Free Diagonal Method Grid Generator
Check framing with the diagonal method instantly with our 100% free grid generator. Upload any photo and four 45-degree lines rise from its corners, their crossings marking where detail draws the eye, then export SVG or 600 DPI in seconds. No signup, no downloads needed. Trusted by 60,000+ artists worldwide.
45-degree lines from each corner
Edwin Westhoff's method: bisectors struck at 45 degrees from all four corners. Where they cross is where the eye settles, which is a different set of points from the thirds intersections and often a better one.
1 control, not a fixed picture
Set also show corner-to-corner X diagonals. Every one is live on the canvas.
Stacks with composition & photography
Overlays combine, each carrying its own transform, so this one can sit under or over rule of thirds, golden spiral and armature (14-line), and 15 more in the same set.
Crossings marked on the print
The intersections export with the lines at up to 600 DPI, or alone on transparency. SVG keeps the 45-degree angles exact.
| Ctrl + Z | Undo |
| Ctrl + Y | Redo |
| Esc | Deselect overlay |
| Alt + drag | Free positioning (no snap) |
| Ctrl + scroll | Zoom canvas |
| Ctrl + P | |
| ? | Show this help |
One from each corner, each running the length of the frame's short side. Where they cross is the whole of the method — four points that move with your aspect ratio in a way the thirds never do.
The diagonal method generator draws one of the 82 composition overlays in Grid Maker Pro, and the rest stack on the same image.
- 4 lines at 45°
- 4 crossings
- SVG + 600 DPI export
- Nothing uploaded
What is the diagonal method?
The diagonal method generator is a free browser tool that draws Edwin Westhoff's construction over any photograph. Four lines are run inward from the four corners at exactly 45°, each the length of the frame's short side, so each ends on the opposite long edge. They cross at four points forming a diamond about the centre — and unlike a thirds grid, those points move with the aspect ratio. On a 3:2 frame they land at one third and two thirds of the width and at a quarter and three quarters of the height. Corner-to-corner diagonals can be shown alongside.

How to use the diagonal method
The four points where 45-degree lines from the corners meet, which is where the eye settles in a frame — a different answer from thirds, and often the better one.
Load the frame
The diagonal method needs a frame: drop in the photograph you are judging. The construction is built from the frame's own short side, so the crossings depend on the aspect ratio and are computed rather than assumed.
JPG · PNG · GIF · WEBP
Read the four crossings
They form a diamond about the centre — two on the horizontal midline, one above and one below. Those four points are what the method actually proposes.
Diagonal Method · line colour · opacity
Add the diagonals if useful, then export
The corner-to-corner diagonals give the frame's main directions alongside the 45° set. Export SVG for vector or a raster up to 600 DPI.
PNG · JPG · PDF · SVG · up to 600 DPI
Three steps, about a minute. No signup, no upload.
What the Diagonal Method Generator does
Exactly 45°, not approximately
The diagonal method is four lines at one angle and nothing else, which makes it the lightest of the 82 composition overlays that places a subject. Sketched by eye a 45° line is easily three degrees out, and every crossing moves with it — the points are where two lines meet, so the error doubles.
Four points that move with the frame
The crossings depend on the aspect ratio, unlike thirds. On 3:2 they sit at a third and two thirds across and a quarter and three quarters down; on 16:9 they spread to 0.28 and 0.72 across. The construction adapts; a grid does not.
Diagonals alongside, optional
The corner-to-corner pair gives the frame's own directions. Together with the 45° set they show where the two families of diagonal agree and where they diverge, which is most of what this construction is useful for.
Over your own photograph
Its own transform, so you can rotate the construction onto a tilted subject or scale it to test a crop without moving the image.
Readable over a busy frame
Per-overlay colour and opacity, and an eyedropper that samples the photo, so four lines sit in the picture rather than across it.
Vector or print size
SVG export keeps four straight lines as vector. Raster export is sized from the paper format and DPI you choose, up to 600.
Diagonal method examples: four ways the construction reads
Diagonal method vs a photo editor's guides vs a downloaded overlay image
| Option | Angle exactly 45° | Crossings computed for your frame | Vector export | Cost |
|---|---|---|---|---|
| Grid Maker Prothis tool | Yes | Yesfrom the short side | SVG + raster to 600 DPI | Free |
| A photo editor's guides | Yesif you place them | Noyou compute them | Yes | Subscription |
| A downloaded overlay image | Whatever aspect it was made at | No | No | Free or paid |
| Drawing it on a print | Only as accurately as you aim | By construction | No | Time |
Who the diagonal method is for: frames whose shape is part of the composition
How the diagonal method works — One angle, four crossings, and what happens to them as the frame changes shape
The diagonal method is the youngest of the 82 composition overlays and the one whose behaviour is easiest to check. Here is exactly what it draws, where its points land on the frames people actually use, and the one aspect ratio on which it stops working altogether.
Take a rectangle. From each of the four corners, draw a line inward at exactly forty-five degrees, and let it run the length of the frame's short side. On a landscape frame each line therefore ends on the opposite long edge; on a portrait frame, the same with the axes swapped.
Four lines, and they cross each other at four points. Those crossings form a diamond about the centre — one above, one below, one to each side — and they, rather than the lines, are what the method proposes as placements.
It is one angle repeated four times. Everything interesting about it is where the repetitions meet.
Where the points actually land
This is the part worth having in numbers, because it is what distinguishes the method from a grid. On a 3:2 frame the two side crossings sit at exactly one third and two thirds of the width, and the top and bottom crossings at a quarter and three quarters of the height.

On 4:3 the side crossings move in to 0.375 and 0.625 while the vertical pair moves out to a third and two thirds. On 16:9 they spread hard: 0.281 and 0.719 across, 0.111 and 0.889 down. The construction is genuinely responsive to the shape of the frame in a way a thirds grid — the same four points on every aspect ratio — is not.
That responsiveness is the honest argument for it. Whether or not you accept anything else about the method, a construction whose points move as the frame gets wider is describing something about the frame, where an imposed grid is describing only itself.
And the ratio where it collapses
On a square, the four 45° lines are simply the two corner-to-corner diagonals, and all four crossings coincide at the centre. The method degenerates completely — it has nothing to say about a square frame beyond pointing at the middle.
That is worth knowing rather than discovering. If you are working square, this construction is the wrong one; use the rabatment generator, the armature grid generator, or a centred system that is designed for symmetry. The overlay will happily draw it, and what it draws will be an X.
What Westhoff actually claimed
The method has one author and one date: Edwin Westhoff, a Dutch photographer, published it in 2006 after examining a large body of paintings and photographs. On his reading, important details sat on these lines rather than on the rule-of-thirds intersections.
That is a correlation claim about a corpus, made by the person proposing the method, and it should be weighed as such. The construction is well-defined and its behaviour is checkable; the historical argument is the same kind of after-the-fact overlay reasoning that dynamic symmetry and the golden ratio both attract, and it deserves the same caution.
The practical position is straightforward. Use it because the placements it suggests produce compositions you like, and because it responds to aspect ratio in a way thirds does not. Do not use it as evidence about what Rembrandt was doing.
How it relates to the constructions next to it
The diagonal method overlay page sets out Westhoff's argument, beside the rest of the composition grid overlays. It is not the golden triangle, though the two get confused. The golden triangle uses one corner-to-corner diagonal plus two perpendiculars onto it; this uses four lines at a fixed 45° from all four corners. Different lines, different points, different behaviour as the frame changes.
It is also not the armature of the rectangle, which is built from corners and edge midpoints and is affine-invariant — its crossings stay at the same fractions whatever the aspect ratio. The diagonal method is the opposite: its crossings move. Those are two coherent and opposite design decisions, and knowing which you want is more useful than knowing either construction.
Using it on a photograph
Load the frame and look at what is near the four points before doing anything else. The method's own claim is about detail placement — an eye, a hand, a highlight — rather than about mass, so the question is whether small important things are near the crossings, not whether the subject is.
Then test a crop. Because the crossings move with the aspect ratio, changing the crop changes where they are, and the overlay's transform lets you see that directly. A frame that has nothing near the points at 3:2 may have something at 16:9, and that is a real reason to crop rather than a rationalisation of one.
And turn the corner diagonals on once. Where a 45° line and a corner diagonal nearly coincide, the frame has a direction that both systems agree about, and that is usually the strongest line in the picture. Where they diverge sharply — on a very wide frame they diverge a great deal — you are being offered two different readings and it is worth choosing.
Practical notes
Line weight matters here: keep it down. Four lines plus two diagonals is six strokes across a photograph, and at full opacity they compete with the subject rather than describing it. The eyedropper is useful here: sample a mid-tone from the image and the lines sit in the picture instead of on it.
For teaching, the aspect-ratio behaviour is the demonstration worth doing. Put the overlay on a square, a 3:2 and a 16:9 in turn and watch the diamond stretch from a point to a wide lozenge. It makes the difference between a responsive construction and an imposed grid visible in about ten seconds.
A note on the corners
One detail of the construction is worth understanding because it decides the whole shape of the result. Each line starts at a corner and runs exactly the length of the frame's short side. On a landscape frame that means the line ends somewhere along the opposite long edge rather than at a corner, and how far along depends entirely on how wide the frame is.
Where the diagonal method gets used

Wide crops. The points move out as the frame widens, which matches what actually happens to a picture when you crop it wide.
I check detail placement with it rather than mass — an eye, a hand, a catchlight. That is what the method is actually about and it is a different question from where the subject sits.
Square, 3:2, 16:9 in a row is the whole lesson. Students see a construction respond to a frame instead of being stamped on it.
Diagonal method: frequently asked questions
The geometry
What does it draw?
Four lines at exactly 45°, one run inward from each corner, each the length of the frame's short side so that it ends on the opposite long edge. They cross at four points forming a diamond about the centre, and those crossings are what the method proposes as placements.
Where do the crossings land?
On 3:2 the side pair sits at exactly a third and two thirds of the width, and the top and bottom pair at a quarter and three quarters of the height. The positions follow the aspect ratio, which is the point: on 4:3 they become 0.375/0.625 across and thirds down, and on 16:9 they spread to 0.281/0.719 across and 0.111/0.889 down.
What happens on a square frame?
It collapses. The four 45° lines become the two corner-to-corner diagonals and all four crossings coincide at the centre. On a square this construction has nothing to say — use rabatment, the armature, or a system designed for symmetry.
Who devised it?
Edwin Westhoff, a Dutch photographer, published it in 2006. He examined a body of paintings and photographs and concluded that important details sat on these lines rather than on the thirds intersections.
Is that claim reliable?
It is a correlation claim about a corpus, made by the person proposing the method, and deserves the caution any after-the-fact overlay argument does — the same caution the golden ratio and dynamic symmetry attract. The construction itself is well defined and its behaviour is checkable; use it because the placements work, not as evidence about old paintings.
Is it the same as the golden triangle?
No, though they get confused. The golden triangle is one corner-to-corner diagonal plus two perpendiculars dropped onto it. This is four lines at a fixed 45° from all four corners. Different lines, different points, different behaviour as the frame changes.
Should I use it for the subject or for details?
Details. The method's own claim is about small important things — an eye, a hand, a highlight — sitting on the lines, rather than about where the main mass goes. That is a different question from the one thirds answers.
Using it
When should I add the corner-to-corner X?
When you want to see how Westhoff's construction relates to the frame's own diagonals — the 45° lines and the true diagonals only coincide on a square. Adding both is a good way to understand the method once; keeping both on makes eight lines to judge four crossings against.
The lines land on nothing in my picture. Does that mean it failed?
It means this picture is not built on the method, which is a real answer rather than a failure. The construction is a claim about where detail tends to fall in pictures that work; a picture that ignores it may still be good, and the grid has told you it is doing something else.
Export and print
Can I print it to check a set of prints?
Export the lines alone onto acetate at your print size and lay it over physical prints one after another. That is how the method was originally tested, and it beats opening twenty files.
Does the 45° angle survive export exactly?
In SVG, yes — the lines stay at exactly 45° as vector paths. In a raster the angle is exact in the geometry and stair-stepped in the pixels, which is fine for reference and visible if you print it large.
Choosing a grid
How is that different from the rule of thirds?
Thirds puts the same four points on every frame regardless of shape. The diagonal method's points move as the frame changes proportion, so the construction is describing the frame rather than being stamped onto it. That is the honest argument for it.
How does it compare with the armature of the rectangle?
They are opposites in a useful way. The armature is affine-invariant — its crossings stay at fixed fractions whatever the aspect ratio. The diagonal method's crossings move. Both are coherent decisions, and knowing which behaviour you want matters more than knowing either construction.
Should I use this or the armature of the rectangle?
The armature grid generator if you want a general-purpose structure; this if you want Westhoff's specific claim about where detail sits. The armature has fourteen lines and many intersections, this has four lines and four crossings — a much narrower and more testable proposition.
Practical matters
Does my photo get uploaded?
No. Everything happens in your browser — the image is drawn to a canvas on your own device and never sent anywhere.
Is the method actually evidence-based?
It comes from Westhoff's own analysis of a large number of photographs, which is an observation rather than a controlled study, and it has not been independently replicated at scale. Use it as a compositional idea worth testing on your own pictures, not as a rule with proof behind it.
Open the tool Diagonal method generator
One angle, four moving points
The diagonal method's diamond stretches as the frame widens. Watch it, then place against it. Export vector or 600 DPI. No signup, nothing uploaded.
Back to the generator →