[GUIDE] Finding the Mathematically Correct FOV & Multimon Settings For ANY Monitor Setup
-
- Posts: 143
- Joined: June 14th, 2021, 2:11 am
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
@BenganJ
Thanks, I'll look into it.
Thanks, I'll look into it.
-
- Posts: 143
- Joined: June 14th, 2021, 2:11 am
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
Well, I dug around a little, but it was difficult for me to find information about how to even make a plugin for this game. I read a bit about the telemetry SDK, and the only data that seems useful is camera position, if I want to adjust the FOV based on that. Although from what I've seen, it only outputs the data, it's not a way to change the data. I also tested the "multimon set" console command, which seems promising, as it immediately updates the view as soon as I enter the command. I can execute multiple commands at once through a script and the "exec" command, which also works in my testing. However, I doubt that I can do programming things within the script (variables, math, logic) or trigger it using an input.
So my conclusion is, if my idea is even feasible, it will most likely have to be coded into a plugin. I probably won't be the one to do it, as I don't have any modding experience and don't know C++. If anyone is willing to try though, I will give you my full support.
So my conclusion is, if my idea is even feasible, it will most likely have to be coded into a plugin. I probably won't be the one to do it, as I don't have any modding experience and don't know C++. If anyone is willing to try though, I will give you my full support.
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
@a1337cookie
Yes, sadly that is what I suspected. I think SCS is deadly scared of people changing game engine
parameter values in real time, hence ONLY outputs to be used to control external things! But your
idea is still brilliant!
Yes, sadly that is what I suspected. I think SCS is deadly scared of people changing game engine
parameter values in real time, hence ONLY outputs to be used to control external things! But your
idea is still brilliant!
-
- Posts: 143
- Joined: June 14th, 2021, 2:11 am
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
Update 5: Added prerequisites section and some info about vertical FOV. Various formatting and wording tweaks. More miscellaneous info.
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
This is a fantastic resource, thanks heaps for putting it together!
I recently upgraded my setup from a desk to a full sim rig, going from triple 24" to triple 32" monitors and getting the multimon settings tweaked correctly has made a huge difference to the over all experience. But because I hate keeping things simple, I am now experimenting with bezel lenses - specifically the Asus bezel free kit. To ensure the image lines up correctly through the lenses, I need to adjust for have a negative bezel adjustment rather than a positive. So rather than hiding the image covered by the bezels, I am instead stretching it on either side so the lenses blend the image over the gap.
I'm not too sure how to modify the formula to account for this but I plan to do some experimenting over the weekend and will post my results if I get any success. However, I'd also be grateful for any thoughts or suggestions on how to proceed correctly!
I recently upgraded my setup from a desk to a full sim rig, going from triple 24" to triple 32" monitors and getting the multimon settings tweaked correctly has made a huge difference to the over all experience. But because I hate keeping things simple, I am now experimenting with bezel lenses - specifically the Asus bezel free kit. To ensure the image lines up correctly through the lenses, I need to adjust for have a negative bezel adjustment rather than a positive. So rather than hiding the image covered by the bezels, I am instead stretching it on either side so the lenses blend the image over the gap.
I'm not too sure how to modify the formula to account for this but I plan to do some experimenting over the weekend and will post my results if I get any success. However, I'd also be grateful for any thoughts or suggestions on how to proceed correctly!
-
- Posts: 143
- Joined: June 14th, 2021, 2:11 am
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
@Foxman
Hey, glad to see the guide getting some use!
The bezel free kit is interesting, it does complicate things indeed. If it works how I think it works, there's probably no way to make it look mathematically perfect for three monitors. This is my thought process.
Here's a diagram of two adjacent monitors. They would presumably be angled, but flat is easier to visualize, and the idea should be the same.
The grid represents the screen image. The bezels are half a grid unit all around.
Now, applying the bezel free kit.
I assume it stretches the images on each side over, so that it lines up in the middle.
To correct for the stretching, the area the kit covers would have to be squished.
Then, when stretched out, it will look correct.
The problem is squishing just a portion of the view. It is totally achievable using the multimon file, but you would have to create a new monitor view for each section. That means, for two adjacent monitors, there would be four monitor views in the multimon file. If there are three monitors, the number of views goes up to seven. As far as I know, you can only have a maximum of four views, so I don't think that will work.
If you don't mind being slightly off mathematically, and just want the views to line up, the bezel correction between monitors can be removed, I can show you how if you want. I don't think you need negative bezel correction, but of course, I could be wrong about how the kit works.
Hey, glad to see the guide getting some use!
The bezel free kit is interesting, it does complicate things indeed. If it works how I think it works, there's probably no way to make it look mathematically perfect for three monitors. This is my thought process.
Here's a diagram of two adjacent monitors. They would presumably be angled, but flat is easier to visualize, and the idea should be the same.
The grid represents the screen image. The bezels are half a grid unit all around.
Now, applying the bezel free kit.
I assume it stretches the images on each side over, so that it lines up in the middle.
To correct for the stretching, the area the kit covers would have to be squished.
Then, when stretched out, it will look correct.
The problem is squishing just a portion of the view. It is totally achievable using the multimon file, but you would have to create a new monitor view for each section. That means, for two adjacent monitors, there would be four monitor views in the multimon file. If there are three monitors, the number of views goes up to seven. As far as I know, you can only have a maximum of four views, so I don't think that will work.
If you don't mind being slightly off mathematically, and just want the views to line up, the bezel correction between monitors can be removed, I can show you how if you want. I don't think you need negative bezel correction, but of course, I could be wrong about how the kit works.
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
Thanks for the reply mate, appreciate it. To be honest, I don't fully understand the way the lenses work but your interpretation is more or less the way I understand it as well. Might as well be magic as far as I'm concerned but they definitely make a difference. I've been experimenting with Assetto Corsa to begin with since it has the ability to adjust monitor settings via the UI so it's quick and easy to get results. Here are a couple of example images:
Positive Bezel adjustment for use without lenses.
Negative Bezel adjustment.
The easiest place to see the difference is the text in rear view mirror top right. With normal 'positive' adjustment, the portion of the image that would be hidden by the bezel shows a clear mismatch in the image. Applying a 'negative' adjustment allows the text to line up more or less correctly. It's not 100% perfect because of slight variations in the angle and position of my left and right monitors relative to each other and my not quite perfect fitting of the lenses so I've sort of 'split the difference' to get the image as close as possible on both sides.
So the challenge now is to replicate that as closely as I can in ATS/ETS2. You're right about there being a limit on the number of views available in the multimon settings so I don't think it's going to be possible to get it perfect but I'm ok with 'good enough'. So removing the bezel adjustment is probably the best option for the time being. Worst case scenario, I drop the lenses and go back to the calculated multimon settings with the bezels intact.
Edit: A short video showing ATS in action with the bezel kits and a little tweaking of the multimon file. The distortion that's visible is slightly more exaggerated than what I see head on due to the angle of the camera but overa all, the effect is pretty good so if I can fine tune it a little further then I'll be a pretty happy camper.
Positive Bezel adjustment for use without lenses.
Negative Bezel adjustment.
The easiest place to see the difference is the text in rear view mirror top right. With normal 'positive' adjustment, the portion of the image that would be hidden by the bezel shows a clear mismatch in the image. Applying a 'negative' adjustment allows the text to line up more or less correctly. It's not 100% perfect because of slight variations in the angle and position of my left and right monitors relative to each other and my not quite perfect fitting of the lenses so I've sort of 'split the difference' to get the image as close as possible on both sides.
So the challenge now is to replicate that as closely as I can in ATS/ETS2. You're right about there being a limit on the number of views available in the multimon settings so I don't think it's going to be possible to get it perfect but I'm ok with 'good enough'. So removing the bezel adjustment is probably the best option for the time being. Worst case scenario, I drop the lenses and go back to the calculated multimon settings with the bezels intact.
Edit: A short video showing ATS in action with the bezel kits and a little tweaking of the multimon file. The distortion that's visible is slightly more exaggerated than what I see head on due to the angle of the camera but overa all, the effect is pretty good so if I can fine tune it a little further then I'll be a pretty happy camper.
-
- Posts: 143
- Joined: June 14th, 2021, 2:11 am
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
@Foxman
Looking at the images, I'm guessing that the way the bezel free kit looks is determined by the angle of your monitors and the width of the bezels. So with your setup, it seems to sort of "overcorrect" for the bezels, which is why a negative bezel correction setting looks correct. It's not a bad thing, since it ensures the bezel is not shown.
For ATS, it would probably be best to shift the side views towards the center view until the image lines up (which is how Assetto Corsa does it too, I believe). How to do this depends on how your multimon file is set up; frustum_subrect_x is the main thing to edit. If you want, let me know what settings you have now or which procedure you followed, and I can tell you exactly how to make the adjustments.
Side note, I can't help but notice that your settings in Assetto Corsa seem strange. Are your monitors actually at 60°? Assuming the screen width and angle are correct, your view distance seems quite far away. Do you sit that far away for ATS too? Just curious, haha.
Looking at the images, I'm guessing that the way the bezel free kit looks is determined by the angle of your monitors and the width of the bezels. So with your setup, it seems to sort of "overcorrect" for the bezels, which is why a negative bezel correction setting looks correct. It's not a bad thing, since it ensures the bezel is not shown.
For ATS, it would probably be best to shift the side views towards the center view until the image lines up (which is how Assetto Corsa does it too, I believe). How to do this depends on how your multimon file is set up; frustum_subrect_x is the main thing to edit. If you want, let me know what settings you have now or which procedure you followed, and I can tell you exactly how to make the adjustments.
Side note, I can't help but notice that your settings in Assetto Corsa seem strange. Are your monitors actually at 60°? Assuming the screen width and angle are correct, your view distance seems quite far away. Do you sit that far away for ATS too? Just curious, haha.
Re: Finding the Mathematically Correct FOV For ANY Monitor Setup
If you're willing to take a stab at it, I'd be hugely grateful! These are the settings I had in the multimon file prior to adding the lenses:
I used the formula for 3 perpendicular screens. Monitor width with Bezel is 714mm, without is 700mm.
As far as the angle of the monitors, they're now actually fixed at 50 degrees but I have been repositioning them so the angle at the time I took those images was somewhere around 62 degrees or so. I sit slightly closer for trucking, 720mm off the top of my head; the exact distance for a given game varies as I move the pedals and/or seat angle to suit the driving style (GT, F1, Truck etc) but the position is consistent each time even if different across games. I'm a bit odd perhaps haha.
Code: Select all
frustum_subrect_x: 0.009804
frustum_subrect_y: 0.009804
frustum_subrect_width: 0.980392
frustum_subrect_height: 0.980392
As far as the angle of the monitors, they're now actually fixed at 50 degrees but I have been repositioning them so the angle at the time I took those images was somewhere around 62 degrees or so. I sit slightly closer for trucking, 720mm off the top of my head; the exact distance for a given game varies as I move the pedals and/or seat angle to suit the driving style (GT, F1, Truck etc) but the position is consistent each time even if different across games. I'm a bit odd perhaps haha.