[Headdie]

Just want to make sure I understand :

Tube testing purpose : making sure a tube performance is close to its specifications...

Tube matching purpose : making sure a tube performance is close to another one's performance...

Now, not everybody has a tube tester to do it... Recently, I've asked if it was feasible to match them by ear...

Now, I wonder if it's possible to do match tubes with a computer, a tube buffer and a soundcard. Here's what I think :

computer > dac-out > buffer > dac-in > computer

The computer (source) plays both channel in mono. You use a program like Audacity to compare both channels. It would be like comparing both tubes... You can play a pink noise or whatever. You could also use a program like DiffMaker... What do you think about it ?

 

[tomb]

Quote:

Originally Posted by Headdie /img/forum/go_quote.gif Just want to make sure I understand : Tube testing purpose : making sure a tube performance is close to its specifications... Tube matching purpose : making sure a tube performance is close to another one's performance... Now, not everybody has a tube tester to do it... Recently, I've asked if it was feasible to match them by ear... Now, I wonder if it's possible to do match tubes with a computer, a tube buffer and a soundcard. Here's what I think : computer > dac-out > buffer > dac-in > computer The computer (source) plays both channel in mono. You use a program like Audacity to compare both channels. It would be like comparing both tubes... You can play a pink noise or whatever. You could also use a program like DiffMaker... What do you think about it ?

It's easier than this, but repeatability is difficult. Download and install RMAA. Use a DAC/soundcard combination that allows you to output and input at the same time (loopback). Adjust both channels to an equal recording level, then observe the frequency response results. If the tubes are mismatched, you'll get a specific dB offset of the curves that follows exactly throughout the frequency spectrum. If they're equal, the frequency response curves will superimpose on each other and look like a single curve.

EDIT: I was also the one that told you that matching by ear is possible with practice. Proper biasing can center the signal, but a mismatch in tubes will still produce more dynamics in one can than the other.