Brun

The post I was agreeing with suggested 1000W. Nowhere did I claim anything like "an 850W PSU is adequate for a 4090 in every instance". I simply stated my anecdotal experience. Recommending a 1500W PSU is still ridiculous.

100% this. 850W has been perfectly fine for the system in my sig, and that doesn't include 2x NVMe, 3x SATA SSD, watercooling pump and a whole bunch of USB devices running directly off it.

After getting very close to pulling the trigger on a VPC collective recently, I had a flash of inspiration and figured I might be able to accomplish something reasonable using my retired Warthog throttle. Don't actually do that much helicoptering so it seemed more reasonable than splashing out several hundred quid. Spent the last few weekends on it and pretty pleased with how it's going so far. Especially the fact that I don't - touch wood - seem to have broken anything in the process. Detail of the throttle assembly, which I'm particularly proud of considering I kind of invented it from scratch. The toothed gear is driven by the teeth on the inside of the throttle handle, which has about 200 degrees of rotation. Because I couldn't work out the tolerances for the gears, the whole thing printed in white has springs underneath and floats vertically to press it against the inner gear. It's currently using the pot from the friction slider because that was the easiest to remove, however I since realised that using one of the hall sensors from a throttle axis would give me the off/idle button so might rethink that. If I do go ahead with it, I'll add some sort of physical button or lock for the idle/off thing. Need to research how these things function in reality. I loved the fact that I had to get my bike bearing press out to put everything together. Felt like I was assembling something very precisely engineered, even though the tight tolerances were really just a fluke. The head was modelled from scratch, loosely using images from Virpil's website as reference (sorry). Kinda limited in options for positioning the three main buttons from the throttle since they're attached to the one board. Lucky I didn't damage anything here since the Warthog throttle doesn't even get recognised by Windows without it connected. Trigger, trim hat and button on the right are all from an old Cougar stick that I was happy to sacrifice. Added a plate on the back to adjust the upper limit of travel. It's moving less than 50% of the original throttle, but changing the calibration using DI View sorts that easily enough so there's no need to mess about with deadzones, saturation etc in control settings. Am quite surprised that maximum friction is enough to hold it in place, although that does feel like it adds some play when making trying to make fine adjustments. The reason I left so much shaft sticking out the back was the intention of adding either a counterbalance weight or damper, which is still the case once I've figured out the best approach. Have to admit, I think I enjoy this sort of thing more than I will actually using it.

I guess the question should have been 'can the HMD and NVGs be used during the same flight?' In contrast to - for example - the Hornet, where you have to tell the ground crew which one to install.

The pin should be easy enough to remove without disassembling anything, it's an M3 thread so just find a suitable screw. Might need something pointy to rotate the empty hole into a better position. Once you've separated the left and right throttles, remove the screw and the pin can be pulled straight out. You can also use an M3 screw as a replacement, however the head may need to be cut off to prevent it fouling the throttle handle. Edit: Just realised it's broken rather than just unscrewed. Guess that makes putting another screw in impossible, but I'd still expect you can release it from the current state and remove the pin using something suitably pointy.

Recommendation for a 1200W PSU seems wildly over the top. In to addition the hardware in my sig, I have 2x M.2 SSDs, 3x SATA SSDs, water cooling pump, 8x 140mm fans, plus a whole load of USB devices powered directly from the PC. All that running perfectly fine with a meagre Corsair HX850. It's plugged into the wall via a power meter and I've never seen it draw more than 650W. People suggest that power meters aren't sensitive enough to detect short spikes in demand, but I've not had a single issue that might have been power related.

It seems a bit weird to deliberately reduce the number of available buttons, but I wanted to simplify one of the F15EX switches by limiting its movement to just forward and backward. A very simple 3D print later, works exactly as I wanted. STL attached if anyone else fancies it. I just used double-sided tape to attach, which seems perfectly sufficient. F15EX_SwitchGuard.STL

Overall latency is a factor of speed and CAS latency (CL). You can compare the two kits using this calculator and see that they're actually identical in this respect. If the extra cost isn't an issue, I'd go for the 3600.

Nick Grey's comments certainly don't inspire any confidence in a 2024 release. Around 25 mins into the video...

Definitely sounds like an application 'capturing' those keys and preventing them being passed though. I'd suggest opening task manager and see if that shows what might be responsible. Force-quit anything that looks like it could be, checking after each whether the keys start working again.

On the calibration page, which currently looks like this?

Having got the impression that this board was at least monitored by WinWing, I thought it worth posting here... While calibrating my throttle it occurred to me that the axis values should be visible on the calibration page. Two reasons... 1. There's currently no way to identify each axis without going back to the 'test' page. Having to do this repeatedly is annoying. 2. It's useful to be able to see raw axis values while calibrating. Simple photoshop mockup...

The most significant light source in DCS is the sun, and that's definitely not static. Neither are the cockpits which it casts shadows into, or aircraft it casts shadows from. While I would agree that there are fewer benefits to ray tracing in DCS compared with other applications, it should at least make shadows more accurate and less prone to issues. It's also important to distinguish between ray traced direct shadows and bounced lighting from diffuse surfaces (aka global illumination or GI), the latter being what baking seeks to address. As far as ray tracing is concerned, they're independent of one another and direct shadows could be ray traced without the performance hit of calculating indirect lighting.

What 'it' are you referring to? Fixed foveated rendering has been part of the toolkit for ages, long before Matt developed the eye tracking stuff. It uses a method called Variable Rate Shading, which is nothing to do with quad views as far as I'm aware. It's been suggested upthread that the foveated rendering of quad views is better than that of the toolkit, even for users of headsets without eye tracking. I'd just like to know what that's based on, or see some evidence for it. Have had a read of the documentation and failed to see anything there, apologies if I've missed it.

Where does he say one is better than the other? I had a look into things and it appears that they use different techniques, but if the quad views method is superior why wouldn't Matt have used that in the toolkit?

It's taken a while, but with the recent arrival of a Bambu Lab P1S I've finally done something about the mess that was visible from everywhere but the pilot's seat. Before... After...

For a headset without eye tracking, what's the difference between this and the fixed foveated rendering in OpenXR toolkit?

They don't take any payment info at all when pre-ordering. On the global store at least.

Going back to the OP, I struggled to fully understand this as well. My Orion2 F-15EX *with finger lifts* arrived today and I think I understand more clearly now. The throttle handles don't connect directly to the throttle levers on the base, the finger lifts fit between each. That being the case, I'm assuming that the 'dual throttle connectors' are just a pair of simple adapters without the finger lift mechanisms.

IFE's comment on the facebook post: "The geometry provided by True Grit is extremely accurate, but we may have to slightly rework the textures for Flight Simulator." I would guess that's in relation to how maps for roughness, metallic etc are handled differently between DCS and MSFS.

I put in a pre-order for throttle + F-15EX grips on August 3rd. Got the email saying they were in stock on the 25th, and they're now on the way. It's worth noting that they don't take any money when you pre-order, it's effectively a no-obligation reservation. The in stock notification gives you 48 hours to complete the order.

New Channel 4 (UK) series about the RAF... https://www.channel4.com/programmes/top-guns-inside-the-raf These things always tend to focus on the people but there's bound to be some decent footage as well.

I think it helps to think of IAS as an indication of aerodynamic performance rather than simply speed. Near the ground they're closely related but as you have noticed, become more divergent with increasing altitude.

I'm also very sceptical that the real aircraft requires such absolute precision. The stick being 'centered' does not explicitly mean both axes have to read exactly zero. Claiming nothing can be done about it despite several people independently suggesting the same perfectly obvious solution is insulting.

Having to compromise flight controls like this just so autopilot can be engaged seems completely unreasonable. It's a workaround rather than a solution. The obvious fix would be for ED to add some tolerance to what is considered 'centered' for the purposes of engaging AP. It can't be that complicated in the grand scheme of things. @BIGNEWY is there a way to make this an official suggestion or otherwise bring it to the team's attention?