Random digging wastes more than shovel durability in Last Breath. A shaft can pass through the right mineral band without opening enough horizontal ground to expose an ore vein, and a successful haul can still be lost when the return path is unclear. The useful question is not simply “how deep should I dig?” It is which depth band to target, how long to work sideways there, and what evidence tells you to move.
Research closed on August 24, 2026, with Update 0.8.0 (August 21) as the current official cutoff. The depth observations below come from a Steam community guide updated for 0.7.9d, while official 0.7.9 and 0.7.9d notes confirm that ore generation, detectors, and depth-weighted resource distribution were recently changed. Treat the ranges as search bands, not guaranteed spawn coordinates; the depth display, detector response, and mineable objects in your current world are the final readback.
Choose a mineral band before breaking ground
Use the HUD depth number as the mine’s coordinate system. Descending without a target only creates a longer return trip. Pick the shallowest band that contains every material needed for the next machine or tool, reach the middle of that band, and then excavate laterally. This gives the generation system more wall and floor area to reveal than a narrow vertical hole.
| Resource | Community-observed search band | Practical first target |
|---|---|---|
| Wood | 0–110 m | Collect it when exposed; do not descend specifically for wood. |
| Stone | 0–110 m | Any working level should replenish the stone spent on mine construction. |
| Plastic scrap | 0–110 m | Treat it as a useful by-product rather than a depth target. |
| Vegetables | Near the surface; the lower limit was not established | Do not build a food plan around an unknown underground floor. |
| Oxerite | About 5–120 m in the guide’s table | Start near 20 m; the same guide encountered a world floor around 108.2 m, so the listed upper end is not a reachable promise. |
| Charcoal | About 5–25 m | Search shallow only if it appears naturally; producing charcoal from wood is the more controllable route. |
| Iron | About 7–30 m | Work the 20 m band before opening a deeper branch. |
| Copper | About 15–50 m | Begin at 20 m, then use a mid-depth branch if demand remains. |
| Aluminum | About 30–60 m | Open a broad chamber around 50–53 m. |
| Graphite | About 40–70 m | Search the same mid-depth chamber used for aluminum and early gold. |
| Gold | About 50–110 m | Try the 50–53 m band before committing to the deep mine. |
| Uranium | About 65–110 m | Use a signed branch around 75 m and confirm with a detector. |
| Titanium | About 70–110 m | Search near 75 m, keeping the final descent reversible. |
The ranges overlap because a depth is a probability band, not a single-material floor. That overlap is valuable. A 20 m chamber can supply the early iron–copper–Oxerite combination; a chamber around 53 m can expose aluminum, graphite, and gold; a 75 m branch can serve gold, uranium, and titanium. One well-marked working level is safer and cheaper than a separate shaft for every ore.
Run three horizontal passes instead of one vertical lottery
Pass one: establish the 20 m workshop. Stop descending when the HUD is close to 20 m. Cut a chamber wide enough to see several fresh surfaces, then extend short spokes in different directions. This is the first useful overlap for iron, copper, and Oxerite. If the chamber produces only common material, widen it before going deeper. A narrow stairwell gives the game little new terrain in which to reveal a deposit.
Pass two: build the 50–53 m junction. Aluminum begins above this point, graphite becomes plausible, and gold enters its observed band. Make this a deliberate landing with storage and a depth label, not a hole passed at sprint speed. Search one direction at a time so a detector contact can be tied to a branch. Return to the landing after each spoke; an expanding circle with no orientation creates a rescue problem later.
Pass three: reserve 75 m for late materials. Uranium and titanium are deep targets, but gold remains possible. Carry enough tool durability and route materials to leave the branch usable even if the first sweep is poor. The community guide’s approximate 108.2 m floor conflicts with its Oxerite table extending to 120 m, so do not plan a “120 m level.” If the terrain stops accepting a downward cut, the correct response is lateral coverage inside a verified band, not repeated attempts to force the world lower.
Update 0.7.9 made underground ores more common and described occasional veins containing two to four deposits. That does not guarantee a vein in every chamber. It changes the best response to an empty spot: expose more terrain at the same useful depth, scan it, and only then abandon the band.
Make the detector answer a decision
Official 0.7.9 notes say detectors were reworked to focus on mineable underground minerals, and the Tier 1 and Tier 2 versions gained depth indicators. Use that information to separate two questions:
- Am I at a plausible depth? Compare the HUD with the target band before trusting any search pattern.
- Is there a mineable target close enough to justify this branch? Sweep from the landing, note the direction, then advance in a controlled line.
- Did the contact end? After mining the object, return to the same scan position and check whether the detector changes target.
Update 0.8.0 fixed a Tier 3 detector fault that could keep it locked to depleted ore and prevent target switching. If a Tier 3 unit still points at an exhausted object, first verify the client build. Re-scan from the landing and test another visible mineable object before extending the tunnel. A stale lock is a troubleshooting signal, not proof of a hidden second deposit behind the empty node.
Do not turn a detector into an instruction to dig straight through everything. Preserve the stair spine and work from branches. When a contact would require cutting away the return route, open a parallel access from the nearest landing instead. The few extra shovel actions protect the only path that still works when the flashlight, tool, or oxygen budget becomes tight.
Tell loose drops from ore boulders before clearing the wall
The community guide distinguishes ordinary spawned resources from underground ore boulders that must be worked with a pickaxe. It also reports a respawn risk: fully removing the surrounding dirt can let a boulder fall and may prevent it from returning. This is a community-observed behavior rather than a promise from the developer, so preserve the object until the current client proves otherwise.
When a mineable boulder appears, stop widening the entire chamber. Clear only the face needed to reach it, leave supporting terrain around or beneath it, and mark the landing and branch. If an earlier excavation already stripped the support, the guide’s recovery method is to use the shovel’s Add mode to rebuild dirt where the boulder sat. Treat that as a repair attempt and verify later that the object actually returns; do not budget future production until the readback happens.
A useful marker records the asset, not merely the depth: 53 m | graphite boulder | east branch. A resource band tells you where a spawn is plausible. The marker tells you where this world produced a reusable point of interest. Those are different kinds of evidence.
Carry an exit budget, not just an inventory budget
A profitable branch becomes a failed trip when every slot is full, the shovel is nearly broken, and no construction material remains to correct the route. Before leaving a signed landing, reserve capacity for the return:
- a usable shovel for opening or repairing the path;
- a pickaxe for confirmed mineable objects rather than loose-dirt excavation;
- a light source that does not depend on remembering an unmarked bend;
- food, water, and oxygen appropriate to the live meters in the current client;
- enough stairs, floors, signs, or other route material to make the next junction readable;
- one free inventory margin so the target ore is not left on the floor.
Update 0.8.0 added a compass to the HUD and a standing torch. They improve orientation but do not solve vertical navigation by themselves. A compass can keep a horizontal branch consistent; a signed stair landing still has to tell you which way leads upward. The same update fixed cases where digging or jumping in caves could teleport a player to the surface. Current clients should not require that bug as an escape plan, and older footage that appears to recover that way should not be copied.
The companion mine layout guide turns those supplies into a protected entrance, measured stair landings, signs, and a return drill. Read it when the ore table is clear but the shaft itself keeps becoming the problem.
Abandon a branch for a recorded reason
“No ore yet” is too vague to guide the next attempt. Before moving deeper, record what failed:
| Observed state | Next action | Why |
|---|---|---|
| Correct band, very little wall exposed | Extend two short lateral spokes | The sample is too small to judge the band. |
| Correct band, broad chamber, no detector contact | Return to the landing and open a new direction | More depth is not automatically better than fresh terrain. |
| Detector remains on a depleted target | Verify 0.8.0+, reset the sweep position, and test another object | The official patch identifies this as a former Tier 3 failure mode. |
| Target material obtained but exit reserve is low | Bank the haul before continuing | A second deposit is not worth risking the first. |
| Near the world floor with no deep target | Work sideways around 75 m | The observed mineral bands already include the deep resources there. |
| Ore boulder has lost its terrain support | Add dirt, mark the spot, and verify a later respawn | The source reports that a fully freed boulder may stop returning. |
A successful ore run ends with three readbacks: the target resource is stored, the branch is labeled with its measured depth, and the return route is still traversable without digging a new exit. If one of those is missing, fix it before opening another level.
Sources and version boundary: the official Last Breath Steam listing establishes the Early Access identity and mining-focused systems. Official Update 0.8.0 controls current compass, torch, cave, and detector claims; official 0.7.9/0.7.9d notes control the recent ore-distribution boundary. Depth bands and ore-boulder handling were cross-checked against the supplied Steam community guide updated for 0.7.9d. Unverified drop rates, fixed vein counts, and permanent coordinates are intentionally not claimed.