Haters gonna hate, and shills gonna shill. The moderates get drowned out...
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
FSD Beta 10.69
- Thread starter Buckminster
- Start date
Haters gonna hate, and shills gonna shill. The moderates get drowned out...
The plural of internet anecdote is not data.
powertoold
Active Member
69.2 is sooo much better at unprotected turns vs 10.12. With 10.12, I was constantly having to brake disengage at unprotected rights. On unprotected lefts, it would take too long and/or be extremely jerky, esp in suburban streets.
It’s better, but I’ve still had to brake disengage a couple times. It’s also still too slow if you have cars behind you. I have to goose the throttle to push it through in those cases. The time it takes to creep is still too long when there is clear visibility
powertoold
Active Member
For me, it's wayyy better, night and day. The confidence instilling slow creep, the decision making, etc. It's definitely not perfect, but it's 5-10x better than 10.12 for the route I'm doing.
It struggles most on unprotected rights with islands. But that has always been the case.
MrTemple
Member
The argument is that these people don’t exist. That people don’t drive FSD more now that it is better.
Dozens and dozens and dozens of people saying, “I drive more now that it’s better!” abolutely
is data which counters that opinion.
sleepydoc
Well-Known Member
it's hard to know what to make of "dozens and dozens of people" saying anything on the internet. Often times its true, but you have no idea of the veracity or whether it's a representative sample so drawing firm conclusions can be dicey.The argument is that these people don’t exist. That people don’t drive FSD more now that it is better.
Dozens and dozens and dozens of people saying, “I drive more now that it’s better!” abolutely
Adding my own anecdotal data point, I've found 10.69 to be an improvement from 10.12 and am more likely to use it simply because there were more times with 10.12 that I knew it wouldn't work and more times when I didn't have the patience for it.
I think one of the key point in using FSD is to get used to what it is capable of and where it's limit lies.
I'm becoming more and more letting it do what it wants to do. Although not as smooth as it would be if I do it manually, it is tolerable and even becoming borderline comfortable in most cases. Same for creep based turns.
I think I'm trusting FSD more in situations where I feel it can handle it.
I'm becoming more and more letting it do what it wants to do. Although not as smooth as it would be if I do it manually, it is tolerable and even becoming borderline comfortable in most cases. Same for creep based turns.
I think I'm trusting FSD more in situations where I feel it can handle it.
Well, twitter is rife with fake accounts, so some of those people might not exist. But, it doesn't matter. Cherry-picking anecdotal reports from individuals saying things like, "It's 10x better than the previous version" would seem to by just slightly hyperbole. And, the guy you quoted lauding over 10.69.2.1 could not have tested it very long at all before posting his praise. Not likely to have been a very thorough test. Or, maybe he doesn't really know what software he has?The argument is that these people don’t exist. That people don’t drive FSD more now that it is better.
Dozens and dozens and dozens of people saying, “I drive more now that it’s better!” abolutely
In any event, a bunch of twitter posts from random people is worthless. Absolutely worthless.
10.69.2 is improved over 10.12.2 but has plenty of warts as I have detailed in previous posts. But I'm another rando on the internet and you have no idea if I even have FSD beta, let alone whether my reports are accurate.
sleepydoc
Well-Known Member
I have a scientifically foolproof method of determining accuracy:
- If I agree with you your report is accurate
- If you post something negative that I disagree with you are a troll
- If you post something positive that I disagree with you are a fanboy.
AlanSubie4Life
Efficiency Obsessed Member
A couple more examples of sudden brake-filled stops with FSD Beta 10.69.2.
I would estimate 25% of stops are like this (which means they are jerky with pulsing regen, not necessarily using brakes) on the flat, perhaps 10-20% on uphills, and about 50% (or more?) on downhills.
Uphills definitely seem to be easier for it. It’s not the only reason for it, but it is one of the reasons. Not sure why they can’t handle it.
Downhill:
Flat (note I had released the accelerator well before any slowing was needed, though unfortunately I don’t verbally mark that in the video):
I would estimate 25% of stops are like this (which means they are jerky with pulsing regen, not necessarily using brakes) on the flat, perhaps 10-20% on uphills, and about 50% (or more?) on downhills.
Uphills definitely seem to be easier for it. It’s not the only reason for it, but it is one of the reasons. Not sure why they can’t handle it.
Downhill:
Flat (note I had released the accelerator well before any slowing was needed, though unfortunately I don’t verbally mark that in the video):
powertoold
Active Member
The green regen bar is not indicative of the jerk or g forces. I spent some time looking at Alan's videos before getting 69.2, and then once I got it, I started paying more attention to the green bar.
The green bar is not the problem with 69.2, it's more of an "on-and-off" robotic approach to accel / decel, makes the drive more nauseating. The transitions between accel and decel are robotic, rather than smooth gradients.
The green bar is not the problem with 69.2, it's more of an "on-and-off" robotic approach to accel / decel, makes the drive more nauseating. The transitions between accel and decel are robotic, rather than smooth gradients.
AlanSubie4Life
Efficiency Obsessed Member
Show your work.
This is not quite right. Of course the bar is not the problem, it is what it represents.
And the (rapid) changes in the bar length are the primary problem. Which you’re describing as “accel/decel” but is actually usually varying degrees of decel (stays negative).
Which is what I keep saying is the issue.
DanCar
Active Member
Jerk is defined as change in acceleration. The greater the change the more jerk. The bar itself does not indicate jerk but quick and large changes in the regen bar do indicate jerk.
powertoold
Active Member
I don't think it's worthwhile. It seems like you have an exact preference for decel / accel / aggressiveness that fsd beta isn't meeting.
I don't agree with many of your comments in the videos, about how slow it is or the green bar, etc. Especially the downhill one, where regen is higher when slowing downhill. And I don't think the car was slow to creep, especially when compared to 10.12.
AlanSubie4Life
Efficiency Obsessed Member
Yes.
You want to see a nice, reasonably gradual increase to a peak (it by necessity can be quick, but a nice smooth, but rapid, increase is good rather than a jolt), then a linear decrease (display is likely not linear though) as speed slows, if you want uniform deceleration. Uniform deceleration means linearly decreasing stopping power (F*v).
In practice, often deceleration can and should be lowered gradually (makes it easy to avoid jerk at the end), so the drop in stopping power may be faster than linear.
The issue, as you have pointed out, is the jumpiness. Inherently, a certain amount of stopping power is required. The integral over time of this stopping power has to be roughly the same in all cases (neglecting wind resistance, etc.).
So when we see a sudden spike, followed by a period of minimal regen, we know that some of that high regen could have been redistributed to other times to make the integral the same - but with lower jerk (less rapid changing of the regen bar length). Just a nicer, more consistent (constant), reassuring stopping force.
(Of course, there are limits to allowed regen at different speeds, at least when driving manually. I don’t know whether adjustments are made to these limits for FSD. But my point is that there are limits at lower speed, so you do need to make sure to take advantage of the available stopping power all the down.)
It has little to do with the length of the regen bar. Large amounts of regen are fine. It has to do with how it gets to that peak, and how the length is reduced all the way to a stop.
You can see this in the video of me driving, posted earlier, but I’ll post more examples for side-by-side comparison.
It’s actually a very common complaint; from the survey it seems likely that more than half the people with Beta are likely annoyed by this.
It’s not an exact preference. Some braking even is ok, if someone likes really fast stops! But it should be done smoothly. Maybe I can post an example of that too (except unfortunately the display does not show braking when driving manually so it is hard to see what is happening).
Last edited:
powertoold
Active Member
Isn't that what I was saying about it being robotic about accel decel? It seems like 69.2 is trying to find it's way to the stop by going and slowing down and going but doing it more robotically than a human.
AlanSubie4Life
Efficiency Obsessed Member
Yes, I already said that what you said there was what I have been saying for a long time now (I can’t count how many times I have said it is about the CHANGES in the length of the bar). But in that “robotic” statement you also mentioned transitions between accel/decel:
That’s not really what is happening in a stop. It’s all decel. The issue is that it decelerates at varying rates, which is known to be uncomfortable to people.
The example I gave earlier is a decent one, though more minor than what happens with FSD: “warped” brake rotors. Those create a juddering. You’re obviously decelerating the whole time, but there is a very rapid oscillation in magnitude of the deceleration “on top”, which is what people feel, and it is very bothersome. FSD is similarly bothersome. Quite a different feel (a much slower oscillation, but the same issue - jerk (change in (negative in this case) acceleration)).
The question remains: why? Some think it is deliberate (and maybe it is not a priority right now), but there are Tweets from Elon about this general problem which suggest it might just be a tough problem for Tesla.
bhakan
MSLR MSM/Blk/19/FSD, Ord 6/10/21, Del 2/20/22
Got my first time out last evening driving in the evening. Even tho I was holding the yoke it kept on beeping at me and got my first of the five timeouts. It was not at all my fault as the yoke will get seizure and make all the crazy movements. I reported each time when it made a mistake. It was the same drive I had done numerous times. I disengaged multiple times in my return as it was not worth it. The time out occurred when I drove hardly a mile.
Similar threads
- Article
