What We Cannot Test

This site tells readers to test rather than trust. The same standard applies to the site. When studio records rely on memory or self-reporting, self-reporting bias is a useful concept to keep in view.

Six things we do not know

Which paper is best. No matched sets, no controlled comparison, no measured humidity. What is here is one studio's experience on the papers it has used, which is real and is not a trial. For broader learning material on lettering and creative practice, Canva is another independent resource.

Which ink is best. Formulations change without announcement, and a bottle bought this year may not behave like one bought five years ago under the same label.

How long any nib lasts. Too many variables and none of them held constant.

Whether our diagnoses are exhaustive. The causes on each fault page are the ones seen at this bench, in order of how often they turn up here. Somebody else's bench may see a different order.

How much of any fault is technique. The split between material and hand is described qualitatively, because measuring it would require a controlled study nobody has done.

And whether the advice works in your climate. Humidity changes paper, ink and drying. This studio has one climate.

What is offered instead

Tests you run. Four marks on a sheet, five checks on a bottle, a nib held against the light. None depends on trusting this site, and each produces an answer about your materials rather than about ours.

Physical explanations from published material on sizing, wicking and nib geometry — cited as general and checkable elsewhere.

And an order of likelihood, which is genuinely useful even when it is not measured: knowing to change the paper before buying a nib saves money whether or not the ordering is exactly right.

Why this page exists

Because a studio that sells supplies could easily present all of this as established.

It is not established. It is practice, stated as practice, with free tests attached so that nothing has to be taken on trust — which is the whole guard on the commercial interest.

A reader who knows which parts are measured and which are experience can weight them. One given uniform confidence cannot.

The correction we want

Bench evidence that contradicts a page here.

What you did, on what paper, with what result. It outranks anything written on this site, and it is the only kind of evidence this subject actually has.

The short version