ESAc_matador Posted April 30, 2016 Posted April 30, 2016 I think, that happened to many people... it is just you needed to change the chip. This is not UH or Mi8!!! Hee
Home Fries Posted April 30, 2016 Posted April 30, 2016 I have a FF stick though, and with no forces it is pretty hard. I hope they implement basic FF (at least center force or something) soon. Check out SimFFB, which runs independent of DCS. The latest version allows for both the classic helo trimmer as well as progressive trim, which is perfect for the Gazelle. -Home Fries My DCS Files and Skins My DCS TARGET Profile for Cougar or Warthog and MFDs F-14B LANTIRN Guide
gospadin Posted April 30, 2016 Posted April 30, 2016 Some heli's are just sensitive. We've had several real life Gazelle pilots test her out and they helped us fine tune the FM to what it is now. What we strive to achieve is realism. So we will continue to work with real Gazelle pilots to tweak the FM to be as realistic as possible. Keep in mind though, the Gazelle is one agile bird. What controllers is your team using to develop and test? If a warthog, what length extension if any? I too feel like the CG is too high relative to the rotors, given how it behaves. It doesn't want to hang on its blades, but feels like it's going to tip over. My liveries, mods, and missions for DCS:World M-2000C English Cockpit | Extra Beacons Mod | Nav Kneeboard | Community A-4E
Mt5_Roie Posted April 30, 2016 Posted April 30, 2016 Each team member has there own setup so I can only speak to mine. I use a saitek x65 and x52. The throttle from the x65 is the collective, and the stick from the x52 is the cyclic. Not sure if any team member has a warthog, but will ask. As for the FM, it's what the real pilots told us along with data we have. I wasn't directly involved with the fm coding so can't really comment on the CG, but I've had no issues flying her. Although based on real world experience there have been plenty of cases where she tipped over - mostly due to pilot error. A lot of people are used to left rudder on helicopters and I the gazelle that can cause a spin which could result in tipping. So right rudder folks! What controllers is your team using to develop and test? If a warthog, what length extension if any? I too feel like the CG is too high relative to the rotors, given how it behaves. It doesn't want to hang on its blades, but feels like it's going to tip over. Coder - Oculus Rift Guy - Court Jester
Oubaas Posted April 30, 2016 Posted April 30, 2016 I think it's a masterpiece. If this is what I can expect from Polychop, I'll be buying every module you make. The flight model seems very believable for a helicopter of this modern vintage. I love the trim system, too. After a day with the Gazelle, I'd suggest: 1). Use a very gentle touch on the controls. Any ham-fisted input will not be well received by the aircraft. Be particularly smooth with the collective. 2). Set deadzones and curves to your personal comfort level. I started out with no deadzones and no curves, flew it for a while, then added just a touch where needed. 3). Learn the trim system. The trim system in this thing is brilliant and if you learn to use it right, flying this machine is almost effortless. In fact, it trims out so well that sometimes I actually turn the trim system off to give her a bit more liveliness. This is a great module, Polychop, and I look forward to purchasing any of your future offerings. Based on the Gazelle, you've gone on my list of favored developers whose work means automatic purchase. Keep up the excellent work! :thumbup:
BronzeYardNo11 Posted April 30, 2016 Posted April 30, 2016 I've heard some not-so-nice "reviews" on reddit about the flight model being "buggy". Is the FM all complete or is there some tweaks coming out between now and the 1.5/steam release? ༼ つ ◕_◕ ༽つ Give OH-6 ༼ つ ◕_◕ ༽つ
Flagrum Posted April 30, 2016 Posted April 30, 2016 I made a tiny test mission myself - just the helo, hot on the ramp at Creech. But I can not reproduce this problem I had in the beginning - experimented with several things, restarted the mission several times, and also restarted DCS in-between one or two times. But aside from the very first 5-10 attempts, everything was okay afterwards. The only significant change was the enabling and also disabling that easy mode in the options and setting and re-setting the cyclic curves. Before it was tWiTcHy as hell and after that, she is still a hand full, but by far not such a beast anymore. :o) After a few hours of sleep I am back at square one. :o( I'ts probably just me ... Flying the Huey is like balancing a 5 feet broom stick on the palm of your hand. Flying the Gazelle the last time was like balancing a cut-in-half, 2.5 ft broomstick on top of a soccer ball in your hand: if you get used to it, it works well ... And now, again, same as when I tried it the first time(s): flying the Gazelle is like trying to balance a soccerball on top of that broom stick on top of the palm of your hand. It feels for me as if the SAS is not really working at all. It makes no real difference for me if it is enabled or not. The nose bobs up and down like crazy if I move the stick even just a millimeter. Same for roll and yaw. This quickly leads to PIO and a helo rolling and tumbling in every direction at once. I keep the stick marker centered in the controls indicator, but as soon as there is just an inch of air under the skids, she ACELLERAAAAAAAAATES already in some random direction... AP channels are all switched on - but as I said, switching them off (and on again) seem to make no difference ... :cry:
msalama Posted April 30, 2016 Posted April 30, 2016 I've heard some not-so-nice "reviews" on reddit about the flight model being "buggy".Well I wouldn't put too much weight on what the armchair experts over there say TBH :music_whistling: And this is an early access release, so one would assume things getting refined with time anyway + let's also bear in mind that the Gaz _is_ an agile chopper! Just hoping they'll update v1.5 soon so I could check her out myself - already bought the licence yesterday... The DCS Mi-8MTV2. The best aviational BBW experience you could ever dream of.
Cykyrios Posted April 30, 2016 Posted April 30, 2016 I haven't flown the Gazelle that much yet (will do more flying today), but one thing I quickly noticed (other than that I was really bad) is the ground effect, which seems to behave differently from the Huey (at this point, at least): in the Huey, once you pull just enough collective to get off the ground, you'll stay at a foot or 2, and will basically stay there without touching the collective, as long as you don't maneuver too hard. In the Gazelle though, pulling the collective slowly makes her want to go forward, so I have to pull aft cyclic. Before I even get off the ground, I'm either going 20+ km/h forward and/or scraping the tail. Once I'm a few meters above the ground, she will go several meters up and down without touching the collective, and even coming to a stop can make her touch the ground instead of "sitting" on the air cushion. Is that just a trait of the Gazelle, which requires pilots to constantly adjust collective even while hovering in ground effect? I also noticed that the torque from the tail rotor affects lift quite a bit, so that might play a role as well.
Flamin_Squirrel Posted April 30, 2016 Posted April 30, 2016 I too feel like the CG is too high relative to the rotors, given how it behaves. It doesn't want to hang on its blades, but feels like it's going to tip over. I agree; it flies more like an aeroplane. There doesn't seem to any lag between the rotor disc moving and the fuselage following. Perhaps worth looking into? That said, if real both the data and Gazelle pilots say that's how it flies then that's fine by me.
Dave317 Posted April 30, 2016 Posted April 30, 2016 I'm using a warthog with saitek combat rudder pedals. I've put a 15% curvature on the stick and pedals and it's so much easier to fly now. I was having a hard time before but it's relatively easy to fly now.
Mt5_Roie Posted April 30, 2016 Posted April 30, 2016 Don't believe everything you read on the internet. I saw the reddit review and found some interesting things out. 1) He took off with left rudder. 2) He lost control fairly quickly after takeoff. 3) He's not a real helicopter pilot (plane pilot yes). So take his review with a grain of salt. The only review I would actually listen to is one by an actual helicopter pilot who's actually flown a gazelle in the real world. Those are the guys we listen to when developing our FM. I've heard some not-so-nice "reviews" on reddit about the flight model being "buggy". Is the FM all complete or is there some tweaks coming out between now and the 1.5/steam release? Coder - Oculus Rift Guy - Court Jester
heloguy Posted April 30, 2016 Posted April 30, 2016 I'm not a Gazelle pilot, but I am a real helicopter pilot. I've flown a Bell 206 some, an MD-500 some, a Blackhawk quite a bit, and an R-22 a bit. Based on my experience so far, I've found that aircraft fly how they look. I know that sounds too much like a generalization, but I think it just has to do with experience, and understanding of aerodynamics. With that, I thought the first iteration of the Huey by Belsimtek was too twitchy, but having not flown a real Huey, I couldn't be sure. Low and behold, it became more stable in subsequent releases. Really, I think what it boils down to is hardware. You guys (developers) are amazing at making these helicopters perform realistically within the confines of our computers, but it is incumbent upon us to do our homework on our joystick setups, and try to make them imitate the real thing as much as possible. This is going to be hard for someone who has no idea what it "should" feel like. It's possible, though, with research, and time. Another bit to think about is this: for some people, it takes 8-10 hours solid to hover for the first time. And I'm talking about a helicopter you have to fly, not one in which you can push a button as long as the AFCS (autopilot) is up and running. This amount of time will only be increased by two factors: lack of instruction with immediate feedback (huge), and unrealistic hardware setups. With all that said, I think what people are saying here is a step in the right direction. Making adjustments to your joystick curves will help offset having a computer joystick that is not built exactly like the real thing. I've only seen a few examples of computer hardware that look like they would act close to the real thing, so if you're trying to fly it with a Thrustmaster Warthog, G940, or MSFFB2 (all of which I own, and I can say must be configured differently), some adjustments will have to be made. I'm of the belief that if you can takeoff, hover, and land in the Huey without crashing, you should be able to do the same with this aircraft if you take the time to setup your equipment, are slow, and deliberate, and don't try to fly it like Chuck Aaron right away. If you made it this far through my ramblings, thanks for reading. I'm going to try and setup my G940 now, and I'll report back what I feel is realistic for settings. 1 i9 12900k @ 4.9ghz, 64gb RAM Nvidia RTX 3090 Windows 11 x64 Pimax Crystal VP Force Rhino w/RS F16GRH, Virpil TCS Rotor Plus AH-64 Collective, BRD F1 Pedals, WH Throttle, FSSB R3 w/WH Grip, PointCTRL v2
westr Posted April 30, 2016 Posted April 30, 2016 I'm not a Gazelle pilot, but I am a real helicopter pilot. I've flown a Bell 206 some, an MD-500 some, a Blackhawk quite a bit, and an R-22 a bit. Based on my experience so far, I've found that aircraft fly how they look. I know that sounds too much like a generalization, but I think it just has to do with experience, and understanding of aerodynamics. With that, I thought the first iteration of the Huey by Belsimtek was too twitchy, but having not flown a real Huey, I couldn't be sure. Low and behold, it became more stable in subsequent releases. Really, I think what it boils down to is hardware. You guys (developers) are amazing at making these helicopters perform realistically within the confines of our computers, but it is incumbent upon us to do our homework on our joystick setups, and try to make them imitate the real thing as much as possible. This is going to be hard for someone who has no idea what it "should" feel like. It's possible, though, with research, and time. Another bit to think about is this: for some people, it takes 8-10 hours solid to hover for the first time. And I'm talking about a helicopter you have to fly, not one in which you can push a button as long as the AFCS (autopilot) is up and running. This amount of time will only be increased by two factors: lack of instruction with immediate feedback (huge), and unrealistic hardware setups. With all that said, I think what people are saying here is a step in the right direction. Making adjustments to your joystick curves will help offset having a computer joystick that is not built exactly like the real thing. I've only seen a few examples of computer hardware that look like they would act close to the real thing, so if you're trying to fly it with a Thrustmaster Warthog, G940, or MSFFB2 (all of which I own, and I can say must be configured differently), some adjustments will have to be made. I'm of the belief that if you can takeoff, hover, and land in the Huey without crashing, you should be able to do the same with this aircraft if you take the time to setup your equipment, are slow, and deliberate, and don't try to fly it like Chuck Aaron right away. If you made it this far through my ramblings, thanks for reading. I'm going to try and setup my G940 now, and I'll report back what I feel is realistic for settings. I think you are correct. The difference for me going from a T-Flight hotas which was very vague. To a TM Warthog hotas was huge. A massive difference more than I could ever have foreseen. My performances with the Huey improved significantly as a result. And with the gazelle I'm able to takeoff put her into a auto hover launch weapons and land. By no means am I as refined as with the Huey but I hope it will come with time. I'm eagerly awaiting a 20cm extension for my warthog that I have ordered I am sure this will help things as well. RYZEN 7 3700X Running at 4.35 GHz NVIDIA GeForce GTX 1080Ti 32gb DDR4 RAM @3200 MHz Oculus CV1 NvME 970 EVO TM Warthog Stick & Throttle plus 11" extension. VKB T-Rudder MKIV
clawz Posted April 30, 2016 Posted April 30, 2016 (edited) Don't believe everything you read on the internet. I saw the reddit review and found some interesting things out. 1) He took off with left rudder. 2) He lost control fairly quickly after takeoff. 3) He's not a real helicopter pilot (plane pilot yes). So take his review with a grain of salt. The only review I would actually listen to is one by an actual helicopter pilot who's actually flown a gazelle in the real world. Those are the guys we listen to when developing our FM. Roie, If you are referring to Helipilot7's first impressions then I believe you need to take another look. He is type rated in a number of helicopters, not fixed wing. Being dismissive of the communities overall feeling (very twitchy and FM oddities) of your recently released module is a recipe for problems. I have watched over a number of your Oculus videos, along with Helipilot7's and there are some glaring differences between the two. First and foremost you either seem to have some very heavy curves in place or are using a different version of the FM when recording. Example is your newest video - [ame] [/ame] While you do not have the control indicator present in the recording you are able to throw your stick around and the Gazelle responds in a predictable manner. This is either "Easy Control", heavy curves or a different flight model at play when compared to the following; - From the second stream Helipilot7 did. Starts off trying to sort out the CG and moves onto showing some other FM oddities. - This segment is particularly telling. 1/3rd power pedal input and the Gazelle becomes a tornado. Once he blacks out and crashes then the air frame continues to spin like a top on the ground. Engine torque is a thing of course but the amount present is unrealistic and indicative of a flight model problem. I agree with other posters views that the Gazelle is going to have a transition time due to how much lighter it is than the other helos we have had access to in the past. Combine that with hardware and curve setting changes and there will be a vast number of things that could cause odd behavior but when a large part of the community is saying something is wrong, I would listen. Everyone has a right to their opinion, wrong or not but dismissing everyone but "real Gazelle pilots" is problematic at best. You may have consulted with them but mistakes can be made when converting them to 1's and 0's for sim usage. That said, you guys have put out a great first effort in the grand scheme of things. I can only hope that the dismissive nature of comments ends and that the energy is focused into continued development of the module to improve upon the foundation already laid. Edited April 30, 2016 by clawz
heloguy Posted April 30, 2016 Posted April 30, 2016 Just tried the G940 with x/y curves of 15, and my Saitek Combat pedals set to 10. It was better, but still needs some tuning. I was able to takeoff, bump, make a decently aggressive turn to come back and approach to a hover, then set it down. BTW I also tried SIMFFB, which worked great as an interim fix to no FFB support. i9 12900k @ 4.9ghz, 64gb RAM Nvidia RTX 3090 Windows 11 x64 Pimax Crystal VP Force Rhino w/RS F16GRH, Virpil TCS Rotor Plus AH-64 Collective, BRD F1 Pedals, WH Throttle, FSSB R3 w/WH Grip, PointCTRL v2
XCrosser Posted April 30, 2016 Posted April 30, 2016 It's not really that bad when you put a curve (only to 15 for me) on your controls and learn how it flies. She's nimble for sure, but can be tamed.
xBuzz Posted April 30, 2016 Posted April 30, 2016 Just can't bring myself to use curves just yet, feels like cheating and the challenge is just more fun!
audax Posted April 30, 2016 Posted April 30, 2016 Isn't it also cheating that you don't really die when you crash in the sim?
emg Posted April 30, 2016 Posted April 30, 2016 Just can't bring myself to use curves just yet, feels like cheating and the challenge is just more fun! The MiG-21 manual actually instructs pilots to use certain curves with that module. Much has been written about short-throw PC sticks vs the real sticks.
xBuzz Posted April 30, 2016 Posted April 30, 2016 (edited) True, that's why put in a "yet" there, as I might in a few days. :p Edited April 30, 2016 by xBuzz
Bond 42 Posted April 30, 2016 Posted April 30, 2016 Isn't it also cheating that you don't really die when you crash in the sim? so true. Funny
BronzeYardNo11 Posted May 1, 2016 Posted May 1, 2016 1) He took off with left rudder. 2) He lost control fairly quickly after takeoff. 3) He's not a real helicopter pilot (plane pilot yes). It's good to know that the FM is all there and it's just some bad player being bad. Soooo looking forward to the 1.5 release! :thumbup: ༼ つ ◕_◕ ༽つ Give OH-6 ༼ つ ◕_◕ ༽つ
Bearfoot Posted May 1, 2016 Posted May 1, 2016 Just can't bring myself to use curves just yet, feels like cheating and the challenge is just more fun! I see it more as calibrating the interface between the simulated world and the real world, like adjusting monitor height/angle, seat etc. It's just math, really, mapping a range of inputs to a range of outputs. No more cheating than using your mouse to flip a switch.
1.JaVA_Platypus Posted May 1, 2016 Posted May 1, 2016 Don't believe everything you read on the internet. I saw the reddit review and found some interesting things out. 1) He took off with left rudder. 2) He lost control fairly quickly after takeoff. 3) He's not a real helicopter pilot (plane pilot yes). So take his review with a grain of salt. The only review I would actually listen to is one by an actual helicopter pilot who's actually flown a gazelle in the real world. Those are the guys we listen to when developing our FM. "I can't fly it, so it must be wrong" :music_whistling::megalol::lol: Happy Flying! :pilotfly:
Recommended Posts