I will test with the humans and other planes and vehicles close to me.
I have been using VR for several years, and my IPD is set to the value given to me by my Optometrist when he measured me for glasses. I can assert that as far as I am concerned, the world scale issue is real when using an original HTC Vive, and changing IPD does not fix it.
In VR try the 152 cockpit view. Look at the yoke in front of you, look at the seat under you and you realize they are sized for children. That can’t be right. Then take place in the Beechcraft Baron and admire the yoke there: It feels right.
I think the IPD discussion is not very helpful here. What is needed is an adjustment tool for scale. Everyone has their own preferences.
It is very important…like !!! MAXIMUM ATENTION !!! not to confuse the in headset IPD with the in game IPD.
This thing should not be read diagonally. It’s like the difference between assets and liabilities… what is an asset for the bank… it’s a liability for the client.
The IPD in headset is a setting that makes the two rendered images for the two eyes be at a certain distance.
The IPD in game is the setting that establishes the distance between the TWO VIRTUAL CAMERAS that take those two images in the definition above!
Let me give you an example:
Suppose you play a VR game where you are a Gulliver in the midgets land.
The IPD in game… which is a hidden setting usually is working on the other part and mimics the IPD of the Gulliver which is as we can imagine GIGANTIC compared with the one of a midget human from the story.
Once the IPD of Gulliver in the game is set correctly… you… the human wearing the headset inside VR with the correct IPD from Optometrist set in the Quest2 or Index or whatever set you may have will see the world at a different scale (not only because of the new height of Gulliver but also because the distance between the eyes of that personage in that world will make the said world smaller. Because you will heave a huge head now. Say Gulliver has a head of one meter in diameter… you can imagine the eyes are not just 6-7cm apart like for a normal human.
Basically, you can understand that to portray correctly Gulliver’s view you need not only to set the height of his head correctly but also the distance between his eyes so you have the “whole picture” sic of his grandeur .
Why are poeple blaming the headset and IPD. Fixing these provides a uniform solution to a problem which is anything but uniform - you can clearly see a different scale being used for different aircraft more often than not it’s too small.
A scale slider isn’t a solution it’s a band aid and may not even be required unless scale is interpretated differently in different headsets. This problem needs to be thrown back to Asobo to make the scale consistent across ALL aircraft.
This is a good explanation, I like the Gulliver analogy!
There is more to it:
@sneaky74 regardless of whether there are inconsistencies in the actual scale of the modeled aircraft, there are definitely inconsistencies between the OpenXR API reported IPD which FS2020 uses for matching a software ICD (actually they shouldn’t do much because OpenXR is giving back the projection matrices directly), and this depends on the headset.
It could be WMR OpenXR error/shortcoming, WMR headset only differences from SteamVR headsets, or even differences in WMR OpenXR and SteamVR OpenXR implementation of the standard.
One thing for sure: there is a bug in WMR for SteamVR today regarding IPD. In effect, it is rounding the IPD to the 1st decimal and we hope this will be solved anytime soon. You might think a precision of 1mm should be enough but in practice it can make the perceived scale quite off enough.
Ah I see @CptLucky8 thanks for clarifying that. So not the users IPD but the interpretation of that on screen is inconsistent. Ok that makes sense. Gosh nothing is simple is it! So that needs to be fixed as well as the inconsistent scale modelling!
Please do understand that there are two IPDs. As I said. Don’t read diagonally that post .
Also… yes… if a plane is twice as big as another yet in the game is quadrupled in size… yes… there is a second problem.
Thank you for your post and link. I missed it altogether!
Just to be clear, FS2020 code is just asking OpenXR: give me the projection matrices for each eyes.
The issue is then the OpenXR driver building those projection matrices from the hardware IPD if any, or the software IPD, and this is where there could be differences. For example:
- WMR Driver running the G2 headset is reporting 64.0 while the lenses are physically 63.0, and is preparing proj. mat. for a wider separation than the lenses with a focal point set to 20m forward (let’s say this is WMR fixed point)
- SteamVR Driver is reporting 64.0 while the lenses are physically 65.0, and is preparing proj. mat. for a narrower separation than the lenses with a focal point set to 25m forward (let’s say this is SteamVR fixed point).
Then FS2020 is rendering each view with their projection matrices, and is sending the rendered view into the API. There the might also be “translating” the rendered image based on the IPD. It shouldn’t with headset with physical IPD adjustment but it should do this for headset with software only adjustment (G1). And there are other headsets with only 3 IPD positions so you’re most likely always above or below.
Add on top you’re using OpenXR <-> WMR bridge or OpenXR <-> SteamVR bridge, that is the OpenXR API specs are translated into the headset driver own API specs. This is at this stage there could also be rounding errors and even 1/10th of a mm difference is enough to see the problem.
You can see for everything to work, the 4 parameters I’ve listed must match perfectly, and they will never match in any case because you’re eyes IPD is varying depending on the vergence (you’d need eye tracking and auto-IPD adjustment like with the Varjo for this to work properly).
All this long text just to further confirm there is a need for a software IPD bias setting, both absolute and relative adjustments, so that you can transform the ICD in the software with this formula:
ICD = (IPD + Bias) * Scale
Bias adjusts for the differences between the lenses IPD and your eyes IPD, whereas Scale could adjusts for the differences between the software IPD and the software ICD.
The percieved scale of a stereoscopic rendered object isn’t only determined by stereo (camera) separation but also determined by convergence (cameras not aligned in parallel). Google EarthVR is an example what can be achieved with auto convergence. Low altitudes give real life scale whereas higher altitudes give toyification effect.
You can also change between Human/Giant mode in their app settings. (I don’t remember the exact setting name)
I got my HP Reverb G2 today and have just briefly played around with it. Perhaps a change in scale implicitly implies some sort of change in focal length but my main impression in only an hour or two of flying around is that the scale change is akin to having the drone camera in 2D MSFS in a more telephoto mode. Distances appear foreshortened and things farther away appear relatively larger as compared to closer objects than they would at a normal 2D viewing focal length. Maybe I’ll grow tired of the effect after a while but for now I like it because the 3D and the larger scale makes it easier for me to recognize streets and buildings on the ground when I fly over them in my home town. But it would be great to be able to tune VR scale/focal length.
Sure something in FS2020 is different to other simulator titles as far as scale is concerned. MSFS is the only title where I hear other players consistently complain about scale being an issue. I never hear these issues from the same people when it comes to DCS, Aerofly or IL2. MSFS is doing something different. I am probably one of the lucky ones where I dont have an issue with scale but many of my fellow simmers have issues with the same headsets (G2’s). The only aircraft I have an issue with as far as scale goes is the C172 as it feels much narrower than the real thing of which I have over 300hours flying time in.
I have a IPD of 65. If I change it to 61 on my headset (OG VIVE) the world becomes bigger, although still smaller than it should be…I can just look at the stick…I would have to fly the plane with just the tip of my fingers…scale is off and that is a fact.
Also, I should not need to adjust my ipd on my headset…yes, I can make the world a bit bigger but also gives me a headache as my brain has to compensate for a wrong distance between my eyes…
It’s not that hard. Just fire up DCS and on the VR setting there is a tiny slider that adjusts world scale…Done!! Everyone happy. If it feels alright to you, don’t touch it. If it feels weird, just adjust it until it feels right. For me I always jump on the A10 and compare the stick to my warthog…when they are the same size I know it is the right setting!
Please implement this as it is very immersion breaking!
These sorts of things have also been discussed in some of the posts above.
Ok, solution found!!
You can use fpsVR program (steam $3) to change your ipd through software.
I have an OG VIVE, kept my physical ipd (65mm) and just changed the ipd offset with fpsVR. You can do this while the sim is running, just need to restart steamVR after each change but not the sim. In my case after testing I needed an offset of -20. Massive change, finally a yoke that a human can use.
Just remember that aircraft are not scaled correctly, meaning, you will have to adjust the offset depending on the aircraft, some are smaller and some are bigger (ex: A320 has a massive cockpit, 747 not so much).
It solves the scale problem but still need ASOBO to do something as I shouldn’t have to pay and use a 3rd party software to solve a msfs problem.
I just now finally tried it on a Quest (1), and everything feels way to small. IPD is adjustable and it feels the same regardless of IPD adjustment. It’s as if I’m stuffed in a slightly scaled down model world.
Elite Dangerous also feels smallish (but not as bad). The only game so far that feels mostly right is No Man’s Sky.
Have you tried Half Life Alyx?
As for FS2020:
[BUG/FEATURE] Cockpit Size and World Scale in VR
I have not tried HLA. Is that on the good or bad side of the scale?