# From simulation to a build

> What 00 Simulator’s model assumes and leaves out, what to enter so the box you build matches it, and how to check the build with a measurement.

Source: https://simulator.00aud.io/docs/start/simulation-to-build

A simulation is a model of the box you describe. It is close enough to choose a design and size a box, as long as you describe the box you will actually build. This page covers what the model assumes, what it leaves out, and how to measure the finished box and compare.

## What the model is

- _Boxes are lumped._ The driver, the air in each chamber and the air in each port are treated as single masses, springs and losses. That holds while the box and port are small next to the wavelength, which is 3.4 m at 100 Hz and shrinks as frequency rises.
- _Horns are paths._ Sound travels along the path as a plane wave, segment by segment, so the model follows the path’s length and area. It does not see its bends and folds.
- _It is linear._ Output scales with power and the shape never changes. Nothing compresses, except the voice coil temperature you can set for the Max SPL tile.
- _The driver is its T/S parameters:_ datasheet values, as the library holds them. Real samples differ from the datasheet and from each other.

## Enter the box you will build

**Volume** in the enclosure panel is the net internal volume: the air the driver works against, “excluding ports, bracing, and driver displacement”. Enter the driver and the bracing under **Box volume adjustments** (**Driver displacement** and **Bracing / other**) and the panel’s **Results** show the **Gross volume**: the space to build inside the walls, port included.

For the [vented example](https://simulator.00aud.io/app?template=vented), 55 L net with one 100 mm port 35.6 cm long, the port itself takes 2.8 L, so the gross volume is at least 57.8 L before you add the driver’s own displacement from its datasheet.

| What to enter | Where | Why it matters |
| --- | --- | --- |
| Driver displacement | **Box volume adjustments** | The magnet and basket take air from the box. Leave it out and the built box is smaller than the model. |
| Bracing and anything else inside | **Bracing / other** | Same: every litre of wood is a litre less air. |
| Port size and number | **Port** | **Port length** follows from the tuning, the port area and the end correction. Check it fits the box before you cut. |
| How the port ends | **End correction**, under the port’s **Advanced** | A port acts longer than it is, and a flanged end, such as one flush with the baffle, acts longer again. With “Two flanged ends” instead of “WinISD default”, the example’s port shrinks from 35.6 cm to 33.1 cm for the same tuning. |
| A finite front baffle | **Baffle step**, **Simulate baffle step** | Off by default. On, it adds the smooth loss that a baffle of your width and height causes as frequency falls, 6 dB at the bottom. It does not draw diffraction ripple. |

**Port model** under the same **Advanced** is **Lumped** by default. **Distributed** treats the port as a duct, which matters when its pipe resonance comes near the band you use. [Ports and end correction](https://simulator.00aud.io/docs/boxes/ports) goes further.

## The top of the response

Above the bass, a real driver’s response is set by its cone and voice coil, which a lumped model describes only roughly. **Settings › Simulation… › Non-horn upper response** chooses how box designs handle it:

- **Estimated · BITS frequency limit**, the default, assumes a flat driver with a 24 dB/octave roll-off at the driver’s published upper frequency limit, when the library has one. The box’s own response and your EQ still apply. Impedance, excursion and phase use the physical model.
- **Physical model** uses the driver’s T/S parameters, including its voice coil inductance, all the way up.
- Horns always use the physical model.

**Upper frequency** under **Settings › Display…** sets how far up the solver runs. Auto is 400 Hz, or 2 kHz when the workspace has a horn.

## What it leaves out

- The room. Its modes and reflections add peaks and dips, often larger than the differences between designs. **Radiation space** uses idealised boundary conditions: a level offset for boxes, and a change in mouth loading for horns. It does not model a room.
- The cabinet itself: panel resonances and standing waves inside a box. A lumped box has the same pressure everywhere inside it.
- Where the sound goes. The model gives one level at the listening distance, not how the sound spreads with angle.
- Behaviour at the limits: a softening suspension and motor near Xmax, a port that compresses and chuffs, a coil that heats up. [Limits: travel, heat and air](https://simulator.00aud.io/docs/start/limits) shows where each one starts.
- Your driver. Measured T/S parameters of your own sample beat the datasheet; enter them with **Edit params** on the **T/S Parameters** tile.

The full list, with how the solver is checked against other tools, is in [Model and limits](https://simulator.00aud.io/docs/reference/model-and-limits) and on the [validation page](https://simulator.00aud.io/validation).

## Measure the build and compare

1. **Measure.** Measure the finished box with a measurement microphone and a tool such as REW. An impedance sweep is worth taking too: it shows the real tuning without any room in it. Note the distance and where the box stood.
2. **Import.** Drop a REW export, an .frd or a .zma file on a graph, or paste a table. Or open **Measured traces**, below the enclosure panel on the left, and choose **Add measured trace**.
3. **Link it.** In the trace’s details set **Linked to** to the enclosure it measures, then fill in **Distance** and **Space**. A linked trace takes a shade of the enclosure’s colour and hides with it.
4. **Line up the level.** **Align…** matches the trace’s average level to the design over a band you choose. Only the offset changes, and **Residual after aligning** says how far the shapes still differ.
5. **Read the difference.** Compare the impedance first: on a vented box the dip between its two peaks is the real tuning, and it tells you whether the port and volume came out as modelled. Then compare the SPL shape, remembering the room.

[Measured traces and notes](https://simulator.00aud.io/docs/guides/measurements) covers the import formats, smoothing and display options.

## Ask your AI

**Ask your AI**

Your assistant can import and compare measurements through the 00 Simulator connector.

> Using 00 Simulator, compare the measured SPL trace in my open workspace with the enclosure it is linked to, over 30 to 200 Hz: tell me the level offset, the RMS residual once the offset is removed, and the frequency of the largest remaining difference.