What's the best approach to testing salvaged germanium transistors?

Even the peak atlas transistor analyser reads leakage current as gain
current and the DMMs I have are completely useless at testing germanium
transistors. Somewhere I have a schematic for a transistor tester that nulls
out the leakage before taking a gain measurement, the leakage is read off
the calibrated null pot, but I'm wondering whether it might be better to
measure the AC gain?

What I'm thinking of is driving the transistor under test with an oscillator
with its output clipped by an inverse parallel pair of diodes and measuring
the rectified output of the TUT to calculate gain.

One particular advantage I'm thinking of, is an amplifier can be added to
evaluate how much hiss the TUT contributes.

The clever bit would be deciding what biasing circuit to use that would
betray the leakage figure by simple voltage measurement under DC conditions.

Any comments/suggestions welcome.

TIA.