haha, those comments made me chuckle. Glad I'm not the only one thinking Nilaus' style of videos is veeery random and unstructured.. (edit: for what they claim to be, that is.)
On the other hand when you are more proficient with wires. you can also have station disabled or otherwise limited that train will not come into the station until there is enough ore(or any other item) in the chests, so that you minimize train waiting time, and thus limit the amount of trains in the network. Now you dont need that big stackers, train throughput suddenly is alleviated and UPS in bigger bases will also thank you. Which means you than want 12 chests per wagon to load said wagon as quickly as possible.
But a train waiting at a station is a good thing because it will not further congest other trains that are moving while maintaining maximum throughout of the station. Only allow the train to dock when there is enough input/output only reduces total number of trains need to be crafted, which is insignificant.
This. I assume "a content creator" is Nilaus, but discussing his loaders without mentioning that his train stations use automation to set limits is absolutely misleading.
yh I had the same, short, informative and to the point. I bought the game a week ago, and since yesterday, I have watched all his vids about it and they are really good. He has a new sub :)
You received a follow right here. Simply because you gave us up front the answer, without making us wait forever until you got to the point. Thank you.
I agree on the 1-to-6 *loader* not being very useful. It's _usually_ (*) better to just use "4-chest loaders". And where that's not enough duplicate them on _both_ sides of the train; Or just double the number of stations; Or both. But for the *unloader,* I don't think there's one _optimal_ design. What's optimal will depend _heavily_ on how many items you need (...per second. And no, knowing that rounded to full _belts_ is usually not enough). Having a *compact* design as your go-to solution will usually (*) serve you best (and again where necessary you repeat it on the other side of the wagons or in a second station). Especially since in some cases, the optimum _will_ be "build two stations/sides that can each handle half of your demand." And _especially especially_ since you'd want that go-to design to not be massively wasteful in a low-throughput station. With that in mind, I think something where the inserters are controlled by circuit logic would be the best go-to design (...for me. Obvious exceptions apply.) (*) "Usually" meaning that if you _know_ that there is a better solution for _your specific_ situation, then use it. If you aren't sure, then go with the solution that is better in most situations, rather than the one that could eek out ever so slightly more performance in a minimal number of situations.
I tend to use a simple 6 chest loader - one blue splitting into two red belts feeding into 3 chests each - for mining outposts on Nauvis. Simply because I'd rather have more miners working as long as possible once the output drops due to corners running out. This way the the mines work more constant. But that is more of a preference than anything else. I do like the unloader design a lot, gonna take that with me for sure. Mostly because I think it looks neat.
This is fantastic advice! I'm new to Factorio and literally just started utilising trains. Thank you so much for making such a clear and helpful video.
Unloading brought me here, and I was satisfied. Haven't gotten stack inserters yet, but never occurred to me to use splitters like that. Definitely will be an improvement to my systems
I usually have one train servicing each outpost and only one depot station without train stacking. It takes a long time to travel to the depot and return, so by the time the train gets back, there is often a full load waiting for it. At the depot station depending on demand, I might be having many outpost trains delivering at the same or similar times, to prevent the trains backing up onto the main line, they need to be unloaded quickly. This is why I end up using six chest, sometimes on both sides of the track, to load and unload my trains, because that quick load and unload are huge for getting trains in and out quickly. I also use circuit conditions to ensure the chests fill up at load stations evenly and empty at unload stations evenly, so all load/unload inserters can work and do their job as quickly as possible when a train comes. Since circuit wires are free, the only cost for doing it this way is the extra chests and inserters and an arithmetic combinator, while barely using any splitters. It then saves space and complexity in the train network.
@@thetute59 as far as I can recall, setting output priority on splitter (but now I cant remember to exactly which side priority should be) gives slight performance increase. And that was tested in 1.x versions, 2.x could have changed their behavior.
The time taken to unload a train becomes critical if you’re going for mega base scale. 12 stack inserters per wagon is basically mandatory when you’re producing a full blue belt of blue circuits.
Sometimes simpler is better. You can factor in time to place, space it takes, range of uses. You can usually spend more in cost and time for the near perfect alternative. But if you can work with something that is adequate and gets you moving to the next project quicker. It's a better choice.
Personally, I just send out 4 belts per side with load balancing so that, once the chests fill up, the train is still unloaded evenly. Before I did that, I would consistently see a train parked for *FAR* too long with only a single car still not empty and the belts leading away containing barely a trickle. (Rather than turning inserters off and on--which introduced *FAR* too much delay--I simply turn the belt for each section on and off. As I'm always unloading more belts than I am sending to the main bus, this gives me completely balanced, compressed belts leaving the station.)
There is one small problem with your assumption of only needing 4 chests per side. And that is that you assume there will be a train to come in once the previous one is loaded. Many players, especially new ones, or players with their starting factories, will not have more than 1 train from an outpost back to the factory. In that case, the loading time is crucial. And having 50% more loading speed makes a big difference.
I use 1 train per loading station because there's no stacker required so I prefer the 6 buffer chest design and in that case the usual bottleneck is the train loading/unloading time.
I think space age DLC has already overridden the unloading into splitters type of setup. This along with the ability for 2 belts meeting up in the middle with an underground entity, it will only make the outer side of the underground direction to be able to continue, even though the outer side is already empty, the inner side won't move. CMIIW.
There is a thing to CPU load when you advance from a casual late game onwards. Belts require less computation than splitters, but a fully saturated belt requires way less than a non-saturated one. So unloading into splitters is better for things you need in vast quantities, like ore. You might want to unload onto belts for things that back up anyway like buildings. However for your own sanity I would always stick to one solution.
Im sold. Ive been using something that works well but takes up far more of a footprint and is unnecessary. I may not rip everything down but new builds and new game.. its ON!
When I replicate the demonstration you do at 2:56-3:00 I don't get that result, both arms take the same time to empty their "inventory", the only difference being that the belts from the splitter are less dense, the "snakes" with items are equally long but stretched on the double belts from the splitter. The arms swing at the same rate. Using 16-size bulk inserters. I think this has been patched... or perhaps the direction matters, i.e. south, west, north, east due to which corner of a tile the inserters use. Edit: nope, definetly patched, does not work. Edit 2: I consitently get less throughput with your method compared to unloading into a simple T-cross instead of splitters.
I love that you are using my blueprint for unloading. Thanks for the math on the loader side. 6 chests is always a pain to route. What's your opinion on skipping splitters on the loader side and just letting the chests at the loader fill unevenly? Its technically inferior as chests load unevenly, but it makes the station thinner, easier to build, and still consumes a full belt (at least with 6 chests, I'll be trying it with 4 next session). Also another unload design to try: Unload both sides of the train, and merge using an underground. Top chest will unload more than Bottom, unloads unevenly but it's still 2 belts throughput, albeit with slightly longer train unload times.
This guide is still relevant and this loader/unloader still great even though i made it pre 2.0. If I come up with a simpler and more efficient design i might make a new video in the future. Skipping splitters? As long as there are no throughput issues go ahead.
I like using logic on my loading stations so that trains don't come to load unless they can be fully loaded from chests. This way more chests are better -- faster loading. On unloading I prefer to have a leeway in my chests until the next train comes. But on unloading I usually try to have backpressure, so, it's not as critical. And for more than 4 belts a train is unloaded too fast -- I'd rather have more unloading stations than trains emptied faster and running for supply much longer than they were unloading.
I perfer A 6 loader due to the speed the stackers allow. Not sure if it really matters but its the only real advantage I see. I do use the 4 unloader with spliters as it seems to be the fastest unload option. Great points tho fo sho!
I'm surprised loading/unloading on splitters works. When I first tried this sort of thing (years ago now) I found the inserters would insert after the split not before, and didn't pick up things after the split. To that end, I'd always have a belt after the splitter. It would make my load/unload builds far larger.
It only works when inserting from the side perpendicular to the direction of the belts. If you insert from "behind" the splitter it will behave like you just described.
I think most of us would eventually come up with something similar by trial and error. But perfectionists need the most efficient system right from the start. I am glad someone is thinking about this stuff. But, I prefer trial and error because it wastes more time. Signed: Time Waster Extraordinaire.
Does this work when you want to ensure that the buffer chests load and unload evenly? Especially load, if you want to call a train precisely when the total amount of material in the chests will fill a train. Then, you call a train with a combinator and that train will load as fast as possible, not linger, and leave again when full. When chests are not even, the train will linger, delaying everything. I've been using Mad Zuri's smart loader system to ensure even loading of six chests per train wagon. But I'm wondering if this would work better. Also, compactness is important for situations like base depots where you want the minimum distance between tracks.
How do you deal with situations where one side of the belt is used less than the other? I always go with a design that balances the belt sides so that I never have to worry about uneven chests
In most situations it doesn't matter. When it does though you can either build a lane balancer or design the factory in a way that will draw evenly from both sides.
@@AVADIIStrategyor you can just ignore that, because the moment you need more than 0.5 of a belt you will start drawing from the other side, its just for aesthetics to lane balance (train unloading might be slower, but at the end of the day material coming through is all that matters, if you needed that train to do something else just add more trains)
Just curious, since from what I understand the major UPS penalty as you scale up is now (as of Factorio 2.0) the computing of collisions, especially when it comes to splitters. At LARGE scale (I never got there, but I am SO curious 😀), did you find this approach limiting in any way? On my current setup, based on MadZuri's averaged chest content (each inserter picks up only when the content of its chest is >= the average content of the chests servicing the train stop, in order to avoid emptying some chests before others) I find out that the belts end up unbalanced on the sides. Now, I get that there's no easy fix for unbalanced use, but... is there a way to fix it? 🙂 Thanks for a very good video!
Unless you want to build a massive multiple thousand spm megabase it wasn't necessary to care about UPS. I made the video for the 95% of the playerbase that doesn't care about that. If you're a megabase builder you don't need this video and you probably avoid inserting into and from chests altogether and use direct insertion. Unbalanced sides don't really matter. If you overproduce just a little bit these issues usually go away. You can also use a lane balancer or make it so that your factory consumes items evenly.
worth noting that if youre shooting for megabased scales you should probably ditch train buffers all together, they tend to be a pretty decent cause of UPS problems and dont actually do anything for throughput beyond delaying you noticing problems
Without buffers throughput is temporarily 0 while a new train is coming in. But yeah you're right: if UPS is the main concern throughput has less priority.
@@AVADIIStrategy alternatively you can use a chest limited to just 1 slot, just long enough for the next train to pull in, as the UPS cost of buffers goes up dramatically with number of storage slots, its not as good as bufferless but preferred over just full buffers, and it makes throughput problems more noticable earlier
@@ivanlagayacrus1891The more buffers the worse the lag?? Why do I hear it for the first time now?? I really do love my buffers, but my ups are also bad. How do I look it up how much performance it cost? F4 and under which point?
Yes, something changed with 2.0. Inserter timings are different now. The 4 wagons to 6 belts design mentioned later in the video doesn't work anymore as well (the one i didn't recommend anyway). If you still want the same throughput unload into green splitters or higher quality inserters. The rest of the belts can stay blue. My unloader is still doing very well and i would still recommend it. Will do an updated video someday.
Not for my blueprint. The one i dislike and don't recommend has the need for stacksize shenanigans. The thing is: it seems that 2.0 changed inserter timings so i don't know if and how that one works now.
My dad taught me KISS as a young boy: keep it simple stupid... I used to watch him "program" an IBM mainframe with wires on a peg board, and after a while, it looked like a multicolored head of hair... so simple means more like Einstein's refrain: keep it as simple as possible... i.e., simpler isn't always better with any combination of metrics, imo, but I'll use you presented design unless another ytuber proves it less effective... now I have to re-review Jar's experiments about this... Idr... but I think he didn't like that dual splitter design... my little brain hurts, thanks, I'm learning stuff.
Well, Jar made his experiments the next day after i showed up in his live chat. I guess he saw my video about it and wanted to make one too. And unfortunately i don't like the video at all because he just counts stuff which is worse than me comparing the swing time and item patterns on the belts. But the main problem i have is: he didn't even offer any opinion about it. Like does he think its a bug? What should the behaviour be? Thinking about it I feel like you haven't seen my video about it so here it is: ua-cam.com/video/3iVuNEElaxc/v-deo.html
@AVADIIStrategy Well, from my perspective of being a great Factorio noob, so like advanced beginner I'm still mapping the problem domain (discovering things that are possible to do, but not in any way being able to do them well), doing an almost random walk in the solution space. I do remember your video linked above, but it was a couple of weeks ago; I rewatched it now, and I did not completely understand it the first time I watched it (!) There was a time in the recent past I didn't know you could even load or unload on an underground or a splitter... that was maybe ~3 weeks ago (!) Time flies while you play Factorio, not just each day, but over weeks I guess... With content creators using and/or reusing each other's content, as long as it's acknowledged, I see no issues. I think Jar did acknowledge some Nilaus stuff, but honestly, Nilaus to me across games is inscrutable until I know it too, and then it's obvious what he's showing. That is an amazing talent, imo, but not one I consider laudatory. Idr him mentioning you... but I could have missed it given I don't have to watch his current playthru minute by minute, given I'm far more advanced now, from where he is atm, (except his train system, which is well beyond my current understanding), waiting until he gets to stuff I'm still working out, so I could compare and contrast approaches. Thanks for responding. I'll double check if I have to rewatch one of your videos before I ask a question from now on. Sometimes I don't realize I missed something...
You didn't mention anything about inserter stack upgrades.imo it would have been useful to know about more early vs late game designs, I assume that late game it may matter less than early game, even if player likely uses weaker belts.
That's the point. This design works for every stage of the game. No matter what your current inserter stack size is. No need for different designs for early/late.
@AVADIIStrategy yea, but you said you hv the blueprint in the description, when i look there, theres no blueprints, thats why i ask, lol but i already make myself by pausing your video in 1080p again, thanks for all the factorio guides, you help my playthrough a LOT - factorio honestly an easy game, but it make things complicated for no reasons instead of releasing QoL features and shout out guides and finding QoL mods is also hard without asking veterans
Or, hear me out, just 6 chests per wagon filled and emptied by bots. No belts. Oooor, megafactory style, just load to/unload from train directly to assemblers. No belts, no bots. Max UPS. Actually with legendary inserters and assemblers throughput is still almost maxed out. Other than that you can just have more trains and more tracks/stations.
This is great till you're looking at a fully saturated green belt with max item stacking in space age... Trains feel far less efficient compared "eh it's only 400 blocks, just belt it in for a gaurenteed 14400 items per minute"
Of course it's company policy never to, imply ownership in the event of a blueprint... always use the indefinite article a blueprint, never your blueprint.
Yes, I've noticed as well. Not sure if this is a bug. Inserters have a weird slower cadence now even though the rotation degree numbers didn't change. This unloader still works very well though.
@@JKnight been a while since I've played it so I'm surprised that they changed that....funny tho, since making that comment I've done about 200 hrs hahhaha To be fair, got a broken leg and can't do much else 😅
OK, I did not like your video, but obviously, you think you're right and I should just leave. Fair. I did delete my large verbal complaint to just leave this. Bye.
bro forgot about the massive amount of space whatever this is takes up and the increase in loading and unloading speed this mindset of complicated == inferior is completely false. with that logic why use trains if you have belts, smh
I'm gonna call it my blueprint now.
Our blueprint ⚒️
@@hi_im_julian_kirsch The royal we.
Our factory must grow comrade @@hi_im_julian_kirsch
@@hi_im_julian_kirsch COMMUNISM
MY PRECIOUSSSSSSS!
Thank you so much! I hate it when others take 45min to explain something that can be explained in 5min.
*cough* *cough* Nilaus
@@googoogaagaayt "HOW TO BUILD PERFECT BASE ON GLEBA"
1.5 h video of him visiting for the first time gleba e trying stuff out
haha, those comments made me chuckle. Glad I'm not the only one thinking Nilaus' style of videos is veeery random and unstructured..
(edit: for what they claim to be, that is.)
I want add that your blueprint also looks way cooler then other unloaders.
Thank you :)
On the other hand when you are more proficient with wires. you can also have station disabled or otherwise limited that train will not come into the station until there is enough ore(or any other item) in the chests, so that you minimize train waiting time, and thus limit the amount of trains in the network. Now you dont need that big stackers, train throughput suddenly is alleviated and UPS in bigger bases will also thank you. Which means you than want 12 chests per wagon to load said wagon as quickly as possible.
But a train waiting at a station is a good thing because it will not further congest other trains that are moving while maintaining maximum throughout of the station. Only allow the train to dock when there is enough input/output only reduces total number of trains need to be crafted, which is insignificant.
This. I assume "a content creator" is Nilaus, but discussing his loaders without mentioning that his train stations use automation to set limits is absolutely misleading.
This was perfect! as a new player i appreciate the shorter to the point format instead of 45 min of rambling that i see other places.
3:07 Opponent :) "Enemy belt detected! Faster!"
Great videos by the way! Kurz, präzise, geistreich! :)
That's just amazing. Thank you! The blueprint for the Nuclear Reactor without heatpipes is also excellent. Need to watch all your videos now.
I cannot conceive of a better UA-cam video at this out of the night.
yh I had the same, short, informative and to the point. I bought the game a week ago, and since yesterday, I have watched all his vids about it and they are really good.
He has a new sub :)
came for learning a little more about trains, the great enigma, stayed for the doggo as to not disappoint them.
You received a follow right here. Simply because you gave us up front the answer, without making us wait forever until you got to the point. Thank you.
Thanks for the video
where can i find your blue prints ?
I agree on the 1-to-6 *loader* not being very useful. It's _usually_ (*) better to just use "4-chest loaders". And where that's not enough duplicate them on _both_ sides of the train; Or just double the number of stations; Or both.
But for the *unloader,* I don't think there's one _optimal_ design. What's optimal will depend _heavily_ on how many items you need (...per second. And no, knowing that rounded to full _belts_ is usually not enough). Having a *compact* design as your go-to solution will usually (*) serve you best (and again where necessary you repeat it on the other side of the wagons or in a second station). Especially since in some cases, the optimum _will_ be "build two stations/sides that can each handle half of your demand." And _especially especially_ since you'd want that go-to design to not be massively wasteful in a low-throughput station.
With that in mind, I think something where the inserters are controlled by circuit logic would be the best go-to design (...for me. Obvious exceptions apply.)
(*) "Usually" meaning that if you _know_ that there is a better solution for _your specific_ situation, then use it. If you aren't sure, then go with the solution that is better in most situations, rather than the one that could eek out ever so slightly more performance in a minimal number of situations.
I tend to use a simple 6 chest loader - one blue splitting into two red belts feeding into 3 chests each - for mining outposts on Nauvis. Simply because I'd rather have more miners working as long as possible once the output drops due to corners running out. This way the the mines work more constant. But that is more of a preference than anything else.
I do like the unloader design a lot, gonna take that with me for sure. Mostly because I think it looks neat.
This is fantastic advice! I'm new to Factorio and literally just started utilising trains. Thank you so much for making such a clear and helpful video.
Going to the point in less than 5min. I wish more of UA-cam was like this :)
oh my god your dog lowkey kinda looks like a fox and (s)he is super adorable, I wish him all the best in life
Unloading brought me here, and I was satisfied. Haven't gotten stack inserters yet, but never occurred to me to use splitters like that. Definitely will be an improvement to my systems
I usually have one train servicing each outpost and only one depot station without train stacking. It takes a long time to travel to the depot and return, so by the time the train gets back, there is often a full load waiting for it. At the depot station depending on demand, I might be having many outpost trains delivering at the same or similar times, to prevent the trains backing up onto the main line, they need to be unloaded quickly. This is why I end up using six chest, sometimes on both sides of the track, to load and unload my trains, because that quick load and unload are huge for getting trains in and out quickly. I also use circuit conditions to ensure the chests fill up at load stations evenly and empty at unload stations evenly, so all load/unload inserters can work and do their job as quickly as possible when a train comes. Since circuit wires are free, the only cost for doing it this way is the extra chests and inserters and an arithmetic combinator, while barely using any splitters. It then saves space and complexity in the train network.
I had to watch it 3 times, 2 times I got distracted by the doggo haha, he sits there so sweet and nice.
Thanks for the Training.
How does this change with the addition of Space age and the 60 items/s green belts?
Beautiful dog!
Very much appreciated, says Floki.
finally I can stop using Nilaus' overcomplicated midwit blueprints, thank you
subbed. that was a 4 minute video packed full with exactly the content I wanted. awesome video.
That was the video I needed a few months ago, but didn't find the answer till now, thank you.
Smart and simple. I'm going to use this when I start playing again!!
If you set splitters output priority to side, closest to inserter, this would give even more performance.
What, really? how?
@@thetute59 that way inserters can fill belts more densely, hence swing faster.
@@Felinaro I dont get how that improves it, but I hope to give it a try next time I do a trainunloading (: thanks
@@thetute59 as far as I can recall, setting output priority on splitter (but now I cant remember to exactly which side priority should be) gives slight performance increase.
And that was tested in 1.x versions, 2.x could have changed their behavior.
Nice idea to showcase gaps in the belt throughput with the sulphur splitter.
Thanks! 👍
Gorgeous dog!
Fantastisch, danke. Mit genug Zeit um rumprobieren, hätte man vielleicht darauf kommen können. Aber: Ain't nobody got time for that...
Well I'm convinced, will be using this from now on. Earned a Sub 👍
Glad I found this. Thank you for sharing!
You just opened my eyes right now. Wow. Thx :)
The time taken to unload a train becomes critical if you’re going for mega base scale. 12 stack inserters per wagon is basically mandatory when you’re producing a full blue belt of blue circuits.
Sometimes simpler is better. You can factor in time to place, space it takes, range of uses.
You can usually spend more in cost and time for the near perfect alternative. But if you can work with something that is adequate and gets you moving to the next project quicker. It's a better choice.
That's a good doggo.
Nice video too
Such a cute pup :3
Wow. I’ve never thought about dropping items onto a splitter to essentially double the unload rate. That’s crazy.
Personally, I just send out 4 belts per side with load balancing so that, once the chests fill up, the train is still unloaded evenly. Before I did that, I would consistently see a train parked for *FAR* too long with only a single car still not empty and the belts leading away containing barely a trickle. (Rather than turning inserters off and on--which introduced *FAR* too much delay--I simply turn the belt for each section on and off. As I'm always unloading more belts than I am sending to the main bus, this gives me completely balanced, compressed belts leaving the station.)
There is one small problem with your assumption of only needing 4 chests per side.
And that is that you assume there will be a train to come in once the previous one is loaded.
Many players, especially new ones, or players with their starting factories, will not have more than 1 train from an outpost back to the factory.
In that case, the loading time is crucial. And having 50% more loading speed makes a big difference.
I use 1 train per loading station because there's no stacker required so I prefer the 6 buffer chest design and in that case the usual bottleneck is the train loading/unloading time.
I think space age DLC has already overridden the unloading into splitters type of setup. This along with the ability for 2 belts meeting up in the middle with an underground entity, it will only make the outer side of the underground direction to be able to continue, even though the outer side is already empty, the inner side won't move. CMIIW.
There is a thing to CPU load when you advance from a casual late game onwards. Belts require less computation than splitters, but a fully saturated belt requires way less than a non-saturated one. So unloading into splitters is better for things you need in vast quantities, like ore. You might want to unload onto belts for things that back up anyway like buildings. However for your own sanity I would always stick to one solution.
Im sold. Ive been using something that works well but takes up far more of a footprint and is unnecessary. I may not rip everything down but new builds and new game.. its ON!
Floki is looking like smart one
When I replicate the demonstration you do at 2:56-3:00 I don't get that result, both arms take the same time to empty their "inventory", the only difference being that the belts from the splitter are less dense, the "snakes" with items are equally long but stretched on the double belts from the splitter. The arms swing at the same rate. Using 16-size bulk inserters.
I think this has been patched... or perhaps the direction matters, i.e. south, west, north, east due to which corner of a tile the inserters use.
Edit: nope, definetly patched, does not work.
Edit 2: I consitently get less throughput with your method compared to unloading into a simple T-cross instead of splitters.
Patched with space age
@justa9560 supected as much. The new stack inserters make belts way better now anyway
I love that you are using my blueprint for unloading.
Thanks for the math on the loader side. 6 chests is always a pain to route.
What's your opinion on skipping splitters on the loader side and just letting the chests at the loader fill unevenly?
Its technically inferior as chests load unevenly, but it makes the station thinner, easier to build, and still consumes a full belt (at least with 6 chests, I'll be trying it with 4 next session).
Also another unload design to try: Unload both sides of the train, and merge using an underground. Top chest will unload more than Bottom, unloads unevenly but it's still 2 belts throughput, albeit with slightly longer train unload times.
This guide is still relevant and this loader/unloader still great even though i made it pre 2.0.
If I come up with a simpler and more efficient design i might make a new video in the future.
Skipping splitters? As long as there are no throughput issues go ahead.
I like using logic on my loading stations so that trains don't come to load unless they can be fully loaded from chests. This way more chests are better -- faster loading. On unloading I prefer to have a leeway in my chests until the next train comes. But on unloading I usually try to have backpressure, so, it's not as critical. And for more than 4 belts a train is unloaded too fast -- I'd rather have more unloading stations than trains emptied faster and running for supply much longer than they were unloading.
Very helpful thank you
“Thus inferior” ❤
love your dog ❤️
Thank you for the guide
I perfer A 6 loader due to the speed the stackers allow. Not sure if it really matters but its the only real advantage I see. I do use the 4 unloader with spliters as it seems to be the fastest unload option. Great points tho fo sho!
Lol i didn’t know that, good job!!
but what if my belts are green? :D
Your dog is super cute
The dlc will probably change a lot about this, no? Faster belts and quality loaders?
Probably. It might still be the way to go until late game.
I'm surprised loading/unloading on splitters works. When I first tried this sort of thing (years ago now) I found the inserters would insert after the split not before, and didn't pick up things after the split. To that end, I'd always have a belt after the splitter. It would make my load/unload builds far larger.
It only works when inserting from the side perpendicular to the direction of the belts.
If you insert from "behind" the splitter it will behave like you just described.
I think most of us would eventually come up with something similar by trial and error. But perfectionists need the most efficient system right from the start. I am glad someone is thinking about this stuff. But, I prefer trial and error because it wastes more time.
Signed: Time Waster Extraordinaire.
Excellent, thanks!
Does this work when you want to ensure that the buffer chests load and unload evenly? Especially load, if you want to call a train precisely when the total amount of material in the chests will fill a train. Then, you call a train with a combinator and that train will load as fast as possible, not linger, and leave again when full. When chests are not even, the train will linger, delaying everything. I've been using Mad Zuri's smart loader system to ensure even loading of six chests per train wagon. But I'm wondering if this would work better. Also, compactness is important for situations like base depots where you want the minimum distance between tracks.
How do you deal with situations where one side of the belt is used less than the other? I always go with a design that balances the belt sides so that I never have to worry about uneven chests
In most situations it doesn't matter. When it does though you can either build a lane balancer or design the factory in a way that will draw evenly from both sides.
@@AVADIIStrategyor you can just ignore that, because the moment you need more than 0.5 of a belt you will start drawing from the other side, its just for aesthetics to lane balance (train unloading might be slower, but at the end of the day material coming through is all that matters, if you needed that train to do something else just add more trains)
Lovely video
Just curious, since from what I understand the major UPS penalty as you scale up is now (as of Factorio 2.0) the computing of collisions, especially when it comes to splitters. At LARGE scale (I never got there, but I am SO curious 😀), did you find this approach limiting in any way?
On my current setup, based on MadZuri's averaged chest content (each inserter picks up only when the content of its chest is >= the average content of the chests servicing the train stop, in order to avoid emptying some chests before others) I find out that the belts end up unbalanced on the sides. Now, I get that there's no easy fix for unbalanced use, but... is there a way to fix it? 🙂
Thanks for a very good video!
Unless you want to build a massive multiple thousand spm megabase it wasn't necessary to care about UPS. I made the video for the 95% of the playerbase that doesn't care about that. If you're a megabase builder you don't need this video and you probably avoid inserting into and from chests altogether and use direct insertion.
Unbalanced sides don't really matter. If you overproduce just a little bit these issues usually go away.
You can also use a lane balancer or make it so that your factory consumes items evenly.
@@AVADIIStrategy Vielen Dank!
worth noting that if youre shooting for megabased scales you should probably ditch train buffers all together, they tend to be a pretty decent cause of UPS problems and dont actually do anything for throughput beyond delaying you noticing problems
Without buffers throughput is temporarily 0 while a new train is coming in.
But yeah you're right: if UPS is the main concern throughput has less priority.
@@AVADIIStrategy alternatively you can use a chest limited to just 1 slot, just long enough for the next train to pull in, as the UPS cost of buffers goes up dramatically with number of storage slots, its not as good as bufferless but preferred over just full buffers, and it makes throughput problems more noticable earlier
Or use cursed wagon to wagon unloading. Better than unloading into ovens directly, I guess.
@@mytiliss682if it works - it is not cursed 😂
@@ivanlagayacrus1891The more buffers the worse the lag?? Why do I hear it for the first time now?? I really do love my buffers, but my ups are also bad. How do I look it up how much performance it cost? F4 and under which point?
I also have a shiba poodle mix. Except mine is grey and white!
Я ставлю 6 сундуков и настраиваю чтобы поезд отправлялся по полному заполнению поезда и сундуков, так больше времени выпить водки
the 8 to 6 doesnt seem to be fully saturated when i try this
Yes, something changed with 2.0. Inserter timings are different now. The 4 wagons to 6 belts design mentioned later in the video doesn't work anymore as well (the one i didn't recommend anyway).
If you still want the same throughput unload into green splitters or higher quality inserters. The rest of the belts can stay blue.
My unloader is still doing very well and i would still recommend it. Will do an updated video someday.
@ yeah I think I’m going to replace my stations with your design
So you said that we have to override some of the inserter stack sizes. Is it some specific number or we have to figure it out experimentally?
Not for my blueprint. The one i dislike and don't recommend has the need for stacksize shenanigans.
The thing is: it seems that 2.0 changed inserter timings so i don't know if and how that one works now.
Me running 1 belt per wagon through a balancer with no additional splitters and chest balancing circuited inserters.
My dad taught me KISS as a young boy: keep it simple stupid... I used to watch him "program" an IBM mainframe with wires on a peg board, and after a while, it looked like a multicolored head of hair... so simple means more like Einstein's refrain: keep it as simple as possible... i.e., simpler isn't always better with any combination of metrics, imo, but I'll use you presented design unless another ytuber proves it less effective... now I have to re-review Jar's experiments about this... Idr... but I think he didn't like that dual splitter design... my little brain hurts, thanks, I'm learning stuff.
Well, Jar made his experiments the next day after i showed up in his live chat.
I guess he saw my video about it and wanted to make one too.
And unfortunately i don't like the video at all because he just counts stuff which is worse than me comparing the swing time and item patterns on the belts.
But the main problem i have is: he didn't even offer any opinion about it. Like does he think its a bug? What should the behaviour be?
Thinking about it I feel like you haven't seen my video about it so here it is: ua-cam.com/video/3iVuNEElaxc/v-deo.html
@AVADIIStrategy Well, from my perspective of being a great Factorio noob, so like advanced beginner I'm still mapping the problem domain (discovering things that are possible to do, but not in any way being able to do them well), doing an almost random walk in the solution space.
I do remember your video linked above, but it was a couple of weeks ago; I rewatched it now, and I did not completely understand it the first time I watched it (!)
There was a time in the recent past I didn't know you could even load or unload on an underground or a splitter... that was maybe ~3 weeks ago (!) Time flies while you play Factorio, not just each day, but over weeks I guess...
With content creators using and/or reusing each other's content, as long as it's acknowledged, I see no issues. I think Jar did acknowledge some Nilaus stuff, but honestly, Nilaus to me across games is inscrutable until I know it too, and then it's obvious what he's showing. That is an amazing talent, imo, but not one I consider laudatory.
Idr him mentioning you... but I could have missed it given I don't have to watch his current playthru minute by minute, given I'm far more advanced now, from where he is atm, (except his train system, which is well beyond my current understanding), waiting until he gets to stuff I'm still working out, so I could compare and contrast approaches.
Thanks for responding. I'll double check if I have to rewatch one of your videos before I ask a question from now on. Sometimes I don't realize I missed something...
@@Logically_Fallacious You don't have to double check, no worries.
I just had a feeling you saw his video first that's all.
You didn't mention anything about inserter stack upgrades.imo it would have been useful to know about more early vs late game designs, I assume that late game it may matter less than early game, even if player likely uses weaker belts.
That's the point. This design works for every stage of the game. No matter what your current inserter stack size is. No need for different designs for early/late.
code for the blueprint?
nice idea, thank you
but where blueprint link? i didnt see anywhere...
That's the point. This loader and unloader is so simple you can just make it yourself.
@AVADIIStrategy yea, but you said you hv the blueprint in the description, when i look there, theres no blueprints, thats why i ask, lol
but i already make myself by pausing your video in 1080p
again, thanks for all the factorio guides, you help my playthrough a LOT
-
factorio honestly an easy game, but it make things complicated for no reasons instead of releasing QoL features and shout out guides
and finding QoL mods is also hard without asking veterans
Or, hear me out, just 6 chests per wagon filled and emptied by bots. No belts.
Oooor, megafactory style, just load to/unload from train directly to assemblers. No belts, no bots. Max UPS.
Actually with legendary inserters and assemblers throughput is still almost maxed out. Other than that you can just have more trains and more tracks/stations.
It is possible to load and unload two full blue belts per train side, so four blue belts per wagon. However, it is unbalanced
How did you know I wanted to know the dog breed? :D This dog has character
This is great till you're looking at a fully saturated green belt with max item stacking in space age... Trains feel far less efficient compared "eh it's only 400 blocks, just belt it in for a gaurenteed 14400 items per minute"
So who's blueprint is it then? 🤔
Awesome! I used to use 6 storage chests per wagon but with your setup 4 should work better. It will save me a good amount of materials.
Do you mean our blueprints?
3 weeks later: low density structure stacks for 50
anyways thanks for the tip that I should insert into splitters
Glad to help! I like the stack size change for LDS.
Cool dog
Flooooki ❤
It's mine now.
Of course it's company policy never to, imply ownership in the event of a blueprint... always use the indefinite article a blueprint, never your blueprint.
where is blueprint link?
@0:35 The point of this video is: it's so simple that you don't need one
@@AVADIIStrategy i eventually copied yeah thanks anyway
Wow that's good ro know
Looks like in 2.0 this cannot saturate blue+red anymore
Yes, I've noticed as well. Not sure if this is a bug.
Inserters have a weird slower cadence now even though the rotation degree numbers didn't change.
This unloader still works very well though.
Я использую логику в манипуляторах, которая распределяет груз равномерно между буферными сундуками.
The Most Important thing allways: KISS (keep it simple stupid). You can make it more complicated by choice but normaly only what is realy needet.
Yeah, im too new to this game for this. Ill be back in 300 hours game time.
yoink
0:41 ok, bye
sadly this unloading is not full balanced, if cunsume only one side of belt chests empty unevenly
Good explanation, but what’s about the freaking dog?!
What about him? He wanted to be in the video apparently :D
It is not your blueprint... But it is out blueprint...
Changes in 2.0 means that this isn't up to date anymore :(
Is that actually true?
@@Tqoratsos666 yh
@@JKnight been a while since I've played it so I'm surprised that they changed that....funny tho, since making that comment I've done about 200 hrs hahhaha
To be fair, got a broken leg and can't do much else 😅
Your dog looks ai generated
Floki doesn't like this comment at all!
OK, I did not like your video, but obviously, you think you're right and I should just leave. Fair. I did delete my large verbal complaint to just leave this. Bye.
bro forgot about the massive amount of space whatever this is takes up and the increase in loading and unloading speed
this mindset of complicated == inferior is completely false. with that logic why use trains if you have belts, smh