- Sonic -

https://store.steampowered.com/app/3824490/Oasis_Driver_for_Windows_Mixed_Reality/ il creatore dei OpenXR ha sviluppato free, su piattaforma steam, dei driver chiamati Oasis, che riportano in funzione tutta una serie di VR che lavoravano solo con WMR. Installando steam e scaricando i driver, si effettua un "unlock" del visore/controller e a quel punto il visore diventa "nativo" per SteamVR, funzionando direttamente con quello. E' il modo con cui io sto' usando il mio Reverb G1, con grande soddisfazione. Funziona? Eccome! Pure meglio, non avendo più WMR come passaggio intermedio.

Se con una combo così vai a 35 FPS vuole dire che il tuo amico ha un'approccio irrealistico con i settaggi di DCS. Per volare al massimo con qualsiasi visore "oneroso" devi avere un computer che manco esiste a livello di potenza per una cosa come DCS che è un buco nero di risorse. Se abbassa i dettagli e leva roba inutile credo che può abbondantemente oltremodo raddoppiarli. Io gioco con dettagli bassi, e con tutte le cavolate tipo ombre etc eliminati...con G1 e un 3080+i7 online vado a 60FPS con risoluzione del visore al 100% (e per SteamVR la risoluzione al 100% equivale ad andare già in supersampling, infati sono molto oltre la risoluzione nativa del visore). Evidentemente deve torvare il suo punto di equilibrio

guarda, io conosco gente che per il pimax ha dalla 4070 Ti in su come GPU...penso per la CPU valga lo stesso ragionamento

se vuoi poca spesa e massima resa, grazie a nuovi driver free Oasis su Steam per la VR ex-WMR, puoi pensare anche al Reverb G2. Ottimo e a meno di 150€ te lo prendi. Io uso il G1 con un a 3080 e va una bomba (sempre con i giusti compromessi)

How can I disagree? Since way back in 1998, with civil simulators, I've always flown in 2D. Then, in DCS, I started using the Trackir and I already saw a big difference, but there was still a "perceptual" limitation. But when I finally switched to VR last year, the leap was epochal: nothing compares in terms of immersion, immensely distant from the Trackir and light years from the 2D monitor. And now that I've also been able to get hold of the Reverb G2 with the Oasis drivers, my experience has improved enormously, both in performance and quality compared to the Quest2 I initially had.

seem planned release via Steam on 30 August

interesting....with ATP in cold start also LGB bombs seem have a accuracy errors

wait...but in AA mode using ATP you can engage enemy aso with AIM120/AIM9 using only the pod?

Thanks to MBucchia!

Absolutely in favor. This is a solid and well-thought-out proposal that would bring real improvements to gameplay, especially in terms of communication efficiency and realism. Adding integrated radios to Tactical Command, Game Master, and JTAC roles—following the FC4 model—is a smart move. It eliminates the need for external tools like SRS and streamlines coordination with air assets. Including three configurable radios covering AM/FM/VHF/UHF is a practical and flexible solution. Extending functionality to the F10 map, with command units or repeaters to simulate signal coverage, is a great way to give currently underused vehicles a meaningful role. The fallback option of full map coverage for simplicity makes sense for initial implementation. The suggestion to link radio behavior in first-person vehicle use to realistic NATO and USSR radio types is a nice touch, adding immersion without overcomplicating things. Overall, this would make Combined Arms more appealing and functional. Fully support the idea—hope to see it implemented soon.

After updating DCS and optimizing settings, the CPU load improved, but the last core still hits 100%, causing occasional CPU-bound warnings. However, DCS runs smoothly without stuttering, while the GPU has room for more load.

After updating DCS to the latest version, removing the "Run as Admin" flag, and adjusting some settings to reduce CPU load, things seem to be improving. On an online server (up to 30 players), stuttering has disappeared (though I need to test more thoroughly). All cores are running well. The internal DCS Frame Rate Counter (locked at 120, as I use a 120Hz G-Sync monitor, 32") sometimes turns red, indicating a CPU bottleneck. The last core appears to be running significantly higher than the others (100%). The GPU is running at 70%, so I assume I could push it further with different settings. I don’t think I can optimize much more—while the RTX 2080 Super has some headroom, the CPU (i5-9600K) seems to be the bottleneck. Maybe I can offload more from the CPU and shift more load to the GPU, but I need suggestions for additional settings

mmm interesting...I will update Dcs and remove admin from start DCS MT

I've to check with new update; tested with version before last after change pagefile (min32-max64). Well, freeze remain a issue. May could be a USB device (I want to change my rudder port, back port of motherboard have issue) and I have to test. Maybe could be AITrack issue? I use a cheap webcam and could be intensive with CPU. I've register the running of session, and how you can se CPU decrease in the same spot ok a freeze: all cores running at 100% and GPU ad 90%...this mean bottleneck (and freeze) I suppose

Hi, I also made this upgrade. I studied a bit, but I can say that in the end, trying and trying again, the result is excellent. I can only recommend these things: 1. be careful to choose the hall sensor compatible with the system (whether it is analog or linear, supply voltage, etc.) 2. be careful with the wiring: the hall sensor has the GND and VOUT inverted compared to how a CTS251B15F104B6NB potentiometer is structured In my specific case, I made the modification partly as an experiment and partly because the potentiometer was starting to have some gaps in the axis reading at the beginning of the throttle (after a bit of forward movement, the reading began). After a couple of attempts (due to what was said above) I also succeeded in the undertaking with excellent results. I used: • Allegro 1324 hall sensors (linear) • 2.54mm 3-pin JST cables • 5*5*5mm cubic N52 magnets • 3D printed support (I attach the STL file) • N52 magnet orientation As I mentioned, the potentiometer (and how it is read by the throttle board) are organized as VCC-VOUT-GND, while the hall sensor is structured as VCC-GND-VOUT. I'll recap with this connection diagram: nullnull So (in addition to being careful with the standard colors that change), the GND and VCC are inverted. For the rest I printed a 3D support that contains the sensor and fits into the metal support of the old potentiometer; to solder I used a simple soldering iron (I put the corresponding sensor and on a paper tape, then I soldered with simple points), covered the solder with heat shrink and placed everything in the support. I glued the magnet in the support for the axis, be careful with the orientation: Then I connected, tested and calibrated and to my great surprise everything works well! The necessary (what is necessary to make 3 new sensors) paid (with shipping) a little less than € 15 from Aliexpress. Two small suggestions: • the STL file may not be perfect; I myself had to file the printed object to make it fit correctly in its seat. • The 3-pin 2.54mm JST connector that I took is different from the one on the peripheral PCB: I also solved this problem by filing some parts to make it fit correctly STL Preview (file attached) the sensor housing has a snap-on cover that you put on top (you can also glue it on if empty); the housing for the magnet has a space in the wide part at the bottom (not visible here) to house the 5*5*5 mm N52 magnet. Use glue to secure the magnet in the housing: braket_hall_sensor_tqs_cougar_V3.stl