I recently added a rendering pass in my engine to compute volumetric shapes via raymarching.
Uncategorized
22
Posts
5
Posters
1
Views
-
@froyok Do you have access to the zbuffer when you do it? Maybe you can do it like msaa and only do full res at edges maybe? Not sure what the best way to to get coherency though…
This only works if edges are the problem though. Upscaling and noise doesn’t play well together sometimes.
@breakin I have the zbuffer yes, but I can't toy with msaa unfortunately. Here the issue is around edges of geometry, noise isn't a problem.
-
@breakin I have the zbuffer yes, but I can't toy with msaa unfortunately. Here the issue is around edges of geometry, noise isn't a problem.
@froyok Ah I was thinking that when you do say half res you could check 4 pixels in zbuffer to see if they are close enough for half res. Otherwise put that pixel id in a buffer. Do you have atomics? Then a second pass does full res for things in buffer. I mean you could redo edge-pixels during upsampling to see if that solves quality first bur sounds slow.
Somewhat like manual msaa.
-
@froyok Ah I was thinking that when you do say half res you could check 4 pixels in zbuffer to see if they are close enough for half res. Otherwise put that pixel id in a buffer. Do you have atomics? Then a second pass does full res for things in buffer. I mean you could redo edge-pixels during upsampling to see if that solves quality first bur sounds slow.
Somewhat like manual msaa.
@froyok Pleanty of tricks like you do volumetric to the closest z value then you can continue from there for you per-pixel rays. I know the edge pixels are the mest expensive ones so maybe a bad idea from that perspective.
-
I'm not sure what I could do better, without temporal anti-aliasing. I'm open to suggestions ☺️
@froyok Have you tried this: https://c0de517e.blogspot.com/2016/02/downsampled-effects-with-depth-aware.html
The important parts are using a min/max depth checkerboard pattern for the downscaled depth and then choosing between bilinear/bilateral upsampling based on depth discontinuity for a given pixel.Also randomly offsetting the rays with blue noise and then running a bilateral blur over the volumetric target before upsampling helps. Allows to get away with huge steps during raymarching.
-
I'm not sure what I could do better, without temporal anti-aliasing. I'm open to suggestions ☺️
@froyok what about variable rate compute shader dispatch? https://youtu.be/mvCoqCic3nE
-
@froyok Pleanty of tricks like you do volumetric to the closest z value then you can continue from there for you per-pixel rays. I know the edge pixels are the mest expensive ones so maybe a bad idea from that perspective.
@breakin Thank for sharing that idea still ! That's sounds interesting ! :)
-
@froyok Have you tried this: https://c0de517e.blogspot.com/2016/02/downsampled-effects-with-depth-aware.html
The important parts are using a min/max depth checkerboard pattern for the downscaled depth and then choosing between bilinear/bilateral upsampling based on depth discontinuity for a given pixel.Also randomly offsetting the rays with blue noise and then running a bilateral blur over the volumetric target before upsampling helps. Allows to get away with huge steps during raymarching.
@blurple I haven't tried the checkerboard, but I tried different combination of bilinear and nearest filtering with custom weights. No luck so far.
-
@froyok what about variable rate compute shader dispatch? https://youtu.be/mvCoqCic3nE
@WaitForPresent Not doable with the framework I'm working with I believe, plus I'm sticking to OpenGL currently. Or I would need to emulate it myself in compute.
-
I'm not sure what I could do better, without temporal anti-aliasing. I'm open to suggestions ☺️
To illustrate a bit, In the worst-case scenario this is how it looks like:
-
To illustrate a bit, In the worst-case scenario this is how it looks like:
I use a 3x3 bilateral filter, each sample is weighted by both spatial distance (sub-pixel position within the half-res grid) and depth
similarity (to prevent light bleeding across silhouettes). -
@breakin Thank for sharing that idea still ! That's sounds interesting ! :)
@froyok This recent link would help if you had sdfs. Maybe some inspiration in sharing work https://pointersgonewild.com/2026-03-06-a-recursive-algorithm-to-render-signed-distance-fields/
-
I'm not sure what I could do better, without temporal anti-aliasing. I'm open to suggestions ☺️
@froyok If instead of 2x2 binning (aka half resolution) would it work to do two versions like 1x8 and 8x1, and then accumulate those into the full resolution final result? Directly inspired by box blur.
-
@froyok If instead of 2x2 binning (aka half resolution) would it work to do two versions like 1x8 and 8x1, and then accumulate those into the full resolution final result? Directly inspired by box blur.
@mirth Hmm, I not sure I see how I could transcribe the effect into this method. What would be the point of making it separable it, outside of the performance cost ?
-
@mirth Hmm, I not sure I see how I could transcribe the effect into this method. What would be the point of making it separable it, outside of the performance cost ?
@froyok Performance and a full resolution if blurry result, potentially solving the upscale question. I know a lot more about vision than graphics so not sure if this would be a sensible trade off.
-
@froyok Performance and a full resolution if blurry result, potentially solving the upscale question. I know a lot more about vision than graphics so not sure if this would be a sensible trade off.
@mirth I will scratch my head a bit about it ;)
-
@mirth I will scratch my head a bit about it ;)
@froyok Trying a version of this rescaling + threshold on a 2D black and white silhouette image I think gives a rough idea of the behavior:
-
@froyok Trying a version of this rescaling + threshold on a 2D black and white silhouette image I think gives a rough idea of the behavior:
@froyok The math would take some thought but it might be possible to sample the vertical and horizontally squished versions at render time, do the threshold, and skip the larger intermediate texture entirely Then the shadows would have fewer artifacts and the overall memory bandwidth used should be smaller.
(this feels like the kind of idea that one of the elder graphics people will chime in and be able to point out some prior art and improvements from 25 years ago)
-
@froyok The math would take some thought but it might be possible to sample the vertical and horizontally squished versions at render time, do the threshold, and skip the larger intermediate texture entirely Then the shadows would have fewer artifacts and the overall memory bandwidth used should be smaller.
(this feels like the kind of idea that one of the elder graphics people will chime in and be able to point out some prior art and improvements from 25 years ago)
@mirth Given the critical part here is the edges, the demonstration you propose doesn't help on this issue.
Then one solution could be compute bits at half-res where there isn't edges (so edges are at full res), but this sounds a bit like the MSAA or Variable rate solution that were proposed in other replies, they are unfortunately difficult to implement in my engine. -
undefined oblomov@sociale.network shared this topic
Feed RSS
Gli ultimi otto messaggi ricevuti dalla Federazione
-
@Gina So true, as a Canadian this was the final straw; protein infused Mac n' Cheese: https://www.cbc.ca/news/canada/kraft-dinner-protein-9.7136154
There is no punishment great enough for this heinous act.
-
@Gina Are you eating the mosquitos again?
-
-
Digital Brain (@yourdigitalbrain.bsky.social)
https://bsky.app/profile/yourdigitalbrain.bsky.social/post/3mhlwvcuegk2s
> The Snow Fairy: Japan's Cutest Bird Weighs Only 8 Grams 🐦❄️
-
Reviving a Cursed Sun SPARCstation IPX
SPARCstation IPX running Solaris 2.6. (Credit: This Does Not Compute, YouTube)
The best part about retro computing is the idea that you’ll save some poor system from being scrapped and revive it to a working state, at which point you can bask in the glory of a job well done. That’s when reality often strikes hard, and you find yourself troubleshooting a maddening list of issues as you question everything about your life choices. Such was the case with [This Does Not Compute] over at YouTube with a Sun SPARCstation IPX that decided to put up a big fight.This is the second video of a series. In the first installment, the PSU was repaired, and a boot failure was diagnosed. The system’s onboard diagnostic led to the assumption that one of the 8 kB SRAM ICs was defective. You can readily get SRAM replacements, so it seemed to be an easy fix. Unfortunately, the fun was only beginning as the system reported the exact same error after the SRAM was replaced.
After flipping a virtual table, the mainboard was swapped with a donor one from a scrapped laptop system. With this, it booted, but the video output showed only vertical lines. Obviously, the solution here was to insert a replacement video card, which not only fixed the display output but also demonstrates once again that you can fix many vintage computing issues by simply replacing hardware.
With the system now seemingly happy, a disk drive was added so that Solaris 2.3 could be installed. This turned into another confusing job. Getting into the GUI was seemingly impossible. Ergo, this virtual table got flipped, too, and Solaris 2.6 was installed. Finally, the system got to the desktop GUI. At this point it was clear that this 40 MHz mid-range SPARCstation from 1991 is no speed monster. Just drawing windows was slower than an X11 remote session over 9600 baud dial-up.
Installing more RAM might have helped here, but the system requires parity FP RAM, which could have been purchased, but at some point, you have to decide how much money you want to throw at an old system like this. Although these are still pretty interesting systems in their own right, it should be clear that they are not easy systems to repair or maintain.
-
🍎$tr€gh€tta🌶💬🦋🍉 (@trilly153.bsky.social)
https://bsky.app/profile/trilly153.bsky.social/post/3mhlwqc4zmk2a
> Da mo…
-
@Gina and fiber is the new protein, what will come next?
-
This post did not contain any content.
