How Can We Help?
Each drawable section is a Unit Part whose job is to carry the artwork:
- Create a Unit Part for the part in question (one per section image you want to be
able to draw — a frame profile, a bead, a bulb, a full block slice…). - Attach the SVG or DXF image of the section to the part.
- Set the part’s description to the image filename/path. This is the convention the
configurator uses to locate the artwork — it is not a free-text description.
The section picture is read from the part’s SVG image field, but it must be a raster
picture — PNG, JPG or GIF. The configurator decodes it as a bitmap; a true SVG (vector) file
cannot be decoded there, so its section is silently skipped. The 7100 series sections in this example are PNGs.
DXF artwork is stored on the part but not drawn on the canvas. Rotation and flip are recorded by the
rules (types 313–317) but are not applied at render time, so supply each image already
oriented the way it should appear in the section; do not rely on rotate/flip rules to fix the
artwork.
Scale the image to the real size of what it shows
A common question: does the image have to be drawn at the pixel size the part has on a
40 x 40 window, so that it can be scaled from there? No. The pixel size of the image does
not matter and is not tied to any window size. What matters is one number recorded on the part:
X, the real-world size of what the picture shows, in the series’ unit of measure (inches
for an imperial series, mm for a metric one). Every time the design is drawn — any size, any zoom,
desktop or web — the configurator works the picture’s size out again from X:
size drawn on screen = image pixels x real size of one pixel x the drawing’s current scale
“Across” is the direction the cut slices through: on a V cut (through the head and
sill) it is the image’s height; on an H cut (through the jambs and vertical bars) it is the
image’s width. The other side of the picture is scaled by the same factor, so the picture keeps
its proportions.
The picture is cropped tight to the artwork, so its full height (V cut) or full width (H cut) is exactly
those 3 inches.
Working out X
Add up everything the picture contains in the “across” direction, using the parts’ own
dimensions:
| Example: frame + sash + glass section | Takes |
|---|---|
| Frame profile face | frame part dim B |
| Frame-to-sash gap | sash clearance (series) |
| Sash profile face | sash part dim B |
| Sash-to-glass gap | glass clearance |
| Glass stub drawn in the image | the length you drew, e.g. 4″ |
| X | the sum — the real height (V cut) or width (H cut) the picture spans |
- Record X on the part as its image scale, in the series UOM: inches on an imperial series
(7100 series example: 3), mm on a metric one (e.g. 76.2 for the same 3-inch picture). - A single-profile picture follows the same rule with a shorter sum: X is that profile’s dim B. If
you leave the image scale at 0, the configurator falls back to the profile’s dim B — fine for a
single profile, far too small for a frame + sash + glass picture. A recorded X always wins. - Keep the head, sill and jamb pictures of the same profile set at the same X, so they line up.
Worked example — the same picture on three window sizes (7100 series)
Head picture: 300 px high, X = 3 in, so one pixel = 3 / 300 = 0.01 in. The drawing scale
comes from the window size (measured on the 7100):
| Window | Drawing scale | Picture drawn | Check |
|---|---|---|---|
| 40 x 40 | 10 px per inch | 300 x 0.01 x 10 = 30 px high, 26 px wide | 3 in of a 40 in frame = 3/40 of 400 px = 30 px ✓ |
| 80 x 60 | 5 px per inch | 300 x 0.01 x 5 = 15 px high, 13 px wide | 3/80 of 400 px = 15 px ✓ |
| 24 x 36 | 11.1 px per inch | 300 x 0.01 x 11.1 = 33 px high, 29 px wide | 3/36 of 400 px = 33 px ✓ |
needed per size; X does all the work.
Resolution
More pixels only make the section sharper — they never change its size. Aim for roughly
300–600 px across so it stays crisp when zoomed in or printed on the cross-section sheets report;
the 7100 pictures (263 x 300) are at the low end of that range.
Pictures that include parts from another level
Installation drawings often show more than the frame: the wall, the rough opening, an install
screw, a receptor or an expander. These belong to a different level of the design — usually the
Facet — not to the frame. The 7100 series head picture is an example: its top third is the
wall and the (yellow) install screw, and the frame only starts about 115 of its 300 px down.
above the red line belongs to the Facet; everything below it belongs to the frame and what it holds.
This matters because the configurator pins the picture by its edge: a head picture hangs down
from the frame’s outer (top) edge, a sill picture hangs up from the bottom edge, and a jamb picture
starts at the jamb’s outer edge. Anything drawn beyond the frame face pushes the profile inward by that
much, so the frame in the picture no longer sits on the frame of the drawing.
frame picture and a Facet picture.
Three ways to deal with it, best first:
| Approach | How | Result |
|---|---|---|
| 1. Split by level (recommended) | Crop the head, sill and jamb pictures so they start exactly at the frame’s outer face (frame, sash and glass only) and record X for that crop. Make the wall / screw / receptor its own X-section unit part with a Facet-level placement rule (see Step 2). |
Each picture lines up on its own edge. The Facet part only appears when it is really in the design, so an option or install method without that screw drops it automatically. |
| 2. Keep one picture and shift it | Record X as the real size of the whole picture, wall and screw included (measure it on the CAD drawing it was exported from). Add an offset rule — ΔY on the head and sill rules, ΔX on the jamb rules — equal to minus the distance from the picture’s edge to the frame face, in the series unit. |
Lines up, but the wall and screw are drawn on every design, including ones installed differently. |
| 3. Crop the extra parts off | Remove the wall and screw from the picture and record X for what is left. | Simplest, when the installation detail is not needed on the configurator or the section sheets. |
A PNG carries no scale of its own. To record X for a cropped picture — or the frame-face offset for
approach 2 — measure on the CAD drawing the picture was exported from (or on a printed true-scale
section). The series frame depth is only a lower bound for a picture that also shows sash and glass.
Common mistakes
its dimension lines.
| Mistake | What you see | Fix |
|---|---|---|
| White margin around the artwork | Section drawn too small and off its dimension lines (Figure 3) | Crop the picture to the artwork’s edges |
| X too small — e.g. 1 instead of 3 (what parts 40–43 had before October 2026) | Section about 10 px high on a 40 x 40 instead of 30 | Record the real X |
| X entered in the wrong unit — mm on an inch series | Section 25 times too big | Enter X in the series unit |
| Picture stretched when exported | Profile looks squashed; the other side is wrong | Export with the real proportions |
| A vector .svg file in the SVG image field | No section drawn at all | Export it as PNG |
| Picture includes the wall / install screw (Facet parts) above or beside the frame | Frame drawn too far in; wall shown on every design | Split by level (Figures 4 and 5) |
| Image scale left at 0 on a frame + sash + glass picture | Section shrunk to the frame face width | Record X |
The rules below only decide where the image lands (ΔX/ΔY offsets from the section’s
origin).