# How to Look at a Painting in a Gallery > A practical method for reading a picture's physical state from the visitor's side of the rope: light angle, varnish, craquelure, edges, frame, and the label read as a document. > The Study - 2026-08-02 > Canonical HTML: https://iconotheca.com/study/how-to-look-at-a-painting-in-a-gallery/ ## Key points - Gallery light is set by conservation limits, commonly 50 lux for works on paper, 150 for oil paintings and 300 for stone and metal, so the raking angle you need must be found by moving yourself rather than the lamp. - Missing cusping along one edge of a canvas is the standard indication that the picture has been cut down, as it was for Artemisia Gentileschi's Self-Portrait as Saint Catherine of Alexandria. - Spike Bucklow's Hamilton Kerr Institute study of over 500 paintings found that crack patterns alone let observers separate art-historical groups with up to 97 per cent accuracy in pairwise comparisons. - The accession number on the label encodes the year the museum acquired the work, which is the last documented event in its provenance and is free to anyone standing in the room. ## Find the light before you look at the picture ## Varnish is the first thing you are seeing, and it is not the artist's ## Cracks record materials and structure, not a certificate of age ## Edges tell you what has been removed ## The label is a document, and its least readable part is the most useful ## Hanging position is an institutional argument None of this replaces the conservator's report, and none of it should be presented as though it did. What it does is stop you from making claims the object cannot carry. Most bad writing about paintings comes from treating a surface as though it had been delivered intact from the studio: praising an artist's cool silvery greys when the greys are yellowed varnish, or an artist's daring crop when the crop was made by a dealer with a saw. Ten minutes spent finding the raking angle, reading the edges against the composition and taking the accession number off the label puts a floor under everything you go on to write, and it costs nothing but attention. ## Questions **Can I judge condition properly without raking light?** No, and you should say so in writing. What the visitor's angle yields is a set of questions: uneven gloss, an edge that looks trimmed, a crack pattern that does not match the claimed support. The answers live in the museum's technical file, its published collection catalogues and, for London, the National Gallery's Technical Bulletin series. **Is craquelure evidence of authenticity?** It is evidence of a material and structural history, which is not the same thing. Bucklow's Hamilton Kerr Institute work shows crack networks correlate strongly with broad art-historical groups, but it reports categories and probabilities, not verdicts on individual pictures. Cracks can be induced artificially, and lining, transfer and heat treatment all disturb them. **What does the accession number actually tell me?** The year the museum took the object into its collection, which is the last documented event in the provenance, plus the key to the object's curatorial file. Quote it in every citation, because it survives website redesigns, retitling and changes of attribution in a way that a URL does not. **Why do galleries keep some rooms so dark?** Because light damage is cumulative and irreversible, and total exposure is lux multiplied by hours. Guidance caps sensitive material at around 50 lux and limits display time, commonly to about three months in every five years for works on paper, so a dim room is a rationing decision made to keep the object exhibitable for another century. ## Sources - [1] Conservation Center for Art and Historic Artifacts, 'Light Exposure for Artifacts on Exhibition' (50/150/300 lux by sensitivity class; UV at 75 microwatts per lumen; lux x hours; display duration guidance). - [2] Canadian Conservation Institute, 'Agent of deterioration: light, ultraviolet and infrared' (cumulative and irreversible nature of light damage; recommended illuminance by material class). - [3] Getty Publications, 'Conserving Canvas', case study on Artemisia Gentileschi's Self-Portrait as Saint Catherine of Alexandria (cusping present at bottom and sides, absent at top; missing palm tip and central pearl; canvas strip added; examination under acute raking light). - [4] National Gallery, London, catalogue entry: Artemisia Gentileschi, 'Self Portrait as Saint Catherine of Alexandria'. - [5] Hamilton Kerr Institute, Fitzwilliam Museum, 'The Classification of Craquelure' (Spike Bucklow; 500+ paintings; four art-historical groups; 30 sorters; ~75 per cent human correlation, >90 per cent confirm/refute, 97 per cent pairwise; 82 per cent and 94 per cent by machine). - [6] National Gallery, London, glossary entry: 'Raking light'. - [7] National Gallery, London, glossary entry: 'Craquelure'. - [8] National Gallery, London, 'Technical Bulletin' (the published series in which condition and technique findings appear). - [9] RKD Studies, 'Counting Vermeer', section 3.6 'Cusping' (cusping as evidence of format and of original tacking edges). - [10] Art Institute of Chicago, 'How to Read a Label' (the first four numbers of the credit line indicate the year of accession). - [11] Mint Museum, 'Deciphering what the information in a museum label means' (tombstone data, credit line, accession number). - [12] Getty Research Institute, 'Categories for the Description of Works of Art': Ownership and Collecting History. - [13] National Gallery, London, 'Caring for the paintings' (varnish discolouration and clouding; the gallery environment).