Kegetys

Are you using the demo from the latest SDK version? Under samples\OculusWorldDemo\Release\OculusWorldDemo.exe. (Oculus website seems down now but I downloaded mine just a few days ago and I dont think there has been any updates to the public SDK for a while).

The difference is visible here (DK1): http://junk.kegetys.fi/oculus_dcs.jpg (DCS beta with the new visualizer.dll) http://junk.kegetys.fi/oculus_tuscany.jpg (Oculus SDK Tuscany demo) There is a large black border in the DCS image - It looks like the required scale factor is not applied to the render target, or the warping parameters are not right. edit: I had a look with GPA and it looks like DCS is using a 1280x800 rendertarget which is not right - The needed scale factor for DK1 is something like 1.7 so it should be around 2176x1360, I guess that is the problem.

Running the latest beta.

Those terms say exactly the opposite: This is directed to Beczl personally - He is obliged by those terms to return our money because he is (apparentely) not able to deliver anything he promised. Leatherneck is not involved in that in any way it seems, and them honoring the preorders done to Beczl is absolutely great. But I would still much rather get my money back from the scam that Beczl ran with that Indiegogo campaign and give them to who actually delivers the product.

Back when I got my Warthog I made a C++ library for controlling those lights if anyone is interested, its available here: http://junk.kegetys.fi/pahkasikaLED.zip (It uses target DLLs). I havent tried if it still works with newer(?) target though.

Well, I guess you could remove the lights. But maybe easier is to just turn them off :) I doubt modifying the shaders would help that much except in special cases. Modifying models so they would be better optimized (proper lod levels, optimized for minimal draw calls) would maybe help but I guess that is a huge task. Better to just wait for EDGE maybe.

I believe this happens because of the way lights are implemented, all geometry lit by the light appears to be drawn for a second time (or more if there is more lights). That combined with badly optimized models results in huge amount of draw calls and everything is bottlenecked by the cpu not being able to push draw calls to the gpu fast enough. Its especially a problem with D3D 9 where every draw call is relatively expensive.

I have thought about something like this as well and I think it would be pretty easy to do: - When mouse is moved physically, do a collision check to the cockpit geometry from the new cursor coordinates. Store nearest intersection point as 3D cursor coordinates in model space. - At every frame, project the 3D coordinates back to screen space and move mouse cursor there. Only issue should be to do the steps in correct place and how to deal with the cursor going outside the screen.

I am willing to do it and I did it for the MiG-21Bis but not the F-35. The MiG-21Bis is among my favourite planes and when the Indiegogo funding for it opened Beczl already had a lot to show. From the screenshots/videos and other material I could already see the project is going well and I felt I could be reasonably confident its going to be finished, and likely to meet the standards I expect. The F-35 I dont really find that interesting as a platform. That doesnt mean I wouldn't buy one as a DCS module but I just dont find it sufficiently interesting to warrant the risk of "investing". Also I felt the F-35 kickstarter really had almost nothing to show - some renders of the exterior model and thats about it. They most definately should have first had something already in-game, including some form of cockpit with maybe a demonstration of some basic avionics stuff working and some kind of flight model before opening the Kickstarter campaign.

Something like Dangerous Waters would be great. It does a great job with a rather simplified interface to the systems (in every platform the sonar etc. work basically the same, you just have a different look/feel to the interface) but still has realistic simulation of all the actual mechanics like sound propagation in water. If there would be a naval module added to DCS, I really hope it would be done at least at the Dangerous Waters level or close to it instead of simplifing everything too much like imo. Combined Arms seems to do at the moment.

is quite interesting hour-long interview as well.

Using the DLL works fine, I "reverse-engineered" it a some time ago. Here is a simple C++ class for using it in case anyone is interested. I havent used it for a while so I dont know if it works with newer drivers etc though... And use at your own risk of course :)

There is an (afaik. undocumented) API for controlling the leds with TmHidControl.dll, so they for example could be used in the lua export script in DCS to react to events... But they're in a place where they are a bit hard to see so I havent found any real use for them.

So is the Su-27SM mentioned in one of the newsletters this FC-level AFM Su-27 or is that another (third party?) project? If its the same thing then I guess there will be at least some new avionics stuff for the mfd displays.

Something like Dangerous Waters would be great, I think it did the simplification of the station interfaces very well (All platforms essentially were the same, but had a different looking interface and different capabilities so they felt different). Co-op with helo, the perry and some subs was great fun, too bad it didnt do very well financially apparently.

I had mine made as two ring bound manuals and I think it works very well. The manual splits nicely between "cockpit controls" and "startup procedures" (both have their own indexes for mine and also some added appendices, I tried to fix the page numbers automatically for the second part also but failed). The two-split just makes it easier to handle in my opinion, you really just need one at a time anyway and the smaller size is then better than one huge book. My only nag with lulu ring binding is that the cover & back are not very thick, it makes the whole manual a bit floppy. Thick cardboard like in the official Ka-50 manual would be better.

Anyone can create those books in lulu, just upload a pdf and make a cover for it and you can buy it yourself. I made a coil bound manual for myself that way (split into two parts) with cover images in theme of the Ka-50 manual, it cost less than 30 euros with shipping using some discount code I found from the internet. I cant offer it for sale though since I dont have permission for the manual or images used in the covers. Hardest part was splitting the PDF into two parts, amazing how hard it can be to do the simpliest of things in Adobe Acrobat :book: :)

From config/graphics.cfg change PrecompiledEffects mode to "USE_PRECOMPILED_EFFECTS_FOR_UNCHANGED_FILES" (Its already there, but commented out) and it should work.

I have Sennheiser HD555 (modded to HD595) with a 5e desktop microphone attached to them with trackclip pro's clip. Works very well, though the three wires makes a bit of a thick cable :)

That F-15C cockpit looks really good, but I do find the price to be too high for me since its otherwise the same old almost-10-year-old stuff again. I would love to get to fly the Lomac planes in DCSW but as I have already paid for them many times it doesnt seem worth it to pay that much yet again just for the compatibility (albeit with some welcome improvements, but not enough to justify the price for me). I am mostly interested about the russian platforms anyway and I can imagine the feeling of testing that nice F15 cockpit and then going into the 10 year old Su27 pit again. I dont know if I would still have much enthusiasm to really fly it that much. I might still buy it some day especially if the price drops, but not really that excited about the announcement.

If the lat/lon conversion works right, it should only happen when flying the Ka-50. In A10C you should see the "old" style coordinates, though I didnt test this through if it actually works as it should.

Nice, it seems to work. It also seems to be quite easy to modify the JTAC communications so that when flying the Ka50 it gives the coordinates to you directly in latitude/longitude. With minimal effort I already got it to say "north 42.6943212 east 41.6683924" by modifying 'Sounds\speech\nato.lua'. Maybe modifying ussr.lua would make it work on russian side as well. Also if you copy around some of the differently named sounds under 'Sounds\Speech\Sound\ENG\DCS Ka-50\Player' so it matches 'Sounds\Speech\Sound\ENG\Common\Player' (many are missing but you can substitute some other sounds) you can hear your checkin as well. The errors of what files are missing come up in the error log/debug output.

This is not entirely true - You can do a wide variety of different projections, what you see here is the result of rectilinear projection. The reason why this is pretty much always used today is because todays 3D accelerators can only "directly" render triangles with straight edges. When restricted to straight lines only, rectilinear projection is the only sane option. With software rendering, (or by some special "tricks"*) you can do other projections as well. See for example Fisheye quake: http://strlen.com/gfxengine/fisheyequake/compare.html * With hardware acceleration, you could render the scene into a cube map and then sample from that to create the final image. But that would require rendering the scene six times for each frame.