Announcing .... the Schiit Freya Einvigi.
An 8-tube 6SN7 amplifier like nothing ever seen before. (*)
Like the Freya+, you can switch between passive, differential buffer, or differential tube gain stages.
So what's an Einvigi?
From
The Vikings of Bjornstad Old Norse Dictionary, Einvigi, is a duel, single combat.
The Einvigi is two Freyas in a box, designed to allow instantaneous switching between any two pairs (or combos) of tubes (or non-tubes.)
Let's imagine you had for the gain (right side) tubes
* 2 matched KenRad VT-231
* 2 matched Melz 1578
And for the (left side) followers
* 2 matched Raytheon 6SN7WGT
* 2 matched Sylvania 6SN7W
You could compare ... again, with no delay, just by flipping switches:
Combos:
* KenRad + Raytheon vs Melz + Sylvania
or
* KenRad + Sylvania vs Melz + Raytheon
or
* KenRad + Sylvania vs Melz + diff-buffer
etc
Hold gain side constant:
* Melz + any two of [Raytheon or Sylvania or diff-buffer or passive]
or
* KenRad + any two of [Raytheon or Sylvania or diff-buffer or passive]
Hold follower side constant:
* Raytheon + vs KenRad w Raytheon
or
* Melz w Sylvania vs KenRad w Sylvia.
Of course, you can also try with Sylvanias and Raytheon as drivers, and/or the Melz and KenRads as followers. You may find you like combos that go against conventional wisdom.
WOW ?? Yes, very neat.
But WAIT .... there's MORE!
There's firmware support for automated blind testing!
Pick your "A" combo, and pick your "B" combo .... which can be any of the combos listed above.
Switches let you pick:
* how many iterations you want (5/10/20/30)
* how long each sample plays (5/10/20/40/80/160/320 sec)
* whether your source is "live" or supplied or whether you feed your DAC a sample stored on the Einvigi
* test pattern: AB paired, AB random, or XAB (**)
Say you pick: 10 iterations, 30 seconds, AB random, and you'd copied a 24-96 FLAC of Yo Yo Ma Plays Ennio Morricone as your listening choice for string music. (A nice choice, but Ennio Morricone playing Ennio Morricone on the soundtrack of The Good, The Bad and The Ugly would also be an excellent choice.)
Hit the CALIBRATE button: Freya Einvigi does an internal sound level calibration to match within 0.05 db.
Hit the GO botton: Display shows "1", and beautiful sounds flow into your ears. After the designated 30 seconds, there's a 2-second gap. On a copy of the Data Form that ships with the product, you write down whether you thought sample 1 was "A combo" or "B combo", and sample 2 starts up.
After listening to all the samples ... 10 iterations x 2 per iteration for an AB x 30 seconds ... so approximately 10 minutes later in this hypothetical example ... you hit the SHOW SEQ button, and Einvigi's minimalist display steps through the random pattern .... 1 was A, 2 was B, 3 was A, 4 was A, 5 was A, 6 was B, etc. You score yourself ... 13 of 20 correct. A handy-dandy table printed on the Data Form tells you that the probability of getting that result randomly is 13.2%. The frequentist interpretation would be that you have not convincingly demonstrated that you can consistently differentiate between what's in Combo A vs Combo B. Whether that's good or not depends on what you were expecting. If the Melz sounded to you like the differential buffer ... hmmm ... let's just say you have ears and/or listening skills that allow you to be a very thrifty audiophile. (This could be me, I don't know, I don't have an Einvigi to help me.)
But WAIT ... there's MORE!
Instead of having a Thunderdome trial of your tubes, you can also AB-Test two different sources ... like 2 different DACS. The Einvigi will send the same stored sample to 2 DACS, and randomly switch between them. (Or you can feed your own source to both.)
But WAIT ... there's MORE!
Most of us don't spend most of our time doing rigorous AB-testing. And don't want to.
Freya Einvigi allows you to switch between two different sets of favorite tubes, without the hassle of pulling hot tubes and putting the other set in.
You like KenRads with some music and NOS TungSols with some other kinds of music, but sometimes it's too much hassle to change? Or you're lucky/clever enough to have some Melz 1578s, but you only use those for very serious listening time? Einvigi is your answer, lazy guy or gal.
Other details:
* In test mode with tubes, all the tubes in play are heated, allowing instantaneously switching.
* but in "normal" operation, you only heat the tubes that are in use.
* internal storage of a sample is necessary ... or at least very helpful ... in allowing you to replay the same tune over and over and over. Per my understanding, that would be Best Practice for AB-testing.
* Einvigi stores several 10,000 strings of (pseudo) random numbers. One for paired AB, one or more for the longer "random but balanced" sequences.
It this some crazy Schiit? Yes it is.
Do you want it? Yes you do.
-------
(*) Not an actual Schiit product announcement. Purely speculative.
(**) I'm not well-studied on AB-testing. I'm using these terms here, which may not be "industry standard":
AB paired: for each pair of tests, the sequence is either AB or BA.
AB random: for say 10 iterations - in total there will be 10xA and 10xB, but they won't necessarily be paired.
XAB: plays A or B as the reference first, then random A/B and random A/B, you have to decide which matches the reference.
to all sacred 
cows 
product? 
