Cockpit bezel lines

[mbmorgan 0]

Folks,The problem that you are seeing is probably not aliasing. If it were then all straight lines on the panel should exhibit the same "staircase" effect whenever not perfectly vertical or horizontal. In fact, looking closely at the original images from the first post in this thread, what I see (with help from a little extra magnification) are perfectly straight lines with the "staircase" effect showing on tiny surfaces along the intersection between bezel faces and frames that are roughly 45 degrees to those faces and frames.I've noticed that, when I fly the JS41 (a fine aircraft, by the way ... thanks, PMDG), the same bezel edges that you are concerned about appear at least partially transparent ... so that I can see external objects "flickering" past through the bezel edges (and a few other edges as well) as though they are tiny little slit windows through the panel to the outside world.Now, I'm not a VC developer, but if a VC is anything like the old 2D panels, there is likely to be a bit mask (or layer) that controls whether an unpainted pixel is either transparent or opaque. I'd suggest that the JS41 panel bit mask that should be controling default transparent/opaque pixels just might be either missing ... or perhaps transparent everywhere so that any unpainted pixels in the VC ... like where bezel faces don't precisely join bezel frames (along those unfortunate "squiggly" edges) allow the outside world to show through. Regards,Mike

[jordanal 70]

Folks,The problem that you are seeing is probably not aliasing. If it were then all straight lines on the panel should exhibit the same "staircase" effect whenever not perfectly vertical or horizontal. In fact, looking closely at the original images from the first post in this thread, what I see (with help from a little extra magnification) are perfectly straight lines with the "staircase" effect showing on tiny surfaces along the intersection between bezel faces and frames that are roughly 45 degrees to those faces and frames.I've noticed that, when I fly the JS41 (a fine aircraft, by the way ... thanks, PMDG), the same bezel edges that you are concerned about appear at least partially transparent ... so that I can see external objects "flickering" past through the bezel edges (and a few other edges as well) as though they are tiny little slit windows through the panel to the outside world.Now, I'm not a VC developer, but if a VC is anything like the old 2D panels, there is likely to be a bit mask (or layer) that controls whether an unpainted pixel is either transparent or opaque. I'd suggest that the JS41 panel bit mask that should be controling default transparent/opaque pixels just might be either missing ... or perhaps transparent everywhere so that any unpainted pixels in the VC ... like where bezel faces don't precisely join bezel frames (along those unfortunate "squiggly" edges) allow the outside world to show through. Regards,Mike

Ah, that's how you spell squiglly, LOL... Yes, I think you are probably very correct in your analysis. I get no AA issues anywhere else on this or any other aircraft. It's just those bezel lines and now that I think about it, they do seem transparent (white), when they shimmy. Extremely noticeable when using my TrackIR5 headset and quickly panning around the VC. Good job, Mike. Thanks for the more in-depth analysis. I hope PMDG can do something about this.

[davidzill 8]

Ah, that's how you spell squiglly, LOL... Yes, I think you are probably very correct in your analysis. I get no AA issues anywhere else on this or any other aircraft. It's just those bezel lines and now that I think about it, they do seem transparent (white), when they shimmy. Extremely noticeable when using my TrackIR5 headset and quickly panning around the VC. Good job, Mike. Thanks for the more in-depth analysis. I hope PMDG can do something about this.

I brought that up in response # 3 or something like that. I believe you are right on. When you zoom in or focus on the lines, it appears as if it is actually a gap in the textures and you can see straight through them. I think it is something worth PMDG developers taking a look at. It is not aliasing because when I manually set my antialiasing to 20X it still has no effect and the bezel lines, but other cockpit edges become ultra smooth.

[mbmorgan 0]

David,You did indeed bring up the possibility of little gaps in an earlier response ... and that is precisely what got me thinking about what I've been seeing and about the possibility of a faulty or missing underlying bit mask (or VC equivalent). My hope is that one of the VC developers just might have one of those Homer Simpson "Doh!" moments, then quietly and easily go correct the underlying mask to be opaque in all the right spots so that those little VC gaps show nothing at all.Mike

[davidzill 8]

I just did a quick test flight at night out of mojave. Against pitch black sky and terrain, the lines are hardly noticable at all. during the day when I prefer the aircraft, i see nothing but light colored transparent like lines. i think the panel mask is a safe bet.

[downscc 1,327]

Against pitch black sky and terrain, the lines are hardly noticable at all

The light lines you are seeing are highlights, and without light they are not noticeable. I don't think the "masking" theory holds up to logic, because not very many users have expressed this as a problem to the degree that you have the problem. My panel looks great in the noon sun.You're not using DX10 for internal shadowing are you? I haven't read anything in this thread yet that convinces my this isn't a "system" problem. I don't know the solution, not having the problem myself:

[jordanal 70]

But Dan, you are running an ATI GPU. Could it be a development issue that rears its ugly head more with nVidia cards?

[Tabs 2,790]

Guys, there is a legitimate issue with the beveling here - it's not gaps/masks or anything, just very small details in the model's edging that don't seem to work out right. (this very well may have worked correctly on older drivers when we made the plane, which explains why it wasn't an issue in beta etc)Here's a shot of it from 3DSMax:If there's an SP2, this will get fixed. I can't guarantee that there will be or that it will be soon, we have a lot on our plate right now with getting the NGX off the ground. For the record, no one has said that the J41 will not receive another update, I'm not sure how that claim got started.I must say though that I find the claim that a little bit of aliasing on the panel makes the plane "unflyable" rather dubious. I see the exact same issue here and while it'd be nice to have it gone, it certainly does not make the aircraft any less flyable to me.

[jordanal 70]

Thanks for the upfront info Ryan. My first thought is, I'm glad this is discovered now as I believe the JS41 was somewhat of a technology platform testbed for the NGX 2.0 which should now not have any of the bezel 'squiggly' lines. :(I personaly find the JS41 a challenge and I am thoroughly enjoying it and can live with these lines for now. Although just not very attractive, it is not stopping it from being a great educational experience and a nice break from my usual big-iron flights. I just hope the possible release of a JS41 sp2 patch doesn't come after the NGX 2.0 as by that point, I will have probably moved on to that new plane, thereby getting most of my attention.

[davidzill 8]

That is great we got an answer to the problem. I guess I will continue to fly the MD-11 before the service pack 2 comes out.

[albar965 1,047]

Thanks for the feedback, Ryan.Certainly more helpful than nitpicking on words. :( I would love to see this fixed in a SP2.Alex

[vscimone 16]

Certainly more helpful than nitpicking on words.

Yeah.. the debate about the use of a word gets a little ridiculous.Oh... and it's "Bevel" :(

[mbmorgan 0]

Technically, the bevel in question is on the bezel ... where the face joins the frame (this is finally getting fun).

[Zachiii 0]

My bezel lines are gone for night flights.In day time operations my bezel lines are gone when J41 is rock steady, but reappearing with movement from steady and magnify with increase in movement.I think we need a PMDG fix.

[WhiteKnight77 0]

Technically, the bevel in question is on the bezel ... where the face joins the frame (this is finally getting fun).

I was gonna say the very same thing.