A carving specified well and a carving specified badly can use the same stone, the same workshop and the same hours, and look nothing alike in the finished room. The difference is almost never the material. It is the sequence in which the decisions get made — and the two variables that most schemes get wrong are light and weight, in that order.

Decide the job before the object

The most common specifying error is starting from a piece rather than from the function it has to perform. A carving in an interior does one of three jobs, and each has a different specification:

  • A focal centre. The room is arranged around it. It is viewed from a distance, from several angles, and often from a seated position. Scale, silhouette and how the material behaves under light matter more than fine detail.
  • A supporting note. It reinforces a scheme without competing: a shelf piece, a counter piece, something on a console. Colour and tone matter more than size, and detail must read at arm's length.
  • A repeated motif. Several pieces through a run of spaces, so consistency across pieces matters more than the quality of any one. This is where colour predictability becomes the deciding factor rather than an aesthetic preference.

Write that sentence down before you look at anything. "A focal centre for the far end of a double-height lobby, viewed from the entrance" already eliminates most of the range, and it eliminates it for the right reasons.

Scale, and the mistake of judging it on a screen

A carving's apparent scale depends entirely on viewing distance. A piece that commands a side table reads as a paperweight across even a modest room, and a piece that fills a domestic alcove disappears in a double-height space. Photographs cannot tell you this, because a photograph has no distance in it.

Two working rules:

  • Specify against the viewing distance, not the piece. Think about where people will stand and sit, and how much of the view the carving has to hold.
  • Use a scale reference every time. A ruler, a hand, a person. This is why we photograph the monumental work with a person in frame — without one, the image tells you nothing about size.

Our standard range runs from 3.8 in (97 mm) to 7.1 in (180 mm), and the working capability for a single piece runs from 5 cm to 150 cm. That is a very wide span, and the reason most projects land near the middle or the very top is the next section.

The constraint that decides feasibility: weight

This is the variable that kills projects late, so it belongs near the beginning of the conversation rather than in the logistics plan.

Weight follows volume, which grows with the cube of size. Double a carving in every dimension and it becomes eight times heavier. Scale a 7.1 in (180 mm) piece to 1.5 m and the arithmetic gives roughly 2.3 tonnes — we treat that as arithmetic rather than a specification, because material availability at that size is what decides the real answer, but the order of magnitude is the point. The full derivation is in how big a single-block carving can be.

In practical terms, that means four things get checked before aesthetics:

  • Floor loading, if the piece sits on a raised floor, a mezzanine or anything not solid ground
  • The route in. A doorway, a lift, a stairwell or a tight turn can rule out a piece before anything else is decided
  • Stability and anchoring. Anything in a public space needs to be designed so it cannot be toppled, and a tall heavy carving has a centre of gravity that has to be planned rather than hoped about
  • Who moves it, and with what. Two tonnes does not get manoeuvred by two people and a trolley

If a scheme depends on a very large piece, raise it early enough that the answer can still change the design.

Light, which matters more than the stone

Polished stone is a curved mirror. It does not have a colour so much as a reflection plus a body colour, and the reflection is usually the larger part of what you see.

That single fact drives most specification errors:

  • A dark polished piece facing a window becomes a glare source. You do not see a black carving; you see a bright rectangle where the window is. If the piece must face glazing, expect to specify a matte background or reposition it.
  • A highly polished surface reads as deeper colour because the polish lets light into the material before it returns. This is why obsidian takes such a black, and why a coarse finish on the same stone looks washed out.
  • Translucent stones change character with backlighting. Clear quartz, amethyst and moss agate glow when lit from beneath or behind, and go quiet when front-lit like an opaque object.
  • Directional stones only perform from a specific axis. Tiger eye and labradorite flash from particular angles and look like grey rock from others. In a scheme with a known principal viewing direction that is an asset; in a room approached from all sides, expect it to read flat from half of them.
  • One light source beats several. Multiple lights flatten the form and cancel the surface. This is as true in a room as it is in a photograph — the reasoning is set out in photographing carved stone, where the same physics decides how a piece sells.

The specification line that follows from all of this is simple: state the principal viewing direction and the light it will receive, and the material choice becomes much easier to defend.

Matching material to scheme

The materials behave differently as interior elements, beyond colour. Our own range, and what each one does in a room:

MaterialCharacter in a schemeWatch for
ObsidianDeep, near-black, mirror polish. Reads as a strong graphic noteShows every fingerprint and speck of dust; glare facing glazing
Rose quartzSoft pink, translucent in places, warm rather than brightColour varies between pieces; value shifts with light warmth
Clear quartzNeutral, high transparency, glow when backlitInterior structure visible through the surface; least forgiving on dust
LabradoriteDark grey-green body with a blue-green flashFlash depends on angle; needs a known viewing axis
Tiger eyeGolden-brown banding with a moving sheenDirectional; can look flat from the wrong side
Tiger ironBanded gold, grey and red-brown with a metallic castDistinct from tiger eye; denser and heavier per length
Agate (crazy lace, Mexican, moss)Warm earth tones, banding and translucencyEvery piece differs — hard to repeat across a run
Red jasper, kambaba jasperOpaque and dense; a satin rather than mirror surfaceFinishes less glossy by nature; some schemes want exactly that
HowliteWhite with fine grey veining, marble-likeSofter look than quartz at a pale tone
FluoriteBanded green, purple and clear zoningA relatively soft mineral — specify for display or occasional handling rather than abrasive cleaning
Bloodstone / dragon bloodstoneDark green with red spotting; bold patterning"Dragon bloodstone" is a trade name — confirm the actual mineral

Two notes that matter more than they look. First, agate and jasper do not take the same finish as obsidian; the material sets the achievable surface, and a supplier promising a mirror finish on every stone is promising something the stone will not do. The reason is in the polishing stage — every surface defect left by earlier grinding shows up under polish, so the achievable finish is decided well before polishing begins, as how crystal skulls are made sets out. Second, trade names are not mineral names — for a commercial interior it is worth asking what the stone actually is, because that is what determines how it ages and how it is cleaned. Our position on this is set out in the material disclosure guide.

Indoor or outdoor

Outdoor specification narrows the material list before anything else is discussed, because weather, frost and ultraviolet exposure all act on the stone over time. If a piece is going outside — a landscape installation, a courtyard, an entrance — say so at the start, because it is a materially different brief rather than a placement detail.

Maintenance, which clients ask about later

Worth deciding now, because it is the thing a client asks six months after handover:

  • Dust is the main enemy of a polished surface, and it is more visible on dark stone than light. Microfibre, no sprays.
  • Nothing abrasive on a soft mineral. Fluorite is comparatively soft; it wants display and occasional dusting rather than being scrubbed.
  • Fingerprints are a real burden on high-gloss dark pieces in public areas. If a piece will be handled, that is an argument for a less glossy surface or a light stone.
  • A recessed or sheltered position ages better outdoors than an exposed one, whatever the material.

Where the public area is a hotel lobby, a guest room or a bar counter, the load is heavier again — cleaning regimes, spills, crowds and local safety rules all start to matter, and some of those bear on the material choice rather than on the design. Carved stone in hotels and bars sets out those constraints.

What goes in the specification

The document that prevents a redesign, in the order that keeps decisions honest:

  1. The job the piece performs, in one sentence
  2. The principal viewing direction and the light it receives
  3. Approximate scale, tied to the viewing distance
  4. Material, by mineral name, and whether any treatment has been applied
  5. Finish expected, and the fact that the material constrains it
  6. Placement constraints — indoor or outdoor, floor loading, access route, anchoring requirement
  7. Tolerances. Natural stone and hand finishing vary; agree a range rather than an exact figure
  8. What counts as a defect as distinct from natural variation — a colour zone is not a crack
  9. The schedule, confirmed in writing once the material is chosen
  10. Maintenance expectations for handover

Item 7 is the one that most often goes unsaid, and it is the one that causes the argument. Stone and hand finishing vary; a specification that demands an exact dimension is a specification for a dispute. Our own guidance on tolerances and remedies is the same as for any commissioned carving: agree in writing before production, and record it in the quotation.

Where we fit

We carve in our own workshop in Donghai, Jiangsu — approximately 300 m², 15 machines, a team of more than 20 — and every finished piece is inspected individually by hand before it enters inventory. Because it is our own workshop, the same people and the same standard run through a 3.8 in piece and a monumental commission, and we can tell you honestly what is realistically available in a given material at a given size rather than promising whatever was asked for.

Monumental work is made to order and specified as a project, since material availability at scale is what sets the practical limit. There is more on the process at manufacturing, and the installation-scale work at monumental works.

If you are specifying a piece for a project, send us the setting, the approximate scale and the light it will receive — with the intended viewing direction, we can tell you which materials will actually do what the scheme needs, and which will fight it.