Okay no way, I was actually just wondering if anyone had made a fully afk-able copper oxidizer earlier today. Of course you are the one who was smart enough to do so, and you went above and beyond in the design and theory behind it. Amazing work as always.

14:56 “Ilmumbo” Damn that’s cursed💀

I love the way by which you can choose the state most blocks will end up as, but as a builder I often let everything oxidize to its maximum state and then just axe it to what I need later. Definitely streamlines the process.

Have to say, that is an ingenious use of powdered snow. Your designs never fail to impress me.

Trying to derive the optimal times analytically (using some values from the wiki) I get the following function to maximise: (chance of block being fully oxidised after a time period, divided by time) (1 - Sum[((1 - 1/4096*64/1125)^(3*20 (60 m)) (-(1/4096)*64/1125)^i Binomial[(3*20 (60 m)), i])/(-1 + 1/4096*64/1125)^i, {i, 0, 3}])/m (m in minutes, mostly because trying to do the same with ticks got me into numerical accuracy issues during calculation) Which gets me an optimal time of around 67.7 minutes, or 81207 gameticks. The difference between the values we got versus what you derived in the video could easily be explained as the difference that your approximation made, though admittedly it doesn't matter too much as the graphs are relatively flat near that maximum. For the other states I got 0 minutes for unoxidised (duh), 1 tick for exposed copper (or 0 in my math but clearly it needs at least a chance to convert), 20 minutes for weathered copper This then means that you optimally need a blockstream of roughly 20*60*67.7/2176 = 37.3 ticks per block (11 ticks for weathered) For reference, the following functions are the chances of finding the copper in each state after m minutes: Copper Block: (71999/72000)^(3600 m) Exposed Copper: (9/71999)^(1 - 3600 m) 20^(2 - 10800 m) m Weathered Copper: 5^(2 - 10800 m) 8^(1 - 7200 m) 9^(1 - 3600 m) 71999^(-2 + 3600 m) (-1 + 3600 m)m Oxidised Copper: (1 - Sum[((1 - 1/4096*64/1125)^(3*20 (60 m)) (-(1/4096)*64/1125)^i Binomial[(3*20 (60 m)), i])/(-1 + 1/4096*64/1125)^i, {i, 0, 3}]) (Or alternately, Oxidised Copper: 1 - the above 3 formulae, its just different notation) I had a lot of fun doing this math, so thank you for inspiring me to do so. Though... It does make me wonder if the derived times could be somewhat off if for example copper blocks don't oxidise while being moved via piston. And if it could be interesting to return the collected blocks back to where you place it so you can give unoxidised blocks a second go through the machine. Optimal times for THAT would probably be different as well...(and likely lower)

Big thanks to Pix for sharing this, such a great video and farm, love the way you broke it down, going to have to rewatch a few times, I suspect

I can’t wait for Docm77 or Xisuma to make this on hermitcraft. This is amazing!

Combining block streams with storage tech sounds like m³'s ultimate fantasy.

Powdered Snow is a great idea and works great if you have a Powdered Snow farm to supply that much. If you're a little short on resources in a survival world though, you can replace it with a simple Stone generator. The Stone or Cobblestone can get blown up in the blast chamber, or if you care about TNT efficiency, transformed into moss at the end of the blockstream and punched out by pistons. This also takes out the need for the closed loop system to recycle the Powdered Snow.

If is 2 in the morning where lm at, glad to finally find this

Genius as always. You’ve gained 20x the subscribers from when I first started watching you, but I say that’s still underrated.

Thanks for a great video once again! A very similar optimization problem is also encountered in other areas of Minecraft. In traditional flushing crop farms the optimal time to harvest and replant is obtained as an 8 stage equivalent of this. Sugarcane with its 15 growth stages is similar, but extra complexity is introduced by the fact that attempting to break a sugarcane that hasn't grown yet will not reset its internal state (btw, explaining the optimal sugarcane harvesting problem would be a great topic for a video). What makes copper different is the need to harvest the intermediate oxidization states as well - something which makes no sense in crop farms. In this farm in particular I can actually think of a similarity to the sugarcane harvesting problem as well. Assume the farm is run in its slow mode and that the buffer chests of partially oxidized copper are full. The copper blocks that enter the silos will now on average take less time to become fully oxidized, because of the intermediate oxidization stages mixed in the input stream. This should increase the optimal input speed from the one that assumes a 100% unoxidized copper input. Solving for the optimal input speed in this case would be an interesting problem. It seems that copper separated by 4 or more blocks oxidizes completely independently of its surroundings, which makes possible an analytic solution derived from binomial/poisson distributions. That said, did you base your curve fits on the experiment shown in the beginning of the video or did you derive them using the oxidization probabilities defined in the code? Did you estimate how much oxidization occurs outside the silos compared to the inside? I'm asking this to understand the extent to which the entire oxidization process is governed by the analytical solution that only describes the sparse configuration of copper inside the silos. Returning to the sugarcane harvesting problem, just as the optimal input rate here for harvesting the first oxidization stage is "as fast as possible", I recall that global maximum sugar cane harvesting rate was also obtained at a tiny (like 10-20 tick) harvest frequency. The reason why this was not "as fast as possible" was that during each harvest there were a couple of ticks on which growth was prevented by the harvesting piston head. Is something similar happening here? While the blocks are being moved forward in the conveyor belts, can the copper block oxidize or is this prevented because the copper block while moving is unable to receive random ticks?

the circuit at 7:20 absolutely blew my mind. along with the rest of the farm lol. amazing content as always !!

The copper block powdersnow wall tech was very clever.

I'm studying chemical engineering and to see you fitting your data using graphs is so satisfying, I've always loved Minecraft especially the redstone side and to see my love for engineering and Minecraft combine into one is so cool, what you do blows my mind

Wowee, this is great! I imagined this kind of thing would be possible, and I even made a video about a machine that would space copper blocks out using powdered snow when copper blocks were first added! I'm glad you've taken this idea to its logical end and built the full copper factory. It's truly a thing of beauty.

Always great to find someone using the term "exponentially" correct :D

absolute genius, enjoying the technical explanantion and line of thought behind these designs

Absolute genius well done man I just love it, your way of approaching this problem !

I cant wait for wavetek next episode!

Wow this is amazing! Also you’ve grown a lot recently congrats!

came here from Pix, and my mind is simply blown!!!!

I don’t know how watching someone talk about block mechanics can make me so happy but it does. Keep up the great work

Just found your channel today… I have to say, I’m completely amazed. You are an absolute genius sir. Thank you for blessing my day with this wonderful video. Sending love your way

Really cool and interesting, thank you!

Lol matlab for minecraft. Truly an engineer

Damn! Well Done Man! Very creative

I love your videos so much ! They are so interesting, keep up the good work !

Amazing design, congratulations :)

I'm disappointed you didn't develop an actual model but just randomly guessed functions. Outrageous!

What a day, log in to KZfaq and a new m³ video. 19/10 this is great

Your redstone mind is insane, how Love your vids

great video as always

Love the timing

Nice job Sean 👍🏼

Absolutely flipping genius.

Wowwwww I freaking LOVE ur video!!!!!

holy shit. the powdered snow stuff and cobble wall is literally so genius I cannot

you really build the best farms!

Well this thing is amazing. Always wondered if you could farm this.

You’ve become my favorite technical Minecraft player in the course of two months and I’ve been watching Minecraft content since 1.12. You do amazing work and I’ve been utilizing your farms ever since I discovered you

You made an error when dividing the exposed copper percentage by time. there is no way you get exp(-kx) from dividing x²exp(-kx) by x. It will have a maximum peak as well, different than 0. EDIT: i retract that. For anyone interested, we can find the formulas for the percentages of the different block stages by solving the following coupled ODE system: a' = -3/4*p*a b' =+3/4*p*a - p*b c' = +p*b - p*c d' = +p*c With some oxidization rate p and abcd the percentages of the different copper stages (a=unoxidized, d=fully oxidized). From that we get the following formulas for the shown graphs: a = exp(-p*0.75*t) b = -3*exp(-p*t) + 3*exp(-3/4*p*t) c = -12*exp(-p*t) + 12*exp(-3/4*p*t) - 3*p*t*exp(-p*t) d = 15*exp(-p*t) - 16*exp(-3/4*p*t) + 3*p*t*exp(-p*t)+1 Only c and d contain parts that look like x*exp(x), which makes them have maxima in abcd/t as well. Notice that a+b+c+d=1 at any given point. EDIT2: I don't get though why cubic would divide by time for the optimal harvest time. The highest percentage of exposed copper (31%) happens at the 23 minute mark (maximum of exposed copper curve), while the highest percentage of weathered copper (24%) happens at the 44 minute mark. Did a plot of the percentage curves and the cumulative curves here: ibb.co/1RnvTGT (i assumed p=1 for the time scale, the real p would be 64/1125/68.27/s ~ 1/1200s) It should be obvious that the optimal harvest point is at the moment where we have the most of the desired block. For example: At 22.5 minutes we don't have the maximum amount of 24% weathered copper but only about 17%.

I love the color scheme that the powdered snow created! Lol it looks like a white tron design 0:06

0:37 the word you're looking for is verdigris btw, the word for oxidised copper colour in the real world :) incredible video as always, I'm always waiting for the next one

This is just amazing in every ways except building it

Never thought about using snow, I rather thought about using for example a cobble gen, separating the cobble from copper in similar way you pickup snow, and then blow up unneeded cobble, or maybe recycle it back using a block stream. If you would recycle the cobble back, it's also a closed system, and it also doesn't require hard to obtain thousands of snow buckets, you just need random filler block. I love the whole finished product tho, it feels so great to see engineering applied on such a simple task, just because automation is the way. Maybe a small suggestion, there could maybe even a mode, which is used to "pull out" all the copper blocks from system, without needing to use more copper blocks to replace them. Basically just have snow being placed continuously until all copper blocks are pushed out.

I've been waiting for your take on this for so long!!! I really wanted to design one of these, but my laptop broke down a couple months ago and I couldn't get started on the project :( Tbh, I don't really like your solution to copper oxidisation. It seems far too complicated for something that doesn't need it. I like the way you thought about the amount of time to wait etc (I wouldn't have thought of that, I would have just brute forced it and checked every minute using world edit or something), but the powder snow system makes the system unnecessarily laggy. That's just my (hopefully constructive) criticism but the video was really fun to watch!

To be honest. This bloody legend right here is mojang material. If gnembon wasn't enough for mojang then this man is a must have.

I was thinking using stone generators to fill the blockstream would be significantly easier and loads cheaper to build. For continuous usage you only need one blast chamber but if you wanted to push input speeds for getting the first stage of oxidation or to use it in the ultra fast mode and just wait you'd need to invest in more blast chambers

Ngl I do love your Saxton Hale skin

super useful

I am informing Docm77 about this right now.😃

Me in math class: I will never need this. Teacher: You play Minecraft?

Nice video

sometimes we wont use the machine for a specific type of copper, because most people would probably want some of each type that they can use when needed. in the graph shown there was a point where the percentage of each type of copper was roughly equal, so there should be a setting where the copper is exploded at around that time.

Imagine making this but you only need like a few stacks.

Your only the second person ive seen do this, but definitely the first to make it this large

Keeping in mind that blocks of higher oxidation state "spread" to neighboring blocks, wouldn't it make sense to place fully oxidized blocks as seeds into the storage area? Great use of powdered snow, btw.

This is so big brain , your approache to solving stuff is genius. Im also a big fan of your music and editing

Found an ilmumbo enjoyer

I am a friggin engineer and I haven't used this much math at work let alone to learn how to create a copper farm in my world

Imagine a caveman watching this. Dude crazy video

For producing maximally oxidized copper, it should be faster to harvest at the maximum placement rate, and instead rely on a sorting system capable of directing the incompletely-oxidized blocks back to the player to be placed back into the input stream a second (or third, or fourth) time. Breaking and re-placing a partially-oxidized copper block doesn't "lose" any progress, but fully-oxidized blocks occupying space in the system _do_ represent an efficiency loss.

waxed lightly weathered cut copper stairs

Nice! I only have one thing to note; can't you just use a dropper to let the buckets fall down instead of having these massive hopper lines?

it's an interesting use of powdered snow, but i think i'll stick to building with unoxidized copper and letting my builds weather.

nice

okay there Saxton Hale

Imagine getting all that powdered snow

All things considered with how long you would have to wait for fully oxydized copper, it would be better to run a 22 minute cycle, and then place down all of theblocks from the middle two stages which have the same math but advanced on or two stages, and harvest your almost pure weathered copper after 30 minutes.

This is a lot of maths and I only understand half of it but I will trust you

How much powder snow would be needed to start this farm?

when i saw ilmumbo, i absolutely busted out laughing

"Lets get in to some technical problem solving" by far one of best thing to do

As anyone noticed ilmumbo in the video

Epic

would you be able to make the system run much faster if you put down already fully oxidized copper in the empty spaces next to where the copper waits to oxidize in the belts?

copper production and oxidation factory sponsored by cubicmetre

Amazing build! Clever use of powdered snow and I loved seeing your thought process behind the math. Do you have Matlab for work or did you buy it for personal use? I get it from school but won't for much longer and I'm wondering if it would be worth buying for personal use instead of learning how to use some python libraries.

super cool video, i hope craftymasterman’s shoutout helps your channel :)

I thought of something similar but then with cobblestone in between (I havent tried it), have you tried it, and ifso why didn't it work or was it inferior to powderd snow?

The NPC is cursed.

at 4:00 it says you recorded this today… and uploaded. how the heck

I kinda wanna see how you made those conveyor belts without bends that are very long

Coud you make it faster if beside the inputstream you woud place oxediced copper blocks?

Woah How can you make the word interface like that? It looks so clean!

i'm actually wondering if just having a normal blockstream full of copper would actually improve the rates, 'cause yea, it's slower, but you got 5 times the amount of blocks oxidizing at the same time, also the feature where the oxidation sort of spreads, may lock all the copper block in one stage and improve the efficiency of it idk

Regarding the waxing the other variants. Instead of placing them down, waxing, blowing them up. Isn't it craftable? Where you put the copper block and one honeycomb in a crafting grid. Is that still a thing or did Mojang remove this?

OK I've looked st this multiple times, but... How do you make the hopper that collects the powdered snow buckets not suck the unfilled buckets out of the dispenser?

That's a lot of fours. NHL it gives modron vibes

everyone gangster til cubicmetre uses exponential regression lines to maximize copper oxidation

You should try to build one of these machines in survival

If someone asks me again, why we do maths in school i`ll just send them this video

I don’t think your model for exposed copper is accurate near 0. Dividing by time should not yield a vertical asymptote. You should be able to derive the actual distributions by using probability (probably with a Poisson or exponential distribution) rather than fitting functions to data points.

Woooooo

Would adding stationary fully oxidized blocks help?

petition to cubic meter to make a cubic centimeter sized farm for new players

Me when I see a cubic meter upload :000

My brain went .exe stopped working when I saw this.... 😂😂😂

can u make an addition so u cut set a mode that reuses all less oxidized blocks then the one u chose when the supply of fresh blocks runs out