I’ve had three false positives so far on the “spaghetti detection” printing this model:
https://www.printables.com/model/209121-cali-dragon
I know about the sensitivity adjustment. Since it’s just this model, I don’t want to fiddle with it (yet). However, is there a way we can give feedback about false positives to the Bambu team? For example, giving additional training data to the AI engine. Or even, is this being done already when you say “Ignore” ??
IDK if you’re aware of the discussions about lighting and spaghetti detection.
I haven’t had the problem (I rarely have used it) but some are reporting that room lighting, especially bright sunlit rooms, can confuse the LIDAR into giving false positives.
I’ve also cautioned against using bright lighting / adding LED strips to “brighten the too dark chamber”, etc.
Could this be a problem for you?
EDIT: My Wife was just asking me to print this a couple days ago … 
I’ve considered room lighting, but this was happening in a dark room with only the LED lights. The spaghetti detection uses the camera, I believe, and not the LiDAR. (Correct me if I’m wrong here!)
And it’s happened on that model only, and three times so far (out of nine). So, I kinda felt the AI could do with some more training data.
Here’s my samples (so far!) I think they’re much cuter than Benchy.
I am not spaghetti! Perhaps cavatappi, but not spaghetti! 
(This photo is from a print job that had just “soft reported” a possible spaghetti, and soon afterwards paused the print - a “hard stop” - for the same reason)
Isn’t the spaghetti detection done by the webcam and not the lidar? I have some troubles with some models too, but usually you just can hit resume and all will be fine.
My spaghetti detection does the opposite of what it is used to do.
When it looks like an Italian restaurant in my printer it continues making more spaghetti. But when I see no spaghetti or “micro-hair” it gives a warning from time to time.
I’m tempted to disable AI for now.
Still not sure how/when/why the spaghetti detection gets triggered.
This also happens when the room is dark with only the build in lightning from the X1C enabled.
It’s AI, meaning noone knows how/why it gets triggered
Humor aside, it’ll all depend on the training data, which is why I hope Bambu is open to accepting more false positive/negative inputs to improve the AI.
Yeah, I do believe that’s right. The LIDAR is UV and the camera of course is wide spectrum. But the “too much light” thing still can be a concern.
Just as a WAG, I could see it being confused by the multiple shadows created from the multiple light sources.
Yep! 
Yeah, it’s “AI” alright. Whenever I hear AI I’m always reminded of the old joke that “Military Intelligence” is oxymoronic. AI is the same.
I printed a couple of calidragons. The (transparent) blue one gave a false spaghetti warning (during the day in a bright room) the others completed fine. I think it was 50/50 for a dark room and a room with lights on. Maybe one other with bright sunlight.
I printed the purple (transparent, TPU) some times and some had heavy stringing (filament got wet real quick and only prints fine when absolutely dry) but no warnings.
So yeah. Its “AI” (Always Interesting 
)
