Correct errors found by review; hedge unmarked inferences

A four-agent review audited arithmetic, cross-document consistency,
epistemic honesty and style. It found substantially more than the
manual pass in the previous commit. Errors that changed conclusions,
not decimals:

Arithmetic
- production-timeline: the headline lead times did not reconcile with
  the stage table on the same page. Launch is 22-49 days (was 21-39),
  mature 11.5-22 (was 14-20), repress 6.5-11.5. Including gates,
  launch is 6-17 weeks, not 8-14.
- Consequence: at 17 weeks the launch configuration is at parity with
  EU/US plants, not ahead of them. The page now compares quote to
  quote rather than a stage sum to a quoted lead time, and says so.
- The galvanics turnaround saving is 1.5-5 weeks, not 3-5.
- scale.md Duplex capacity 300-600k/yr did not reproduce from its own
  stated basis; corrected to 175-540k. production-timeline disagreed
  with scale.md on the same fleet; both now agree.
- "300 units is under an hour of pressing" was out by an order of
  magnitude on the semi-auto line: 5-8 hours.
- project-timeline carried two totals for the same quantity. Now
  27-33 months throughout, with the Gantt's two ordering breaks fixed.
- scale.md: 300 kWp covers 28-45% of consumption, not 20-35%; roof PV
  stated as both 150-200 and 247 kWp for the same roof; parking canopy
  dimensions and PV disagreed with the RFQ.
- RFQ: fire tank stated 200 m3 where both its geometry (254 m3) and
  its own mass (250 t) say 250. Drainage now gives total sealed area
  66%, not the 46% paving figure, since that is what attenuation needs.
- Galvanics payback used only the top of the saving range: 1-5 years,
  not 1-4.

Inference marked as inference
- The 22,000 THB stamper figure is an invoice RESIDUAL. It carries
  lacquer cutting, test pressings, setup and margin, so it is an upper
  bound on the stamper component, not a measurement. Everything derived
  from it now says so, and a line-itemised invoice is the top question.
- The saving was also benchmarked against the wrong supplier: CNX buys
  from Stamper Discs, not a Thai reseller. Stated as 175-665 USD until
  that quote exists.
- The stamper-bottleneck hypothesis now names its alternatives. Jacket
  print runs would produce the same symptom and would constrain CNX
  equally, which would remove the competitive read entirely.
- Competitor capacity relabelled as nameplate times an assumed OEE,
  with the assumed shift pattern stated; all rows on one 300-day basis.
- The fixed/variable decomposition keeps the shape and drops the point
  estimates: the solve point inside the flat 1000-3000 tier is
  arbitrary and moves the fixed component by nearly half.
- "Nobody in Thailand plates" reduced to "we have found no evidence",
  with a desk check named as the action.
- Cut the USD 450-700k valuation of a competitor's presses entirely.
  It extrapolated a semi-auto quote onto a different machine, ignored
  the used market, and supported no decision.
- The collaboration payback divided capex by revenue where the same
  page uses contribution elsewhere: 5-10 years, not 1-3.

Missing caveats added
- No financing is committed or modelled anywhere in this repo;
  project-timeline previously asserted money was not a constraint.
- Leasing cannot pull the launch forward on its own, since the press
  chain governs and construction has float against it.
- The whole competitive analysis is one buyer, one plant, one
  afternoon, and the primary evidence is not archived in the repo.
- ResurRec is a competitor and may decline to buy from a new entrant.

docs/finance/phase1-construction-rfq.md

@@ -126,7 +126,7 @@ PIR.
 | **Paving total** | **4,387 m²** — see note below |
 | Car/scooter parking | Roofed canopies: **27.7 × 7.0 m (195 m², 10 cars)** + **13.7 × 5.5 m (76 m², 30 scooters)**, steel canopy carrying **~40 kWp** PV |
 | Guard house | **12 m² (4 × 3 m)**, insulated/AC'd, with toilet; sited at the gate, north-west side |
-| Tank foundations | Ring foundation for a **9 m dia × 4 m** fire water tank (~200 m³, ~250 t full) sited in the **north-west corner of the plot**, with the **fire pump house adjacent to it**; plus a **5 m dia × 3 m** raw water tank. Tanks themselves are vendor-supplied bolted panels — foundations only |
+| Tank foundations | Ring foundation for a **9 m dia × 4 m** fire water tank (**~250 m³, ~250 t full** — the geometry gives 254 m³ gross; if 200 m³ usable with freeboard is intended, the foundation still carries the full 250 t) sited in the **north-west corner of the plot**, with the **fire pump house adjacent to it**; plus a **5 m dia × 3 m** raw water tank. Tanks themselves are vendor-supplied bolted panels — foundations only |
 | Fire main | Ring main runs from the north-west corner to the buildings — quote the **length**, not just "a ring main" |
 | Drainage & sanitation | Site stormwater drainage, domestic septic/treatment for ~25 staff — **see note below** |
 
@@ -142,8 +142,12 @@ things follow:
   which will be dug up again for Phase 2 piling and foundations. **Option to
   price:** compacted sub-base only now, surfaced after Phase 2 completes.
 
-**On drainage.** 4,387 m² of pavement seals about **46 % of the 9,567 m²
+**On drainage.** 4,387 m² of pavement is about 46 % of the 9,567 m² plot,
-plot**. Runoff increases materially over the agricultural baseline, and
+but the figure that matters for attenuation is the **total sealed area**:
+adding the main building (1,500 m²), utility yard (178 m²), guard house
+(12 m²) and parking canopies (271 m²) gives **6,348 m², or 66 % of the
+plot**. Quote the contractor that number, not the paving one.
+Runoff increases materially over the agricultural baseline, and
 stormwater attenuation — quite possibly a detention pond — may be required
 for the factory permit. Ask the contractor to state what they have assumed.
 

docs/finance/revenue-model.md

@@ -20,14 +20,24 @@ market range from multiple engineers.
 | Item | THB | USD | Note |
 |---|---|---|---|
 | Audio mastering | 5,000 – 20,000 **per song** | 150 – 605 | Varies by engineer |
-| Father / mother / silvering (stamper set) | ~22,000 **per title** | ~665 | Derived from the full-service vs pressing-only differential |
+| Father / mother / silvering (stamper set) | ≤22,000 **per title** | ≤665 | **Residual** between the full-service and pressing-only invoices — see below |
-| All-in per 12" record | 650 – 950 | 19.70 – 28.80 | Includes mastering, stamper, packaging and pressing |
+| All-in per 12" record | 650 – 950 | 19.70 – 28.80 | Believed to cover pressing and stamper; whether mastering and jacket printing sit fully inside it is unresolved |
 
-**These are domestic Thai prices with no freight or import duty in them**,
+**The 22,000 THB is a residual, not a price.** It is the difference
-which makes them directly comparable to what CNX would charge. They are
+between two invoices, and that difference carries everything the two
-one customer's experience with one plant, so treat them as a strong signal
+engagements did not share — lacquer cutting, test pressings, setup and
-rather than a survey. **The 650–950 THB figure is the price at 300 units**;
+changeover, project management, and margin on the bundle — not only
-the full volume ladder is in point 3.
+metalwork. It is therefore an **upper bound** on the stamper component.
+Every figure derived from it below inherits that, and the fix is cheap:
+ask for a line-itemised full-service invoice.
+
+These are domestic Thai prices with no separately-billed freight or
+import duty, though the stamper component carries inbound freight inside
+it. They are **one buyer's retail experience at one plant**, on a scope
+that includes a printed jacket — indicative of the Thai price level, not
+directly substitutable for what CNX would quote a label placing 2,000
+units. **The 650–950 THB figure is the price at 300 units**; the full
+volume ladder is in point 3.
 
 ## What the ResurRec factory tour showed (August 2026)
 
@@ -40,23 +50,29 @@ picture more than the invoices do:
   temperature, humidity or PM control.
 - **Pressing only. They do not plate.** The stamper set behind the
   22,000 THB differential is bought in and resold, not made in-house.
-- **Two presses, both Newbilt** (confirmed August 2026): a **Classic 12"
+- **Two presses, both Newbilt**, identified on the tour as a **Classic
-  fully automatic** — 3-column hydraulic, built-in 15 kW extruder,
+  12" fully automatic** plus one **semi-automatic**. This is a real
-  automatic patty and label handling, Siemens PLC, **28 s cycle** — plus
+  capital fleet, not a hobby setup. The Classic 12" specification —
-  one **semi-automatic** 4-column book-mould press. This is a real capital
+  3-column hydraulic, built-in 15 kW extruder, automatic patty and label
-  fleet, not a hobby setup; on the EUR 200,500 quoted for CNX's own
+  handling, Siemens PLC, 28 s cycle — is from Newbilt's datasheet, not
-  Single 2.1, the pair is plausibly USD 450–700k of press equipment.
+  from anything observed running. No nameplate was photographed and the
+  age or provenance of either machine is unknown.
 
 Four things follow, and one of them corrects the read in the previous
 section.
 
-**Nobody in Thailand appears to plate.** If the country's established
+**We have found no evidence of plating capability in Thailand.** The one
-plant buys its stampers in, then every Thai title currently makes an
+plant we have seen does not plate, and buys its stampers in. That is one
-overseas round trip for silvering and electroforming. That makes in-house
+observation, not a survey: it does not establish that no Thai title is
-galvanics a genuine first rather than a cost saving, and it means the
+plated domestically, that ResurRec is the only Thai plant, or that no
-three-to-five-week turnaround advantage in
+electroforming capacity exists outside the record industry. **Run a desk
-[production-timeline.md](../production-timeline.md) is measured against a
+check of Thai electroforming and metal-finishing capacity before "a
-competitor who cannot close that gap without building the same line.
+genuine first" goes in front of an investor.** If it holds, it would make
+in-house galvanics a capability rather than a cost saving, and the
+turnaround advantage in
+[production-timeline.md](../production-timeline.md) would be measured
+against a competitor who cannot close the gap without building the same
+line.
 
 **The 22,000 THB is a resale price, not a plating cost.** It embeds
 ResurRec's margin and inbound freight, so the underlying plating cost is
@@ -65,26 +81,32 @@ buyer pays for a stamper set today, which is exactly the right basis for
 what CNX could charge or avoid paying. But it should not be read as the
 cost of running a galvanics line.
 
-**They out-press CNX's launch phase by two to three times.** The Classic
+**On nameplate they out-class CNX's launch fleet by two to three times.**
-12" runs a 28 s cycle — **129 records/hour** from the vendor's own figure,
+Newbilt specifies a 28 s cycle for the Classic 12", which is ~128
-not an estimate — against 60–80/hour for the semi-automatic. The pair
+records/hour with no stoppages — a vendor figure for a machine in
-nameplates at **~190–210/hour**. On one shift, 250–300 days, at the 60–70%
+isolation, not a floor rate anyone measured — against 60–80/hour for the
-overall effectiveness this repo applies everywhere, that is **~230,000 to
+semi-automatic. The pair nameplates at **~190–210/hour**.
-350,000 records a year**, and roughly double on two shifts. At Newbilt's
-claimed ≥95% uptime and yield it would be higher still.
 
-| | Records/yr |
+Everything past that point is assumption. Their shift pattern, working
+days, operator count, order book and maintenance regime were all
+unobserved; the table below applies one 8-hour shift over 300 days at the
+60–70% effectiveness this repo uses for its *own planned* fleet, which
+could be wrong in either direction. Read it as installed capability, not
+as output.
+
+| | Nameplate × assumed OEE, records/yr — **not observed output** |
 |---|---|
-| ResurRec today, one shift | ~230k – 350k |
+| ResurRec, assuming one shift over 300 days | ~275k – 355k |
 | **CNX launch** (1 semi-auto, 1 shift) | **90k – 135k** |
-| CNX Duplex (2 semi-auto, 1–2 shifts) | ~300k – 600k |
+| CNX Duplex (2 semi-auto, 1–2 shifts) | ~175k – 540k |
 | CNX mature (2 semi + 2 automated, 2 shifts) | 1.0M – 1.5M |
 
-CNX does not overtake them on volume until the Duplex expansion, and does
+All rows are on the same 300-day basis so they can be compared. CNX does
-not clearly lead until the mature fleet. Nameplate is not sold output —
+not overtake them on installed capability until the Duplex expansion, and
-their actual utilisation is unknown, and an unconditioned floor will not
+does not clearly lead until the mature fleet. The fleet is real and
-hold top-end effectiveness through the wet season — but the fleet is
+should not be discounted — but nothing here is a measurement of what they
-real and should not be discounted.
+actually ship, and an unconditioned floor may not hold top-end
+effectiveness through the wet season.
 
 What remains true is the **quality and capability gap**, and it is where
 the whole differentiation now sits: no plating, no climate control, no
@@ -103,24 +125,52 @@ A plant that can press 230,000–350,000 records a year:
 - and needs **two to four extra weeks** for anything over 3,000.
 
 None of that is press capacity — 3,000 units is roughly three days of
-pressing on that fleet. Every one of those behaviours is explained by the
+pressing on that fleet. The most economical single explanation is the
-**stamper being a purchased import**: it is the irreducible fixed cost that
+**stamper being a purchased import**: it is the only cost that is both
-makes small runs uneconomic, the cost that does not fall with volume so
+large and volume-invariant, which would make small runs uneconomic, keep
-the discount ladder stays flat, and — at 1,000–2,000 pressings per stamper
+the discount ladder flat, and — at 1,000–2,000 pressings per stamper —
-— the reason a 3,000-unit run needs two or three sets and therefore
+mean a 3,000-unit run needs two or three sets and therefore another
-another overseas order.
+overseas order.
 
-**They have the presses. They cannot feed them.** That single sentence
+**But that is a hypothesis, and it is not the only one that fits.** An
-explains the entire observed price list, and it is the market position
+apartment-sized plant is constrained in several ways at once, and any of
-CNX's galvanics line is aimed at.
+these would shape a price list the same way:
 
-It also reframes the collaboration. If the constraint is metalwork rather
+- **Printed jackets.** Their price carries one, and a jacket print run
-than machines, a CNX galvanics line could sell stampers **into their
+  takes 10–20 working days from approved artwork. A 3,000-unit jacket
-existing spare press capacity** — revenue against the galvanics capex
+  order is a bigger, slower print job — and this constraint **CNX would
-before CNX presses anything itself. At 300 titles a year of their work
+  share**, which would remove the competitive read entirely.
-that is ~6.6M THB (~200k USD); if faster stamper supply let them grow into
+- **Floor space.** Storing, finishing and shipping 3,000 records is an
-their fleet, more. Against a 320–650k USD line that is a one-to-three-year
+  area problem before it is a metal problem.
-payback **from a single customer**.
+- **Finishing labour**, which this repo separately calls labour-bound.
+- **Lacquer cutting**, presumably also outsourced — a second overseas
+  dependency, distinct from plating.
+- Deposit and cash-flow terms on a larger PVC and jacket buy.
+
+The tour did not distinguish between these, and the 1,000–2,000
+pressings-per-stamper figure the inference leans on is itself unsourced
+in this repo. The stamper hypothesis is the one CNX can act on, which is
+a reason to test it, not a reason to believe it. It is also a single
+point of failure: the collaboration case, the "pull galvanics forward"
+argument and the "their weakness at volume is structural" claim all rest
+on it together.
+
+It also reframes the collaboration, though more modestly than it first
+appears. If the constraint is metalwork rather than machines, a CNX
+galvanics line could sell stampers **into their existing spare press
+capacity** — revenue against the galvanics capex before CNX presses
+anything itself.
+
+Sizing it honestly: 300 titles a year at their retail 22,000 THB would be
+~6.6M THB (~200k USD) of revenue, but a competitor buys at wholesale, and
+revenue is not contribution. On this page's own 25–35% marginal-cost
+assumption, and at a wholesale price well below retail, 300 titles might
+net **60–100k USD of contribution — a five-to-ten-year payback** on the
+line from this customer alone, before galvanics opex, the building, the
+clean room, ZLD or depreciation. It also assumes they would move their
+entire stamper supply to a new entrant who is also their rival.
+
+Treat it as a useful offset against the capex, not as a business case.
 
 Which raises a sequencing question the roadmap does not currently ask:
 **should galvanics come first rather than fifth?** It is a fraction of the
@@ -147,15 +197,26 @@ argument instead:
 - Airborne particulate embeds in the record surface. That is the classic
   source of clicks and pops, and it is the complaint that gets a pressing
   rejected by a label.
-- Thailand's burning season makes this seasonal as well as chronic — a
+- Thailand's burning season makes this seasonal as well as chronic, so a
-  plant without filtration produces measurably different records in March
+  plant without filtration would be expected to produce different records
-  than in September.
+  in March than in September.
 
-**Newbilt is now vendor-validated in Thailand, and that is worth money.**
+These are the standard process arguments for climate control, not
-The incumbent runs Newbilt kit, which means the machines demonstrably
+measurements anyone has taken here — no records were tested. And the
-survive Thai humidity and Thai power, and that somebody has already walked
+obvious counter belongs next to them: **the incumbent sells commercially
-the import, commissioning and spares path CNX is about to walk. Two things
+from that room today**, which bounds how much the Thai market currently
-to ask on the back of it, both free: ask **Newbilt** for their Thai
+penalises the absence of climate control. The quality case is real, but
+it is a bet that buyers will pay for a difference they are not currently
+demanding.
+
+**There is a Newbilt install base in Thailand, which is worth a phone
+call.** Seeing two Newbilt presses in a room establishes that they exist
+there — not their age, provenance, downtime, spares availability or
+service history, and not that they were bought new from Newbilt rather
+than used or through an agent. If they were bought new, somebody has
+already walked the import, commissioning and spares path CNX is about to
+walk. Two things to ask on the back of it, both free: ask **Newbilt** for
+their Thai
 install base and a service-response precedent — the northern-dispatch SLA
 question in [equipment-suppliers.md](../suppliers/equipment-suppliers.md)
 is much easier to answer with an in-country reference — and, if the
@@ -172,26 +233,43 @@ currently shipping to Europe, the US or Japan.
 
 ### What the invoices say about the model
 
-**1. The galvanics saving is bigger than modelled.** The model credits
+**1. The galvanics saving may be bigger than modelled, but it is being
-in-house galvanics with removing 175–460 USD per title of external stamper
+measured against the wrong supplier.** The model credits in-house
-cost. The observed differential is **~665 USD** — 1.4 to 3.8× higher, and
+galvanics with removing 175–460 USD per title of external stamper cost.
-it is a resold price rather than a plating cost, per the tour notes above.
+The ResurRec residual is **≤665 USD**.
-On a galvanics line at 320k–650k USD, and assuming in-house marginal cost
+
-(nickel, silver nitrate, chemistry, labour, QC, water treatment) runs
+Two problems with treating that as the saving. First, **CNX's launch plan
-25–35% of the external price, the net saving is roughly 430–500 USD per
+does not buy stampers in Thailand** — it buys from Stamper Discs in
-title:
+Sheffield, so CNX's avoided cost is the Stamper Discs price, which is
+nowhere in this repo. Second, neither 175–460 nor 665 has a source: the
+first is an industry-norm estimate, the second is a residual carrying
+cutting and setup. Until a Stamper Discs quote exists, the honest
+statement is that the saving lies **somewhere between 175 and 665 USD per
+title**, and getting that quote is the cheap way to narrow it.
+
+The table below takes the optimistic end. On a galvanics line at
+320k–650k USD, and assuming in-house marginal cost (nickel, silver
+nitrate, chemistry, labour, QC, water treatment) runs 25–35% of the
+external price — an assumption with no basis stated — the net saving
+would be 430–500 USD per title:
 
 | Titles/yr | Annual saving | Payback on 320k | on 650k |
 |---|---|---|---|
 | 300 | ~130–150k USD | ~2.1 yr | ~4.3 yr |
 | 600 | ~260–300k USD | ~1.1 yr | ~2.2 yr |
 
-A one-to-four-year payback is a materially better case than the model
+Every cell above divides capex by the **top** of the saving range. At the
-implied, and it is driven by **title count, not record volume** — which
+bottom the same table gives 2.5 and 5.0 years at 300 titles, 1.2 and 2.5
-is an argument for pulling galvanics earlier than its current Phase 5
+at 600 — so the honest span is **one to five years**, and the adverse
-slot in [roadmap.md](../roadmap.md).
+corner stretches past six once the ±30–50% band that
+[total-investment.md](total-investment.md) puts on the galvanics capex is
+applied. It is nonetheless driven by **title count, not record volume**,
+which is a real argument for pulling galvanics earlier than its current
+Phase 5 slot in [roadmap.md](../roadmap.md) — on a benchmark that still
+needs the Stamper Discs quote behind it.
 
-**2. Per-title services are understated by roughly 3 to 4×.** The model
+**2. Per-title services may be understated by up to 3–4×, but the
+evidence is internally inconsistent.** The model
 bundles mastering, cutting, stampers and test pressings into 19,000–57,000
 THB per title. At observed market rates a ten-track album masters at
 50,000–200,000 THB *before* the 22,000 THB stamper set — so the real
@@ -202,9 +280,15 @@ against the 0.17M–0.69M in the table below. Two things cap it, and both
 are real:
 
 - **Mastering is labour-bound.** An engineer masters perhaps 2–4 songs a
-  day, so 600–1,200 songs or 60–120 albums per engineer per year. The
+  day — 500–1,000 songs, or 50–100 ten-track albums, over a 250-day year.
-  300–600 titles/yr in this model would need **3 to 5 mastering
+  The 300–600 titles/yr in this model would therefore need anywhere from
-  engineers**; [scale.md](../scale.md) currently plans for 1 to 2.
+  **3 to 12 mastering engineers** depending which corner you take;
+  [scale.md](../scale.md) currently plans for 1 to 2. Even the friendly
+  corner is short.
+- The upper bound above rests on a ten-track album at the 20,000 THB/song
+  top rate. That is a figure this page doubts further down, so the 4.0M
+  USD end of the range should be treated as an artefact rather than a
+  forecast.
 - CNX would not necessarily capture full boutique retail rates as a plant
   rather than a studio.
 
@@ -221,32 +305,57 @@ for **300 units**. The published discounts:
 | | | *+2 weeks to 1 month lead time* | |
 
 **The curve is remarkably flat.** A tenfold increase in run size buys only
-30% off. Solving the first three tiers as a fixed-plus-variable structure
+30% off. Whatever the exact split, that shape means a large per-title
-(the fit is good to ~2%) gives:
+fixed cost that does not fall with volume — and it is the shape, not any
+solved number, that is informative.
 
-- **Per-title fixed: ~34,000 – 50,000 THB** (1,030 – 1,515 USD)
+Solving the first three tiers as a fixed-plus-variable structure gives
-- **Per-unit variable: ~535 – 780 THB** (16.20 – 23.65 USD)
+roughly 34,000–50,000 THB fixed and 535–780 THB variable, **but that
+answer should not be leaned on**. Three reasons:
 
-Two readings fall out of that, and both matter.
+- The 1,000–3,000 tier is a flat band, so the second solve point is an
+  arbitrary choice inside it. Taking 1,000 instead of 3,000 moves the
+  fixed component from ~32,500 to ~41,800 THB on the low ladder alone.
+- Two parameters fitted to three points has one degree of freedom, so a
+  small residual is guaranteed by construction and validates nothing.
+- These are not three observations. They are published discount
+  percentages off a single reference price; only the 300-unit tier is a
+  transaction anyone has seen. Fitting a line to a discount schedule
+  recovers the schedule, not the supplier's cost structure.
 
-**The variable price is four to seven times the model's assumption.** This
+Two readings still fall out of the shape, and both matter.
+
+**The variable price is roughly three-and-a-half to seven times the
+model's assumption.** This
 model prices a record at 115–150 THB (3.50–4.60 USD) blended. The Thai
 domestic market pays **535–780 THB** for the same per-unit scope, and even
 at the deepest >3,000 discount never drops below ~455 THB. Some of that
-gap is scope — their figure carries a printed jacket where this model
+gap is scope: their figure carries a printed jacket where this model
-passes packaging through separately — but nothing like all of it. Either
+passes packaging through separately. But a jacket, inner sleeve and
-the pricing assumption here is far too low for this market, or ResurRec is
+shrink at retail is on the order of 100–150 THB, which does not close a
-pricing at the import-parity ceiling that a sole domestic supplier can
+gap of roughly 400 THB.
-hold. Probably both, and a second plant would compress it. **It still says
-there is far more headroom in domestic pricing than this model assumes,
-and that is the most consequential finding in the data.**
 
-**The fixed figure does not reconcile, and that is informative.** At
+Two readings remain, and they **pull in opposite directions**:
-34,000–50,000 THB it barely covers the 22,000 THB stamper set plus one to
+
-three songs of mastering. So the 650–950 THB either does not fully carry
+- The pricing assumption here is too low for this market, and there is
-album mastering, or albums are mastered far below the 5,000–20,000 THB
+  genuine willingness to pay.
-per-song headline. Worth resolving — it is the difference between point 2
+- ResurRec is holding an import-parity ceiling that only a sole domestic
-above being a 3–4× understatement or a much smaller one.
+  supplier can hold — in which case **CNX's own entry is what compresses
+  it**, and the headroom evaporates on arrival.
+
+Nothing in one supplier's price list separates those. It suggests more
+headroom in domestic pricing than this model assumes, from a single
+published list in a market with one supplier. It is the finding most
+worth testing next, not the finding most worth relying on.
+
+**The fixed component looks small against the stated inputs.** At
+34,000–50,000 THB it barely covers a ≤22,000 THB stamper set plus one to
+three songs of mastering — though at the other end of the solve range
+(~41,800 THB low ladder) the tension eases, so this may be an artefact of
+the endpoint chosen rather than a real gap. Either the 650–950 THB does
+not fully carry album mastering, or albums are mastered far below the
+5,000–20,000 THB per-song headline. Worth resolving: it is the difference
+between point 2 above being a 3–4× understatement or a much smaller one.
 
 **The >3,000 lead-time penalty is probably a stamper problem, not a press
 problem.** At nameplate the two presses clear 3,000 units in about two
@@ -257,13 +366,14 @@ stamper sets** — and with no plating in-house, every additional set is an
 overseas order. "Blocking out the entire factory" reads like scheduling
 around an imported-metalwork wait.
 
-If that inference is right it is the sharpest competitive fact yet found:
+If that inference holds it would be the sharpest competitive finding yet:
-**the incumbent's lead time degrades with run size precisely because they
+the incumbent's lead time degrading with run size *because* they cannot
-cannot plate**, while a plant with in-house galvanics pulls a fresh
+plate, while a plant with in-house galvanics pulls a fresh stamper off an
-stamper off an archived mother in a day or two (see the repress path in
+archived mother in a day or two (see the repress path in
-[production-timeline.md](../production-timeline.md)). Their weakness at
+[production-timeline.md](../production-timeline.md)). But see the
-volume is structural and is the exact thing the Phase 5 galvanics line
+alternatives listed earlier — jacket print runs in particular would
-fixes.
+produce the same symptom and would constrain CNX equally. Confirm the
+cause before treating their weakness at volume as structural.
 
 **And the floor of the market is unserved.** A 150-unit minimum that is
 "rarely accepted because the cost is way too high" means sub-300 runs are
@@ -284,11 +394,26 @@ that CNX would not have.
 
 ### What is still missing
 
-- Is the 22,000 THB stamper set for a complete 12" (both sides), or per
+- **What else sits in the 22,000 THB differential besides metalwork?**
-  stamper? This drives the two-to-three-sets inference above.
+  Ask for a line-itemised full-service invoice. This is the highest-value
+  question on the list: the residual is the base of the galvanics case.
+- Is the stamper set for a complete 12" (both sides), or per stamper?
+  This drives the two-to-three-sets inference above.
 - Does the >3,000 lead-time extension come from stamper supply, press
-  scheduling, or PVC procurement? Worth asking directly — it decides
+  scheduling, jacket printing, floor space or PVC procurement? It decides
-  whether their constraint is one CNX escapes.
+  whether their constraint is one CNX escapes or one CNX shares.
+- What is ResurRec's actual annual output and title count? The
+  collaboration sizing is guesswork without it.
+- What does the Stamper Discs quote say? It, not the Thai resale price,
+  is CNX's real avoided cost at launch.
+
+Two standing caveats on this whole section. **The evidence is one buyer,
+one plant and one afternoon** — the invoices and tour notes are not
+archived anywhere in this repo, only summarised here, and for an
+investor-facing document they should be. And **ResurRec is a competitor**:
+every collaboration scenario above assumes they would buy a critical
+input from a new entrant, which they may simply decline to do, and which
+would also accelerate a rival.
 - Does 650–950 THB include jacket printing, or is packaging billed
   separately?
 - Is the 5,000–20,000 THB/song mastering rate for general audio mastering,
@@ -362,7 +487,7 @@ boutique-heavy mix pushes the title line up and the volume down.)
 | Stamper amortization | 0.03 to 0.07 | 1.1 to 2.3 | in-house stamper over 1,000 to 2,000 pressings |
 | Maintenance, spares, consumables | 0.12 to 0.23 | 3.8 to 7.6 | estimate |
 | Reject allowance | 0.12 to 0.17 | 3.8 to 5.7 | 5 to 10% reject rate carried by good units |
-| **Total direct cost** | **~1.45 to 2.75** | **~48 to 91** | |
+| **Total direct cost** | **~1.45 to 2.77** | **~48 to 92** | |
 
 ## Margin per record
 
@@ -411,10 +536,11 @@ proves the market before the big blocks are bought.
   power there is unproven and is the single biggest assumption in this
   model. The turnaround claim is quantified in
   [production-timeline.md](../production-timeline.md)
-- **In-house galvanics** removes **~665 USD** per title of external
+- **In-house galvanics** removes **somewhere between 175 and 665 USD** per
-  stamper cost — the observed ResurRec differential, up from the 175–460
+  title of external stamper cost — the low end an industry-norm estimate,
-  USD previously estimated — plus three to five weeks of turnaround,
+  the high end an upper-bound residual from a Thai reseller who is not
-  feeding both margin and the sales story
+  CNX's actual supplier — plus 1.5 to 5 weeks of turnaround. **Get the
+  Stamper Discs quote before this number is used for anything**
 - **Energy**: solar plus heat recovery pushes the energy line toward the
   low end and hedges Thai tariff rises
 - **The currency mismatch is a live exposure.** Revenue is quoted to

docs/production-timeline.md

@@ -21,23 +21,39 @@ divide by 5 for weeks, and note that Thai public holidays run to
 
 | Configuration | Plant-controlled | Including customer gates |
 |---|---|---|
-| **Launch** (bridge stampers, no in-house cutting or galvanics) | ~21–39 days | **~8 to 14 weeks** |
+| **Launch** (bridge stampers, no in-house cutting or galvanics) | 22–49 days | **~6 to 17 weeks** |
-| **Mature** (full chain in-house) | ~14–20 days | **~5 to 10 weeks** |
+| **Mature** (full chain in-house) | 11.5–22 days | **~4 to 11.5 weeks** |
-| **Repress** from an archived mother, no changes | ~7–11 days | **~2 to 3 weeks** |
+| **Repress** from an archived mother, no changes | 6.5–11.5 days | **~1.5 to 2.5 weeks** |
 
-Reported EU/US plant lead times currently run roughly **10 to 16 weeks**
+These are straight sums of the stage table below, plus 9–35 days of
-for a standard black LP, down from the 6-to-9-month peak of 2021–22. An
+customer gate time for the first two rows. They are wide because the
-Asian label using a European plant adds freight on top of that in both
+stages are wide, and they are not a quotable lead time — see "What to
-directions. So the mature configuration is a genuine regional advantage
+promise a customer" at the end, which adds queue and buffer on top.
-and the launch configuration is merely competitive — which matters,
-because the launch configuration is what the first two years actually
-sell.
 
-**Bringing cutting and galvanics in-house takes roughly 3 to 5 weeks out
+Reported EU/US plant lead times run roughly **10 to 16 weeks** for a
-of the quoted lead time.** That is the number behind the sales story,
+standard black LP, down from the 6-to-9-month peak of 2021–22. That
-and it is the commercial case for the Phase 5 galvanics line stated in
+figure is an industry impression gathered in conversation, not a
-time rather than in the ~665 USD per title of stamper cost that
+surveyed one, and it should be tested at Making Vinyl alongside the
-[revenue-model.md](finance/revenue-model.md) already counts.
+price. Two cautions on the comparison:
+
+- **It is not like-for-like.** The EU/US figure is a *quoted* lead time
+  carrying queue and buffer; the CNX figures above are unbuffered stage
+  sums. The honest comparison is against the quote table at the end of
+  this page — 7–9 weeks mature, 10–14 launch.
+- **On that basis the launch configuration is at parity, not ahead.**
+  Its 10–14 week quote sits inside the EU/US 10–16 week band, and its
+  unbuffered worst case of 17 weeks is beyond it. The regional advantage
+  comes from freight and time zone at launch, and only becomes a
+  manufacturing advantage once cutting and plating come in-house. That
+  matters, because launch is what the first two years actually sell.
+
+**Bringing cutting and galvanics in-house takes roughly 1.5 to 5 weeks
+out of the lead time** — replacing a 13–27 day outsourced metal block
+plus freight with 4–7 days of in-house cutting, plating and stamper
+prep. The spread is wide because the outsourced block is the part
+nobody has measured. That is the commercial case for the galvanics line
+stated in time rather than in the stamper cost that
+[revenue-model.md](finance/revenue-model.md) counts separately.
 
 ## Two tracks run in parallel
 
@@ -123,7 +139,7 @@ overall effectiveness already applied there.
 | Fleet | Good records/day | 300 units | 500 | 1,000 | 2,000 |
 |---|---|---|---|---|---|
 | Launch: 1 semi-auto, 1 shift | ~300–450 | <1 day | 1–2 | 2–3 | 5–7 |
-| Duplex: 2 semi-auto, 2 shifts | ~1,000–1,500 | <1 | <1 | 1 | 1–2 |
+| Duplex: 2 semi-auto, 2 shifts | ~1,150–1,790 | <1 | <1 | 1 | 1–2 |
 | Mature: 2 semi + 2 auto, 2 shifts | ~3,300–5,000 | <1 | <1 | <1 | <1 |
 
 Read these as **press-days, not calendar days**. At mature scale no
@@ -138,9 +154,13 @@ decision.** A plant running at 90% quotes 16 weeks with the same
 equipment that quotes 6 weeks at 50%. Any turnaround promise made to
 the market is implicitly a promise to hold spare capacity.
 
-Small runs invert the arithmetic. At 300 units the run itself is under
+Small runs invert the arithmetic — though not as sharply as it first
-an hour of pressing and the entire timeline is setup, approvals and
+looks. At 300 units on the semi-automatic line the run is **5 to 8
-assembly — which is exactly the boutique/color work the semi-automatic
+hours**, most of a shift, once the 60–70% effectiveness above is
+applied; it is roughly an hour on the full mature fleet. What dominates
+a small order is not the pressing but the setup, the approvals and the
+assembly, all of which cost the same at 300 units as at 3,000 — which
+is exactly the boutique/color work the semi-automatic
 line exists to serve, per
 [pressing-line.md](equipment/pressing-line.md). Small runs are not
 faster to deliver; they are only cheaper to press.

docs/project-timeline.md

@@ -13,11 +13,18 @@ governs the whole schedule.
 
 ## The answer
 
-**First paying release lands around Q4 2028 to Q1 2029 — roughly 27 to
+**First paying release lands around Q4 2028 to mid-2029 — roughly 27 to
-30 months from today**, on the buy-and-build path with a BOI-funded
+33 months from today**, on the buy-and-build path. The top of that range
-machinery order.
+is the honest planning figure while the press build time is unquoted; the
+Gantt below draws the optimistic end.
 
-The critical path is not construction and not money. It is:
+The critical path is not construction. Whether it is **money** is outside
+this page's scope, and that gap should be stated plainly: no financing is
+committed and none is modelled anywhere in this repo, which prices
+1.7–2.9M USD for the launch phase and 4.7–8.9M for the full vision
+without saying where any of it comes from. This calendar assumes capital
+is available when each block is ordered. The critical path it does model
+is:
 
 > **BOI certificate → press order → press build → sea freight →
 > commissioning**
@@ -38,14 +45,14 @@ gantt
     section Gates (Phase 0)
     Market validation / Making Vinyl   :a1, 2026-08-01, 120d
     BOI consultant engaged             :a2, 2026-08-01, 60d
-    BOI application prepared           :a3, after a2, 90d
+    BOI application prepared           :a3, after a2, 75d
     60 rai title check                 :a4, 2026-08-01, 90d
 
     section BOI
     Application filed                  :milestone, m1, 2026-12-15, 0d
     BOI review and analyst meeting     :b1, 2026-12-15, 150d
     Promotion certificate              :milestone, m2, 2027-05-15, 0d
-    Company incorporation              :b2, after m2, 45d
+    Company incorporation              :b2, after m2, 30d
 
     section Land and building
     Zoning and title verification      :c1, 2026-09-01, 90d
@@ -92,8 +99,9 @@ for a built-to-order press system from a small German manufacturer, and
 it is carrying about **40% of the entire schedule**.
 
 If the real figure is 15–18 months — entirely plausible for a shop of
-Newbilt's size with a full order book — first production moves to late
+Newbilt's size with a full order book — first production moves to
-2029 and the whole plan shifts by half a year.
+mid-2029 and the whole plan shifts by six months, or nine at the widest
+corner.
 
 **Action, and it is free:** [roadmap.md](roadmap.md) item 4 already
 calls for a Newbilt re-quote, since the 2024 offer has expired. Add
@@ -177,16 +185,20 @@ only money:
 |---|---|---|
 | Upfront capital | ~0 | 1.0–2.1M USD (34–69M THB) |
 | On the critical path? | No | Yes, with ~3–5 months float |
-| Earliest first production | **~Q1–Q2 2028** | ~Q4 2028–Q1 2029 |
+| Earliest first production | ~Q4 2028, same as build | ~Q4 2028–mid 2029 |
 | Slab rating, 3 t/m² | Must be verified in an existing building; most Thai light-industrial stock will not have it | Specified from the start |
 | 500–1,000 kVA supply | Existing connection, or a PEA upgrade with its own lead time | Designed in |
 | BOI land ownership privilege | Unused | Used |
 | Galvanics effluent, Phase 5 | Landlord's permission needed, and may not be given | Ours to permit |
 
-Leasing pulls the launch forward by roughly **three quarters** and frees
+Leasing frees a seven-figure sum but, on this page's own critical path,
-a seven-figure sum, at the cost of a building that probably compromises
+**cannot pull the launch forward on its own** — the press chain governs,
-on the two hardest requirements (floor loading and power) and may block
+and construction has only 3–5 months of float against it. Removing an
-the Phase 5 galvanics line entirely. Locating in the Northern Region
+item that is not on the critical path does not move the end date. Leasing
+only advances the launch if the **press order** also moves forward, which
+is the BOI question above; the two are one lever, not two. The cost is a
+building that probably compromises on the two hardest requirements (floor
+loading and power) and may block the Phase 5 galvanics line entirely. Locating in the Northern Region
 Industrial Estate, which [boi.md](boi.md) already favours for its shared
 central effluent treatment, is the version of leasing that keeps the
 galvanics option alive.

docs/scale.md

@@ -10,7 +10,7 @@ page as real quotes and the site decision land.
 | Phase | Fleet | Rough output |
 |---|---|---|
 | Launch | 1 semi-auto press (Newbilt Single), 1 shift | ~90,000 to 135,000 records/yr |
-| Duplex | 2 semi-auto presses, 1 to 2 shifts | ~300,000 to 600,000 records/yr |
+| Duplex | 2 semi-auto presses, 1 to 2 shifts | ~175,000 to 540,000 records/yr |
 | Mature | 2 semi-auto + 2 automated (50/50 fleet), 2 shifts | ~1.0 to 1.5 million records/yr |
 
 Basis: the automated AD12 presses ~140 records/hour each (vendor datasheet);
@@ -20,16 +20,27 @@ realistic overall effectiveness of 60 to 70% applied. The theoretical
 ceiling of the mature fleet at 100% uptime is roughly 1.9 to 2.1 million
 records per year.
 
-**The 60–70% effectiveness assumption is worth challenging.** Newbilt
+**The 60–70% effectiveness assumption is worth challenging, but not yet
-publishes **≥95% yield and ≥95% uptime** for the Classic 12" automatic at
+changing.** Newbilt publishes ≥95% yield and ≥95% uptime for the Classic
-a 28 s cycle (~128 records/h), and quotes ~30 s average for the Duplex.
+12" automatic at a 28 s cycle (~128 records/h), and quotes ~30 s average
-Even discounting the vendor's optimism and multiplying availability by
+for the Duplex. Three reasons not to adjust anything on that basis:
-yield by a realistic performance factor, that implies an OEE nearer 80%
+
-than 65% — which would make the capacity table here conservative by
+- **Machine uptime is not plant OEE.** A vendor uptime figure excludes
-roughly a fifth to a third. Ask both press vendors for the **basis** of
+  stamper and mould changes, label reels, colour changeovers, PVC lot
-their uptime and yield figures (measured over what period, at what run
+  changes, first-article QC and breaks — precisely the losses that
-length, counting which stoppages) as part of the re-quote. It is a free
+  dominate a short-run boutique plant. They are different quantities.
-question and it moves the whole capacity model.
+- It would apply only to the **automated half** of the fleet. The
+  60–80/h semi-auto figure already embeds operator handling, so applying
+  an uplift to that half would double-count.
+- 0.95 × 0.95 is 0.90 before any performance factor; reaching 80% OEE
+  needs a performance factor near 0.89, which nobody has estimated.
+
+Ask both press vendors for the **basis** of their uptime and yield
+figures — measured over what period, at what run length, counting which
+stoppages — as part of the re-quote. It is a free question. If the
+figures survive scrutiny they would argue for a higher OEE on the
+automated fleet only, and the same logic would apply to any competitor
+running the same machine.
 
 ## Building
 
@@ -71,8 +82,10 @@ production, so either the main building shrinks by ~126 m² or the program
 grows into it. Decide which before the building is priced.
 
 Staff parking is roofed, in the front yard clear of the loading route:
-**10 car bays** (25 × 6 m canopy) and **~30 scooter spaces** (14 × 5 m),
+**10 car bays** (27.7 × 7.0 m canopy, 195 m²) and **~30 scooter spaces**
-carrying a further **~32 kWp** of PV on their canopies. At 4.1 m and 3.5 m
+(13.7 × 5.5 m, 76 m²), carrying a further **~40 kWp** of PV on their
+canopies — dimensions per the modelled layout in the
+[construction RFQ](finance/phase1-construction-rfq.md). At 4.1 m and 3.5 m
 they sit below the main building's 7.0 m low eave, so they cannot shade it.
 
 ### Expansion headroom: a second building
@@ -147,15 +160,17 @@ Grid connection: a medium-voltage connection with a transformer in the
 
 ## Solar and storage
 
-- Roof of a ~1,500 m² building fits roughly **150 to 200 kWp**; reaching the
+- The phase 1 roof carries **~247 kWp** (425 modules on the 1,523 m²
-  illustrative 300 kWp in the utilities CSV needs carport or ground-mount
+  roof plane, per the [construction RFQ](finance/phase1-construction-rfq.md)).
-  on the plot (about 1 rai of ground per additional 100 kWp, estimate)
+  This supersedes the older rule of thumb of 150 to 200 kWp on a
+  ~1,500 m² building, which the massing model has overtaken
 - Northern Thailand yield ~1,400 to 1,500 kWh/kWp/yr, so 300 kWp produces
   roughly **420 to 450 MWh/yr**
 - Plant consumption at mature two-shift operation: roughly **1,000 to
   1,500 MWh/yr** without electric steam, up to **~2,500 MWh/yr** with it
-  (estimate), so 300 kWp covers on the order of **20 to 35%** of
+  (estimate), so 300 kWp covers roughly **28 to 45%** of consumption
-  consumption; a bigger array shifts that up
+  without electric steam, falling to **~17%** with it; a bigger array
+  shifts that up
 - Battery: **271 kWh** Gobel stack (peak shaving and ride-through, roughly
   two to three hours of the chiller's electrical draw)