Introduction
I’m still new-ish to the tube world, but bought one of those cheap Chinese tube buffers back in August 2019 along with upgraded tubes and was really impressed with the results. After hearing from a variety of sources, including @ZeosPantera, that the Darkvoice 336SE was the budget-ish standard bearer for what tube amplification should sound like (which is not saying it’s the BEST tube amp, just a good example of tubey-ness), I started getting really curious if this little tube buffer offered a true taste of tube. In other words, does this little tube buffer give a listener an accurate representation of tube sound? If so, it gets a user into the tube world for much less money than a Darkvoice AND leaves that user with a pure solid state option while still staying firmly in the budget realm. Well, when Drop offered the Darkvoice at $185 on the most recent drop, I snagged one. It arrived shortly before the COVID lockdown and I’ve been listening to it a lot for about 3ish weeks now. I’ve even had a chance to try some different tubes!
With this sorta-review-shootout post, I hope to answer 2 questions:
-
Do these little Chinese tube buffers give a true tube flavor to the sound?
-
What are the sonic differences between a tube buffer and the Darkvoice 336SE, if any?
Notice I intentionally did not use the word “better” here. I’m not really interested in “better” between the 2 in this post, just if the buffer gives a legitimate tube experience.
For the impatient, the answer to Question 1 is YES. The tube buffer, as compared to the Darkvoice, did inject a legitimate tubeyness into the music. The answers to Question 2 are that there are several subtle differences. In general, the Darkvoice with the tubes listed below had more midrange detail and clarity resolving the reverb on things like vocals, acoustic guitars, and pianos with more definition. The tube buffer arguably did more “tubing” of the sound by smoothing out, but not removing or losing, those subtle details. The tube buffer also had greater frequency extension in both directions, almost always providing more sub-bass impact. More detailed findings will follow for those still curious. Or, you can scroll down to my discussion section where I talk about my recommendations for what potential users might think about if they’re looking to try tubes and are on a budget. Fair warning, you may be reading for awhile…
Gear and Method
- Darkvoice 336 SE headphone amp with Tung-Sol 6SN7GTB pre tube and RCA 7613 power tube.
- Douk Audio Dz Tube “Preamplifier” (really buffer) under Nobsound branding: https://www.amazon.com/Nobsound-Vacuum-Preamplifier-Digital-Control/dp/B01NBM5CTN/ref=sr_1_8?crid=30RRWEE13CGML&dchild=1&keywords=nobsound+tube+preamplifier&qid=1585945026&sprefix=nobsound+tube+pre%2Caps%2C201&sr=8-8 and with GE JAN 5654W tubes: https://www.amazon.com/gp/product/B01BFJDTW0/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
- JDS Labs Atom headphone amplifier
- SMSL SU-8 DAC
- SMSL SP200 headphone amplifier
- Beyerdynamic DT880 600𝞨 headphone
- Massdrop + Sennheiser HD6XX headphone
- Massdrop + HiFiMan HE4XX headphone
There were 3 “layers” of testing. The SU-8’s unbalanced signal was split with one signal going to the Darkvoice and the other going through the tube buffer and into the Atom’s 3.5mm input. The Darkvoice’s pre-out was also connected to the Atom’s RCA input. The first “layer” of testing was tube sound vs solid state. The SP200 (connected to SU-8 via balanced) and the Atom when not playing the tube buffer sound served as the example solid state sounds for comparisons of tube sound vs. solid state. The second layer: I listened to the DT880 and the 6XX powered directly off the Darkvoice and the buffer+Atom. I also did a hybrid test with the Darkvoice pre-out into Atom for a more direct comparison of “tube pre vs. tube pre” using the 4XX. Call this the third layer. The general method was to listen to a song all the way through, then switch amps and listen to the same song all the way through again - taking notes on the similarities and differences I heard. In the case of the hybrid test, the same method was used except I used the Atom’s input switch to switch between the DV preout and buffer signals. The tube buffer here has tone controls. For all testing I left those controls at the 12 o’clock positions. I also did my best to level-match in all circumstances. I don’t have fancy measuring equipment for this, so I did my best by ear. But, this is a limitation in my method. The songs I used will be listed at the very end for those curious, as well as what I was listening for with those tracks.
Answer to Question 1: YES. And here’s why…
Both tube solutions - even in hybrid mode with the DV - introduced a “holographic” effect to the spatial presentation. When compared to pure solid state, namely the Atom and the SMSL SP200, both the DV and buffer fill out the soundstage and tighten up the imaging. The biggest change is on the 6XX, followed by the DT880. When going from solid state to either of these two tube options, the sonic spaces between left and center and right and center fill in. It’s like there were holes in the imaging in those areas and the tubes filled those holes, completing the soundfield. This filling-in happened to varying degrees by headphone and then by headphone + amp combination, with no consistent “winner” here, but the point is they both did it. Since this “holographic” nature is often associated with the “tube sound”, we have a feather in the cap of the buffer for doing tube things.
Both tube solutions consistently “warmed” the sound and increased bass extension. Again, this happened to varying degrees depending on headphone and amp and combination, but the point is again both tube solutions provided a deeper, richer, fuller low end than solid state. To my ear, this extended to the lower mids, causing vocals to sound more natural and less honky or shouty than on solid state amplification. The 6XX again, not surprisingly, benefitted the most from this extension and demonstrated the largest magnitude of bass increase. A second feather in the cap for the buffer doing legit tube things.
Both tube solutions “goo-ified” the sound, which is to say they sounded more relaxed. Solid state sounded “exacting” in comparison. They both sounded less sharp and defined when compared to solid state. They both tended to blend sounds together a bit more than SS. Feather in the cap 3 for the buffer doing tube things.
From my understanding, these are the 3 draws to the tube sound in the audiophile community: a more holographic spatial presentation, a warmer sound, and a more relaxed sound. Check, check, and check for both the Darkvoice and the little tube buffer. I thus claim the tube buffer does give “the tube sound.”
Question 2: What are the sonic differences between the DV and the Tube Buffer?
First the caveats. These comparisons were done without the stock tubes. The differences I’m about to describe will likely change with different tubes, and the stock solutions are likely to be different too. I don’t think these tube rollings change the answer to question 1, but they likely will change the answer to this question more significantly. With that said, on with it…
In general the Darkvoice with the Tung-Sol driver tube and RCA power tube sounded more resolving than the tube buffer + Atom. The DV did a good job of combining solid-state-like resolution with tube-like warmth and character. Guitar string plucks came with the sound of the pluck plus the sound of the vibration. The tube buffer did too, but there was less distinction between the pluck and the vibration. The DV also resolved reverb more clearly than the tube buffer. I’m avoiding using the word “detail” here because I think to say that one is more detailed than the other is misleading. Both tube devices reproduced the same sounds. I really never noticed that the DV could reproduce something the buffer couldn’t. To me, that’s detail - having something the other is lacking. The buffer just didn’t separate the different sounds as well. The different sounds blended together more. I think in audiophile terms this would be described as “fast”. The DV had a “faster” sound. The drivers returned to their initial positions just a bit quicker. The buffer had longer decays to the sounds. Returning to Question 1 for a moment, there is an argument here that in these tube configurations that means the buffer sounds more tubey than the DV. This difference was subtle, though. I really only noticed the differences in decay length for about the first 10-20 seconds a song played the second time through.
The tube buffer produced more bass and treble extension than the DV, and added more oomph to the bass. As consistent as the slight difference in resolution was, the tube buffer routinely added sub-bass that the DV couldn’t match. Simply put, the buffer hit deeper, harder, and rumbled more. That is not to say the bass was sloppy or uncontrolled, just that there was more of it than the DV produced. This difference was most apparent on the DT880. I think the DV struggled with the low end on the 600 ohm can. The DV definitely added bass extension to the DT880’s sound, but could not match the output level the buffer + Atom could manage. The difference was less on the 6XX. Arguably the DV’s bass was more powerful down to about 60 Hz on the 6XX, but then the buffer really brought the sub-bass out more. In the treble the buffer was more extended. This treble extension was both good and bad. On the DT880, that could still inject some sibilance back into the sound. The Darkvoice almost never sounded sibilant with the DT880s. The 6XX though had a bit more sparkle with the buffer. I think the DV also extended the 6XX’s treble response beyond its solid state performance, but not as much as the tube buffer. I’ll again remind here that I left the buffer’s tone controls at 12:00 for these tests. Adjusting those would obviously change some of the differences described here.
The hybrid test (with the DV’s pre-out) provided similar results on resolution and frequency extension. There were three additional differences of note, though. The first is that the DV’s pre-section is much noisier than the tube buffer. With no music playing and the Atom on high gain and cranked to max volume, there’s a noticeable hum coming from the Darkvoice. At normal listening levels it’s just barely audible in a completely silent room, but it is there. The tube buffer had a high frequency hiss that was quieter than the DV’s hum, but not zero. The second difference was that sibilance was occasionally audible with the buffer and the Atom-powered HE4XX. I never noticed sibilance with the DV as a pre, but did with the buffer. So the sibilance I heard with the DT880 was not because of the Atom’s op-amp topology, it was an artifact of the buffer likely not removing sibilance-causing frequencies like the DV pre did. The third difference is the spatial presentation. When able to rapidly switch back and forth by simply pushing the Atom’s input button, the perceived vertical position of the soundfield changed slightly. The DV had a spatial image that extended more vertically than the buffer did, and also sounded like it was physically lower relative to the height of my ears. It was strange. When shifting back and forth it was like the position of a center-placed voice moved up or down an inch when I switched between the two. I didn’t notice this at all when I had to physically pull a plug out of one amp and stick it in another. But it came out here with the 4XX. Again, though, after 30 seconds of a song playing I didn’t think about it again for the remainder of the track.
Discussion
I think these little tube buffers are underrated and need to be paid more attention by the audiophile community. Josh Valour’s review of the FX-Audio Tube 03 was the first review from any of the reviewers we commonly discuss on HiFi Guides paying attention to one. I paid about $50 for the tube buffer I used for this post. The upgraded GE tubes were about $35 at the time. I see they’re cheaper now. The Atom amp costs $99 and already has two inputs. That means I paid $185 for an amp, a tube buffer, and new tubes. That’s essentially two very different and still very good sounding amplifiers for $185. I paid $185 for the Darkvoice on Drop, which was the cheapest non-EBay scam price I’ve seen for it. I spent another ~$65 tube rolling on the DV. The Darkvoice currently lists on Amazon for $289.99. For that price, you can add an SMSL M100 DAC to the Atom + buffer + tubes and still pay less than you would for the Darkvoice alone with stock tubes. The value proposition here is huge.
Buuutttt, wow is there something special about that Darkvoice sound with the Tung-Sol and RCA tubes. It’s ability to combine tube-like sound with a level of resolution and speed more reminiscent of a solid state sound is intoxicating. The 6XX in particular sounds amazing. The DT880 sounds awesome too, just with not quite as much bass punch.
I’ll be keeping both the DV and the buffer. I prefer the 6XX and sometimes the DT880 when they are driven pure tube all the way like the DV does. But as a pre-amp, I think the buffer does a better job. Overall, I like what it did the the 4XX more than the DV. Plus, the tone controls, which I didn’t even use for this test, offer so much more customization. More bass? Can do. Less treble sharpness? Got that covered.
So, if you’re on a tight budget, already have a solid state amp, and want to try the tube sound, you can get one of these tube buffers and be confident you’ll get a reasonably good tube flavor, plus still have that solid state sound available when you want it. If you have a bit more to spend and want a wonderful blend of quality tube sound but without sacrificing all that solid state speed and precision, the Darkvoice with tube rolling is a great choice.
Thanks all for reading this far. Enjoy the music!
Tracks for Detailed Testing:
Everywhere by Fleetwood Mac - the intro has some brush-hit cymbals. Great for hi frequency detail tests. Also, the vocals can sound a bit sibilant. Not a sibilance torture test, but a borderline sometimes it’s there and sometimes it isn’t test.
Pain by The War on Drugs - If you’re not listening to this song, you should. Very active and punchy bassline. Also just a very full, well recorded song.
Lady in Black by Uriah Heep - Lots of reverb in this song, on the vocals, guitars, drums, just about everything. A mid-range “speed” test.
1812 Overture (Tchaikovsky) conducted by Erich Krenzel and performed by Cincinnati Pops - this version has live cannon fire, a vocal choir, and a bell choir. Great for dynamic range, sub-bass detail and rumble, and also imaging. It’s a full orchestra and you can hear where each group of instruments is located on a good system.
Mountains by Hans Zimmer from Interstellar soundtrack - It’s a go-to for dynamic range, but also has prodigious sub-bass. Plus, can your system resolve anything beyond those pipe organ and brass blasts? Find out with this track.
Telephone by Lady Gaga ft. Beyonce - Basically, can it do pop music? But, the female vocal duet is another good midrange test.
Nowhere Now by Steven Wilson - Wilson is awesome at mixing and mastering. His records routinely sound wonderful. This track has quiet ambience and foot-tapping rock and roll all in four minutes. Plus, Wilson’s voice can sound shouty to me on many systems. To my ear his voice is a shouty/honky acid test.
In My Time of Dying by Led Zeppelin - Bonzo’s drums sound somewhere between very good and great on all Zeppelin cuts. On this track though, they are startlingly good. They sound huge, snappy, and intimate. This track is my go-to for drum realism.
I also did a lot of general listening in the various configurations. These were just the tracks I chose for careful comparison.
bravo, nice write-up.
more or less. Because for whatever stupid reason I wanted tubes but a perfect frequency response. But doing this does… the opposite of what you said. Like, all is “in your ears”, clearer than ever, but nothing is in front of you. That’s like the anti-tube amp. Lol. Putting the bass knob like you did, at 12:00 
fixes it, and wow, here’s the “holographic-ness” of tubes (especially Voskhod 6J1P-EVs – apparently), again. Oh well! 
and the treble “off”… that’s a lot of mids, but that works too.