With the design where you unload on belts from left and right and merge them, for example 4th from the right, you should try and play around with hand size. Because of how the inserters swing, if you turn on maximum stack size, they will wait until they are fully empty and then they will swing back. The issue is that at maximum stack size what happens is that their timing will be off which means that the inserter will have to swing back at wrong time and leave a gap. For bulk inserters you could unload with 4 inserters and make a fully stacked blue belt by limiting the hand size to 8 (or 6, i dont remember). I don't know how it is with the green belt and stack inserters but there is something you can test
@@DavidLindes these are some interesting concepts to try... especially because the examples with only 4 inserters tend to keep their boxes more balanced automatically, it could be worth getting them to work optimally
saturation alone doesn't really indicate how good a design is, since an empty belt and one that's 90% full are counted the same way, also, a belt with horizontal (using the video as perspective) gap in items would be considered unsaturated for less time than a belt that had a vertical gap on one side of the belt, even if the throughput ends up being the same, still an interesting test nonetheless
Thanks hypercrafting! You're definitely right, which is why I also did the count test, maybe I can devise some more intricate test where it counts the missing items... anyways, interesting results and great comment!
@@TheJarGames Saturation only matters as a roundabout measure of throughput anyway. When unloading a train, there's only two factors that you should actually care about: How much material is making it into my factory (throughput/item count), and is the train unloading evenly (is the unloading going to slow down when one chest empties before the others). If you can measure throughput directly (as you have), then there's no point measuring saturation. And actually, even evenness of unloading only matters so far as it affects average throughput over time. So the ultimate test is this: Build actual unloading stations, with trains that have to empty, depart, and come back, with a proper train stacker, and variations with and without buffer chests. And count total item count over time. Highest total throughput is the best unloader, and only UPS can argue otherwise.
would probably also be interesting to know how even the unloading is, testing by unloading into chests first and then on belts. for example befor 2.0 you couldf use the 4th design to get an even unloading but with the change for inserting on splitters it has become quiete uneven
if you're talking about endgame, you might aswell test it out with quality inserters :P otherwise you wont ever be able to get a turbo belt fully saturated with 6 stack inserters (reason ist that stack inserters drop 4 stacks of 4 items compared to bulk which drop 12 items per swing)
If rare stack inserters can saturate a turbo belt, maybe test uncommon quality as well. Would be nice to know the exact threshold required. Late game, rare quality items are not hard to get, but getting uncommon items is almost trivial so if that gets the job done it would be very easy to retrofit all your stations to get that throughput. We’re talking 24 inserters per 1-4 station (double if you use buffer chests) and by late game you probably have dozens of stations, so the lower quality you can get away with, the better.
Try: Yellow belts with red splitters; Red with Blue; and Blue with turbo. My anecdotal observations is the splitter is a bottleneck when unloading. So if I want a fully saturated red belt going into my smelter, I'll set up the unloader with blue belts feeding into a blue splitter onto a red belt.
this is content that touches the very edge of the venn diagram of total technical content housed by factorio. there are entire massive rabbitholes for each and every component in factorio, you can do so many things. I only just figured out theoretically how to assign colors to flash based on different signals coming in to a single lamp. you can make a lamp count down from green to yellow to red, just ONE lamp now.... theres SO much stuff to do. keep it up man, Im watching these little tech vods, i know quite a few others like em as well. the thruster efficiency value is such a nice little door into technical factorio. not a requirement, but if you figure it out you are rewarded. I love it.
the faster belt we use - the more important inserters thoughput become. add counterpart on ther side of the vagon for design n4 and you'll get fully saturated green belt since there will be 8 inserters working
yup! that's a great idea. I also have a video testing quality stacks to see what is the lowest quality necessary to full saturate a green belt with just one side unloading... stay tuned!
Interesting that, if you aren't looking to mix/max for a specific situation, "design 1" seems to be a great default. Especially considering the fact that most unloading stations will eventually be upgraded into the stack/turbo combo, I might just go for simplicity (and compact footprint!) by default in all my blueprints from now on.
6:30 Loading onto spitters was broken in 2.0 unless they fixed it in the last month. Could also add 2 more inserters to have 6 like the others instead of only 4. Curious if the extra 2 splitters would help it. Ive been using your "winning" design so cool to see it tested and that its on the right track for unloading speed.
it is aktualy not broken but the intended way now, it was broken befor 2.0 because the inserters where putting the item on the split instead of befor the split, which resulted in instant splitting
@Odysseus0000 Each color of belt functions differently though which is generally a sign of not working right. Iirc blue was the worst and basically the same as loading onto a regular belt.
@@Silv3rDragon that is because now the inserter puts the item on the belt part befor the split, the devs confirmed it in the bug report forum that this is intended behavior, that was "fixed" with 2.0
the 2nd unloader design unequally puts more on the right side of the belt as inserters put items on the right side when facing the direction of the belt. in the wagon-wide design it works as 3 inserters end up putting on each side. on the 2nd design it ends up being 4-2 so it's quite imbalanced. the reason it's bad is that the 4 side may end up filled and not able to take the load from the wagon leading to longer unloading times in the station
I wonder if saturation and throughput could be improved by prioritizing the main splitter output (right side by belt direction), adding another output from the splitter to keep overflow moving, and dumping that overflow back onto the main belt...
try putting 4 splitters in the same direction, one connected into the next, merge side of last splitter into output line. Unload on those splitters. With some shenanigans, you can unload 2 wagons into 6 full blue lines this way. Not had a chance to test Space Age yet
Really been finding your videos informative, I'd love to see quality inserters tested to finally get full saturation! I've been using a modified version of the far left design in each group of 4. Only difference is it has splitters bracketing the output belt much like the old go to design. Not sure if it really helps to add those but I might need to do some testing now myself
lmao that's a unique design, I Think it would be much wider... but I've been thinking about using a wagon for a buffer, since it would effectively auto-balance, as opposed to 6 chests... but the storage buffer would be so much less
will have to make another vid with 2-sided, test how full the chests get (if they stay balanced) and also using quality! The quest for the best continues!
the leftmost design(s) could be improved by reducing inserter capacity that makes the outer one move earlier while the inner ones have enough stock to compensate for the outer ones swinging || there was a mistake: i wasn't following yet ... i fixed that!
so, one of my thoughts is that that should be unloading into chests. I assume what will happen is that some of the designs will unload unequally so some of the chests will empty near the end which will slow down throughput in a longer timeline.
Didn't see my preferred unloader in the mix, basically a cross between to the far right and far left with only one set of facing splitters instead of two and then a belt behind each for more pushed through each splitter. It can easily keep a blue belt fully saturated with four bulk inserters and is nicely compact (5x2 tiles), but have not stress tested on green yet or with stack. Seems all of them struggle with green plus stack.
NGL I don't like that design simply because it's not symmetrical. Is there anything to gain by adjusting the stack size in the stack inserters? Perhaps changing one in each half to eight items instead of 16?
Probably covered but Design One (left to right) requires a single inserter to load fewer items faster to make for full saturation. Designs 2-4 are not good for UPS due to Splitter usage. Though Design Two is the best outta the worst. All are valid designs up to 6.5-8k SPM base.
I didn't realize there was a Fallout 4 kinda console command "~" that can get you to an unmodded Factorio Blueprint like scenario (where you can even save you current work state to come back to it and work on it some more), where you can prototype stuff without the mod Blueprint Sandboxes, which now I wonder, why is that even a thing!? Now I understand your confusion with not even seeing the scenario of Sandbox on the creation of a Factorio new game page. Ty for presenting videos where I learn Idk what Idk Idk, but I know Idk a lot (about Factorio this time...!)
@@TheJarGames You aren't using the /editor command? That's what it looks like you're using, that I wasn't aware of... I'm trying to confirm that, cause that's what it looks like to me.
I want to love your videos, but please increase your bitrate of your videos and/or move to 1440p or 4k recording, this is lacking so much detail it hurts to watch. That said I want more of this kind of content, I just want it in a reasonable resolution :)
there are a couple 4 inserter designs, they don't perform as well but they can balance their boxes more. Look out for the next video where we look at quality and the benefits of some of the 4 inserter designs
my dumbass made an unloader like the best one you concluded in the video but i put a 3 to 1 balancer without thinking 💀 (i didnt think it wouldnt have effect)
idk, dont you need to swap out the train every 40 stacks unloaded? and in that time the belt is empty anyay. the only thing that matters is how many inserters are unloading and if they work at max capacity.
In your third design, delete the first row and change every odd conveyor belt to an underground out belt, that way you get two straight pieces without a corner piece, and the two inserters can fill two separate sides. It's effectively the same design but one tile shorter, more compact! But I just did an unloader design and decided that it's best to use the splitter out one, your fourth design, but on both side, and then I can bring in the other side with an underground belt, that way I have 8 inserters loading instead of 6, and it my balancer at the station can stay simpler too. (And it's fairly compact too, on the other side of the wagon it's just 3 wide (1st row: inserter/undergound in the middle, 2nd/3rd row: splitter)
I made the design shorter as per your suggestion, love it! still, if you have that blueprint of your 2-sided unloader, let me know! I just tested quality and the results are surprising!
I set my train unloading station with 3 belts per 2 trains with 4 2+2 4 this ensures maximum balanced unloading at all times and I can use other techniques down the line to compress the belts. I am very disappointed in your best unloading suggestion because it biases the unloading of one side. Balanced unloading ensures that the transfer inserters between the train car and the chest are at max capacity. If one chest fills up that one 6th of your throughput is gone. Resulting in slower train unloading.
With the design where you unload on belts from left and right and merge them, for example 4th from the right, you should try and play around with hand size. Because of how the inserters swing, if you turn on maximum stack size, they will wait until they are fully empty and then they will swing back. The issue is that at maximum stack size what happens is that their timing will be off which means that the inserter will have to swing back at wrong time and leave a gap. For bulk inserters you could unload with 4 inserters and make a fully stacked blue belt by limiting the hand size to 8 (or 6, i dont remember). I don't know how it is with the green belt and stack inserters but there is something you can test
was coming to say something similar. There are some weird quirks and (arguably) non-intuitive results with hand size limits.
@@DavidLindes these are some interesting concepts to try... especially because the examples with only 4 inserters tend to keep their boxes more balanced automatically, it could be worth getting them to work optimally
Yeah, 12 on the outside and 8 on the inside. Full saturation, even on green belts.
saturation alone doesn't really indicate how good a design is, since an empty belt and one that's 90% full are counted the same way, also, a belt with horizontal (using the video as perspective) gap in items would be considered unsaturated for less time than a belt that had a vertical gap on one side of the belt, even if the throughput ends up being the same, still an interesting test nonetheless
Thanks hypercrafting! You're definitely right, which is why I also did the count test, maybe I can devise some more intricate test where it counts the missing items... anyways, interesting results and great comment!
@@TheJarGames Saturation only matters as a roundabout measure of throughput anyway. When unloading a train, there's only two factors that you should actually care about: How much material is making it into my factory (throughput/item count), and is the train unloading evenly (is the unloading going to slow down when one chest empties before the others). If you can measure throughput directly (as you have), then there's no point measuring saturation.
And actually, even evenness of unloading only matters so far as it affects average throughput over time. So the ultimate test is this: Build actual unloading stations, with trains that have to empty, depart, and come back, with a proper train stacker, and variations with and without buffer chests. And count total item count over time. Highest total throughput is the best unloader, and only UPS can argue otherwise.
How about if each Belt would output to a Chest or few, then we compare the contents of the Chests?
@@toddkes5890 We can already count belt throughput as a pulse per item. I believe that's what TheJarGames has already done.
would probably also be interesting to know how even the unloading is, testing by unloading into chests first and then on belts. for example befor 2.0 you couldf use the 4th design to get an even unloading but with the change for inserting on splitters it has become quiete uneven
if you're talking about endgame, you might aswell test it out with quality inserters :P
otherwise you wont ever be able to get a turbo belt fully saturated with 6 stack inserters (reason ist that stack inserters drop 4 stacks of 4 items compared to bulk which drop 12 items per swing)
yes! good call! Next time we will test some rare, epic, and legendary quality stack inserters
If rare stack inserters can saturate a turbo belt, maybe test uncommon quality as well. Would be nice to know the exact threshold required. Late game, rare quality items are not hard to get, but getting uncommon items is almost trivial so if that gets the job done it would be very easy to retrofit all your stations to get that throughput. We’re talking 24 inserters per 1-4 station (double if you use buffer chests) and by late game you probably have dozens of stations, so the lower quality you can get away with, the better.
@@TheJarGames 6 uncommon would be enough for full green belt
@@AndrewBezdenezhnykh yup!
@@TheJarGames and 4 legendary is great for stacked green
Try: Yellow belts with red splitters; Red with Blue; and Blue with turbo. My anecdotal observations is the splitter is a bottleneck when unloading. So if I want a fully saturated red belt going into my smelter, I'll set up the unloader with blue belts feeding into a blue splitter onto a red belt.
this is content that touches the very edge of the venn diagram of total technical content housed by factorio. there are entire massive rabbitholes for each and every component in factorio, you can do so many things. I only just figured out theoretically how to assign colors to flash based on different signals coming in to a single lamp. you can make a lamp count down from green to yellow to red, just ONE lamp now.... theres SO much stuff to do. keep it up man, Im watching these little tech vods, i know quite a few others like em as well.
the thruster efficiency value is such a nice little door into technical factorio. not a requirement, but if you figure it out you are rewarded. I love it.
the faster belt we use - the more important inserters thoughput become. add counterpart on ther side of the vagon for design n4 and you'll get fully saturated green belt since there will be 8 inserters working
yup! that's a great idea.
I also have a video testing quality stacks to see what is the lowest quality necessary to full saturate a green belt with just one side unloading... stay tuned!
Interesting that, if you aren't looking to mix/max for a specific situation, "design 1" seems to be a great default.
Especially considering the fact that most unloading stations will eventually be upgraded into the stack/turbo combo, I might just go for simplicity (and compact footprint!) by default in all my blueprints from now on.
Yeah good call
I think the design with one splitter can be optimized. Currently for inserters fill the right lane and only two the left.
6:30 Loading onto spitters was broken in 2.0 unless they fixed it in the last month. Could also add 2 more inserters to have 6 like the others instead of only 4. Curious if the extra 2 splitters would help it. Ive been using your "winning" design so cool to see it tested and that its on the right track for unloading speed.
it is aktualy not broken but the intended way now, it was broken befor 2.0 because the inserters where putting the item on the split instead of befor the split, which resulted in instant splitting
@Odysseus0000 Each color of belt functions differently though which is generally a sign of not working right. Iirc blue was the worst and basically the same as loading onto a regular belt.
@@Silv3rDragon that is because now the inserter puts the item on the belt part befor the split, the devs confirmed it in the bug report forum that this is intended behavior, that was "fixed" with 2.0
the 2nd unloader design unequally puts more on the right side of the belt as inserters put items on the right side when facing the direction of the belt. in the wagon-wide design it works as 3 inserters end up putting on each side. on the 2nd design it ends up being 4-2 so it's quite imbalanced.
the reason it's bad is that the 4 side may end up filled and not able to take the load from the wagon leading to longer unloading times in the station
I wonder if saturation and throughput could be improved by prioritizing the main splitter output (right side by belt direction), adding another output from the splitter to keep overflow moving, and dumping that overflow back onto the main belt...
try putting 4 splitters in the same direction, one connected into the next, merge side of last splitter into output line. Unload on those splitters. With some shenanigans, you can unload 2 wagons into 6 full blue lines this way. Not had a chance to test Space Age yet
If you can get me a blueprint, I'd love to test it!
Really been finding your videos informative, I'd love to see quality inserters tested to finally get full saturation! I've been using a modified version of the far left design in each group of 4. Only difference is it has splitters bracketing the output belt much like the old go to design. Not sure if it really helps to add those but I might need to do some testing now myself
I wonder if you could stack up vertical cargo wagons next to the sides of the unloading station.
lmao that's a unique design, I Think it would be much wider... but I've been thinking about using a wagon for a buffer, since it would effectively auto-balance, as opposed to 6 chests... but the storage buffer would be so much less
You should show a test for designs using 12 inserters and underground belts for each car
will have to make another vid with 2-sided, test how full the chests get (if they stay balanced) and also using quality!
The quest for the best continues!
the leftmost design(s) could be improved by reducing inserter capacity that makes the outer one move earlier while the inner ones have enough stock to compensate for the outer ones swinging || there was a mistake: i wasn't following yet ... i fixed that!
so, one of my thoughts is that that should be unloading into chests. I assume what will happen is that some of the designs will unload unequally so some of the chests will empty near the end which will slow down throughput in a longer timeline.
Didn't see my preferred unloader in the mix, basically a cross between to the far right and far left with only one set of facing splitters instead of two and then a belt behind each for more pushed through each splitter. It can easily keep a blue belt fully saturated with four bulk inserters and is nicely compact (5x2 tiles), but have not stress tested on green yet or with stack. Seems all of them struggle with green plus stack.
How about counting total ticks, or making it run for a set amount of time
great idea!
There was a bus fix recently w
On the side splitter merhod that "fixed"the rate that inserters can drop on them
NGL I don't like that design simply because it's not symmetrical. Is there anything to gain by adjusting the stack size in the stack inserters? Perhaps changing one in each half to eight items instead of 16?
Probably covered but Design One (left to right) requires a single inserter to load fewer items faster to make for full saturation.
Designs 2-4 are not good for UPS due to Splitter usage. Though Design Two is the best outta the worst.
All are valid designs up to 6.5-8k SPM base.
I think if stack inserters were uncommon you actually can get full saturation on the turbo belts
very informative video thank you ! keep up the good work
Thanks Walid2875!
I didn't realize there was a Fallout 4 kinda console command "~" that can get you to an unmodded Factorio Blueprint like scenario (where you can even save you current work state to come back to it and work on it some more), where you can prototype stuff without the mod Blueprint Sandboxes, which now I wonder, why is that even a thing!? Now I understand your confusion with not even seeing the scenario of Sandbox on the creation of a Factorio new game page.
Ty for presenting videos where I learn Idk what Idk Idk, but I know Idk a lot (about Factorio this time...!)
oh cool I didn't even know that!
@@TheJarGames You aren't using the /editor command? That's what it looks like you're using, that I wasn't aware of... I'm trying to confirm that, cause that's what it looks like to me.
I want to love your videos, but please increase your bitrate of your videos and/or move to 1440p or 4k recording, this is lacking so much detail it hurts to watch.
That said I want more of this kind of content, I just want it in a reasonable resolution :)
Very interesting test. Gonna do some new blueprints...👍
What about 4 inserter designs?
there are a couple 4 inserter designs, they don't perform as well but they can balance their boxes more.
Look out for the next video where we look at quality and the benefits of some of the 4 inserter designs
can you test which quality stack inserters can saturate green belt?
3 legendary stack inserters are required to saturate a green belt.
my dumbass made an unloader like the best one you concluded in the video but i put a 3 to 1 balancer without thinking 💀 (i didnt think it wouldnt have effect)
idk, dont you need to swap out the train every 40 stacks unloaded? and in that time the belt is empty anyay. the only thing that matters is how many inserters are unloading and if they work at max capacity.
In your third design, delete the first row and change every odd conveyor belt to an underground out belt, that way you get two straight pieces without a corner piece, and the two inserters can fill two separate sides. It's effectively the same design but one tile shorter, more compact!
But I just did an unloader design and decided that it's best to use the splitter out one, your fourth design, but on both side, and then I can bring in the other side with an underground belt, that way I have 8 inserters loading instead of 6, and it my balancer at the station can stay simpler too. (And it's fairly compact too, on the other side of the wagon it's just 3 wide (1st row: inserter/undergound in the middle, 2nd/3rd row: splitter)
cool!
any chance you can blueprint it and paste it here? would love to see it and test it out myself!
I made the design shorter as per your suggestion, love it! still, if you have that blueprint of your 2-sided unloader, let me know! I just tested quality and the results are surprising!
I set my train unloading station with 3 belts per 2 trains with 4 2+2 4 this ensures maximum balanced unloading at all times and I can use other techniques down the line to compress the belts.
I am very disappointed in your best unloading suggestion because it biases the unloading of one side.
Balanced unloading ensures that the transfer inserters between the train car and the chest are at max capacity. If one chest fills up that one 6th of your throughput is gone. Resulting in slower train unloading.
When you are late game what matters most is UPS cost. And for unloader, the simpler the better. Belt filling is issured by slight under consumption.