Maintain Position/Ballast Bug?

[shrekstyle]

Before the update the ballast would correctly work if you flood an airlock the ballast would keep the sub neutral if it could, with maintain position.

One of my subs has 2 airlocks that are different sizes and either one (not both at the same time) could be flooded and make the ballast go to around 5-10% full to keep neutral buoyancy, now when i flood the larger airlock the sub starts to sink quickly and the Ballast is about 25% full, with maintain position on. (it can just "only just" keep the sub neutral with the smaller airlock flooded)

If i go to manual control and make the sub move up it empties the ballast tank and makes the sub rise then i can fine tune it to make it neutral. but once i switch to maintain position it adds more to the ballast and makes it sink again.

i have checked the ballast numbers of the ballast hull rooms and they are still set in the nav-terminal correctly.

I have tested it in the editor with a test sub built just for this and it remains a problem where maintain position wont keep neutral buoyancy if it should be able to, but manual control can easily.

short version : the sub wont maintain position if the airlock is too big even if there is enuff ballast to allow it.

[daddy O]

Bug Report: Airlock breaks Autopilot

After the Autopilot was turned on, set to "maintain position", and managed to stop the ship, toggling an airlock that fills with water will make the sub fall indefinitely. The autopilot adjusts the ballast a little bit for the filled airlock, but it is not enough (there is water left to drain in the ballast!). Please fix this, it is very annoying and can be game breaking.

[daddy O]

I second this. I have an airlock that has 2 states: open and filled, closed and empty. When autopilot is on and managed to stop the ship, toggling the airlock (opening and filling it) will make the ship fall indefinitely, even though the autopilot is set to maintain position. The autopilot adjusts the ballast a little bit to accomodate the filled airlock, but not enough. Please fix this bug.

[thesecretbookofsaxon]

Having the same issue. Ballast will not fully empty any more when airlock floods.

[G0D]

Publicado originalmente por Deceptive Pastry:

Something is definitely wrong with the autopilot maintain position. I have a mini-sub shuttle with an airlock small enough that it can be full of water yet the sub can remain neutrally buoyant, which it did so before the update. If I fill the airlock and set to maintain position, the ballast will only empty to about 4% and the sub will sink at a rate of 2 km/h, however there's still an extra 4% of ballast that it isn't utilizing. If I switch to manual and drag the arrow fully upward, the ballast will empty completely to 0% and the sub will stop sinking, actually still having enough ballast left to RISE even with the airlock full. For some reason autopilot is not utilizing the full capabilities of the ballast.

Good explanation, I hope the dev sees this.

[J]

Can confirm this bug as well. Was planning a weekend session with some friends, but this is absolutely game breaking.

[Zenthar]

On our Humpback campaign, we've noticed our ship feels much heavier after the new update. Going up and down feels much harder and slower.

[Amberbaum]

- Basically maintaining auto-position leads to the ships slowly drifting down. Doesn't matter if we checked if the airlock is full of water or not. We were constantly second guessing if we had too much heavy stuff on board.

- Diving down is crazy fast, ascending is at snail's pace. We were on a trip with 3 people, two of us died, so our captain had to drive the ship to the next outpost on their own. He was right in front of the airlock of the outpost and activated the autopilot to dock. One crash at the rear was enough, after the reactor went offline due to not enough fuel, to cause the ship to sink down 1000m in 15 seconds or so. The crash happend after the reactor went online again, btw. The sweet pea is small so our captain was able to stuff a fuel rod into the reactor and fix the back of the ship fast enough. The amount of water that came into the ship should have never caused our our ship to drop 1km that fast. It took him 5 minutes or more just to dock at the station again.

[Master-Miner]

Seems like the setTargetLevel is capped between -75 and 75. So I added a multiply by 1.33 component to it and now in hold position mode its draining the tank again like I expect it for my sub with side opening ballast.
Manual control seems to not care with the values >100 and <-100 so this is a quick fix to get back to "normal" behavior.

[G0D]

Publicado originalmente por Dynamoon:

Hi! I believe this is the same thing I just responded to here https://steamcommunity.com/app/602960/discussions/1/3130541122082036828/#c3130541122084108349 – in short, it's not a bug, but a fix. However, we unfortunately forgot to include a mention of it in the changelog.

Publicado originalmente por Dynamoon:

We've chaged how ballast pumps work (and it seems we also forgot to include this change in the update changelog – sorry about that.) Previously the pumps wouldn't always utilise the full volume of the ballast which caused the sub to be less responsive than it should, and prevented the sub from maintaining position (the sub moved up and down around the position it should maintain).

If this change is proving inconvenient for you, please give feedback as always – it certainly shouldn't mess with default subs in any way, so if your'e having problems, please describe them in detail and we'll take a look!

Okay at this point we have quite a few people saying they are having problems. Any chance of it getting tweaked or anything?

[daddy O]

There seems to be a relation to whether the command terminal is in use or not when the airlock is toggled. When Im at the terminal, and airlock was toggled, ship stops, as soon as I leave the terminal the ship starts dropping again. As if the TargetLevel was influenced by command terminal being used or not.

[shrekstyle]

Publicado originalmente por daddy O:

There seems to be a relation to whether the command terminal is in use or not when the airlock is toggled. When Im at the terminal, and airlock was toggled, ship stops, as soon as I leave the terminal the ship starts dropping again. As if the TargetLevel was influenced by command terminal being used or not.

Yep this happens as seen in my test when the airlock has water in it.
If you put a water sensor in the ballast and link it to a text display near the terminal you can see what level the ballast goes to when not viewing the command terminal.
If you have a status monitor linked to the command terminal it will do the same thing when viewing the status monitor.

[Deceptive Pastry]

I think I see what's happening. I believe it does have something to do with them extending the set_targetlevel range from -50/50 to -100/100 but they didn't properly update the autopilot to use the new range. I hooked up text displays to the nav terminal y output, as well as a water detector in the ballast to see what is going on. I also set the neutral ballast to 0.5000 so neutral output is 0 and the autopilot isn't making adjustments in either direction.

When NOT viewing the terminal and maintain position is on, as in dragging the cursor far in one vertical direction and immediately leaving the nav terminal, the output range is only limited to -50/50, which gets you a ballast range of 25% - 75%. It will not go any further in either direction.

When you ARE viewing the terminal and you do the same thing, the output range is limited to -75/75 (actually -74/74, perhaps rounding issue). This gets you to a ballast range of 13% - 87%.

So basically if you have maintain position on and aren't actually viewing the terminal, your ballasts can only empty to a max of 25%, if your ship needs less ballast to stay buoyant it will sink.

[Deceptive Pastry]

Publicado originalmente por Master-Miner:

Seems like the setTargetLevel is capped between -75 and 75. So I added a multiply by 1.33 component to it and now in hold position mode its draining the tank again like I expect it for my sub with side opening ballast.
Manual control seems to not care with the values >100 and <-100 so this is a quick fix to get back to "normal" behavior.

This does seem to be the case when nav terminal is in use. When nav terminal is not in use the range is further limited to -50/50.

Your fix actually seems like a decent idea for now. You may want to try experimenting with multiplying by 2, in which case things will be almost how they worked before, where in manual control your ballasts max out at 50% input and dragging the arrow any further vertically is wasted. Though that may be to worth it to keep the ship floating when parked and not using the terminal.

[Deceptive Pastry]

shrekstyle, your test sub has neutral ballast set to 0.4254. So the nav terminal is outputting -14.92 (roughly) by default on manual control to achieve neutral ballast at 43% (rounding). Formula to figure out percentage is ((output + 100) / 2). Max range when NOT viewing terminal is -57.46/42.54, which results in a ballast range of 21.27%/71.27%, or 21%/71%. You'll notice, -14.92 / 2 = -7.46, which is roughly what the autopilot range is being offset by. It seems like when NOT viewing the terminal, the autopilot is dividing all output by 2 which would have been necessary with the old set_targetlevel range of -50/50. What is causing -75/75 when you ARE viewing the terminal, I'm not sure.

[shrekstyle]

Publicado originalmente por Dynamoon:

This may change the behaviour of ballast-related setups you've had in the past. Do you think that would explain what you're experiencing?

Hey Dynamoon this is a great explanation of whats happening.

Publicado originalmente por Deceptive Pastry:

shrekstyle, your test sub has neutral ballast set to 0.4254. So the nav terminal is outputting -14.92 (roughly) by default on manual control to achieve neutral ballast at 43% (rounding). Formula to figure out percentage is ((output + 100) / 2). Max range when NOT viewing terminal is -57.46/42.54, which results in a ballast range of 21.27%/71.27%, or 21%/71%. You'll notice, -14.92 / 2 = -7.46, which is roughly what the autopilot range is being offset by. It seems like when NOT viewing the terminal, the autopilot is dividing all output by 2 which would have been necessary with the old set_targetlevel range of -50/50. What is causing -75/75 when you ARE viewing the terminal, I'm not sure.

Cheers for the help Deceptive Pastry in testing this. I hope to see it fixed soon.