Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by e345spd

  1. Years later and you guys are still improving the experience for msffb2 and g940 ...Looking forward to BOK, that's for sure. I still haven't spent nearly enough time with what we already have, so no big rush haha Edit* Are the point sizes are off for the forums or is it just me? 12pt 14pt 18pt doesn't look right. Not that it matters though haha
  2. Disabling direct mode is the only way to run it for now, yes. Honestly, I think you're being a bit silly. The devs need reports on various platforms and nothing bad happens outside of direct mode. Do you expect hardware failure? They probably can't get direct mode working reliably yet. I would find that completely understandable. I think people might miss this, so I just wanted to repeat it.
  3. Well f12 didn't seem to be working for me either, so num5 is a good tip, thanks!
  4. must have an option to disable direct mode for OR as well, debug tool maybe? so far for vive it's a bit finicky getting the game to actually display in the headset. Will need to mess with video settings too, whatever settings it was on felt like ~ 30 FPS. Will test more later tonight. I remain hopeful, AC is amazing in VR, so it is possible to retro-fit support onto a fairly demanding title created before any real practical vr discussion and still get excellent results.
  5. It's just launching Steam/OpenVR and seems to loop back to trying to open SteamVR again, I'm going to mess with it some more. Edit****New update? The launcher is downloading 161MB
  6. HAHAHA I literally was about to test a new VR demo, saw an update for BOS/BOM and thought 'damn, I really hope the devs are looking into OpenVR support...'. About to go test now, very exciting!
  7. Really looking fantastic. You guys are monopolizing my available flight sim time these days! I was running a mission last night and fell asleep once I gained cruising alt. Woke up ~7 min. later at the action point, same relative altitude ​ ​ - Flyable Ju-52m3 Supporter
  8. Haha these are great, I think I need to re-purpose some old controllers.... The twist rudder on msfb2 just lacks the necessary precision.
  9. Really appreciate the info! Nice surprise when I came home and saw an update had download. I know what I'm going to be flying later
  10. I think some of you are forgetting the actual dimensions of the radiator and the card itself: Anyways, they will be releasing air cooled versions in the near future, plus the tiny nano. Also, the Fury non-x specifically is the one to look at for relative value at the high end if you aren't attracted to playing with the full die. The cut stream processors have a minimal effect on performance in most situations and result in in card with fantastic overall efficiency per relative minimum frame time. Shader-heavy fury x is more aligned with dx12 in terms of hardware resource balance. Overclocking is still waiting for 3rd party tool support to really figure out. Apparently, getting software control over the voltage regulators is always rather complex for AMD cards, or so I read. HDMI 2.0 is supported through a conversion attachment. Frankly, the displayport version on those cards is superior to hdmi 2.0 in terms of flexibility and performance. I see no reason to waste space on hdmi 2.0 specifically. The majority will be using displayport.
  11. Hololens is all about agumented reality. Overlaying objects and information into the real world. You are correct, not applicable for VR experience. That said, the potential is amazing. I'm concerned the platform will remain heavily compute-performance constrained for a long time, though. Simply improving the accuracy of the rendered objects to match the true environment will be a long-term pursuit iin basic compute performance, just like it is with games. Eventually raycast photon simulation from something like luxrender, maybe capturing lightmaps from a 360 degree sensor in real time. Augmented reality alone is a very cool concept, but convincing and complex augmented reality will be astounding.
  12. Yeah, new extensions that aren't utilized by the majority of software and smaller processes that have less physical surface area to dissipate heat can have limited real-world value. Aside from efficiency, of course. At this time, Intel has no incentive to release significantly greater compute performance for anything less than an exorbitant amount of money. Their best, highest overall compute performance work doesn't even apply 99.99% of games (a higher count of slower cores from xeon), the software still isn't capable. It's really an annoying time for x86.
  13. You mean you see ~10% uplift from 15.7 on 8.1? As far as I understand, some of the long-term experimental work from the wddm 2.0 branch has been brought into the 8.1 branch. Working very well with windows 10 and I'm trying to figure out how much of the uplift is due to driver specifically.
  14. Well, 1074 (a bad build in general) with this driver (a combination that AMD specifically says to avoid) yielded a little more single GPU performance than previous builds/combinations. Crossfire was also a little unhappy with that specific combination. The WDM 2.0 driver has been exhibiting gains here and there throughout the windows preview process, so I think this 'final unification' of sorts has brought in some significant changes. Better performance than I got with 8.1 and the latest betas (at the time... 15.5 or something)/ ~ march release modded wdmn 2 branch for 8.1 You should be seeing gains in 8.1 due to 15.7 I would think. Should apply to a pretty big range of GPUs, but windows 10 is still likely going to provide a better environment, at least for this system. I really like the re-worked audio kernel, too. No more windows notification sounds with nasty compression, better control over sample and bit rates. It's good stuff. Now, one tiny issue remains, but I know how to fix it. When you first load into the cockpit, you have to look around your plane to clear up a few initial frame drops. When I was running 8.1, moving the game data to SSD improved on the problem. Moving the game data to a ram drive pretty much took care of it entirely as far as I can remember. It also seemed to even out the frametime differences between heavy action and solo flight a bit. Maybe assets aren't being streamed in and out with necessary preemptive buffer. Most games do not exhibit this sort of behavior, usually you only get improved loading times. So yeah, keep game data cache speed in mind, might be a little performance uplift available. Haven't run a ram drive on 10 yet, the experience I'm getting on a 7200 rpm is hard to complain about.
  15. Yes this is a good explanation. Also, someone mentioned the carb cutting out when upside down. The AI may be maintaining just enough force to prevent the carb float from flopping towards earth gravity. It doesn't necessarily take much and it might be pretty subtle from an external point of view.
  16. It's all about angle of attack. High enough angle of attack will overrule the airfoil (why airfoil doesn't necessarily matter).
  17. I use no curve and .7 speed. All I do is bump the relative scale up and down with the arrows as needed. For the gunsight, I just lean forward. Feels really nice to me, easily controlled but you can still lean out of the canopy with a pretty natural range of motion when you're on the ground or your canopy is smoked. Wondering how much distance from camera to sensor changes our experiences. Mine is probably closer than most.
  18. Latest build plus the 15.7/drivers installed by default with latest build are yeilding a very respectable performance increase. Not just in BoS, either. So far, everything I've tested has higher minimums and better scaling. No problems to report. Little something to look forward to with windows 10. Release is almost here.
  19. Well there's no question about dx9 support as it exists in standardized, defined form. If someone is talking about a vendor-specific extension that doesn't actually exist in the definition, then that would be different, as no one else who supports the standard would have a license for using the extension. I know this happens all the time with openGL and less aware developers, It (basically) doesn't happen with dx9 aside from sometimes both intentional and unintentional platform targeting through shaders, not dx9 itself. That's also not an issue of support for the standardized shader model used by dx9, it's simply standard optimization for more than one architecture, as is always the case. The 'targeting' can be especially troublesome and 'unintentional' for any cross-platform/vendor optimization if you decide to say...Implement a package of shaders that someone sends you, all contained within in a pre-compiled binary that you don't have any access or control over . Anything else without frametimes and repeatable results is simply anecdotal evidence. As such, I will provide anecdotal evidence. My hardware runs all of those titles excellently at high native resolution. Crossfire works with any dx9 game that has basic functional multi-gpu support, including CloD . If people believe they have noted a trend in performance differences in any games that I care about, I've certainly not seen any valid evidence of it. When properly configured, I would wager that I get superior multi-GPU scaling 9 times out of 10, too. Anyways, the real problem is that we don't have access to any decent benchmarks for titles we actually want to play. I assume most of us on here have a similar selection of titles, many of which are never in reviews. Plus, the reviews are next to useless anyways without frametimes or at least a hint at meaningful minimums (some drops can be completely unnoticeable if fast enough). Someone should start a 'sim hardware review' website that actually looks at important things with technical understanding, not just this input accessory/game review nonsense.
  20. Yes, those are good points. This looks like it would be pretty comfortable in left exactly as it is now in for me when using two sticks. I completely missed the detaching HAT block and handrest, nice and modular. Everything I've seen from these guys looks really well engineered. The button box kit definitely included. Sort of off topic: Did you modify an x45 into a button box?
  21. I've seen trackIR crash a few times when it started trying to open in a tiny, collapsed window, though it was very infrequent and not with BoS specifically. Expanding the window and restarting the explorer process stores new window size settings. It's strange you're seeing actually seeing corruption after the crash. Do try running without a profile, I just use speed .7 with no curves for everything and just bump the axis scaling up and down with the arrows as needed.
  22. Yes, thanks for keeping us updated. I see the game in the profiles. Haven't tested yet though. I keep mine on exclusive and don't have any BoS specific issues, so I doubt I will see any changes.
  23. Actually, based on those last images, they could just put another button on the right in the same spot as the one on the left and call it done. Looks comfortable enough, bit of ergonomic tweaking for both in mind going on, maybe? Just flare out the hand rest on the left a bit more.
  24. Ok. I've been wondering why someone doesn't do this. It's not like the components are necessarily all that expensive. Just needs a good engineer and cheap mass production. Doesn't sound like an ffb2 replacement (that would be an unreasonable expectation, given the competition), but I really, really like the input design and layout. Compact and functional with inputs for everything you need. This hardware takes care of any non-clickable cockpit concerns. Really looking forward to seeing this progress. Would be neat if they can come up with some way to switch between left/right, not just for those who are dominant left, but also for dual-joystick games like starcitizen. I do sometimes! It is indeed frustrating and I'm somewhat ambidextrous compared to some. They seem pretty set on replication. Either I'm being stupid, or there's no easy way of making this design work for both. Maybe they will consider another version along the lines of the 'Saitek Cyborg 3D'. Looks weird but works surprisingly well. I guess then there's the problem of button and throttle placement, but detach could solve that.
  25. The first one looks pretty durable to me. The metal one looks pretty nice, but note that all the wiring is exposed, the extra incidental flexing from things touching the wires will probably lead to eventual failure before the LEDs burn out. I think I would try out the plastic one. Building them is definitely $25 worth of effort for a less robust result. One thing though. Are all of these arrays in a straight line when viewed from the front? I know that's how I built mine, but the official trackIR clip actually has an offset. Seems to help prevent point collision at certain angles. If you can find a clip that has offset, you might get a bit better result. Metal one appears straight/flat, too. Maybe there's still incentive to build your own after all, haha
  • Create New...