If it's the Quake/Source engine running somewhere under the hood, I'd wonder if there's someway to access demo recording and playback functionality, because then you might be able to simply replay the recorded demos from old speed runs. playdemo <file.dem> or timedemo <file.dem> or whatever. Can't remember if there's launch options to avoid having bring up a console to play a demo.
Digital Foundry Article Technical Discussion [2022]
- Thread starter BRiT
- Start date
- Status
If it's the Quake/Source engine running somewhere under the hood, I'd wonder if there's someway to access demo recording and playback functionality, because then you might be able to simply replay the recorded demos from old speed runs. playdemo <file.dem> or timedemo <file.dem> or whatever. Can't remember if there's launch options to avoid having bring up a console to play a demo.
Riddlewire
Regular
Re: Other RTX Remix titles,
I wonder if RTX Remix (or other ray tracing implementations) would break the gameplay of Thief: The Dark Project. Surely the lighting system they used is integrated into the AI. That's not a simple swap. Either the original system would have to still be running, but invisible, or the AI would have to be re-coded to be aware of the new RT luminance on Garrett.
Or maybe it's not as complex as I'm thinking it is.
In any case, that would be a great game for ray tracing.
I'd expect that the lighting visualization is separate from the lighting logic and AI logic -- it's not like the game would be rasterizing from the perspective of the AI to determine what areas are light and dark. Doing the game with RT would probably make you want to rewrite the game from the ground up, as there'd be tons of features and details you could do that would be more interesting.
Your motion would move the air causing a draft like displacement of air, reaching a point where you are both pushing the air in front of you while also being pushed by the air that you are pushing forward, because it's traveling through the portal and hitting you in the back.
Cardassian Bob
Newcomer
Air is a fluid like water. A body displaces fluid as it moves through.
You would have to be moving at supersonic velocity to compress the air in front of you.
Drag force cubes with the doubling of speed, so it will overcome the acceleration of gravity causing you to reach a terminal velocity long before you break the sound barrier.
In the case of falling into portals you get a compounding effect because you're continually accelerating from the force of gravity. Light travels at a constant velocity, so the act of entering and exiting a portal doesn't add any energy. In a naive RT/PT model of light propagation, a mirror doesn't behave any differently than a portal -- light gets absorbed and then reemitted. If a portal were to cause some weird accumulation artifacts that break energy conservation than so should a mirror. A portal/mirror would behave analogously to a perfectly elastic bounce, absent of gravity or air resistance. You throw a ball at a wall it bounces off at the same velocity, if there are a series of walls then it will rebound indefinitely but not increase in speed. If it were recursively going through portals it would likewise continue at the same velocity.
Perhaps there'd be an issue with improper energy accumulation if you're using some kind of surface caching for GI such that surfaces have a form of latent memory of 'old' light that then gets recursively added to with reflections/porting? If you're not actually tracing back to the original emitter then you might not be getting a statistically-valid/ground-truth sample for your monte carlo sampling?
Perhaps there'd be an issue with improper energy accumulation if you're using some kind of surface caching for GI such that surfaces have a form of latent memory of 'old' light that then gets recursively added to with reflections/porting? If you're not actually tracing back to the original emitter then you might not be getting a statistically-valid/ground-truth sample for your monte carlo sampling?
Henry swagger
Newcomer
@Dictator and john trolling the console fanboys " cache scrabbers " lol
davis.anthony
Veteran
I do like seeing this more chilled back and jokey side of DF.
I think they're being way too kind to AMD in regards to RDNA3, and equally too harsh on people for having higher expectations. This idea that the product is fine and it's just a case of people being overhyped? This might have been the case with very early RDNA3 rumors, especially with all the 2xGCD+2x performance rumors, but it was not unreasonable after we learned that it was only gonna be 1xGCD and AMD themselves were claiming 50% improvement in performance per watt and 50%+ improvement in overall performance uplift from 6900XT. There was no reason to doubt these claims. AMD did not have some history of hugely inflating these sorts of numbers - quite the opposite, they've tended to be reliable, even to the point of caution at times. Combined with all the factors involved here in terms of architectural upgrades, the long development period, and a large process node jump, a mere 35% improvement is terrible and is not 'normal/standard', nor should it have been all we expected. I still think it's extremely clear that something is wrong with it.
Also disagree completely with Alex that RDNA3 is some 'transitory' architecture, at least anymore than any architecture is.
This is funny and not trolling some people think it helps have a more powerful GPU.
Conservation of energy. The photons that go through a portal didn't hit anything else. No light is amplified.
It's not that the light is amplified, it's that it's sampled multiple times in the model, to infinity. In the real world, you could take an object and shine a light on it. And then you could use a mirror to reflect more light from the same source onto it. Therefore, the object will be brighter. The problem they were describing in portal is that you can have a light shin on an object that is bouncing light through a portal and onto itself. Since you can look into a portal and see the portal you are looking into, that means you can see the object that light is hitting, and reflecting onto itself. That means that the now brighter object is reflecting now more light onto itself. It's not clear to me if they are saying that in the primary view of the world, ie where the character is located, if the object in question would be infinitely bright or not, but it does make sense that if you look through the portals, in the first portal it must be, because it's getting the light from the primary world, and is reflecting into the second. That means that the 3rd instance would be even brighter, because the second instance would reflect more light. But to trace that back, because the portals loop, would the object in the primary world have light shining from the portal from itself? And if so, wouldn't it be bouncing more light on itself through the portal?
I love that you guys are trying to disprove Portals in universe physics by using real world physics. Little metal things on your legs wouldn't stop you from getting hurt falling for 900 feet either. Regardless, you don't have to be moving at supersonic speeds to move air. If you jump through a set of vertically aligned portals, and you are being affected by drag force, then you must be displacing air. Meaning that you must be forcing air through that portal. Meaning that you are being slowed by drag force while also being pushed by the air that's rushing behind you to compensate for the displacement. You will also be affected by your own draft. Your velocity with each loop should lower the air's resistive effect on you every time you make a loop, as the air in the path would also start moving through the portal at faster speeds as well.
Consider a closed cubical room with a light source inside. 100% of the light will be absorbed and some portion of that reemitted back onto itself, but the room won't get perpetually brighter; it will obviously reach some equilibrium where the surfaces will be of a certain constant brightness. Now imagine that room but half of the surface area is covered with either: a) perfect mirrors or b) portal entrances and exits. In either case what you've done is remove 1/2 of the opaque absorbing surface area and replaced it with surfaces that can't absorb. The result, unless I'm missing something obvious, is that the remaining opaque surfaces will just be getting 2x as much energy as they were previously, and will be appropriately that much brighter (not 2.5x, not 3x, not arbitrarily-more-x).
Latest hotfix
this is going to be a very interesting video. I played the game for a few minutes yesterday. Setting the max framerate to limited, the game was running at about 90-120fps (naked eye numbers, MSi Afterburner wasn't working).
In the end, the DX12 version has RT options greyed out for Intel GPUs, which happens, I imagine, 'cos they are investigating an issue with low performance on those. So I am putting the game on hold. I got used to TW3 RT and Doom Eternal RT and the more RT games I play -save for Shadow of the Tomb Raider, which only changes the shadows- I am starting to discern when RT is enabled by the naked eye. In Doom Eternal the textures look flatter without it, and in TW3 -without RT- the lighting, everything is just a flat lie.
I mean, you can see "patches" of unnatural shadows and "patches" of lit areas that look good but you don't have that indescribable feeling of "this looks natural" at all. You know, by how everything is set up, that the lighting is artificial. Since I completed the game and the expansions several times with the vanilla experience, I don't mind waiting for a new patch that enables RT.
The extra light that is hitting some surfaces is the exact same amount of light that went through the portal amd thus didnt reflect off that surface. The sum total of light in the scene is still the same. You've just concentrated it more in specific areas.
Silent_Buddha
Legend
Sure if there is something in that cube that can absorb energy it's maximum potential will be limited by the amount of the entering energy eventually being absorbed. But what if all surfaces are perfectly reflective?
Going back to portals, obviously a light source will not emit photons with each going exactly parallel to each other thus almost all of the portals will eventually leave this system as their angle of travel will take them out of it after some number of loops (0 to near infinity). We have two portals setup such anything exiting the first portal enters the second portal and anything entering the 2nd portal exits out of the first portal. Now, let's say that there's no atmosphere to potentially alter the path of the photons although I don't think this really matters due to the properties of photons. There will occasionally be 1 photon emitted by the light source that is perfectly aligned such that it'll enter and re-enter the portals infinitely many times.
Then another photon is emitted that will also enter and re-enter the portals infinite many times. So, now there's 2 photons in an infinite loop, then 3, then 4, then 5, etc.
Because of the properties of photons they can theoretically all occupy the same space which means in theory you could have a infinite stream of infinite photons accumulating photons for infinity with every individual "space" in that stream containing an infinite number of photons.
Ridiculous? Yes. Theoretically possible? Eh, yeah, sorta, possibly.
Regards,
SB
- Status
Similar threads
