Shanghai GP3 100

Shanghai GP3 is the Chinese answer to budget panchromatic film. It comes out of the Shanghai Jiancheng Film factory and has been in roll production since the early 2000s, with on-and-off 35mm availability depending on what the factory feels like making in a given year. The 35mm version returned in 2019 with a pilot run of 10,000 rolls after a long gap, which is why a fresh-stock review economy around it has only existed in the last few years.

The comparison most reviewers reach for is Kodak Plus-X, the long-discontinued ISO 125 stock that disappeared from the Kodak lineup in 2011. GP3's grain is older-school cubic crystal, more pronounced than FP4 Plus or T-MAX 100 at the same speed, and the tonal response carries the slightly contrasty midtone snap Plus-X used to deliver. The rumor that the formula is a Plus-X derivative has never been confirmed.

Developer choice changes the film. Rodinal 1:50 for 12 minutes at 20C gives sharp, well-defined grain with strong midtone separation. Xtol 1:1 produces a smoother negative with less character. Stand development in Rodinal 1:100 for an hour with light agitation at 30 minutes gives a compensating effect that helps the narrow latitude. A stop of overexposure prints flat, and shadows block faster than HP5 Plus or Tri-X.

Handling quirks. The base is unusually thin and curls aggressively after drying, which makes scanning a fight. Anti-Newton glass or weighted clips help. Reticulation appears with even small temperature swings, so match the wash water to the developer.

Available in 120 reliably, in 35mm intermittently, and in 4x5 plus custom-cut sheet sizes direct from Jiancheng. The 120 is how most people first encounter the stock because the larger frame hides the slightly soft sharpness that shows up in 35mm enlargements.

Reciprocity exponent is 1.31. Zone Light Meter applies the correction past one second. A 30-second meter reading climbs to about 90 seconds at the negative, and a two-minute exposure stretches closer to seven minutes. Published reciprocity data is thin, so bracket the corrected reading by half a stop either way.