Matching? For driver tubes, that depends on who you ask, with these Reflector 6N3P-E true NOS tubes the variation, triode to triode and tube to tube, is likely to max at 15%, and usually much closer than that. So buy a few and listen to them, swap tubes if necessary, the goal is the accurate, centered, 3D aural projection. Even if you have a perfectly 'matched' pair, it may fail to deliver the solid 3D projection. Anyway, some tube dealers provide Platinum Grade at "typically 10% or tighter, but may be higher depending on available stock," and Gold Grade at "typically between 15% and 20%..."
That's the easiest, quickest, and least costly answer.
TL;DR: Only for the obsessive without tube testers.
One step further, and again referring to triode driver tubes (voltage gain stage), we are told:
The three parameters that are relevant in matching tubes are Gm (Transconductance), Rp (Plate Resistance), and Mu (Gain).
And, Mu = Gm x Rp. And, for triode driver tubes in a voltage gain stage, the important number is Mu. Check the following for more detail:
http://www.tubeaudiostore.com/tubmatdem.html
http://www.augustica.com/Tutorials/Tubes/matching_triodes_and_pentodes.htm
https://www.audioantiquary.com/en-us/categorie/valvole-nos-nib/
So, if Mu (voltage gain), is the important number, and it can be calculated from test measurements, then it's easy.
But if you don't have a tube tester, or you bought a DIY tube tester and haven't bothered to finish it
...
Why not just measure the voltage output directly, triode to triode, tube to tube, in the circuit being used, and match accordingly?
(In the Schiit hybrid amps, rectification and power output are solid state, and assumed to be consistent and well matched channel to channel,
so you are not dealing with other tube stage variables.)
I'm only suggesting this for true NOS tubes from the same batch, box, and/or production run.
You are then only comparing gain between new/unused tubes built from the same component pieces, assembled in the same way, at the same time.
With the RFL 6N3P-E this is possible because these are still available in true NOS pristine condition in any quantity you choose.
You need a voltage meter and a tone file. Voltage meter can be expensive or modest. Tone file can be created or downloaded.
You can use 1kHz as a typical default, or you can use more than one tone to check response at different frequencies.
If you want to check the individual triodes, you can use a hybrid amp like the Vali2 which uses one triode per channel.
Remember these will be relative values, meaning they are only relevant when comparing values using the same circuit, signal level and volume pot setting.
And yes, volume pots can have some small channel imbalance at either extreme, but you can check this.
Or, you can do two readings, one at full open and one at a typical listening level.
Remember that even between tube testers, the numbers are not going to be completely consistent.
So you can match your tubes to 1%, 5%, 10%, whatever you choose.
How long does it take for a new tube to reach a reasonable stability for reliable numbers?
Another question would be: How long did the tube dealer give your matched pair of tubes?
With the RFL 6N3P-E, my experience is that ~30 minutes will usually give numbers that (may) fluctuate within 10% for the first ~60 hours of burn.
And then settle to something close to the original readings. But in the first few hours, after ~20 minutes, there is very little or no fluctuation.
Bigger question is: Can you actually hear the difference between a 1% or 5% match and a 10% match?
Most importantly, you are actually listening to each tube to check sound quality first, something that the measurement numbers can never tell you.
Again YMMV. Not saying that this method is any where near complete, or even adequate, just a pragmatic and useful approach.
For me, not only pragmatic and consistent, but with results equal to, or surpassing, my expensive dealer-matched tube pairs (which I don't listen to anymore
).
(BTW, first thing I did with my DMM was to test the V output on those dealer-matched pairs, and yes they were very close.)