AI-generated illustration, not a photograph of a specific product
Read two product pages for Japanese N scale models and you will see two different ratios. That is not an error on either page, and it is not something you get to average out.
Why one gauge carries two scales
KATO and TOMIX both define N the same way: 9 mm gauge, with Japanese conventional-line prototypes reduced at 1:150 and Shinkansen and foreign prototypes at 1:160. That looks contradictory until you separate three words that people use interchangeably:
| Term | Answers | For N |
|---|---|---|
| Gauge | How far apart the inside faces of the rails are | 9 mm |
| Scale | How many times smaller the model is than the real thing | 1:150 conventional / 1:160 Shinkansen |
| Compatibility | Whether two things will actually run together | Cannot be derived from the two rows above |
The reason both ratios land on 9 mm is that Japanese conventional-line prototype track is narrower than standard gauge. Reduce a narrower prototype by a smaller factor and you arrive at the same model gauge. So "both are N scale" and "both are the same scale" are genuinely different statements. If you are weighing N against a larger scale entirely, see N vs HO.
Real ↔ model converter
Convert in either direction, with metric or imperial units on the input and the output.
Our take: the useful part is not treating Shinkansen as 1:150
Most of the time the 6.67% difference is invisible — you are not going to notice it looking at one car. Where it stops being invisible is anywhere you add lengths together. A station, a passing loop, a fiddle yard, a shelf. Eight cars at the wrong ratio is about 67 mm out, which is roughly a quarter of a KATO platform section.
So the practical rule is narrow and worth remembering: match the denominator to the stock that will actually stand there. When you size a platform with the platform length calculator, pick 1:160 for a Shinkansen set and 1:150 for a conventional formation, and the arithmetic will not quietly drift.
A 20 m car, both ways
| Cars | At 1:150 | At 1:160 | Difference |
|---|---|---|---|
| 1 | 133.3 mm (5.25 in) | 125.0 mm (4.92 in) | 8.3 mm |
| 4 | 533.3 mm (21.0 in) | 500.0 mm (19.7 in) | 33.3 mm |
| 8 | 1,066.7 mm (42.0 in) | 1,000.0 mm (39.4 in) | 66.7 mm |
| 16 | 2,133.3 mm (84.0 in) | 2,000.0 mm (78.7 in) | 133.3 mm |
Scale arithmetic from a nominal 20 m prototype car; coupler spacing is not included. Note the sixteen-car difference is exactly one car length — that is not a coincidence, it is the 6.67% compounding.
Rescaling a 3D print between scales
If an STL is already at one scale and you want it at another, you do not divide by the new scale — that would treat the file as though it were full size. You need the ratio between the two:
Rescale percentage = source denominator ÷ target denominator × 100%
| Source | Target | X / Y / Z scale | Change |
|---|---|---|---|
| 1:160 | 1:150 | 106.6667% | 6.6667% larger |
| 1:150 | 1:160 | 93.75% | 6.25% smaller |
Those two percentages are not symmetric, which catches people out — going up 6.6667% and coming back down is not "minus 6.6667%". Use the formula rather than reversing a remembered number.
This is a geometry tool, not a compatibility check
One boundary worth stating plainly: this converter answers how big something is and how much to scale it. It does not answer whether two things will run together. Two products can both say N scale, both sit on 9 mm gauge, and still differ in wheel flanges, couplers, minimum radius and current pickup — and none of that is derivable from a ratio. Those are per-product questions, answered by the instruction sheet. For couplers specifically, see the coupler compatibility guide; for what a vehicle will take on a curve, the radius guide.
Sources
Last checked 2026-08-01. Gauge and scale definitions follow KATO's and TOMIX's own published descriptions of N gauge. The conversion and rescale figures are arithmetic, shown so you can check them. For any specific product, the stated scale on that product page takes precedence.
- TOMIX beginners page — 9 mm gauge and the scales used on it
- KATO FAQ: N gauge — 1:150 with Shinkansen and foreign prototypes at 1:160
Now put the number to work
Scale conversion: common questions
Why does N scale have both 1:150 and 1:160?
Because gauge and scale are different things. KATO and TOMIX both define N gauge as 9 mm between the rails, then model Japanese conventional-line prototypes at 1:150 and Shinkansen and foreign prototypes at 1:160. The track is shared; the reduction ratio on it is not. Japanese prototype track is narrower than standard gauge, which is why a different ratio still lands on 9 mm.
How different are 1:150 and 1:160 in practice?
About 6.67 percent in every linear dimension. A 20 m prototype car is 133.3 mm at 1:150 and 125 mm at 1:160 — a difference of 8.3 mm on one car. Over a long formation that adds up: eight cars differ by about 67 mm, which is most of a platform section. It is small enough not to look wrong on its own and large enough to matter when you are sizing a station.
What is the difference between gauge and scale?
Gauge is the distance between the inside faces of the rails, which for N is 9 mm. Scale is how many times smaller the model is than the prototype — 1:150 or 1:160 here. A third thing, compatibility, follows from neither: whether two vehicles can run together depends on wheel standards, couplers, minimum radius and electrical arrangements, and has to be checked per product.
What percentage do I use to rescale a 3D print from 1:160 to 1:150?
106.6667 percent, applied equally to X, Y and Z. The formula is source denominator divided by target denominator times 100 — here 160 ÷ 150. Going the other way, from 1:150 to 1:160, is 93.75 percent, a reduction of 6.25 percent. Note the two percentages are not symmetric, which is why you should use the formula rather than assuming the same number works both ways.
Can I just divide by the new scale to rescale a model?
No, and this is the common mistake. Dividing by the target scale treats the file as though it were full size. A file that is already at some scale needs the ratio between the two scales — source denominator divided by target denominator. Getting this wrong by dividing a 1:160 file by 150 makes it more than a hundred times too small.
Is a uniform rescale safe for a 3D print?
It preserves the geometry's proportions and nothing else. After scaling you have to re-check minimum wall thickness, hole diameters, clips and tabs, and supports, because all of them scaled too. Scaling a body shell up also scales the wheel openings and coupler positions, so it will not automatically fit 9 mm gauge running gear.
Does the same scale mean two models will run together?
No. This converter answers how big something is, not whether it will run. Two products can both say N scale and both sit on 9 mm gauge while differing in wheel flanges, couplers, minimum radius and how they pick up current. Those are per-product questions answered by the instruction sheets, not by a ratio.
Which scale should I use for a station or platform calculation?
Match it to the stock that will actually stop there. Use 1:150 for Japanese conventional-line formations and 1:160 for Shinkansen. On a long train the choice is worth most of a car length, so a platform sized with the wrong denominator can leave the last vehicle hanging off the end.
