hans, i like your writing/reviewing style. 

Well, it is 2 delta-sigma DACs with similar filters (standard amount of pre and post-ringing) compared to 2 R2R NOS DACs, one of which uses unknown chips and supposedly some special glue logic to make them work right for audio. I wasn't surprised the Mini-I and X-Sabre sounded more alike than comparing either against the Hex or NOS1704 (makes sense), but I have to say I was surprised at the differences between the NOS1704 and Hex. I'm very glad I revisited the 1704. I'm not sure if the 1704 chips are just well-suited for NOS mode or if this is a particularly good implementation, because it measures well compared to most other NOS DACs and seems to show in listening tests.

Personal preferences are fascinating to me, and I hope my writing is clear enough that even if someone has completely difference preferences from me, they'd still be able to easily get an idea in their head what they might think based on their preferences relative to mine. I did indeed read your Mini-I review after I gave it a listen and wrote down my thoughts. It's good to see we largely agree and yet interesting to see how that plays into what we're personally looking for when it comes to music playback.

What I'd really like to do is get my hands on some DACs that pirates would generally consider poor sounding to see if my hearing is actually any good or not.

Gamma2 - Various Filter Impressions and Thoughts

I have OJ's Gamma2 on hand (thanks, OJ!). I was told the hardware upsampling is bypassed and that USB is limited to 16/44.1 and 48 (also has a sweet 16/32 mode, according to Windows, which is a nice change from this HD audio baloney). I was excited to try this DAC out not just because I had heard good things about it, but because it lets you select from 3 different digital filters on the fly. I wanted to test the filters out before comparing with other DACs, both out of personal interest and to get a good feel for the Gamma2. As of writing this, I still have not compared it against other DACs.

Now, I will admit that I had already looked over the Gamma2 product page, including the filter details, in the past. I believe the last time I looked was around the first time OJ mentioned loaning me the DAC, so, roughly 2-3 weeks ago. I had been there a few times prior in the few months before that as well. I did retain some of that information in my mind, but I couldn't remember exactly which filter corresponded to A, B, or C when I plugged in the Gamma2 (was roughly 65% sure I remembered the "mapping" in my head), nor could I say with 100% confidence I remembered what one of the three filter types was (was 75% sure it was minimum phase, post-ringing only, but did remember the other two).

Therefore, this test did have some level of bias involved. Keeping that in mind, here are my thoughts and impressions that I gathered while listening to the three filters. I tried my best to forget what I saw on the website leading up until then! 

Filter A

Filters A and B were a bit hard to differentiate, though C was noticeably different than both. A sounded the smoothest, most laid-back of the three, though occasionally mushy sounding in comparison to B and especially C. At times, it had the most bass warmth to the sound. It was the least resolving of the three. A and B had the best tone/timbre compared to C, IMO. A and B also had the best sense of body and presence to the sound. Both A and B had a better sense of rhythm, slam, and attack than C but were definitely more laid-back up top (less pop). Drums overall sounded best on A and B, but C did a better job letting cymbals come through. I found A and B to be more engaging than C, along with being easier to listen to over time. A was the most condensed/cohesive sounding filter in terms of space, soundstaging, and layering.

Filter B

Again, it was often hard to pick up differences between A and B, as the differences were subtle and didn't always show. B had a better sense of clarity and detail than A. It was better at resolving details than A, but not as well as C. It had a slightly better sense of layering and spacial characteristics than A, but not as much as C. In terms of smoothness, tone, timbre, slam, bass, etc., it is much closer to A than C. With rhythm, transients, and attack, B sounded the least sloppy of the three in most cases, in that A just sounded a touch soft and C had a flatter, less dynamic sound with a slight sense of artificial space or echo (IMO) in everything. This was similar to what I noticed when comparing the D/S DACs to the R2R NOS DACs earlier, but I'm really just speaking of relative, not absolute, differences between filters.

I think B had the best bass texture and pitch. A sounded a touch less lean than B, but also not as clear. Vocals sounded the best on A and B, but which of those two sounded best at the time largely depending on the source material. B had cleaner sounding vocals.

Filter C

Filter C was the best at resolving details, espe cially those in the background, and had the best sense of space and layering. It was a noticeable difference compared to the other two. It was the most lean and artificial sounding filter as well. It lacked body and presence relative to the other filters, which made drums sound a touch thin and small in comparison. It did well with cymbals, though, but I often preferred the timbre of cymbals on A and B. C was clean sounding, but a bit flat in terms of dynamics, attack, and transients compared to A and B. I found C to be the least engaging and most fatiguing filter of the bunch.

The best way I can describe the difference with attacks and transients again deals with drums. On many tracks, B had the strongest, cleanest sense of attack with snare drums. A was similar, but a touch softer and less defined. C had the softest attack. On the other hand, C did the best job presenting and sustaining the upper details of the snare drum post-attack, and that too could give one a better, but different, sense of attack at the expense of the drum itself sounding thinner overall and having a softer initial attack. With cymbals, they most often seemed to benefit more from the C filter than A or B, unless hit really hard and suddenly against blackness, but A and B seemed to have better transient control of everything not in the upper spectrum. So, you can think of it as the A/B vs. C filters being more and less laid-back than each other, but in different ways...sort of.

Remember, differences here are ultimately very subtle, to the point where much of it could potentially be placebo.

Going Back to the Gamma2 Site...

There was no way I could have done this test in a completely non-biased fashion by myself, since I had viewed the specifications online prior to knowing I'd be able to get my hands on a Gamma2, but I did what I could aside from giving myself drugs to make me forget what I knew about the device. Perhaps I missed a fun opportunity there, but I'll live. So, for what it's worth, here is what the specifications show online:

Filter A (short pre and post-ringing, but at higher amplitude)




Filter B (No pre-ringing, longer/stronger post-ringing than A but about as long as post-ringing on C)




Filter C (Longer pre and post-ringing than A, but at lower amplitude)



Take this all with what you will. I ended up slightly preferring Filter B over the others, though it was a close call with A. It will certainly make comparisons with other DACs interesting/more difficult, but I love having the filter options available.

Gamma2, Mini-I, and X-Sabre Comparison
For this comparison, I used filter C (linear phase, pre and post-ringing) on the Gamma2 to try and match it as best as I could to the Matrix DACs, which both use pretty standard linear phase filters. I will have to double check when I take measurements, but I think the Gamma2 might have less ringing. Either way, I didn't have the patience to compare with all three filters, but I figured someone could take my comments on the various filters, combine that with my comparison here, and get a rough idea how the Gamma2 would compare to the Matrix DACs on all three filter settings.

The first aspects I focused on were differences in treble presentation. Out of these DACs, the Mini-I has the most “hash” to its treble, and it often seemed as though it had a touch more energy up top than the Gamma2. The X-Sabre was very clean and smooth sounding overall, especially in the treble, but it had a sort of hardness to the treble that I grew sensitive to rather quickly. I preferred its tone and timbre the least. The Gamma2 often had a slight bit more of a “hashy” sound to the treble over the X-Sabre, but it also had a very smooth, fatigue-free quality to it. It was natural sounding relative to the other DACs, though it may have been slightly worse at resolving details compared to the Mini-I…maybe.

The Gamma2 had the least amount of air and was the darkest sounding of the DACs. I thought it had a nice, black background, but it was less spacious and more condensed sounding than the other DACs. This was to my preference, though, as it simply sounded most natural to my ears. I do want to emphasize again that I found the Gamma2 to be noticeably more natural and smoother sounding overall than the Mini-I or X-Sabre.

What surprised me the most about the Gamma2 was how dynamic and powerful it sounded despite its size and running off USB power. This thing could really kick hard! I thought it had a better sense of body, presence, slam, and bass texture/pitch than the other two DACs, and I also found it to do best with transients and attack (filter B still wins out here). The X-Sabre was the leanest of the bunch, but comparable in dynamics to the Gamma2 (hard to tell). The Mini-I was overall the flattest and most laid-back sounding DAC of the three, though, again, it did sometimes have more energy up top than the Gamma2. The Gamma2 was by far the most engaging to listen to.

Gamma2 vs. Hex
I set the Gamma2 to filter B. It was my preference, and I thought it would be the best matchup against the Hex. I’ll be a little be sparse here, because this was a relatively quick test, and I didn’t analyze every possible aspect I could.

Hmm, these two have a surprising amount in common! Without using upsampling on the Hex, it has a smoother, but less detailed, sound than the Gamma2. The overall tone and timbre are pretty close, along with many other characteristics, but the Hex sounds a bit more relaxed and out of focus than the Gamma2. The Gamma2 seemed to have tighter bass, and while it also benefitted from having a tighter grip on the treble, the Hex did have a sort of liquidity to it that the Gamma2 lacked.

Now, you can make the Metrum DACs sound pretty similar to other delta-sigma/OS DACs by utilizing software upsampling. As I’ve shown, this can impart direct, measurable improvements in the frequency domain. Even upsampled, the Metrums will always retain a certain characteristic to the sound, though. I knew I could bring better treble to the Hex through software upsampling, but I wanted to retain the transients and full-bodied nature of the sound. However, I also wanted to match it up with the Gamma2. A minimum phase filter was an obvious choice, and HQPlayer has plenty of options to pick from! I ended up going with the basic IIR filter, as its post-ringing characteristics measured the least impact on the Hex.

With this form of software upsampling, the Hex and Gamma2 may as well have been brothers. The bass on the Hex tightened up and kicked a touch harder and cleaner, and treble details came through more clearly. Overall, it had a more focused, clearer sound. The Gamma2 was a bit more spacious, but the Hex had a natural presentation in that area. The Hex retained that liquidity and brought a smoother sound to the table than the Gamma2. The Hex’s ability to resolve details was now much closer to what the Gamma2 could do, though it did have a tendency to smooth off that last bit that would make them nearly equal. I slightly preferred the Hex here, but, clearly, neither DAC is meant to be a resolution monster. Overall, the upsampled Hex primarily sounded like a smoother, perhaps more natural version of the Gamma2 (B filter). I love the sense of life both DACs give to music.

Maybe I Should Get One of Those…
I really like the Gamma2. It doesn’t have the looks, build quality, or built-in HP amp of the Mini-I, but it sounds damn good. I really like the selectable filters too. Well, you might not like it the DAC’s sound as much as me, but I encourage you to strongly consider the Gamma2 if you want a relatively inexpensive, smooth-sounding, small-with-big-sound DAC. Oh, and this was just with USB. I’m not sure what other tweaks OJ may or may not have done to this besides bypassing the hardware upsampler. I just wish the Gamma2 was easier to come by, because I could probably live with just that (add a nice PSU and cut off USB power, maybe some other tweaks/mods...).

Great work hans!

Listening to the gamma2 against the Gungnir (amongst others) was quite an experience. I think that says a lot about its great sound, and its great price/performance ratio. It's definitely not a DAC I'm going to leave any time soon. In fact, I'm looking into some sort of USB-to-I2S converter board (XMOS perhaps?) so I can use the gamma2 with 24/96 material, since I still have a bit of that in my own library. And I actually do have a dedicated power supply (sigma11) on hand.

How do you think the gamma2 compares in terms of retrieving low-level information (detail, plankton, whatever you want to call it)? An in depth comparison there would be interesting for myself