Aries144

Has gun spread/accuracy been examined by the team yet? Has Belsimtek reviewed the issue and made a determination? I know they're probably loaded with work, but it's a bit of a glaring issue. The current modeling really doesn't reflect the Northrup document for gun accuracy. If lack of an original source for what a "mil" equates to is an issue (it's surprisingly hard to find!), I'm including a link to a US military sniper's manual that states that 1 mil is equal to 1m diameter at 1000m (or 10cm diameter at 100m, 50cm at 500m, etc). The relevant information is in the paragraph and picture on page 43. Currently, the F-5's gun spread is about 30-40 mil (equivalent to a 300-400cm diameter group at 100m). According to the Northrup document, 80% of shots should fall in an 8 mil area (equivalent to a 80cm diameter group at 100m). I hope this information is helpful!

What you're feeling in the roll axis is the aileron roll limiter. Users of non FFB sticks have to press a button to get the full roll rate, while FFB users just have to push through the bump of extra pressure- just like the real plane. ;)

I honestly haven't had that issue flying the Mig-21 on your server man. I feel like visibility sucks in the Mig-21 and tend to fly low level and keep my speed high near mountains. I don't often turn with F-5s because they have a turn advantage at low altitude that's hard to counter in the Mig-21. I had no issue killing around 6 F-5s in a row over the course of an hour or so a few days ago. I think the real difference is the blue paint on the F-5 and that it produces less smoke than the Mig-21. I typically spot the Mig-21 in a turn or at close range because of its smoke. That big black smear in the sky is a dead give away. I'm playing with an Oculus Rift too. Not exactly an advantage for visibility. ;) Maybe try a different paint scheme with the F-5s? Maybe one of the darker camo ones like the Mig-21 is using? You could also always do the all-black "Mig-28" scheme if you can't find any other way to stop losing against the F-5. ;)

Wow! That new Jetseat model looks great!

I get a huge FPS reduction when using the stock cockpit textures. If I use an English cockpit mod, my FPS is significantly better.

I'll see if I can get details from Drexx on how he rolled back his Oculus from the beta driver to the current stable.

Is there any more info on what things the different "scene.m:setArgument" numbers control? Specifically, I'd like to set what pylons are hanging on the F-5s wings.

Try it. Both myself and Drexx think it is.

Set the PD in DCS VR settings to 1. Set the Pixels Per Display Pixel Override in the SDK to the highest your hardware allows. Typically between 1.3 and 2. Higher is better, but requires more powerful hardware.

I found it didn't work for me. After I hooked a monitor up to the hdmi port, I found it was still set to limited in Nvidia Control Panel. Also, it only sets the dynamic range, it doesn't change any other settings like vibrance.

Essentially, before the changes the outside lines of aircraft were so blurry, and their colors so indistinct from sky and ground colors, that I found myself at quite a disadvantage in far visual range and even near visual range fights compared to using a monitor and trackir. Add in some motion blur when frame rates dipped too low and it was really impossible to be competitive. Now the outside lines of aircraft seem much more defined and the slight differences in color between aircraft and their surroundings are more apparent. Before, it was often extremely difficult to stay padlocked on a target, even if it was moving across my field of view, until it was well within a mile. Now I can spot it if I look hard enough and get lucky, just like with a monitor. It's like going from 20/60 vision to maybe 20/30. Here's a download link for the SDK. For the SDK to work, you can't use the Beta Oculus driver. I'm using 1.13. I set the in-game PD setting to 1. I'm using an nvidia 970, so I set the SDK value to 1.3. The 'HDMI trick' to adjust digital vibrance is as simple as unplugging your Oculus from the HDMI port, plugging in a monitor to the same port, and changing the following in Nvidia Control Panel for that monitor: Adjust Video Color Settings Under Advanced tab, change Dynamic Range to "Full" Adjust Video Image Settings Try increasing edge sharpness to 60 (not sure yet if this works or not) Adjust Desktop Color Settings Increase Digital Vibrance to something around 60-70 After you plug the Oculus back in, those changes will still be in effect for the Oculus. I had to buy a DVI to HDMI adapter, as my monitor doesn't have an HDMI input.

A friend showed me this trick tonight. Anyone else try it yet? I'm so far finding that using a setting of 1.3 in Pixels Per Display Pixel Override in ODT looks significantly less blurry than 1.3 PD using the in game setting, and it appears to have less frame hit as well. Combined with a small bump in digital vibrance using the HDMI trick, my ability to spot nearby aircraft against the ground, sky, or horizon finally seems to be approaching what is was with my 23" 1920x1080 monitor.

Was there an update I didn't catch?

Yeah, this doesn't seem correct. What's the point of a backup that experiences the same error all the time?

Whoa, slow down man. 8 mil = 8m at 1000m = 80cm at 100m. Remember, that's the factory minimum for the test (more than 80% of rounds outside of an 8 mil circle and the gun fails the test, probably to be rebarreled or to have the ammunition lot tested) and it's also when firing full auto at a very high rate of fire. The same gun firing a single shot at the same distance would likely show far smaller group sizes, as Cartridges of the World seems to indicate. It would be interesting to try and contact some of the maintenance personal working on F-5Es in the US Aggressor squadrons and obtain more detailed information about group sizes when the guns are tested. Maybe they tend to shoot better than that until they near maximum wear?

Yeah, I've seen the same thing first hand firing various small arms. However, the thing is we have manufacturer data about the precision standards of the weapon when fired in full auto, on the ground, during weapon sight zeroing. The above Northrup document states that the accuracy standard is 8 mil, 80%. There are actually documents (Cartridges of the World I think?) that show the M39 cannon shooting 1 MOA (aprox 0.25 mil) when it is securely mounted and single fired on the ground, so 8 mil sounds about right for high ROF full-auto. The issue is that, all pilot error aside, the cannons are currently spraying rounds in a pattern that is so large it would actually be impossible to replicate in real life without the cannon rounds being completely destabilized ,tumbling end over end, and curving wildly all over the place like a US baseball pitcher throwing curveballs. ETA: Completely jealous of your experience firing a 30mm on an IFV. Fun memories! :D

The dispersion of the F-5's gun in game looks to be around 40 mil, just watching the dispersion in the reticle. The M39 is supposed to be 8 mil, same as the M61. Also, is there any information about what size the target in the Swiss video is and what range the aircraft were firing from?

Completely normal. :) I find it advantageous for fine control, particularly when hovering in a helicopter.

Can we get any word on whether or not this subject has been noted by Belsimtek? This is really an annoying bug.

I just want to express my public thanks for Belsimtek's outstanding work on the F-5 module. It is an absolute joy to take a break from life and fly the F-5. I really appreciate the details, like the utilization of force feedback to simulate the aileron limiter "wall" that can be pushed past and force feedback trim. I have absolute confidence that they'll get the little issues, like gun dispersion, fixed in time because they've gotten so many other details so correctly precise. Belsimtek can count on me to buy every module they make!

I want to add that if you're hosting a server and use the pylon jettison, it disconnects all players connected to you. Pylon jettison disconnects me if I'm on a multiplayer server as well. What a weird bug!

You're correct, I'm used to thinking in terms of 0.1 mils. Still, 80cm at 100 meters sounds almost absurdly large for any firearm...

It's great to see Belsimtek is going to be addressing the accuracy of the F-5's guns. The current 20-30ish mil performance is unrealistic in the extreme- to have that level of inaccuracy, rounds would have to be failing to stabilize and tumbling through the air or the guns would have to be very loose in their mounts. 8 mils = 8cm at 100m and 80cm at 1000m.

mkiii, I have a fix for the axis reversal in opentrack: cover the camera with your hand for a second or two. If the camera loses sight of your lights/points opentrack gets confused and may reverse axis. Covering the camera, adjusting so that the points will be visible to the camera again, and removing the hand from in front of the camera solves the issue for me 100% of the time. You can prevent the issue entirely by ensuring that your tracking points don't become obscured by each other, other objects, or leave the camera's field of view.

Debolestis, Thank you very much sir! You've made this so much easier. Again, thank you! Sokol, Thank you very much for your contributions as well. Your efforts are very much appreciated!