2
  • Theater Room projector: 50-100 cd/m2
  • High Dynamic Range: 600-2000 cd/m2 (based on commonly available TVs as of July 2019)

I replaced d65-d50 with grade in my venom chain, seems to work fairly fine. You have to edit the preset manually in notepad or some text editor for this to work properly though.

I also remove the color profile pass.

1 Like

100 nits is part of the SD standard, so that should be our minimum. It’s absolutely fine to go brighter than this, but at a certain point you’re going to get clipping. I would consider the 100 nits as a minimum after adding all scanline and mask effects. 150 nits is probably a good maximum. It really depends on the ambient light. I’d say the ideal is the highest level at which you don’t get any clipping OR any eye strain with bright scenes.


This is what I do as well; I’m too lazy to re-number all those shader passes lol.

Any other shaders you don’t want to use can also be replaced with “stock.slang.”

1 Like

@rafan very interesting reference from Calman. On my researches on dim surround and display brightness I also found this:

 ISO 3664:2009 establishes a monitor whitepoint brightness level between 80 (dim surround -64 lux-) and 120 cd/m2, depending on the controlled ambient conditions of the work space.
 Grade 1 Monitor: 70 to at least 100 cd/m2 (dim, master target)
 Grade 2 Monitor: 70 to at least 200 cd/m2 (daylight)

From what I could read and despite the ITU.709 spec, HDTVs rarely reach a gamma of 2.4. I measured about 2.2 on mine and only lowering the gamma setting I reached to 2.3x but things got ugly from there on, you really need a display designed for gamma 2.4 (grade level)

1 Like

@rafan @Dogway

For CRT emulation we should set brightness as high as possible while avoiding clipping or eye strain.

Then, we gradually increase the mask/scanline strength until we have at least 100 nits when a white screen is displayed (or higher depending on ambient light and preference).


@Nesguy - yup, man. We seem to agree on most important features nowadays, which is a good sign, in the sense that it means that both of us have come a long way and learnt quite a bit from each other and everybody else :slight_smile: This thread and its siblings and offspring have become some of the (if not the) most useful and informative discussions on the topic that one can find online, far surpassing my initial expectations and it’s great to be part of it. Oh and yes, it’s easymode, and also I think you are right regarding the presence of scanlines in white areas. And I must definitely try Dogway’s. It’s getting so much love around here, and well deserved I’m sure.

Regarding the gamma talk, 2.2 is a very good reference and starting point, but more often than not it needs to be tweaked based on content. I also find that it’s very strongly linked to colour preferences. I tend to desaturate old games quite a bit, and when doing so I need to lower the gamma levels to make up for the lost contrast and vibrancy. That’s one of the reasons why it is so useful to have per channel gamma control in color mangler.

@BendBombBoom - those are good choices for bright and dark content. You certainly improved SFA3’s gamma, but the shaded screen looks a bit too warm and saturated, I reckon. As for Vagrant, the cooked shot looks actually worse than the raw one. You are losing a lot of detail and crushing dark areas.


I agree. My pattern is to tweak a shader, think I have it right, use it a day later to play a game and say “What the hell is this?” And then I mess with that shader, forget to play the game and the pattern repeats.


Actually, in regards to brightness, it might be a good idea to set this at the level where you just start to see some clipping w/color bars, that way you have as much headroom as possible to work with. The clipping should go away when you add the mask/scanlines if it wasn’t too severe to start with. @dogway @rafan

@BendBombBoom Lol, I think everyone here can relate. For the most part, guest-dr-venom+grade w/the default settings should give you an image that will pass all the tests you can throw at it. The only thing that really needs to be adjusted per-game is CRT gamma, but 2.4 is fine most of the time. The defaults are fine for the vast majority of content produced in Japan.

Also bear in mind that almost all of the shader settings will affect saturation, contrast, brightness, etc, which is why I typically turn most effects off. Just adding scanlines and the mask adds a lot of darkness to the image, which can be a good thing if your display is calibrated to have enough headroom. Bloom will add white to the image, which can introduce clipping if you aren’t careful. Be sure to set guest-dr-venom’s gamma to 2.2/2.2 so it’s not doing anything further to the gamma after grade.

1 Like

Haha I feel you! I think we all do

I’m not so sure about that. Are you using a colorimeter? I’m ‘meticulous’ in calibrating all of my displays and I don’t really see any issues like you are mentioning. Is that a particular game? Here’s my SFA3 CPS2 I see a little bit of crush on the 1 bar for blue and passable for red but at least you can tell it is not 0. I don’t get banding. This is 7, 6 and 5 have less crush and still no banding.


The screen cap I posted is from Fudoh’s 240p test suite running in SNES9x, I’ve gotten the same results on three four different displays with those masks including my iphone (which should be very accurate). The banding / posterization should be clearly visible in the first shot I posted.

Please try 240p test suite in SNES9x and see what you get so we can see where the discrepancy lies.

Maybe somebody can help me here. I have a 4k monitor by LG, and it has a setting “picture mode.” Custom, Vivid, Rec709, SMPTE-C, and a few others. I’ve been trying to find out about Rec709 and SMPTE-C, and I just don’t understand what these are. They seem to be a type of broadcasting signal, but for some reason they’re also monitor settings. I have two questions:

  1. When a shader has a SMPTE-C setting in it’s parameters, a la Dogway’s, is that specifically for a certain type of device to display it, or is it just going for a look? or should I say: should I set my monitor to SMPTE-C as well as the shader?

  2. What setting should my monitor be at for shaders in general? Are they all made with Rec709 in mind, or Vivid, or what?