Could or should Frontier enhance the FSS or add in and incorporate an optional ADS-like expansion module?

Regarding precision, I found that the small rotaries on my x56 throttle have just enough precision to work well for tuning. I frequently find myself making single "tick" adjustments, but the ticks are small enough to do the job and the setting is stable over time. I think those are the least precise of the four rotaries on that model.

I think it probably depends on how a HOTAS implements the rotary control. It seems many of them have too much resolution, rather than too little. I'll need to get into the technical aspects for that to make sense.

A USB game-pad controller provides up to six "analogue" controls: LX, LY, LZ, RX, RY, RZ. Each of these controls can have any range. The game-pad specifies the minimum and maximum range of the numbers it will be providing. It may, for example, say that it will supply numbers in the range 0-1000. The game software then has to map those numbers onto the thing that the control is controlling.

There are two sorts of control: analogue and digital. Analogue controls are cheap (a few pence) and digital controls are expensive (a few pounds to many pounds). With an analogue control the gamepad reads a voltage off the control and converts that to the number range it has said it will be using. With a digital control it will get signals when it is rotated up or down. It adds these numbers to a running total to produce the current position of the control. It follows that you can't move it by less than one position and that limits the practical resolution because if you want a control with 1000 positions per revolution you will be spending £50 or more for it. The result is that you have to spin the control to get anywhere.

An analogue control on the other hand has a fixed range of movement -- generally 270 degrees -- but reading an analogue voltage is not a precise activity. There will be noise in the system which generally means that you have to take great care if you want more than about three digits of accuracy. Even with three digits you will see the effects of noise -- the position of the joystick will jump around a little. There are ways to mitigate this in both hardware and software but they will limit either the resolution or the responsiveness of the control. As a rule of thumb, a range of 0-255 should be stable enough but 0-16383 would always be unstable. 0-1000 is probably acceptable on an expensive enough joystick. There is also a question whether the human fingers can control with that kind of accuracy.

It's possible to implement a range of 0-16386 with an analogue control that you've given a resolution of 0-256, simply by multiplying by 64 but that means the gamepad will be producing numbers with holes in. If you move the stick by a tiniest amount the number will jump by 64. This might confuse the game software.

The FSS tuning control I built uses a digital control with a range of 0-255. It works very well with the FSS. One detent on the control is often the difference between being in tune and being out of tune, but it never skips over the tuned position. On the other hand my cheap HOTAS joystick (Thrustmaster USB) which I think has a range of 0-16383, is very fiddly to use. You have to get the positioning spot on and it tends to jump around. But the control itself isn't that bad; it doesn't jump around by 64 so it should be as stable as a control with a 0-255 range, probably even 0-1000. I believe that ED adapts the tuning requirements according to the resolution advertised by the control. As a result it is easier to tune with a low resolution control than with a high resolution control.
 
Is it not odd that some VR users think the FSS is an abomination yet declare the ADS as the Bee's Knees? As a VR user myself the ADS was an abomination in VR and the FSS a delight...
[JOKE]You must be a masochist then. ;)[/JOKE]

I never said that the ADS is the bee's knees and nor has anyone else in essence - all that has been said is that the ADS is better than the FSS on one-hand (subjective) and on the other facilitates an experience that was lost with the FSS (factual). Further more, the greatest objection is the removal of choice - the FSS is too pervasive and badly implemented to be legitimately considered a good move for ED. FD avoided such a fractious move with mining but failed to do so with exploration.

This is however mostly moot - what is clear is FD needs to rework the FSS to address the broad critique of it. Some aspects require more urgent attention than others and these aspects should not require that much effort (i.e. are on the most part supplemental tweaks).
 
Last edited:
I think it probably depends on how a HOTAS implements the rotary control. It seems many of them have too much resolution, rather than too little.
The problem is lack of resolution - not too much of it, already tried it

FTR - with kit I have to hand and have tried
  1. Thrustmaster Warthog (Throttle unit) - 14-bit Throttle, 10-bit Trim, 10-bit Mini-stick
  2. Logitech G X56 (Throttle unit) - 10-bit Throttle, 4 x 8-bit Rotary, 8-bit Mini-stick
  3. Thrustmaster T-Flight HOTAS X - 10-bit Stick primary axis, 8-bit Twist, 8-bit Throttle, 8-bit Rocker
The 10-bit Trim control on the Warthog is not really good enough for absolute control of the tuner, 14-bit might be sufficient for absolute control but the overriding point is that it still would not address the underlying problems of lack of tuning memory and control re-use.

[EDIT]As for the stability of input from ADCs of varying resolution ALOT depends on the implementation and higher-resolution devices need not be that unstable (there are means to avoid this). The main factors will be line noise on one hand (stability of voltage being key) and mechanical construction on the other. Where hall-effect devices are concerned, there is the additional complication of stability of the surrounding magnetic field which can be affected by other powered electrical equipment near by. The EMC regulations help to moderate this effect but it can not be completely avoided - the nominal counter being calibration.

Overall though, the upshot of this is that ensuring tuning memory is properly and consistently implemented at the software level is the right thing for FD to do - special or custom hardware should not be required.[/EDIT]
 
Last edited:
I really like the FSS, but man it's still buggy and half-baked. One thing I don't understand with Frontier's UI in general is why they have to flash giant messages in our face that covers up important information, when those messages just end up in the info panel anyway, WHERE THEY BELONG.

And then there's the data on the right side that STILL doesn't always show, despite multiple patch notes that declares this as fixed.

And of course the insane POI scan times.

And the utterly useless and infuriating Analysis Mode....

To be honest, while I like the FSS way more than the ADS, at least the ADS worked the way it was designed. I actually find myself missing the simpler days of the ADS. That's too bad, because the idea of the FSS resonates with me, but the execution of that idea leaves much to be desired.
I almost completely agree. I don't miss the old ads. I like the fss enough not to want the ads back. But really fix the POI loading times! And those other things Duck mentioned, but mostly the POI thing.
 
Last edited:
The problem is lack of resolution - not too much of it, already tried it

FTR - with kit I have to hand and have tried
  1. Thrustmaster Warthog (Throttle unit) - 14-bit Throttle, 10-bit Trim, 10-bit Mini-stick
  2. Logitech G C56 (Throttle unit) - 10-bit Throttle, 4 x 8-bit Rotary, 8-bit Mini-stick
  3. Thrustmaster T-Flight HOTAS X - 10-bit Stick primary axis, 8-bit Twist, 8-bit Throttle, 8-bit Rocker
The 10-bit Trim control on the Warthog is not really good enough for absolute control of the tuner, 14-bit might be sufficient for absolute control but the overriding point is that it still would not address the underlying problem of lack of tuning memory.
Granted about the tuning memory. It doesn't worry me the way I work, but I get that it does you.

On the other hand, I have 8-bit resolution and absolutely no issues. It tunes perfectly every time and I often don't have to change it to pick out a succession of ice worlds.
Here it is in use. It's not the most exciting system and I'm not trying to be efficient. I don't know why I hesitate at the beginning and zot the tuning around, maybe to test out the acceleration. But it shows how precise and repeatable the tuning is.
 
On the other hand, I have 8-bit resolution and absolutely no issues. It tunes perfectly every time and I often don't have to change it to pick out a succession of ice worlds.
8-bit is fine for relative tuning but not absolute, that is even with a rock-solid stable 8-bit rotary switch.

Try doing extended USS hunting sessions and you will start to see the issues with lack of tuning memory.

[EDIT]I suggest you read my recent edit of my post that you quoted.[/EDIT]
 
8-bit is fine for relative tuning but not absolute, that is even with a rock-solid stable 8-bit rotary switch.

Try doing extended USS hunting sessions and you will start to see the issues with lack of tuning memory.
My control and that video use absolute control.
Since that control is only used by the FSS and the free camera, I effectively have tuning memory most of the time.
 
My control and that video use absolute control.
Since that control is only used by the FSS and the free camera, I effectively have tuning memory most of the time.
I have no doubt an 8-bit absolute value will be fine for some cases - but it is not good enough for ALL cases. The tuning bar obviously has greater resolution than that and greater than 8-bit accuracy is required in at least some cases, there will be points of convergence where co-incidence plays in the favour of an 8-bit device but overall there are cases where it is insufficient.

The long and the short of it is that specialist hardware should not be required for tuning memory.
 
Last edited:
I have no doubt an 8-bit absolute value will be fine for some cases - but it is not good enough for ALL cases. The tuning bar obviously has greater resolution than that and greater than 8-bit accuracy is required in at least some cases.
No. The bar appears to adapt to the resolution offered by the controller. My controller advertises the control as -127 to +127. I know because I programmed it that way. I never, ever have any problems tuning. In fact it appears to require less precise positioning than it did when I used the HOTAS throttle. I do do a fair amount of FSSing, both in and out of the bubble. If there were problems then I would have noticed and extended the range.

If a HOTAS has an 8-bit control but advertises it to the USB driver as, say -512 to +512 (10-bit) and scales the value accordingly, that may very well give you problems tuning because one tick would change the value by 4.
 
No. The bar appears to adapt to the resolution offered by the controller.
Unlikely and does not match onto reality - I have looked at the raw input from the rotary controls on the X-56 for example and it is clearly an 8-bit device where the inputs are concerned. The same goes for the trim control on the Warthog.

Ultimately, this side-discussion is pretty moot FD should not require a physical device for tuning memory - period.
 
Really? I use the large cheap plastic turning things on the x56 throttle for tuning and have no problems with it. There's quite a bit of allowance where it will still work, you don't need precision?

The act of tuning like the rest of the fss though feels like filler. There's no sense of accomplishment from doing it.. its about as complicated as walking to the door.
 
Unlikely and does not match onto reality - I have looked at the raw input from the rotary controls on the X-56 for example and it is clearly an 8-bit device where the inputs are concerned. The same goes for the trim control on the Warthog.

Ultimately, FD should not require a physical device for tuning memory - period.
I agree you should need to have a particular set-up to get tuning memory, but 8-bit absolute works better than any other method for me. I don't use it to get tuning memory, that's just a nice-to-have. If existing HOTAS hardware cannot get the same ease-of-use effect that I get then that is a shame and a bug in the FSS. FD should not require that their users are real-time-control programmers willing to put together bespoke hardware.

The fact that the physical device is 8-bit doesn't guarantee that it is provided to the USB as 8-bit. I can't find any diagnostic tool that will show the details of a game-pad device.

Have you watched my video? I'm not saying you're doing it wrong, but maybe we can agree that 8-bit absolute works well with my hardware?
 
Another option: Frontier could add some optional ADS-like mechanics to fleet carriers. (Especially so that "Exploration support ships" wouldn't just be mobile UC stations.) If your carrier is within 500 ly, have the ability to do an ADS honk - perhaps consume something to do so.
But really, anything that would give back no longer existing functionality that could be tied to fleet carriers would be good. What would the argument be against it anyway? Either "I don't like that fleet carriers can do something my ship can't", which they will be able to do anyway, or "I don't want people to use anything remotely similar to the ADS even if I won't have to", which very few people say anyway.
As a bonus, adding back functionality elsewhere instead of reworking the FSS allows Frontier to save face too, by pretending that all's right with it.
 
The act of tuning like the rest of the fss though feels like filler. There's no sense of accomplishment from doing it.. its about as complicated as walking to the door.
Don't you understand? People shouldn't be allowed to replace doors with archways! It gives them an unfair advantage walking between rooms!
 
[JOKE]You must be a masochist then. ;)[/JOKE]

I never said that the ADS is the bee's knees and nor has anyone else in essence - all that has been said is that the ADS is better than the FSS on one-hand (subjective) and on the other facilitates an experience that was lost with the FSS (factual). Further more, the greatest objection is the removal of choice - the FSS is too pervasive and badly implemented to be legitimately considered a good move for ED. FD avoided such a fractious move with mining but failed to do so with exploration.

This is however mostly moot - what is clear is FD needs to rework the FSS to address the broad critique of it. Some aspects require more urgent attention than others and these aspects should not require that much effort (i.e. are on the most part supplemental tweaks).
Why should any user be given a 'choice' when the developer implements new features? Your opinion of the FSS (only opinion remember) is yours alone - as is mine - fortunately I am able to cast aside what was and to work with what is...

Perhaps it was a bad idea for mankind to come down from the trees after all :)
(with apologies to Douglas Adams :))
 
Unlikely and does not match onto reality - I have looked at the raw input from the rotary controls on the X-56 for example and it is clearly an 8-bit device where the inputs are concerned. The same goes for the trim control on the Warthog.

Ultimately, this side-discussion is pretty moot FD should not require a physical device for tuning memory - period.
I use the yaw function on my X52 Pro for tuning, not had any issues to date...

As for the bolded bit - why? Period.
 
Ok, this thread is far too long to read, but my 2 pence is:

How about if the ADS was reintroduced, but as an expendable resource.

So pilots, could stock up on ADS (rounds? or something similar) the user could then use the FSS or decide to use one of their (precious) ADS rounds.

Perhaps on exploration trips you could synthesize the rounds, but generally the user would have to use the FSS to discover stuff.
 
Top Bottom