Had no idea Tom had done measurements on the T2. I was wondering when some would be available.
I've built and listened to the T2 and as much as I appreciate what Tomchr has done, people need to understand that he comes from a very different and strongly opinionated design perspective.
Right now I have an Asgard 3, Project Sunrise III (tube hybrid), the T2, and modified cheap JLH-based headamp. I've had all of them in rotation during the build and listening to the T2. All listening done with Soekris 1021 DAC and HD6XXs. My observations (Be warned, *subjective comments ahead. gasp.*):
1. It sounds very good stock. 'Mid-fi' tube amp sounds about right as a descriptor. Classic Class A sound. If you like that sound, you'll like this amp. Definitely not the lowest noise floor or separation - but I think that can be improved upon (see below).
2. I don't experience a turn on thump.
3. Output is indeed lower than typical but I think Mark Johnson's reason for doing so is an interesting choice - minimizing the channel balance impact of volume pots. With my HD6xx/650s this means I only have about a quarter turn or so to blow my eardrums vs. three quarters of the typical headamp. Big Deal.
4. I found a linear power supply (simple LM317 with an LC 'Noise Nuke') to be preferable to the SMPS - although the SMPS does sound very good. The LPS is smoother to me.
5. Since this is an inexpensive DIY and the board is nice and simple, there's plenty of room to try different things to see if you can improve what you hear/tailor the sound to your expectations. Ex: Personally I found the 3.3uf input cap to make the low frequencies a little congested/muddy, so I lowered until I found a value I liked (2.2uf). This opened the sound up for me and widened the presentation/stage.
Secondly, a member in the T2 diyaudio thread recommended bypassing the 3 IN4004 diodes with a large (3300-4700uf) cap to lower noise. I did this and found it did indeed have an audible impact for the better. The combination of reduced input cap value and the 1N4004 bypass IMO improved things considerably (to my taste). Furthermore, the layout and simplicity of the board allows you to play around. Skeptical of the bypass? Bloop, just drop the cap at each end of the diodes. Takes 10 seconds. Listen. Compare. Done. Same with any other parts on the board.
So I guess I would say I am somewhere in the middle in my assessment. Stock, it ranked about on par with the cheap but good sounding JLH-based amp (which makes sense since they share similarities). The Sunrise, which is a *great* amp, was clearer, lower noise, more extended in high frequencies, tubey yum yum. Asgard 3 I would consider more to Tomchr's camp - excellent micro-detail, low noise and distortion, 'neutral' (whatever that means). But after I made those couple of changes, the T2 is about equal with them. Just a different presentation - sort of a mix between the two.
Finally, I think the T2 is another important lesson in how low distortion is not a holy grail. Is it nice to be able to hear an ant walking along the studio floor during the recording? Sure. But it doesn't make or break an amp/speaker/headphone for me.