|
RabbitWizard posted:The colors in Dectorio are really bright (and there is no white floor) so I'm not sure I'll continue painting my base. An effect of letting players (and mod makers, and humans in general) set 8 bit color values in RGB is that people who don't know how color works tend to think things like "I want blue! So I'll set B to 255 and RG to 0!" This unfortunately results in garish fluorescent garbage. The resulting Maximum Blue color is physically impossible to paint or print; if you try to use it for something that's supposed to be a natural or painted surface your brain will just fail to make sense of it and think it looks weird. Fortunately(?), Factorio's lighting is also garbage and doesn't really do colors correctly which may cover your crimes, but using physically impossible colors doesn't help as a rule. I think that mod author might have tried to desaturate a bit but clearly not enough because ow my eyes they bleed. In case anyone cares why RGB(0, 0, 255) isn't possible to paint or print: the brief explanation is that what you consider "blue" light is actually light comprised a bunch of different wavelengths that happen to tickle the "blue" cone cells in your eye. Your "red", "green" and "blue" cone cells -- assuming they all work for you -- each has their own spectrum of wavelengths that they detect and, crucially, those spectrums overlap. Most of the photons that activate a "red" cone would also activate a "green" cone. When you say that the color of some material is RGB(0, 0, 255) then you're basically saying that material is reflecting All Of The Blue Wavelengths but, because the blue-cone and green-cone spectrums overlap, to do so a real material must also reflect Some Of The Green Wavelengths. The only way you could get a pure blue color from reflecting white light is that if the material reflects more of the blue-only wavelengths than there are in white light to begin with, which is basically what fluorescence means. It gets a little messier in reality because computer-blue and eye-blue are not defined by the exact same spectrum of wavelengths and because "white" is a matter of definition, but without going Full Math with optimal colors, MacAdam limits and Pointers gamut that's basically the idea. The rule of thumb is that plausible material colors -- both natural and artificial -- are never going to be both bright and saturated. In practice - Avoid extreme values at or very near 0 or 255. - In particular, never have G close to 0 or B close to 255. - If one of RGB is very large, its neighbor(s) cannot be very small.
|
#
?
Nov 12, 2020 14:54
|
|
|
|
| # ? May 27, 2024 17:16 |
|
|
Xerophyte posted:An effect of letting players (and mod makers, and humans in general) set 8 bit color values in RGB is that people who don't know how color works tend to think things like "I want blue! So I'll set B to 255 and RG to 0!" This unfortunately results in garish fluorescent garbage. The resulting Maximum Blue color is physically impossible to paint or print; if you try to use it for something that's supposed to be a natural or painted surface your brain will just fail to make sense of it and think it looks weird. Fortunately(?), Factorio's lighting is also garbage and doesn't really do colors correctly which may cover your crimes, but using physically impossible colors doesn't help as a rule. I think that mod author might have tried to desaturate a bit but clearly not enough because ow my eyes they bleed. Well this is certainly useful and explains why my colours always look like poo poo.
|
|
#
?
Nov 12, 2020 18:21
|
|
|
Could the tones in the mod's lua files be modified manually?
|
|
#
?
Nov 12, 2020 20:57
|
|
|
Xerophyte posted:In case anyone cares why RGB(0, 0, 255) isn't possible to paint or print: the brief explanation is that what you consider "blue" light is actually light comprised a bunch of different wavelengths that happen to tickle the "blue" cone cells in your eye. Your "red", "green" and "blue" cone cells -- assuming they all work for you -- each has their own spectrum of wavelengths that they detect and, crucially, those spectrums overlap. Most of the photons that activate a "red" cone would also activate a "green" cone. The rest of your post is good advice for getting natural looking colors but this is not why at all! The difference is additive (emitting colored light) versus subtractive (using pigment to absorb wavelengths) coloring. You can't print #0000FF because your printer doesn't actually contain blue ink. It contains Cyan. It also doesn't contain Red (Magenta is used) or Green (it has Yellow!). Not to mention that we really don't have any actual pigments which absorb everything but 445nm EM radiation, and the substrate you put it on or the medium you mix it in has its own absorption profile which would impact the visible color. The entire visible color spectrum is always your brain interpreting the signals from 2 or 3 of your cones, if there was no ability for your brain to see #0000FF, that would be the case for both on the screen and in real life. Dietrich fucked around with this message at 21:28 on Nov 12, 2020 |
|
#
?
Nov 12, 2020 21:07
|
|
|
There's also the fact that RGB(0, 0, 255) isn't actually a defined color, so "printing" that means "print the most blue thing you can" which is going to vary from printer to printer. It also varies from monitor to monitor - a specific RGB value on one monitor will look different from another unless they've both been calibrated to the same spec. In any event, those colors on the concrete aren't actually at 0, 0, 255 because they would be completely well... blue and have no detail. They are the normal concrete sprite multiplied by that value, which basically just removes the red and green channels from the sprite. It still looks like garbage, but not for that reason. KillHour fucked around with this message at 22:38 on Nov 12, 2020 |
|
#
?
Nov 12, 2020 22:36
|
|
|
Dietrich posted:The rest of your post is good advice for getting natural looking colors but this is not why at all! Guess we're going a bit into the weeds... First: a mea culpa, my memory was that saturated sRGB blue was outside the MacAdam limit. It's not (for a D65 illuminant) so you're right that there exists a theoretical perfect blue ink that could represent #0000FF in white light, quite unlike what I said. The MacAdam limit certainly exists, though! It constrains the printable range of less poo poo color spaces: for instance, Adobe RGB contains colors that are outside the limit and therefore cannot be printed with even a theoretical perfect ink. A saturated red Adobe RGB color can still be perceived and displayed on a monitor, but it cannot be printed. That's not because of some limitation of the human visual system, it's just that a white (D65) illuminant of intensity 1 doesn't contain that much radiance in the red spectrum to begin with so there's no way to make a pigment that reflects it. You can still get a color that red by changing the illuminant to something else, but the entire point of a reflectance spectrum is that it's defined for a white illuminant of intensity 1. That it's very nearly impossible to make a #0000FF blue ink is still why a maximally blue color looks wrong on a train in a video game. Printers aren't really the point: the reflectance spectrums of actual materials are not limited by the gamut of CMYK pigments either and they're still never that blue (they're within Pointer's Gamut). The reason they're not that blue is because in practice if you want something to reflect a lot of the "blue" wavelengths you have to also reflect some of the "green" wavelengths, regardless of precisely what matching functions or XYZ primaries you use to define blue and green. E: I guess I'm still being a little gauche with using color in both the sense of "reflectance distribution" and the sense of "radiance distribution" here, plus handwaving a lot of other terms. I just wanted to give a layman's explanation on why using really saturated colors looks weird in the silly factories and train game without getting into the CIE standard observer, the linearity of tristimulus color spaces and the difference between an E and a D65 illuminant... Xerophyte fucked around with this message at 22:47 on Nov 12, 2020 |
|
#
?
Nov 12, 2020 22:37
|
|
|
what the gently caress kind of layman explanation is that Roughly zero of your information is actually relevant. All three 'bright' colors in that screenshot show up in nature because setting something's color to (0,255,0) in Factorio does not end up creating a single pixel of the color (0,255,0) on your screen, they're not pushing the limits of light reflection/absorption by any means. You're applying theoretical information without paying attention to the screenshot itself at all, so you're talking about something which is true (that modders sometimes use pure diode colors that look hosed up and eye-searingly unnatural) in vastly overcomplicated terms in a situation where it does not even apply. The colors in that screenshot look unnatural because they would quite simply not show up that way in a scenario that approaches realism--you could paint something any of those colors for a pittance but if you paint the floor that color in an industrial zone in the middle of a desert that color won't last long enough for the paint to dry, as dirt and other particles blow in from the desert and industrial work will inevitably cause heat/scraping/etc damage. e: Actually I remember you can get an absolutely horribly colored train so I think the textures of things you can change the color of do have parts that show the pure color, but the point remains that it's not actually relevant to why that mod's paints look garish. Count Uvula fucked around with this message at 05:39 on Nov 13, 2020 |
|
#
?
Nov 13, 2020 05:36
|
|
|
Is there any way to get rid of pollution over water? I’m concerned the pollution from a copper mine is going to prompt a bunch of attacks from way, way far away because it seems to just be growing across the water with no mitigation. Is there any way to scrub pollution generally? My map is luckily heavily forested which makes pollution mostly a non-issue but efficiency modules only go so far and base defense is tedious.
|
|
#
?
Nov 13, 2020 05:54
|
|
|
jokes posted:Is there any way to get rid of pollution over water? I’m concerned the pollution from a copper mine is going to prompt a bunch of attacks from way, way far away because it seems to just be growing across the water with no mitigation. Not without mods (of which there are plenty). Also, water does absorb pollution better than clear land.
|
|
#
?
Nov 13, 2020 10:27
|
|
|
Anyone have an idea of why bugs don't seem to attack my single turrets surrounded by walls? They'll just forever run into the walls and stand there whilst getting shot, or occasionally they'll damage the wall a tiny bit. They go ham on buildings though. Five hours into my game and I'm really not liking these ore spawns, to the point where I might restart. They're all within spitting distance and are a whopping 4mil each. I had it set to railworld but clearly that isn't enough. I wanted my ore to be spread out way more than this, like one of these 4 mil patches in the entire map screen (even the black uncovered parts), with the next patch being an equally large area away. Anyone know how to fix the railworld settings? As it is, I can easily just grab two or three of these patches without needing a large rail network and be good for 20 hours.  I think railworld may have overruled RSO settings? I used the console command to show a huge part of the map, and it's just tonnes of ore patches everywhere. Can I use a console command to change the world gen settings in a current game? Qubee fucked around with this message at 13:43 on Nov 13, 2020 |
|
#
?
Nov 13, 2020 10:41
|
|
|
E: ^^^ no idea, sounds a bit weird. Maybe you could try making those turret outposts bigger? --- Megabase time! I've done distributed/modular rail bases, so this time I wanted to challenge myself by producing a main-bus-fed 1000 sci/min base, and without using bots for any bulk cargo. I like playing in expensive mode for some reason, so I guess this equates to 2000+ sci/min. In the end I've ended up constrained by rail-junction throughput, and ripping up all of my rail & redesigning all of the junctions is far too painful at this stage. Dear Wube, please please please implement train tunnels/overpasses!  This is what's left of my starting base. My bigger base grew around it (as you can see) and I didn't have the heart to remove this completely. So it just sits there, reminding me of simpler times   The base of the bus for the bigger 60 sci/min base, and its terrible-but-works sushi-belt-mall:   And here's the tail end of the bus:  Here's all of this base. I used a mod to give me an island start, with a continent just to the north (you can see the rails stretching off-screen). Even after moving to the proper base, this one still serves as my arms, rocket fuel and nuclear fuel manufacturer.  Now on to the main event! Here's the main bus, ~36 blue-belts wide. To the right is a tiny fraction of the Low Density Structure manufacturing, and because I enjoy compact bases you can see an out-going rail line too.  Science packs & research. I had to change the sushi-belt arrangments a few times to increase throughput and stop it seizing up, but got it into a satisfactory state.  For this base I used many mods, but the key one was Built-In Beacons - this essentially lets you build faster Assemblers, etc. that act as though they're surrounded by beacons, when in fact they're built into the producer. I don't really like beacons as they're currently implemented, so I would thoroughly recommend this mod. Here is the full base. You can see the top of my starting island just at the bottom, and one of several nuclear plants at the top. And off to the west, part of the sprawling rail network and some of the terrible junctions I've bodged together. This base does *all* of the smelting & manufacturing, save for petrochem, arms and nuclear fuel. You can also see most of the earlier defensive wall running down the left-hand side, and where I had to remove part of it for more green circuit production.  Speaking of power, early on I produced this hodge-podge nuclear processing & plant. Surprisingly, it produces enough U for 20+ GW of nuclear power.  After a few iterations, I designed this 23-reactor nuclear plant (there are more turbines off the edges of the screen). It's not the most efficient, but I'm happy with it. It was actually very tricky to get a sufficient steam flowrate to all of the turbines. In the end I had 6 of these dotted across the map.  And finally, here's the whole map. Bottom-right is the vestigial starting area; just above that is the actual base. To the north-east is half of my petrochem processing, with a second plastic production facility in the south of the map. Everything else outside the base is mines, nuclear plants, train depots and artillery outposts. 
FnF fucked around with this message at 17:37 on Nov 13, 2020 |
|
#
?
Nov 13, 2020 17:32
|
|
|
I want snow, how'd you get snow!
|
|
#
?
Nov 13, 2020 17:50
|
|
|
Alien Biomes mod adds a ton of new environments.
|
|
#
?
Nov 13, 2020 18:10
|
|
|
Nice updates into today's FFF https://factorio.com/blog/post/fff-363 They are adding directional arrows for alerts so you can get a hint about where they are coming from without having to go to the map. Also some UI changes which look nice (easy way to add icons to station names and such), getting rid of the build/character/logistics tabs and just putting it all on one page so you don't have tabs within tabs.
|
|
#
?
Nov 13, 2020 18:57
|
|
|
Squibbles posted:getting rid of the build/character/logistics tabs and just putting it all on one page so you don't have tabs within tabs. You can already do that, they're just making that' setting the default now.
|
|
#
?
Nov 13, 2020 19:03
|
|
|
Squibbles posted:They are adding directional arrows for alerts so you can get a hint about where they are coming from without having to go to the map.
|
|
#
?
Nov 13, 2020 20:06
|
|
|
I like the new night time lighting!
|
|
#
?
Nov 13, 2020 21:24
|
|
|
FnF posted:Base screenshots. Lookin good.
|
|
#
?
Nov 14, 2020 10:37
|
|
|
necrotic posted:I like the new night time lighting! Electric Furnaces mod has been doing that since 2017  
|
|
#
?
Nov 14, 2020 11:32
|
|
|
necrotic posted:I like the new night time lighting! Yeah, I can't wait to see it in action. It looks beautiful!
|
|
#
?
Nov 14, 2020 15:05
|
|
|
FnF posted:
Is it normal to put all your lab packs together on a single belt? I'm starting on my 5th lab pack and not sure if I should keep dedicating lanes to packs or just let them sushi like yours.
|
|
#
?
Nov 14, 2020 19:40
|
|
|
the tingler posted:Is it normal to put all your lab packs together on a single belt? I'm starting on my 5th lab pack and not sure if I should keep dedicating lanes to packs or just let them sushi like yours. Having science on a single sushi belt is abnormal, but abnormal can be fun. For vanilla it's arguably easiest to have 3 belts running parallel to your labs and using 1 normal and 2 long inserters, which is enough for 6 pack types. You can move packs between labs with inserters, so another option is to have a field of labs with filter inserters to move one set of science packs e.g. left-to-right while another set goes right-to-left.
|
|
#
?
Nov 14, 2020 20:15
|
|
|
Regarding color chat (which was very interesting): https://www.youtube.com/watch?v=FTKP0Y9MVus @FnF: Very nice base! So I had another great idea, why build individual chunks of furnaces that only make one item? What a waste of time and capacity, I don't need all the items all the time anyway. Let's just make some smart furnaces that produce all the items. ~2-3 hours later:   Awesome! 5 minutes later: Oh no. Bricks and steel need a specific amount of items, so sometimes items remain in the furnace and the new thing doesn't get produced. Well, let's add a little bit logic, can't be that hard. ~5 hours later, including a dozen redesigns and testing, I finally did it.  Just connect a red and green wire from the control unit:  It works!  One question came up while building this: Why can't you transmit a zero on a logic channel? Did I miss anything? The "each" function is great. But it is useless if I want to - for example - add the number 5 to all things going in (number of bricks/plates/whatever), because if I'm out of thing X it won't output X=5 and instead give me nothing at all because a 0 doesn't get transmitted  I can't be the first person to have this problem, so can anyone explain whyyyyyyy??? Working around that took quite a while.
|
|
#
?
Nov 14, 2020 21:49
|
|
|
RabbitWizard posted:One question came up while building this: Why can't you transmit a zero on a logic channel? Did I miss anything? The "each" function is great. But it is useless if I want to - for example - add the number 5 to all things going in (number of bricks/plates/whatever), because if I'm out of thing X it won't output X=5 and instead give me nothing at all because a 0 doesn't get transmitted From the wiki: quote:All unused channels have the value zero. All signals are always transmitting 0 until you explicitly transmit a value other than 0. You don't see this because they filter out 0s on the mouse over that shows signals. I believe all of the virtual signals exclude signals with a value of 0, as this is treated as the "not set" value. edit: Ah, no, just each: quote:Each: This combinator can use the 'Each' signal for both input and output, in which case all non-zero input channels will have the combinator's operation performed and broadcast on the output side necrotic fucked around with this message at 22:04 on Nov 14, 2020 |
|
#
?
Nov 14, 2020 22:00
|
|
|
I understand that part. I don't understand why. I feel like if I'm sending values on 4 different channels into a logic block, values on all 4 channels should come out, even if they are zero. Most basic thing is figuring out if my logic is correct. If my logic is wrong, nothing gets transmitted on a channel I expect to transmit something. If the value somehow becomes zero, also nothing gets transmitted. Which one is it? Always an adventure to figure it out. Basically, just a difference between "nothing at all and there never was" and the value zero due to some math. Instead of using one logic unit to see which of the 20 things I'm tracking are empty so I can request new ones, I have to use 20. Is this a programming thing from the game so it just can't deal with that?
|
|
#
?
Nov 14, 2020 23:00
|
|
|
Circuit networks aren't like traditional variables, as matching signals are always summed. If you want to keep 20 each of something, have you considered using a constant combinator to send -20 of each of those items, and then only taking action on the signals that are less than zero?
|
|
#
?
Nov 14, 2020 23:24
|
|
|
RabbitWizard posted:I understand that part. I don't understand why. I feel like if I'm sending values on 4 different channels into a logic block, values on all 4 channels should come out, even if they are zero. Could you add a separate constant combinator always sending zeroes after the output from your logic? E: or what gotLag said
|
|
#
?
Nov 14, 2020 23:26
|
|
|
I beat the game standard and now I'm doing a run with LTN and most/all the Angel / Bob's mods installed. One thing that's really annoying is when I pickup a building/inserter/etc it just dissassembles it to raw resources in my inventory and I have to re-craft it. Is there a way to change this? I didn't see it under Mod Settings. Simultaneously, is it possible to assign things to my hotbar and then have it start crafting the building/whatever by pressing it on the hotbar if I don't have any in my inventory? This would be SUPER helpful because its a pain having to go through the crafting menus constantly and if I could just hotkey lots of them and just press it on the hotbar to craft it, it'd save so much time. Xaris fucked around with this message at 06:41 on Nov 15, 2020 |
|
#
?
Nov 15, 2020 06:38
|
|
|
Xaris posted:I beat the game standard and now I'm doing a run with LTN and most/all the Angel / Bob's mods installed. One thing that's really annoying is when I pickup a building/inserter/etc it just dissassembles it to raw resources in my inventory and I have to re-craft it. Is there a way to change this? I didn't see it under Mod Settings. I think that's an angel's industries feature, I'm sure there was an option last time I played it.
|
|
#
?
Nov 15, 2020 13:24
|
|
|
the tingler posted:Is it normal to put all your lab packs together on a single belt? I'm starting on my 5th lab pack and not sure if I should keep dedicating lanes to packs or just let them sushi like yours. Normal for me! Using dedicated lanes is 100x simpler even for all 7 science-types, but it always annoys me when I go from 4 to 5 types 'cause I'll have to rip stuff up and make more room. Plus, sushi-belt arrangements feel more compact, and more compact == more efficient == more good to my broken brain  When I inevitably pick this game up again, I'll finally implement a proper one-out-one-in system so that more science packs are fed in to the labs only when they consume. Which should theoretically never jam! FnF fucked around with this message at 17:14 on Nov 15, 2020 |
|
#
?
Nov 15, 2020 17:11
|
|
|
I always go with the interlocking ring for science which allows each lab access to up to 8 belt-halves https://steamcommunity.com/sharedfiles/filedetails/?id=2288291705
|
|
#
?
Nov 15, 2020 17:31
|
|
|
I do a similar thing, except I don't make them into rings. One thing I do do is put a corner buffer for each research just before merging, allowing me to sit on ~1k science. I've never built a base big enough that I cared about the size of my research facility. It's one of the smallest things, the only reason to space-optimize it is for fun.
|
|
#
?
Nov 15, 2020 17:47
|
|
|
What’s wrong with just an infinitely long series of two parallel belts (4 belt halves)that have a long inserter and a yellow inserter pulling off it? And then the same thing on the other side if you want up to 8?
|
|
#
?
Nov 15, 2020 18:45
|
|
|
 You can toss a belt along the top when you get space science.
|
|
#
?
Nov 15, 2020 18:59
|
|
|
uPen posted:
But then you hit the problem where, with enough upgrades, you're burning through the vials of science faster than the inserters can transfer them, so that topmost lab never spends any time researching, as the arm is constantly shuttling things down to the southernmost lab... :\
|
|
#
?
Nov 15, 2020 19:50
|
|
|
XkyRauh posted:But then you hit the problem where, with enough upgrades, you're burning through the vials of science faster than the inserters can transfer them, so that topmost lab never spends any time researching, as the arm is constantly shuttling things down to the southernmost lab... :\ someone do the math and figure out how many labs could be serviced by 3 stack inserters going down the chain
|
|
#
?
Nov 15, 2020 19:53
|
|
|
Evilreaver posted:someone do the math and figure out how many labs could be serviced by 3 stack inserters going down the chain Even stack inserters will only move one *type* of science at a time. If you're doing one of the end-game Infinites that require six different vials, that's going to be your bottleneck. But yes, I suppose 3 stack inserters would mitigate this issue a lot better than a single fast inserter, as the screenshot I quoted!
|
|
#
?
Nov 15, 2020 20:01
|
|
|
Your bottle neck there is the stack inserters feeding the first labs from the belt.
|
|
#
?
Nov 15, 2020 20:10
|
|
|
This is the science part of my last megabase, it is a bit overbuilt and eats 1 blue belt of science. Replacing the filter inserters with stack filter inserters makes the setup less efficient.
|
|
#
?
Nov 15, 2020 20:15
|
|
|
|
| # ? May 27, 2024 17:16 |
|
|
XkyRauh posted:But then you hit the problem where, with enough upgrades, you're burning through the vials of science faster than the inserters can transfer them, so that topmost lab never spends any time researching, as the arm is constantly shuttling things down to the southernmost lab... :\ Under ~100 SPM it works great. If you're getting into beacons it obviously won't work as there's no room for beacons.
|
|
#
?
Nov 15, 2020 20:24
|
|