Or if possible use the Toslink or Coaxial Modi inputs.
Toslink has the most jitter and you have to have a device which outputs Toslink Optical.
Coaxial output is impossible to come by unless you have a $200 sound card in most cases.
There is only one $70 sound card which has Coax output if you install dual PCI-E plates in the back of your desktop (requiring you to have a desktop that has extra slots).
Or you could spend $60 on a SMSL xUSB and then output to Coax but.... all of that adds considerably to the cost when you could just buy a D30 that is done right and forget about it.
Here's a tip on a nice DAC that people don't often consider. I was at the Schiitstorm HeadFi meet in Gaithersburg MD a few months ago. I brought my Focusrite 2i4 2nd gen USB audio interface. Most people there had not thought about using one as a "DAC" as it is normally a pro-sumer audio input interface for recording music. Turns out it has a DAC that can go high speed (192k) and sounds great. It even has single ended and balanced outputs. Never pops. A real volume knob, and a basic headphone out (small amp) with separate volume knob. Plus, it can record as well. Anyhow, many HeadFiers there listened to my amps on it and said it sounds really good. Up there with some of the best. Since it has inputs - I could measure its performance in loopback. Noise floor is -130dB, -85dB stereo separation, THD is -110dB second order dominant with a bit of third, fourth and fifth at much lower levels (-120dB). I use the balanced outputs to drive two of my XRK Audio Pocket Class A amps (standard single ended amps), to power a balanced headphone. Works great!
Doesn't cost much either. Nice footprint and built like a tank. Cool red anodized aluminum box too.
Those cost around $150 and it does have quite a low end Dac Chip. I do have a friend who has one and he likes it though.
I'm not sure where you got your info on or have even heard the MHDT Pagoda and Metrun Musette but these are not "very warm" sounding and are quite extended in the highs with great resolution. I didn't list it but we also used a Bimby and D30 (my friend got it with the A30) as well but the results were the same with the O2 and A30 being brighter and thinner sounding than the Magni3. The most notable difference was with the HE560 where the Magni3 had similar sub-bass that was much like when its driven by the LC and Lyr2 but was lacking with the O2 and A30. This makes me think that you might have a defective Magni3 or another issue somewhere in your audio chain to make the Magni3 seem brighter and have less sub-bass.
I read on another forum about the Pagoda and Musette. Its quite well known that NOS DAC's have a warmer sound.
Lyr2 is a tube amp, also something known for being warm and modifying the sound (like any tube amp).
The Magni 3 is definitely brighter than these other 2 Amps and it has a sharp response in the high end.
I tried it with 2 DAC's and 5 pairs of headphones, I'm not sure how you would have literally the opposite experience from me. Unless you are so used to warm sounds that you don't know what neutral sounds like, and even in that case you would still notice the Magni having sharper highs vs the other 2 Amps.
It also depends on how you have the Amps wired, because if you are using anything via the Magni's Pre-Amp outputs then the volume will be attenuated basically always to whatever is next in the chain unless Magni 3 Volume is at 100%.
I went from the DX7 -> A30 -> Magni 3 -> JDS o2
I did D30 -> Magni 3
D30 -> A30 -> Magni 3
DX7 -> Magni 3
and DX7 -> A30 -> JDS o2
They all came back with the same results. But it is worth noting that the A30 especially has a higher output impedance which could definitely affect frequency response depending on which headphones you are using.
When I use my A30 with my Denon D7200's it definitely changes the sound a bit, but this doesn't happen on any of my other headphones (Denon AH-MM400, Panasonic RP-HD10, 1More Quad, 1More Triple, Yamaha YH-2).
But then when I use the JDS o2 it sounds fine as it has a very low output impedance.
The D7200 has alot of sub-bass and it was noticeably reduced immediately with the Magni 3.
Its also worth noting that my A30 is Rolled with a LM4562 vs the stock OPA2134.
So its possible it has to do with headphone pairing specifically, or the Magni 3 tolerance per unit could be semi-high where they have different frequency responses.
Or it could be the output impedance of the DAC's you used, although the D30 has a higher impedance also so the result would be similar to my result if not exactly the same...