Quote:


Like there is just one tolerance . Look at a tube data sheet and there are many specs: capacitances, currents, voltages and transconductance. Transconductance (g_m) is what tube testers test, but it is not correlated to "sound." In other words tubes with very different g_ms can sound the same, and tubes with very similar g_ms can sound very different. In this context very different means greater than 20%.

This circles back to the fundamental magic of tubes, and you can generalize it to amps if you want. What do you measure to tell what it sounds like? You figure that out, and you've really accomplished something. And then we can start measuring tolerances.

My experience with tubes is "it depends." For a lot of tubes from the golden age of tube tech, the "same" tubes, i.e. tubes that were made the same, and look the same in detail, usually sound pretty much the same. Usually but not always. Russian tubes that I've bought in quantity, from the same factory and the same date code, can be all over the map. I think a big factor is that Western tube factories did real QC, and so the factory brand tubes generally performed pretty consistently. Yes, tubes were a commodity, but they were a highly competitive commodity and quality mattered. You could say it was a simpler time.