Method: how the monitor selector computes its results

What is compared

The selector evaluates fifteen formats, the combinations of diagonal and resolution that account for nearly all monitors sold: from 24-inch Full HD to 49-inch super ultrawide, including 5K and 6K screens. For layouts with two screens it also evaluates pairs of identical 24 and 27-inch monitors. The complete list, with dimensions and densities, is in the comparison of formats.

It does not evaluate models. Two monitors of the same format differ in stand, ports, calibration and price, all things that change with every product generation and that a specification sheet shows. The format is the decision that depends on your desk and your eyes.

From your answers to the parameters

Your viewing distance is the starting point. If you do not give it, it is estimated from the depth of the desk: depth minus 5 cm on a stand, plus 5 cm on an arm (the reasoning is in screen size and desk depth). The time spent reading sets the sharpness target: 40, 50 or 60 pixels per degree. The use sets the weights of the criteria, the budget the price level beyond which a format is penalised, and the system the scaling rules.

The six criteria

Each format receives six marks between 0 and 1.

1. Fit in the field of view

The horizontal viewing angle, θ = 2 · atan(width ÷ (2 × distance)), is compared with the comfortable band of the family: 40 to 53 degrees for a 16:9 screen, 44 to 60 for a 21:9, 58 to 80 for a 32:9, 72 to 100 in total for two screens. The mark is 1 inside the band and decreases linearly outside it, to 0 at 22 and 68 degrees for a 16:9 screen. Origin of the bands: viewing distance.

2. Sharpness

The pixels per degree at your distance, PPD = PPI × distance (in inches) × 0.01745, are compared with your target. The mark is 1 at the target and above, 0 at 62% of the target and below. Exceeding the target earns nothing: beyond what you can see, pixels are a cost, not a quality. On macOS the target is raised by 4 PPD, because text is not smoothed at sub-pixel level, and formats between 125 and 185 PPI lose 15% of their mark, because they are displayed resampled.

3. Workspace

The width of the workspace after scaling (resolution ÷ scaling, the scaling being the multiple of 25% nearest to PPI ÷ 110). The mark goes from 0 at 1600 units of width to 1 at 3840.

4. Motion

A judgement per format, from 0 to 1, of its suitability for games: refresh rates available on the market, load on the graphics card, response of the panels sold in that format. 27-inch 1440p has the maximum mark, 5K and 6K formats the minimum.

5. Colour

A judgement per format of the offer in screens for image work: gamut coverage, calibration, uniformity of the models available. 4K, 5K and 6K formats on 27 and 32 inches have the highest marks.

6. Price

Each format has a price level from 1 to 5. Up to the ceiling of your budget (2, 3.2 or 5) the mark is 1; above, it falls by 0.45 per level.

The weights

The marks are combined according to the use.

Use	Fit	Sharpness	Workspace	Motion	Colour	Price
Office work	33%	32%	10%	0	0	25%
Programming	25%	35%	20%	0	0	20%
Photo, video, design	22%	25%	13%	0	25%	15%
Gaming	25%	15%	5%	35%	0	20%
Trading, dashboards	25%	20%	35%	0	0	20%
A bit of everything	30%	22%	10%	10%	8%	20%

A tight budget multiplies the weight of price by 1.7. A generous budget removes 80% of it and gives the freed weight to sharpness and to the quality the use calls for. Four adjustments follow:

the total is multiplied by a factor between 0.55 and 1 according to the fit mark: a screen that does not fit the field of view cannot be rescued by its other qualities;
two screens lose 5 points, for the central bezel and the double cabling;
when you leave the choice of layout to the selector, wide formats lose 4 points: a flat 16:9 screen is the default;
a format whose interface is too small at 100% scaling, or too tall for the distance, loses 6 to 10%.

The format with the highest total is recommended. The “cheaper” option is the best-placed format at least half a price level below; the “step up” option, the best-placed one above. The score out of 100 measures the match with your situation, not the absolute quality of a format.

Panel, refresh rate and connection

These three indications are not scored: they are rules. The panel follows the use and the budget (IPS for text, wide-gamut IPS for images, fast IPS or OLED for games). The refresh rate follows the use and the load of the format. The connection is the least capable link whose payload carries the image uncompressed at the suggested rate, computed with reduced blanking; the figures are in connections.

Limits

The comfort bands are conventions drawn from ergonomic practice, not physiological constants. Some people are comfortable closer or farther; the gauges of the result show where you stand.
Eyesight varies. The sharpness targets assume 20/20 vision, corrected if need be. Sharper eyes benefit from more; presbyopia without suitable glasses changes everything.
The motion and colour marks are judgements about the offer of a format at the date of this page, not measurements of a model.
Prices are levels, not amounts: promotions and new generations move them.
The distance estimate has a margin of some 5 cm. Measure if you can.

The handbook explains each notion so that you can disagree with a weight in full knowledge of the facts, and the calculator lets you redo every computation.

Independence and funding

The site is free and requires no account. It is funded in two ways, both identified as such on the pages.

Affiliate links. Where a partner shop exists for your language, a button offers to search it for the recommended format. If you buy after following it, the site receives a commission at no extra cost to you. The link is created after the computation, from the format alone: no shop, brand or commission enters the scoring, and the method on this page is the whole of it.

A sponsor card, labelled “Sponsored”, may appear under the tools. It is served by the site itself, without third-party scripts or advertising cookies.

The site counts, as anonymous daily totals, the number of recommendations computed and of clicks on these links. It keeps no IP address, sets no advertising or analytics cookie, and loads nothing from third-party servers.