A3DR updated Honda S800 RSC with a new update entry:
RefractionFX and more
Read the rest of this update entry...
RefractionFX and more
Read the rest of this update entry...
Cars Honda S800 RSC 2.5
- Thread starter A3DR
- Start date
Thanks a3dr and aphidgod for the update,
A few new things to note that could have escaped testing, most go for the newer lod/light additions:
- if not intended, the lr interior's and lodB rear mirror wasn't removed so it pops back: Object179_LR and Object105_LR are missing from the hide= list, so as honda_s800_rsc_lodB.kn5 from the file= list of that section to complete its removal. IRL RSCs seem to have kept the mirror but at an higher position, but I guess it was already discussed in this thread.
- Object125_LR (interior windscreen on lr cockpit) is missing the is_transparent flag while using alpha blend, so it can cause broken render distortion when seen from the side window and the camera moves (this obj could probably just be removed from the cockpit_lr, or just added to the config's hide list) (img)
- The new removal of front+rear turning lights lens+reflectors from lodB doesn't look good, as it makes forced/closer loda->b transitions vs. AI grids more noticeable, and will cause holes on the rear lights section from some 3 quarter angles. Not to mention their emissives will stop working after lodA. (img)
- It should be kept in mind that csp is prone to replace some Bs with Cs vs an AI grid. C previously had the issue of missing the lower black car floor, and might be worth to add it too keep the silhouette, so the car doesn't suddenly gain clearance from the ground (please see the comment/image at the end of this old post).
- The included honda_s800_rsc_lodD.kn5 is now redundant
Some things that could be revised, mostly for refractions:
- For as much as csp additions can be great when well done, rear lights now look like very translucent glass/gelatin instead of semi opaque 60s plastics, specially on a >50 year old car, and their nm looks very different for the squared shapes that should look smoother. (img) (img)
- Front lights can almost look like a thick mirror glass without details seen behind (img) edit: endless floor looks like a worst case scenario, as they can look better in other conditions
- Cast lights on the ground for rear lights and all turning lights can look too strong for a 60s car, while the rear lights casing itself (reflector+outer plastic lens, that should diffuse light) becomes barely illuminated other than the lamp itself, making it hard to be seen at night if not very close. (img)
As always, this is more in-line with my usual beta testing reporting and not user feedback, a great part of the above won't be noticed/cared about by end-users.
Thanks if you mind looking into any of it.
A few new things to note that could have escaped testing, most go for the newer lod/light additions:
- if not intended, the lr interior's and lodB rear mirror wasn't removed so it pops back: Object179_LR and Object105_LR are missing from the hide= list, so as honda_s800_rsc_lodB.kn5 from the file= list of that section to complete its removal. IRL RSCs seem to have kept the mirror but at an higher position, but I guess it was already discussed in this thread.
- Object125_LR (interior windscreen on lr cockpit) is missing the is_transparent flag while using alpha blend, so it can cause broken render distortion when seen from the side window and the camera moves (this obj could probably just be removed from the cockpit_lr, or just added to the config's hide list) (img)
- The new removal of front+rear turning lights lens+reflectors from lodB doesn't look good, as it makes forced/closer loda->b transitions vs. AI grids more noticeable, and will cause holes on the rear lights section from some 3 quarter angles. Not to mention their emissives will stop working after lodA. (img)
- It should be kept in mind that csp is prone to replace some Bs with Cs vs an AI grid. C previously had the issue of missing the lower black car floor, and might be worth to add it too keep the silhouette, so the car doesn't suddenly gain clearance from the ground (please see the comment/image at the end of this old post).
- The included honda_s800_rsc_lodD.kn5 is now redundant
Some things that could be revised, mostly for refractions:
- For as much as csp additions can be great when well done, rear lights now look like very translucent glass/gelatin instead of semi opaque 60s plastics, specially on a >50 year old car, and their nm looks very different for the squared shapes that should look smoother. (img) (img)
- Front lights can almost look like a thick mirror glass without details seen behind (img) edit: endless floor looks like a worst case scenario, as they can look better in other conditions
- Cast lights on the ground for rear lights and all turning lights can look too strong for a 60s car, while the rear lights casing itself (reflector+outer plastic lens, that should diffuse light) becomes barely illuminated other than the lamp itself, making it hard to be seen at night if not very close. (img)
As always, this is more in-line with my usual beta testing reporting and not user feedback, a great part of the above won't be noticed/cared about by end-users.
Thanks if you mind looking into any of it.
Last edited:
The right instrument gauge (speedometer) is plain black atm. Though I have driven this car a lot, I never noticed that before. I'm using CSP 0.1.75, and noticing that other versions than 0.1.73 could give some issues I reverted back to that version, but the gauge is still black. I downloaded two earlier car versions (from the history), but still the same.
Was I dreaming and was this gauge always black ? I doubt it, but I'm not sure either ?
Was I dreaming and was this gauge always black ? I doubt it, but I'm not sure either ?
It's not black, it's missing, has been on earlier versions of the car too.
(from https://www.honda.co.jp/Racing/gallery/1968_2/01/)
Not really a bug, It's the sort of detail that depends on the works car used as a reference.
You should be my official QAThanks a3dr and aphidgod for the update,
A few new things to note that could have escaped testing, most go for the newer lod/light additions:
- if not intended, the lr interior's and lodB rear mirror wasn't removed so it pops back: Object179_LR and Object105_LR are missing from the hide= list, so as honda_s800_rsc_lodB.kn5 from the file= list of that section to complete its removal. IRL RSCs seem to have kept the mirror but at an higher position, but I guess it was already discussed in this thread.
- Object125_LR (interior windscreen on lr cockpit) is missing the is_transparent flag while using alpha blend, so it can cause broken render distortion when seen from the side window and the camera moves (this obj could probably just be removed from the cockpit_lr, or just added to the config's hide list) (img)
- The new removal of front+rear turning lights lens+reflectors from lodB doesn't look good, as it makes forced/closer loda->b transitions vs. AI grids more noticeable, and will cause holes on the rear lights section from some 3 quarter angles. Not to mention their emissives will stop working after lodA. (img)
- It should be kept in mind that csp is prone to replace some Bs with Cs vs an AI grid. C previously had the issue of missing the lower black car floor, and might be worth to add it too keep the silhouette, so the car doesn't suddenly gain clearance from the ground (please see the comment/image at the end of this old post).
- The included honda_s800_rsc_lodD.kn5 is now redundant
Some things that could be revised, mostly for refractions:
- For as much as csp additions can be great when well done, rear lights now look like very translucent glass/gelatin instead of semi opaque 60s plastics, specially on a >50 year old car, and their nm looks very different for the squared shapes that should look smoother. (img) (img)
- Front lights can almost look like a thick mirror glass without details seen behind (img)
- Cast lights on the ground for rear lights and all turning lights can look too strong for a 60s car, while the rear lights casing itself (reflector+outer plastic lens, that should diffuse light) becomes barely illuminated other than the lamp itself, making it hard to be seen at night if not very close. (img)
As always, this is more in-line with my usual beta testing reporting and not user feedback, a great part of the above won't be noticed/cared about by end-users.
Thanks if you mind looking into any of it.
Thanks for such detailed feedback! will take care of these and post an update sometime this week.
You should be my official QA
I didn't give the man full team credits on the WRX for nothing.
And yeah, race cars have very little need for a speedo. Tach is really all you have to pay attention to. No surprise it'd be removed... saves weight!
Last edited:
Hi, thanks for the update! I love the entire S line by Honda!
I don't know if it was already said but it seems something is not right with the glass (including windscreen) depending on the angle
I don't know if it was already said but it seems something is not right with the glass (including windscreen) depending on the angle
Last edited:
<- for the other posts aboveAtrupelador: Could you please share your replay file and extra fx settings? I can't tell if it's the diffuser getting also the edges unfocused, taa + motion blur ghosting there, pp glare, or just clip compression.
I'll then try to reproduce the issue, as I can't atm.
Could be just a matter of adding the problematic stuff to MASK_GBUFFER = . Which btw, the config is using SKIP_GBUFFER on line 83 for mat:EXT_diffuser mat instead of MASK_ , and that can cause issues with sslr. I ended updating my previous post some pages back after bringing it to x4fab, who warned that skip_ should be used instead for very transparent surfaces, and not as reflective. He also suggested the inner side to be set as a ksWindcreen instead to get neat droplets from driver view.
The weird thing is that diffuser is affected on both sides/objs, but it's setup differently. So until someone with the slightest idea chimes in, I fear it's just extra fx crapping out and having a bad time with that external glass piece.
Last edited:
Atrupelador: Could you please share your replay file and extra fx settings? I can't tell if it's the diffuser getting also the edges unfocused, taa + motion blur ghosting there, pp glare, or just clip compression.
I'll then try to reproduce the issue, as I can't atm.
Could be just a matter of adding the problematic stuff to MASK_GBUFFER = . Which btw, the config is using SKIP_GBUFFER on line 83 for mat:EXT_diffuser mat instead of MASK_ , and that can cause issues with sslr. I ended updating my previous post some pages back after bringing it to x4fab, who warned that skip_ should be used instead for very transparent surfaces, and not as reflective. He also suggested the inner side to be set as a ksWindcreen instead to get neat droplets from driver view.
The weird thing is that diffuser is affected on both sides/objs, but it's setup differently. So until someone with the slightest idea chimes in, I fear it's just extra fx crapping out and having a bad time with that external glass piece.
replay file wasn't saved, sorry, but here it is my fx tab
in the video tab, MSAA, Anisotropic Filtereing and FXAA, I tried every combination and simply nothing changes.
If you guys aren't seeing in your PCs, what the video shows, it may be because I'm using Camtool with extreme close ups. Could that be it?
I've just tried those settings to no use, but using a less dynamic camera could change things (I only tried spining in front of a static camera or with track cams). The chance of being an issue with tight moving fovs and those extreme extra fx settings (prob a ton of taa+motion blur+sslr) could be to blame. I can only suggest you to try first without extra fx, then enable each checkbox one by one to see what effect triggers it.
Update: Using your settings and the interior camera, I got something similar happening permanently to the inner side of the deflector and only in some exterior tv angles while moving. Extra fx Motion blur seems to cause it when enabled over taa and sslr here. This is a last resort solution, but could you backup your a3dr_honda_s800_rsc\extension\ext_config.ini, and on the lines 82/83 replace:
with:
to see if the transparent diffuser gets better from both sides when using motion blur in your clips? A fair warning, other effects will also be skipped on it (refer to this wiki and try if anything else works better for you). GL
Update: Using your settings and the interior camera, I got something similar happening permanently to the inner side of the deflector and only in some exterior tv angles while moving. Extra fx Motion blur seems to cause it when enabled over taa and sslr here. This is a last resort solution, but could you backup your a3dr_honda_s800_rsc\extension\ext_config.ini, and on the lines 82/83 replace:
Code:
[EXTRA_FX]
SKIP_GBUFFER = material:EXT_diffuser
Code:
[EXTRA_FX]
DELAYED_RENDER = Object13223, Object202
Last edited:
I will try this later! Thanks for the technical help my friend!
@flyingsaucer
Just in case my previous comments on the rear lights were obtuse, I'll add some info if it's of any help being that lights are newly scratch made - this is none but to help you improve them if you find any of these points of criticism fair:
- Before moving to the next points, please re-check if the real lights are wired differently. The outer circle could be just a completely opaque reflector (instead of a brake light) without any wiring behind it (common in 60s cars), with brakes working as just the rear light going stronger. Clip with opaque reflector close-up link1, works rsc #25 braking link2. On modern and usually very accurate s800 representations, as in GTS, the upper lamp is also used for braking link3. Please disregard this comment if the mod as presently depicted uses a modified works car as reference that doesn't match the museum #25.
- When I referred to an inadequate nm and lack of opacity, it was not just as a csp glass maskpass/thickness mult or equivalent properties concerning refraction (I'm far from versed on these), but also to the inner profile of the rear lights shapes with varying thickness not present on the nm which gives a less translucent look to these lights IRL, as csp refractions greatly depend on them for realistic looks.
Atm, you are using a nm that looks like this, just as a grid with flat squared see-through shapes in between it.
If you check the real lights, the grid shapes seem to be the thicker part of them while inside the squares there's a concave shape with varying thickness to help diffuse light, which should be nm'd to represent it. Please check clip link1 above, other yt clip close ups as img, or better, in GTS how they solved them via nm img1 img2 img3 (this is from the previous clip), hopefully showing those curves in more detail than possible IRL on aged plastics. The presence of these should drastically affect how clearly you see the lamps behind them, so you don't end with a flat piece of translucent glass.
Edit: These references might help confirm the square sizes, as they seem to fit one extra square in the narrower section between the reflector and reverse.
- Mapping/area: on real lights it looks like that these shapes should only be seen on the rear facing surface, and just a bit past the curved border (their inner depth). Atm the mapping is distorting them to wrap around the whole light (img). This could be an issue if you update the nm to a more detailed one (filling the squares), so not only the criss cross stretches in thin lines, but then you can get wide stripes in different tones.
- Regarding light diffusion on the casing: if not possible to tweak it via refraction params without blowing lamps, one solution would be to include a baked emissive (for bounced light), so the light spreads more uniformly near the reflector (see works car braking: img). If brake positions end up changed, this could contribute to make them more visible ingame while racing during day, or for a better look at night. In all fairness, except for when braking, the irl rear lights can look pretty dim and concentrated on the lamp during day (clip).
- OT: Could the large dash light on the racing car have been repurposed as low fuel pressure warning instead of a parking light? (clip)
Best regards
PS: imgur seems dead atm, if it doesn't go back online in a few hours I'll rehost/replace the broken img links. Sorry for the inconvenience.
Update: Moved img links to postimg.cc
PS2: The feedback above was left at open in this thread so anyone else can correct it, refute it or (hopefully) advice better practices. My hands-on experience with refractions is nill for proper guidance.
Just in case my previous comments on the rear lights were obtuse, I'll add some info if it's of any help being that lights are newly scratch made - this is none but to help you improve them if you find any of these points of criticism fair:
- Before moving to the next points, please re-check if the real lights are wired differently. The outer circle could be just a completely opaque reflector (instead of a brake light) without any wiring behind it (common in 60s cars), with brakes working as just the rear light going stronger. Clip with opaque reflector close-up link1, works rsc #25 braking link2. On modern and usually very accurate s800 representations, as in GTS, the upper lamp is also used for braking link3. Please disregard this comment if the mod as presently depicted uses a modified works car as reference that doesn't match the museum #25.
- When I referred to an inadequate nm and lack of opacity, it was not just as a csp glass maskpass/thickness mult or equivalent properties concerning refraction (I'm far from versed on these), but also to the inner profile of the rear lights shapes with varying thickness not present on the nm which gives a less translucent look to these lights IRL, as csp refractions greatly depend on them for realistic looks.
Atm, you are using a nm that looks like this, just as a grid with flat squared see-through shapes in between it.
If you check the real lights, the grid shapes seem to be the thicker part of them while inside the squares there's a concave shape with varying thickness to help diffuse light, which should be nm'd to represent it. Please check clip link1 above, other yt clip close ups as img, or better, in GTS how they solved them via nm img1 img2 img3 (this is from the previous clip), hopefully showing those curves in more detail than possible IRL on aged plastics. The presence of these should drastically affect how clearly you see the lamps behind them, so you don't end with a flat piece of translucent glass.
Edit: These references might help confirm the square sizes, as they seem to fit one extra square in the narrower section between the reflector and reverse.
- Mapping/area: on real lights it looks like that these shapes should only be seen on the rear facing surface, and just a bit past the curved border (their inner depth). Atm the mapping is distorting them to wrap around the whole light (img). This could be an issue if you update the nm to a more detailed one (filling the squares), so not only the criss cross stretches in thin lines, but then you can get wide stripes in different tones.
- Regarding light diffusion on the casing: if not possible to tweak it via refraction params without blowing lamps, one solution would be to include a baked emissive (for bounced light), so the light spreads more uniformly near the reflector (see works car braking: img). If brake positions end up changed, this could contribute to make them more visible ingame while racing during day, or for a better look at night. In all fairness, except for when braking, the irl rear lights can look pretty dim and concentrated on the lamp during day (clip).
- OT: Could the large dash light on the racing car have been repurposed as low fuel pressure warning instead of a parking light? (clip)
Best regards
PS: imgur seems dead atm, if it doesn't go back online in a few hours I'll rehost/replace the broken img links. Sorry for the inconvenience.
Update: Moved img links to postimg.cc
PS2: The feedback above was left at open in this thread so anyone else can correct it, refute it or (hopefully) advice better practices. My hands-on experience with refractions is nill for proper guidance.
Last edited:
awesome car! one on my favorites! just one thing though.. for me when i have RefractionFX on, the car wont open in game so i have to turn it off to use this car?
Scroll to the top of page and click the History tab.
Under the "History" tab you can download previous releases if you scroll down.
Edit: Ninjad
Last edited:
YW fella. I looked at my reply after a couple hours and thought it looked a little abrupt. Appologies for that, it wasn't my intention. I hope you arrived at a good solution to the missing in car mirror? All the best.
Latest News
-
2024 Formula One Miami Grand PrixFrom new liveries to groundbreaking team personal changes, here's everything that's gone down in...- Connor Minniss
- Updated:
- 3 min read
-
2003 NASCAR Cup Cars Scanned For iRacingDale Earnhardt Jr. Has been at it again, teasing future iRacing content that will bring those...- Thomas Harrison-Lord
- Updated:
- 2 min read
-
Automobilista 2: 2024 F1 Liveries For New F-Ultimate Gen 2 CompleteJust in time for the 2024 Miami Grand Prix, the current season's liveries for the reworked...- Yannik Haustein
- Updated:
- 2 min read
-
PISTA Motorsport: New Dev Log Announces Early Access, Confirms ModsPISTA Motorsport developer REG Simulations is back with another dev log for their project...
- Yannik Haustein
- Updated:
- 3 min read
-
World of Outlaws: Dirt Racing 24 is iRacing’s Dirt Oval SequelA sequel to the critically well-received World of Outlaws: Dirt Racing will slide into view this...
- Thomas Harrison-Lord
- Updated:
- 2 min read
-
Nordschleife Now On Console For Assetto Corsa CompetizioneAfter a relatively short wait, the Nürburgring Nordschleife can be driven on PS5 or Xbox Series...
- Thomas Harrison-Lord
- Updated:
- 2 min read
-
Fanatec Launches Second F1 Direct Drive BundleFollowing the recent extension of a licencing deal, here comes a CubSport DD and F1 wheel...
- Thomas Harrison-Lord
- Updated:
- 2 min read