Nota Bene (it is latin):
This is trial and error and you usually change the values in an opposite fashion: raise blue/ lower red (or vice-versa @KevyKevTPA this one is Latin too )
The neat thing about these registry keys is that you can see the change immediately: every time the headset sensor is detecting presence, it reloads the registry values, so that you can quickly remove HMD, change values, put back HMD and compare.
I’ve posted the above to compare and document each one experience against the 3 possible outcomes in the end: sharp, sharp but with fringes, blurry.
Let’s try to compile some feedback.
Do most of you with a G2:
- Usually need reading glasses?
- See sharp like the 2nd picture (no fringes)?
- Or like the 3rd picture (with fringes)?
[updated with the question about reading glasses]
PS: Thank you for any one participating and I hope this will in the end reveal it is not a widespread problem but just mine because in this case, I’d be more than happy sending it for RMA and not thinking I’ve a useless piece of VR equipment, especially since with the latest SteamVR beta I can run the Index comfortably at TAA100 SS124/128 (EFIS/GA) and at this level I can read small green EFIS lettering without any issue in the Index.
Closer to pic #2 for me
Yep, same for me. Picture #2 would be the visual clarity with my G2.
I’m still rummaging around on comparative tests (have to keep wife happy doing real world things!). I notice there is a thread on chromatic aberration in MSFS and the difference in 2D vs. VR settings: Here's how you turn off post processing effects! (Chromatic aberration for example)
When I look in the VR section of my UsrCfg.opt file, here’s what I find:
The OP post in the thread that I cite says that setting FRINGE to 0 removes the “artistic” chromatic aberration effect, at least in hangar scenes. What are your VR UserCfg.opt settings, @CptLucky8 ? Mine is 0.
I did my own color “calibration” image. hah.
Opening that image on paint and then using desktop view at the cliffhouse I can easily change the registry values and see how they behave in the headset. When using the default values for all color channels I get a massive blue offset from the red and green channel, but yellow channel is more or less aligned. Tip: Close one eye, move close, look at the center dot. Then move your view left and right repeatedly. You shoud see a blue halo separating from the line as the view changes from left to right.
That speaks of a huge color misaligment bug in the headset. Makes me wonder what else is fubared.
The good thing is these issues can be fixed on software updates.
Also, @CptLucky8, I don’t seem to have any of the Color Distortion registry keys you cite. I have also blithely disregarded your advice and am running Nvidia driver 461.09 with GeForce Experience(!), if that might make any difference.
@JALxml The UserCfg.opt settings are just for artistic rendering in 2D.
@JALxml You have to manually create them, they’re undocumented otherwise (thus not present)
@Altsak Thank you for this, it tells me I’m not mad yet! This blue shifted copy is what I can see and I’ve reported very early indeed: Use a G3000/1000 aircraft, look at any purple text, there is a visible blue copy of the same text overlapping which can go up to 1/2 letter size depending on distance to center.
The blue copy might not be anti-CA, but uncorrectable light leakage instead:
When you see bright text on a black background, you’re going to see several “ghost” images of that text due to light leaking across color channels on the panels. You’ll see the same thing on a Quest 2. This particular light leakage isn’t correctable since the neighboring pixels are black or near black, so there’s no light to use to compensate for the leakage.
PS: I got this one from Alex Vlachos via a WMR contact so I believe he knows about this!
The thread that I cite claims that there are TWO separate settings sections, one for 2D, one for VR, and when I look in my UserCfg.opt file, I do see two separate PostProcess sections and the one I quoted in my post above is under the VR setting. And the settings values are different for 2D vs. VR Why call it “VR” if it only applies to 2D? When in MSFS, I also seemed to find different settings for Settings, General, Graphics, depending on whether I was wearing my G2 or not (the pressure sensor in the G2 would trigger the switch between the two, so I actually set my graphics settings in MSFS while wearing my headset, which is hard to do since I can’t see any reference settings in a browser windows (maybe now that I’ve switched to windowed MSFS I can figure out how to do that).
On the undocumented, I thought you might mean hidden registry keys so I ran psexec.exe from Windows SysInternals, i.e. " psexec.exe -s -i regedit.exe" at an admin command prompt but didn’t see any hidden registry keys under the ExtendedComposition key, just to report that adventure. Same keys, same DWORD values. Since I’m not having a problem with my headset/my eyes, I don’t think I want to add the undocumented keys but I’ll try tilting the headset as you previously suggested to see if I can induce chromatic aberration (it’s a bit of an operation to set up MSFS in VR for me, uncoil the cable, move chairs, turn on USB hub switches, etc.).
I believe it is appearing in the VR section only because they didn’t fundamentally changed the load/save settings code. They’ve just provided 2 slots (2D + VR) and this saves both slots settings. It is only there but it is not used when in VR. As for the registry they are undocumented. You won’t find them in your search engine either (well now you’ll start to get some since the cat is out)
As for shifting the eye position, it is on one hand producing the blue/red fringes I’ve illustrated, but more importantly to me, it turns the white blurry text and eye focus effort to make it clearer, into sharp text surround by color fringe with no effort to see clearly (well a little bit of effort, enough to start feeling some fatigue after 1h). This is the position I’ve to put the headset in order to make it clearer, hence, the need to also compensate the fringes with the undocumented registry keys.
Here’s a simple test for chromatic aberration. I see none in normal viewing position unless I displace my G2 from my normal wearing position. The test image is a Windows 3D Builder BLACK cube with a WHITE lightning bolt on its front face and since this site doesn’t take .3MF uploads, I’ve linked to the file on my OneDrive account. In my normal G2 viewing position, I only see a clean white/black edge, say, along the border of the bolt. If I tilt the bottom of my headset out and up from my face, I see a yellow red birefringence along a border. If I tilt the top of the headset out and down from my face, I see a blue birefringence instead. But otherwise in anywhere near a normal viewing position, I see an EXCELLENT image.
OK. The link will open in either Edge or Chrome and show you the 3D cube. You can download it.
In the Sky Loft view in VR, click the Windows button on your WMR controller, when the Windows interface pops up, click All Apps on the right side, then 3D Viewer app, then from the 3D Viewer menu at the top, pick FROM FILE, and open the Cube with White Lightning. Shoot the front edge and pull the cube in 3D as close to your “eyes” as you want to view the object. When right in front of my face, I have NO CHROMATIC ABERRATION with an HP Reverb G2. I will further explore my settings but as mentioned above, I’m running the Nvidia driver 461.09.
So, in answer to @CptLucky8 , yes, if I want I can induce chromatic aberration in my G2 but normally I have as close to zero as I can tell. The CA I get by tilting my headset off a proper face mount as above is obnoxiously obvious.
One thought where chromatic aberration comes from in “normal” viewing is just that the sufferer’s face has a facial tissue distribution such that the vertical plane of the mask goggles is tilted relative to what the goggle designer intended (or perhaps a similar defect in the construction/assembly of a particular individual headset so the plane of the goggle lenses isn’t right?). Maybe goggle designers need to include an up/down forward tilt mechanism to modestly adjust the viewing angle? I remember some poster remarked that he/she got the sharpest image with the G2 if the bottom of the mask was tilted slightly away from his face.
I guess there are at least some advantages to being confined to a hospital bed. No moving furniture, no uncoiling cable, no turning on hub switches (I do have and use one, but only for one port for my primary flight stick, and it’s always attached/on anyway), etc.
Took me awhile, but I found at least these advantages!!
Eyesight based calibration would be nice, don’t like having to wear contacts or glasses with a headset on. I’m quite pleased with the performance and graphics settings I’ve achieved but the smallish sweet spot strains the eyes, feel like my retina will detach if I fly in VR for too long. Having depth perception on taxi and landing is fantastic but flying on my Ultrawide monitor is overall a much more enjoyable experience.
Just noticed that these values change the color position in the Z direction in the image plane. Does anyone know if there’s a coefficient that would offset the colors in the X,Y plane? Just to check things out.
@Altsak the change is in a form of a pincushion/barrel distortion of the individual colour planes, therefore they change size in both directions on the plane perpendicular to your eye sight.
@JALxml Thank you for the additional reports!
Just in case I wasn’t clear in my explanations, when I wear the G2 as it is supposed to, I don’t have any CA either, but in this case the center disk of clarity is not sharp and it requires some efforts to focusing the image. When I wear the G2 purportedly “displaced” to show CA, I get a sharp image in the center, albeit with additional CA, and for which I must use the registry keys to compensate and reduce the fringes. I believe the very wearing displaced makes the light passing though the lenses with a slightly different focal length due to the curvature of the lens and therefore helping focusing farther away.
NB: I can get sharper image in rotating the headset nose-out / eyebrows-in (and in reverse) but with quite a pronounced angle and in this case, this creates CA too in the same way when I’m shifting slightly above/below.
So for now with the reports in this discussion and my experiments, I can only conclude:
I have developed a new eye sight problem and I’ll need to wait I can get my vision checked. This is unlikely because I would see other problems outside VR with just my reading glasses everyday, and in any case, I can see clearly with the Index for many hours in a row (let alone it has edge-to-edge clarity up to about 85% off the center!)
My G2 is wrongly calibrated and the way I’m wearing it “shifted” is in fact where I should be wearing it, thus the additional manual calibration via the registry keys I must do which would have been already set for me otherwise.
My G2 is wrongly assembled and the way I’m wearing it “shifted” is just a workaround but it will never be as clear as expected even with the registry keys correction.
PS: no surgery, no alterations, just age induced need for reading glasses.
My G2 is wrongly assembled
Then HP has a bigger issue because a lot of the headsets are wrongly assembled.
The next thing we’ll need a brave soul who’s willing to open up his/her headset and see if there’s anything out of the ordinary. There’s a chance one could swap the lenses around to see if that has any effect on the issue.
Whilst it doesn’t bring much to this conversation, I have certainly noticed that I feel the need to push the “nose/cheekbone” bottom edge of the headset into my face fairly firmly when putting it on in order to get the best image, so maybe this is a result of the hinge mechanism.
Perhaps HP should have put some kind of catch/lock to ensure the angle of displays/lenses is optimal once in use?
You might want to follow this:
And for those wondering why they are using Fresnel lenses (comment from Valve employee):
Thanks @CptLucky8 - both very interesting reads!
And interesting to note that different people have very different perception and nauseating effects, etc. with different lenses, so the manufacturers have to go for the “broad middle ground” which isn’t necessarily what they might consider the best image they can produce, even when it’s not a commercial decision! (which is what most people, myself included, think that these things normally boil down to)