Resource Scanners - How to Use/Interpret?

[Srpadget]

There's obviously something I'm not "getting" in regards resource scanners.

I now have a survey scanner in the obligatory high polar orbit and a narrow-band scanner in a low equatorial orbit over Minmus. (The latter is where it is because I'm lazy and I intend to put my mining operation A: on flats, and B: near the equator. I want to pick the least-bad spot that meets those criteria.)

I am reasonably clear on the overall concept: the survey scanner gives a broad global overview of resource concentration, but isn't terribly accurate or finely-grained. The narrow-band scanner only provides data for the location it's overflying, but is supposed to be more accurate and thus supplements/improves the overview data obtained with the survey scanner. So, no need to explain that part.

Now here's my issues:

1) I see no way anywhere in the UI to STORE the narrow-band results. It doesn't seem to be updating the survey scanner results... am I expected to get out paper and pencil and make hardcopy notes about the lat/lon of superior locations? Or is there some way to get the game to store the narrow-band results for later use, and I just haven't stumbled across it yet?

2) Near as I can tell, the two instruments' readings have NOTHING WHATSOEVER to do with each other. They're not even VAGUELY similar. The survey scanner claims there are wide swathes of Minmus with Ore concentrations over 60%--there's even one area of Flats, on the equator, that it claims is somewhere in the 50%-60% range. But the narrow-band scanner? It shows NOPLACE with Ore concentrations above 8%; and it appears to be providing values that are tightly tied to Biome type. ALL areas of Slopes have concentrations of 5.65%, all Midlands are 2.75%, the entire Greater Flats is 2.0%, and so forth. Including locations where the survey scanner is claiming 80%-90%-ish concentrations. What's up with that? Clearly, (at least) one set of measurements is utter garbage. Or I'm getting hit by a bug (No! A *bug* in KSP? Say it isn't so!), or I'm grossly misunderstanding something, or or or....

I haven't sent a rover or hopper with a surface scanner yet...which I am given to understand provides highly-precise results for the exact location it is on/over, and is really the one to use for site selection. But I'd kind of HOPED that, y'know, the orbital scanners might be of SOME value in helping me find a decent spot, and then land a prospecting rover there to locate/verify a specific rich location. Without hours upon hours of driving a whole fleet of rovers across every meter of flat ground on the whole moonlet.

So. Educate me. What incredibly-obvious thing am I missing here...?

Edited October 10, 2015 by Srpadget
fix typo

[Clipperride]

As I understand it, the percentage you see from the low resolution survey scanner is based on the planets maximum ore value. So, if the Moon had a maximum of 10% at any location, then a reading of 80% means 80% of that maximum or 80% of 10% (ie 8% actual ore) for that biome. I've found surface reading to be the only really useful guide to placing a mining rig. I've also found that moving just a few metres can sometimes make a big difference in yeild, with some "sweet spots" being quite small.

As for logging information from orbital scans, I used a pen and paper so would also be interested if there's an in game way to do it.

Edited October 11, 2015 by Clipperride
Typo

[KerikBalm]

I'm experimenting with modifying the Overlay.cfg file for the scanner results.

In the past versions, the dots had a different "interpolation" level than the others.. 3 vs 4.

I had the impression that the dots were more accurate then.

Now, in 1.04 they are all set at 4.

I tried taking it down to 1, and comparing it to the others... the result is very different.

I think its much more accurate.

Roverdude did things to "average" and "smear" the results in the overlay... I think the NBS is the only reliable one.

If you don't mind modifying the text files, you can modify Overlay.cfg, and set 3 different interpolation levels for the 3 modes.

I did 1, 2, and the default 4...

Then compare which mode is most consistent with the NBS results.

If you find one that is very consistent, maybe use that mode for convenience (so far, due to the problems you've discussed here, I haven't used the NBS).

At the moment... I have absolutely no idea how the surface scanner works, or what parameter it changes to make the results more accurate.

I think 15% is the maximum ore concentration you will find anywhere.

Indeed Ore.cfg has this:

GLOBAL_RESOURCE

{

ResourceName = Ore

ResourceType = 0

Distribution

{

PresenceChance = 100

MinAbundance = 1

MaxAbundance = 15

Variance = 50

Dispersal = 3

}

}

So... all bodies have ore (except the custom defined ones: Sun, Jool, and it also defines custom ore for certain biomes.... ie no ore in liquid covered places), and the concentration can be as high as 15%.

Clearly, some areas have 0%, despite MinAbundances being set for 1... so that must be for the whole body, not anywhere on that body (perhaps that means every non-custom defined body will have a spot with at least 1% ore).

I'm not quite sure how variance and dispersal work yet.

I'm fairly sure that the % displayed in the overview is showing you the % of 15%...

So if the surface concentration was 15%, the overview would show 100%. If the surface concentration was 7.5%, the overview would show 50%... that is before taking into account the smearing.

[vasco]

The system is designed in a such way that you have to use all three scanners if you want to find a good spot.

First is survey scanner in polar orbit, the result of the scan is completly useless but it has to be done to enable other scanners.

Narrow band scanner - shows average concentration for each biome in range but you can't save the result, so use pen and paper or ScanSat.

And concentration in different places can vary by few percent from the averarage for the biome so you need to use short range scanner, preferably on something flying low as it's less convinient on rovers. And I belive like NBS it doesn't update planet's overlay so again, pan and paper or ScanSat...

Edited October 11, 2015 by vasco

[maltesh]

For a darned good guide, see this post:

Paydirt: FInding Good Ore Deposits, by Geschosskopf

For what I actually do (in collusion with SCANSAT):

1) Use the (orbital) Survey Scanner. THis will give very, very general information about the ore in the biomes.

2) Throw multiple Surface-Scanning Module carrying probe landers, one into each Biome I'm considering landing in. Running an analysis in the biome corrects the orbital scanner values (Especially the ones in the SCANSAT Map).

3) Drop a rover with the Narrow-Band Scanner and the Surface Scanning Module into the best candidate, and drive around looking for a good flat area with high ore.

4) Put the mining base there.

[DMagic]

Near as I can tell, the two instruments' readings have NOTHING WHATSOEVER to do with each other. They're not even VAGUELY similar.

For maximum confusion the percentage used for the cutoff isn't the same as the actual resource concentration. I'm not entirely sure if the cutoff scale just uses the maximum resource concentration value from the config file, or if there is some variation, but given the relative uselessness of the stock overlay it doesn't really make a difference. The cutoff percentage is best used as a way to ignore the areas with very low resource concentrations, set it to 30-50% or so and only pay attention to areas above that cutoff level.

The SCANsat wiki has a section on stock scanning, it covers stock mechanics as used without SCANsat. It also has sections on everything you can do for resource scanning with SCANsat (much more than stock ).

I'm experimenting with modifying the Overlay.cfg file for the scanner results.

In the past versions, the dots had a different "interpolation" level than the others.. 3 vs 4.

I had the impression that the dots were more accurate then.

Now, in 1.04 they are all set at 4.

I tried taking it down to 1, and comparing it to the others... the result is very different.

I think its much more accurate.

Roverdude did things to "average" and "smear" the results in the overlay... I think the NBS is the only reliable one.

If you don't mind modifying the text files, you can modify Overlay.cfg, and set 3 different interpolation levels for the 3 modes.

I did 1, 2, and the default 4...

Those interpolation values determine how much of the may is based on actual resource readings and how much is based on interpolations between those readings. Setting it lower makes the map more accurate, but slower to generate (the game just pauses for a second or so when generating the map at low interpolation values). You can also increase the map size to increase the accuracy and quality of the overlay, but this also makes it slower.

Setting it to 1 will probably make the overlay very slow to use; I wouldn't go below 3.

SCANsat has similar settings for its overlay maps, by default they use the same map size (512*256 textures) but a lower interpolation value (calculates every 8th pixel, instead of every 16th as stock does when the value is set to 4). Since SCANsat has a more efficient map generation method I always switch them to 1024*512 textures and an interpolation level of every 4th pixel, this makes the overlay far more useful without being too accurate; these values can easily be changed in-game to see the difference.

[Srpadget]

Okay, I think I'm beginning to see how these different scanners work (or *don't*, as may be...) and it seems that the only way to get actual decent info is to drop a whole dang fleet of surface-scanning hoppers or rovers onto every candidate biome...and physically traverse every flattish square meter of Minmus's surface.

Related question: I've seen a few analyses (usually concerning when it makes sense to power a mining operation with fuel cells that use a fraction of the rig's LF/O to power the facility) that provide tables in terms of % concentration of Ore. They generally run concentrations all the way up to 100%. I guess I should assume that "100%" on those tables is actually 15% concentration as measured by the NBS or ground scanner?

[Goomblah]

Finding a good mining spot with good ore concentration is really just a three-step process; you don't need to spend a ton of time scanning every single biome. To get a good spot (which may or may not be the best on the planet/moon, but it will be well above average) you really only need to do three things:

1) Use the orbital survey scanner to get a general idea of where ore is located. Find a spot with relatively high concentration that's in a good spot (close to the equator and on a flat surface).

2) Send a lander equipped with the surface scanning module to the area you selected. Precision isn't terribly important; it just has to be in the same biome as the target zone. Do the surface scan; this will give the NBS more accurate data for that biome.

3) Use an orbital NBS passing over the target, or drop a rover with it in the area. Make a few passes over the area to find a good ore spot; you can click around on the UI to find a good place. Write down those coordinates, then drive a rover over there to function as a landing target. If your rovers are manned, plant a flag.

I find that, when searching for good mining locations, it's not worth your time finding the absolute best spot, or scanning every biome trying to find the best ore. Just locate a location that looks good, then scan the immediate area for the best nearby spot.

You actually have almost all the scanning you need done; just send a lander with a surface scanning module to the spot that had 8% ore on the NBS. Then you will get more accurate data - that 8% is a biome average, so you can probably find a spot with even higher concentration once you have the surface scan done.

Edited October 12, 2015 by Goomblah

[Srpadget]

2) Send a lander equipped with the surface scanning module to the area you selected. Precision isn't terribly important; it just has to be in the same biome as the target zone. Do the surface scan; this will give the NBS more accurate data for that biome.

3) Use an orbital NBS passing over the target, or drop a rover with it in the area. Make a few passes over the area to find a good ore spot; you can click around on the UI to find a good place.

If I'm interpreting those 2 steps correctly, then, it seems that if I drop a surface scanner onto a biome, then the narrow-band scanner will start providing "local" numbers when overflying that biome, rather than a single constant "global" value for the entire biome? Because as it stands now, the narrow-band scanner is pretty well useless for nailing down a location of interest--ALL areas of "Highlands", for instance, are shown as 6.52%. I take it that will change when I land a surface scanner someplace in a Highlands biome, and the NBS will start showing me that certain Highlands areas are better than other Highlands areas?

[Goomblah]

If I'm interpreting those 2 steps correctly, then, it seems that if I drop a surface scanner onto a biome, then the narrow-band scanner will start providing "local" numbers when overflying that biome, rather than a single constant "global" value for the entire biome? Because as it stands now, the narrow-band scanner is pretty well useless for nailing down a location of interest--ALL areas of "Highlands", for instance, are shown as 6.52%. I take it that will change when I land a surface scanner someplace in a Highlands biome, and the NBS will start showing me that certain Highlands areas are better than other Highlands areas?

That's correct - the NBS will just give you a biome average until you do a surface scan in that biome. One you have done that, it will start reading different values for different locations in that biome. The UI's color coding will not change, but the data values will no longer be uniform.

[Ripper2900]

The UI's WILL change after doing surface scans. Both the survey overlay and the shades of the NBS view. But usually not by much.

I've just scanned and mined Moho, and in my case, the hottest biome was the "Minor craters" - a number of craters scattered around the surface. After landing in one and doing the surface scan, the orbital overlay changed to highlight some of the craters over others.

The best was at high lattitude : bad for mining. So I focused on the equatorial ones, and found that the western part of one gave me 6.9%, while others had less than 4%.

I like to use one of the non-monochrome highlight colors for the overlays. They provide a much clearer image of what's going on, than the default pink color.

[Srpadget]

... scanned and mined Moho, and in my case, the hottest biome was ...

Heh. I see what you did there!