Randommer/Interestinglier/Weirder - now more of everything for the same low price!

[majestic]

 

[Gorth]

Having tinkered around a bit with Stable Diffusion (and tearing out my hair in frustration at times), I've also been keeping an eye out for interesting videos on the subject of AI image and video generation. Ran across this earlier today, being impressed by the AI algorithms interpretation of the characters of one of my favourite Animes.... in a what if they had been in a live action movie (without the at times exaggerated  proportions and stylistic faces of Anime).

edit: My own observations and constructive criticism of whoever fed the information to the AI

Starks red iris' are overdone. Needs desaturation

Eisen is showing too much of his face

Flamme is missing her two "signature" earrings, that Frieren wears after her death

Ubel is showing too little armpits

Denken looks too tall. He's supposed to be short and very stocky (and he's missing his signature monocle)

Kanne is the same as Stark, the iris colour is overdone. Needs desaturation

Methode looks too "youthful". She is the curvy, "older woman" adventurer that Sein always dreamt of meeting (ironically leaving the party just before the introduction of Methode)

Himmel his face is slightly too angular. He has noticeable "soft" facial features to give him that cute baby look in the Anime

 

 

[Raithe]

 

[Malcador]

47 strikes in NYC.

[Gfted1]

Face of real Santa Claus reconstructed with 3D tech after 1,700 years

[Gfted1]

SpaceX | Wishing you a warm holiday season and a lit New Year | Instagram

Go full screen and turn on volume.

[rjshae]

 

[Azdeus]

I hope it goes well for him

[majestic]

On the one hand, yeah, honey is a scam run by PayPal, no less, on the other hand, anything that scams scummy influencers and scummy shops out of their money can't be so bad. Then again, that scammed money is going straight to PayPal, so that's bad, but on the other hand, it's money that would otherwise go to scummy influences and scummy shops, but... error, maximum loops reached. Aborting.

That's a tough one.

Edited January 6 by majestic

[Azdeus]

12 hours ago, majestic said:

On the one hand, yeah, honey is a scam run by PayPal, no less, on the other hand, anything that scams scummy influencers and scummy shops out of their money can't be so bad. Then again, that scammed money is going straight to PayPal, so that's bad, but on the other hand, it's money that would otherwise go to scummy influences and scummy shops, but... error, maximum loops reached. Aborting.

That's a tough one.

If it only were scummy influencers and not legit content creators aswell getting cought up in **** like this

[Malcador]

2 hours ago, Azdeus said:

If it only were scummy influencers and not legit content creators aswell getting cought up in **** like this

Acceptable collateral damage

[Gorth]

Let the conspiracy theories begin… who could have an interest in an airplane crash remains unsolved?

Both black boxes are suddenly missing the data from *before* the bird strike until the crash

https://www.bbc.com/news/articles/cjr8dwd1rdno

 

[Zoraptor]

"Flight data and ****pit voice recorders from the South Korean passenger plane that crashed last month stopped recording four minutes before the disaster," is not surprising or suspicious so long as that isn't before the reported bird strike. Which it wasn't, supposedly the bird strike was 5 minutes before the crash.

(the plane certainly seemed to be landing in an uncontrolled state- way too fast (?looked like no flaps deployed so no hydraulics?), too steep, too far along the runway plus most obviously no landing gear deployed, which suggests a very significant adverse event that could well cut off power to the tail section- iirc where both the cоckpit and data recorders are located. Seems almost certain the hydraulics were out, and if my memory of Mayday/ Air Crash Investigation is correct that system has a lot of redundancy built in to it. If i had to speculate, blown turbine disc. That's basically a big bunch of shrapnel flying off at high speed)

[Gorth]

1 hour ago, Zoraptor said:

"Flight data and ****pit voice recorders from the South Korean passenger plane that crashed last month stopped recording four minutes before the disaster," is not surprising or suspicious so long as that isn't before the reported bird strike. Which it wasn't, supposedly the bird strike was 5 minutes before the crash.

(the plane certainly seemed to be landing in an uncontrolled state- way too fast (?looked like no flaps deployed so no hydraulics?), too steep, too far along the runway plus most obviously no landing gear deployed, which suggests a very significant adverse event that could well cut off power to the tail section- iirc where both the cоckpit and data recorders are located. Seems almost certain the hydraulics were out, and if my memory of Mayday/ Air Crash Investigation is correct that system has a lot of redundancy built in to it. If i had to speculate, blown turbine disc. That's basically a big bunch of shrapnel flying off at high speed)

No idea how long it was in the air after the bird strike. I was lazy and just took the time from the pilot reporting the bird strike (8:59) to the crash (9:01) and concluded 3 minutes

https://www.bbc.com/news/articles/ckgzprprlyeo.amp

Weirdest thing I do remember from the 3 videos I’ve seen is the one from the ground showing the actual bird strike also shows flaps and gear deployed, the two different angles showing it crashing both showing it having power on the damaged engine (producing thrust and thrust reverser deployed) but the supposedly undamaged engine being inactive 

[Zoraptor]

That BBC article has the actual crash at 9.03, the crash is when you'd expect the data to cut out so it is 5 minutes. They've certainly worded it somewhat confusingly though. The Indie has a less confusingly written timeline confirming the 5 minute gap.

(There was certainly something odd going on, but it could have been something as simple as both engines getting hit in the same strike- like the plane that ended up landing on the Hudson in New York. Without hydraulics/ power flaps will retract 'naturally' since the aerodynamic forces will always try to push them back flush with the wing. Similarly, while gravity will try to push landing gear and its bay doors down aerodynamics will push them back up. I'd say that a reverser working and the flaps not might be unusual but while I know landing on one reverser is safe I'd imagine landing with only one wing's flaps deployed would be extremely dangerous if possible at all due to the asymmetric lift it would supply)

 

Edited yesterday at 05:45 AM by Zoraptor

[Gorth]

3 hours ago, Zoraptor said:

That BBC article has the actual crash at 9.03, the crash is when you'd expect the data to cut out so it is 5 minutes. They've certainly worded it somewhat confusingly though. The Indie has a less confusingly written timeline confirming the 5 minute gap.

(There was certainly something odd going on, but it could have been something as simple as both engines getting hit in the same strike- like the plane that ended up landing on the Hudson in New York. Without hydraulics/ power flaps will retract 'naturally' since the aerodynamic forces will always try to push them back flush with the wing. Similarly, while gravity will try to push landing gear and its bay doors down aerodynamics will push them back up. I'd say that a reverser working and the flaps not might be unusual but while I know landing on one reverser is safe I'd imagine landing with only one wing's flaps deployed would be extremely dangerous if possible at all due to the asymmetric lift it would supply)

 

I don't have a link at hand, but I wonder if that matches up with the time flightradar24 lost track of the plane (which was some minutes before the crash, it disappeared and could no longer be traced)

 

edit: https://www.flightradar24.com/blog/jeju-air-2216-muan/

"It appears that ADS-B data was either no longer sent by the aircraft or the aircraft was outside our coverage area after 23:58:50 UTC."