Sure.
So, if I am in the blue zone (not throttle) and maintain the blue zone, the [up] thruster output is more effective? I go [up] quicker?
Is there a decent video or post that details the mechanic?
FA On vs. FA Off Data
- Thread starter Frenotx
- Start date
Sure.
So, if I am in the blue zone (not throttle) and maintain the blue zone, the [up] thruster output is more effective? I go [up] quicker?
Is there a decent video or post that details the mechanic?
All tests were done with 4 pips to ENG. Though I do normally fly with analog thruster control, I used digital commands for these tests to make sure the inputs were pure and crisp.
I'm confident that, at least with FA on, the acceleration curve is non-linear. It's pretty clear when you're moving through the video frame-by-frame, as you have to click a lot more to get through those last m/s.
It's a bit harder to say with FA off, though. Hitting the speed cap with FA off seemed more abrupt. The power curve is definitely I'm interested in seeing. After that, it's just ship-by-ship boost power we need, and we'd have a pretty comprehensive grasp of thruster performance.pI simply record myself using shadow play (built-in recording solution that comes with the GPU software), and then returning to 0. So, from a stand still, go to full throttle and wait until I hit my top speed, then go to 0 throttle, and wait until I stop. Then I turn off FA, and repeat the movement. Full forward thrust until I stop accelerating, then full retro thrust until I start moving in reverse a little.
Once I've done all the movements, I turn of the recording software, and load the video into Windows movie maker. In that program, you can pause, and move forward and back frame-by-frame. I find the frame that a given movement starts at (right when the throttle is set to full forward, for instance), and note the timestamp of that frame. I then find the frame in which that movement ends (the frame I hit my top speed), and note its timestamp. Subtract the former from the later, and you have how long that movement took. Divide the max speed attained by that calculated time, and you get the m/s/s acceleration.
Last edited:
I repped you without even reading your post first.
I wish FD were far more transparent about the game stats in general. They leak out tidbits of information when asked or pressured on the forums but there's very little official info and why work like yours will always get all the rep I can dish out.
I wish FD were far more transparent about the game stats in general. They leak out tidbits of information when asked or pressured on the forums but there's very little official info and why work like yours will always get all the rep I can dish out.
Heh, that's exactly what I do when I'm coming in hot in my trutter. D rated thrusters are pretty drifty
Same with my FAS and AspX. I just do it for funsies though, not any technical reasons. Same as doing donuts in a empty snowy parking lot. Speaking of which...
I believe FD leave the Theory Crafting to the players. There is always a group, or collection of players, that dive in a cut up the data, why take that activity away from those that enjoy the work?
@Frenotx:
I will have to spend some time working on the elements of the process. I do have Shadowplay and all of that but, I have never had an interest in it before. When this testing is done, do we want un-modded data, or Data appended with any mods involved?
@Frenotx:
I will have to spend some time working on the elements of the process. I do have Shadowplay and all of that but, I have never had an interest in it before. When this testing is done, do we want un-modded data, or Data appended with any mods involved?
I did this with my modded viper IV, since I was mostly collecting the information for my own benefit. If we were trying to detect trends among ships, then mods wouldn't really be a problem. If we were trying to deduce specific performance data for specific ships, then factory a-rated thrusters would probably be the best choice.
I use exactly the software Frenotx describes (Nvidia Shadoplay to record, Windows Movie Maker to play back, consider times, edit etc.)
A couple of potential wrinkles:
- Nvidia Shadowplay does not come with older cards (and obviously not on non-Nvidia cards) - however you can obtain good free recording software on the web regardless.
- Windows Movie Maker is one of those things that Microsoft like to pretend doesn't exist now. It is bundled with Windows 7 but not 10. You can still download it from Microsoft for free but they make it hellishly awkward to find (for some reason they want everyone to use something newer that's worse, I think) - however most ordinary playback stuff should still be good enough for you to find the start and end of a movement.
Give up on Windows Movie Maker. It's perfectly decent but rather than play hide and seek with Microsoft for software that will never be updated again, just use one of the superior free ones from the web with modern features and things like native H.265 support.
Just Google "best free video editor 2017" and you're likely to get quality software like Lightworks and Shotcut, or super simple stuff like VideoPad which is not that different to WMM but just up to date and easy to get hold of.
Just Google "best free video editor 2017" and you're likely to get quality software like Lightworks and Shotcut, or super simple stuff like VideoPad which is not that different to WMM but just up to date and easy to get hold of.
I tried a few other movie editing software solutions when I was working on putting together that divebomber how-to, and was left frustrated and irritated by all others before settling on WMM. Software that works and does its job doesn't really need updates.
Last edited:
How did a useful and interesting thread like this turn into a discussion on the relative merits of Windows Movie Maker?
I fly almost exclusively with FA off for the same reason I have a well-groomed beard, skinny jeans and winklepickers. It's not because it's more effective in combat (that's qualitatively and now quantitatively not the case!). I do it because, well, I'm in space, and want to feel like I'm flying a spaceship.
Thanks for the data, Frenotx. I had come across Alexander the Grape's video just tue other day and had started investigating it myself. You've saved me a lot of work.
I fly almost exclusively with FA off for the same reason I have a well-groomed beard, skinny jeans and winklepickers. It's not because it's more effective in combat (that's qualitatively and now quantitatively not the case!). I do it because, well, I'm in space, and want to feel like I'm flying a spaceship.
Thanks for the data, Frenotx. I had come across Alexander the Grape's video just tue other day and had started investigating it myself. You've saved me a lot of work.
Every year or so I try to find a newer / better free video editor. Every year I end up using WMM anyway.
...back on topic, I'd like to see some more tests that check FA off accel similarly for other ships. I'll do my own if I find the time.
I look forward to your results.
Data table for engineered Keelback: https://docs.google.com/spreadsheets/d/1VSkUFSvdZ5sOm1sTw8uYlIBRIs3USElO34NE86nrilU/edit?usp=sharing
My findings support the idea that FA off and FA on, in general, mostly have the same acceleration parameters for single-axis acceleration. The exceptions for lateral deceleration and FA off trichording have already been discussed. A new one that hasn't:
In FA on, there's a "surge" effect for acceleration where it takes some time to fully ramp up.
I accelerated from a stop to full speed with both FA off and FA on. With FA on, the acceleration from 0-50 averaged to 23.5 m/s^2, and from 50-250 averaged to 27.8 m/s^2. With FA off it was much closer to even (and if anything, slightly faster at the low speeds: 28.9 m/s^2 for 0-50 and 27.8 m/s^2 for 50-200). So going from 0-50 is faster in FA off, but continuing that acceleration from 50-250 is the same for FA off and FA on. And when you average things out, FA off forward acceleration is always going to be slightly faster (especially with trichording in the mix!).
This effect seems to hold for lateral thrusters as well. "Average" acceleration is about the same, but with FA on that accel takes a second or two to ramp up fully.
EDIT: Table formatted by request.
All values in m/s^2. I didn't measure from the very max speeds since there's an acceleration falloff right at the last few m/s of velocity that I want to ignore. And naturally there's some measurement error in exactly which frame I select for start/end times calculated.
My findings support the idea that FA off and FA on, in general, mostly have the same acceleration parameters for single-axis acceleration. The exceptions for lateral deceleration and FA off trichording have already been discussed. A new one that hasn't:
In FA on, there's a "surge" effect for acceleration where it takes some time to fully ramp up.
I accelerated from a stop to full speed with both FA off and FA on. With FA on, the acceleration from 0-50 averaged to 23.5 m/s^2, and from 50-250 averaged to 27.8 m/s^2. With FA off it was much closer to even (and if anything, slightly faster at the low speeds: 28.9 m/s^2 for 0-50 and 27.8 m/s^2 for 50-200). So going from 0-50 is faster in FA off, but continuing that acceleration from 50-250 is the same for FA off and FA on. And when you average things out, FA off forward acceleration is always going to be slightly faster (especially with trichording in the mix!).
This effect seems to hold for lateral thrusters as well. "Average" acceleration is about the same, but with FA on that accel takes a second or two to ramp up fully.
EDIT: Table formatted by request.
All values in m/s^2. I didn't measure from the very max speeds since there's an acceleration falloff right at the last few m/s of velocity that I want to ignore. And naturally there's some measurement error in exactly which frame I select for start/end times calculated.
| FA On | FA Off | Diff | |
| Forward 0 to 50 | 23.47 | 28.90 | 5.43 |
| Forward 0 to 250 | 26.80 | 28.00 | 1.20 |
| Forward 50 to 250 | 27.78 | 27.78 | 0.00 |
| Reverse 250 to 50 | 20.83 | 20.68 | -0.15 |
| Reverse 250 to 0 | 20.83 | 20.71 | -0.12 |
| Reverse 50 to 0 | 20.83 | 20.83 | 0.00 |
| Lateral 0 to 50 | 18.3 | 20.83 | 2.45 |
| Lateral 0 to 200 | 20.08 | 20.83 | 0.75 |
| Lateral 50 to 200 | 20.72 | 20.83 | 0.12 |
| Lateral 200 to 0 (with counter thrust) | 71.43 | 20.83 | -50.60 |
| Boost 0-200 | 111.11 | 114.94 | 3.83 |
Last edited:
Tannik, this is truly amazing data. Thanks indeed. I will get on uploading my own asap.
If you have a spare moment, is there any chance that you could put it into an official forums table, for ease of comparison...?
Cheers!
Done, post edited.
The lateral 200 - 0 stats seem off. Is the difference in acceleration really that pronounced?
Absolutely. Try it yourself! FA on gets magical deceleration powers for lateral/vertical thrusters.
There's a really old video I did here that demonstrates the effect (ignore the second half, they tweaked the physics and it largely no longer applies).
Did some quick tests with a Taipan as well, got about the same "surge" results. Results are a bit noiser since acceleration was overall much faster, so off-by-a-frame measurement errors have a larger impact.
| FA On | FA Off | Diff | |
| Forward 0-50 | 30.12 | 55.56 | 25.44 |
| Forward 0-250 | 46.04 | 53.19 | 7.15 |
| Forward 50-250 | 53.05 | 52.63 | -0.42 |
| Reverse throttle 250 to 50 | 45.45 | 46.51 | 1.06 |
| Reverse throttle 250 to 0 | 45.79 | 46.30 | 0.51 |
| Reverse throttle 50 to 0 | 47.17 | 45.45 | -1.72 |
Yes indeed, this is one of the things I was referencing earlier in this thread about evasion. If a fixed kinetic guy is leading you as you thrust laterally in one direction with FA-off, you can frequently make him miss just by re-engaging FA-on and thrusting in the opposite direction.
Basically the shooter will see a fast moving object stop dead and miss ahead of it, in its original direction of travel.
It's techniques like this combined with g5 dirty drives that make the 1,600 mps of fixed multis frequently inadequate to secure much time on target. Hence the considerable movement in the PvP community towards gimbals, post 2.1. To achieve the same time on target with fixed kinetic that you would have had before 2.1 you're effectively forced to use the long range mod.
