Holo Audio - Spring DAC - Level 3 - "Kitsune Tuned Edition" - Impressions & Reviews

IME, it's less "break-in" with DACs as opposed with leaving them one for a period of time, at least two days, or a week. I never turn off my DACs. What break in does do over time, is decrease the warm up time requirement, in some cases significantly. The 20 year old Theta Gen V I had required almost no warm up time to sound good. Also, most of the changes take place early on, not at the tail end.

While there can be too much "detail", there is no such thing as not enough microdetail. In terms of the micro stuff, to put things bluntly, the Holo is simplifies and drops musical information. I'm used to certain level or performance, and the lack of micro stuff causes my ears to strain. At the end of the day, I can't really say what is more natural because this is different for different people.

In this case, I don't find the Holo to be necessarily more "listenable" like the old UltraAnalog DACs of yore, the best implementations of had a super liquid and fluid quality like vinyl. Despite the Holo's sweeter treble quality, it does have slightly brighter presentation and straighter and cleaner lines. The nature of these DACs are even more apparent with speakers.

BTW I moved the DACs to my most resolving speaker setup. At most 10 seconds off for each DAC. Re-warming things up. System consists of custom EC 45 amp, Moth Cicadas + Tweaked Hsu STF subwoofer, Theta Data III CDT. The Theta is feeding the Holo via AES. The Gungnir Multibit only takes coax.

 

@Marvey It's a sort of intangible thing for me that is near impossible to explain if you don't fall into the broken-brain camp I do. I'd say with 99% of oversampling DACs I've tried, there's something there that fatigues me given enough listening time, even the ones that are classified as easy to listen to. That SFD I have does it, for example. I get about 1-2 hours in with OS, maybe 3 if lucky, and want to quit. It's at like a subconscious level, not something you can subjectively explain. NOS doesn't do that to me, and this is after experimenting with various headphones, amps, and sources. It always comes back to that. It's not a matter of brightness either, because high sample rate content on NOS DACs still allows me listen for however long I wish.

Interestingly enough, the Gungnir Multibit is one of the few OS DACs where I actually felt fairly comfortable listening for long periods, just still not with the same level of ease as I get from NOS. Something about OS, something I can't explain, keeps me on an uncomfortable mental edge.

Same story from me over and over. Trying to find the best compromise I can so I can actually enjoy music for as long as I'd like, rather than feel like there's a limit for me, even if that means sacrifices in some areas. And almost no one else falls into that camp, so I try to steer basically everyone to OS DACs.

 

I bet it's information overload - spatial information - in conjunction with headphone use. Like how my playing Clash of Clans in the car gives me a headache.

 

The differences I've cited are not huge. These DACs sound more similar to each other than not. Thought I should mention this so people didn't get the wrong idea. Spider chart forthcoming in a few more days will make this more clear.

 

It really must be, and I've had "it," this mysterious fatigue, happen a few times in, say, really loud college bars with a million things going on and a million people talking all at once. However, loud concerts, no issues, likely because I can focus on the loud ass music and count on that to drown everything out that would otherwise split my attention too many ways.

I can do speakers with oversampling just fine, and I enjoy or even prefer it for short term, super critical headphone listening at times too. I expect very few will have this issue. I don't encourage people to go NOS because of salesmen, buzzwords, hype, claims of it sounding like vinyl, etc.

Well, that and there are some particular NOS traits that I genuinely enjoy, even at the expense of others. It works out OK for me.

 

One last commentary before I revisit in another day or two while the DACs stablize their resistor and transistor temps.

On soundstage, nothing unexpected from NOS. I couldn't tell jack on the HD650 and Jotunheim. Both these components don't portray depth well (the HD650 + good tube p + good vinyl does.) Essentially, from the headphone rig (using the Marantz CDT, which is not as good as the Theta Data III), headstage was the same more or less between these two DACs: upfront with little depth. Keep in mind that I am a headstage retard anyway.

On speakers and my custom amp with the 45 DHTs with the HF heaters, the difference is very noticeable. I've mentioned before, the Gungnir Multibit stage is upfront. You are in row 5. The stage is fairly but not overly wide and there is a great sense of depth. (I strongly disagree with people who say the Yggdrasil / Gungnir Multibit lacks stage depth). Combined with ambient cues, you can hear the size and get a really good sense of the venue of the recording.

With the Holo (NOS), you are in row 1. Width is slightly narrower, but depth is flat. Totally flat. It's like a flat 2D stage. There is no in front or behind layering. All the instruments are on the same plane. There is also no ambient information. No up, down, way back wall, ceiling, etc.

Although the audience here is mostly IEM and headphone users, I figure I'd include my speaker observations because I know they translate for you guys with higher headstage IQ than me.

 

That also goes back to a common NOS trait I mentioned. Staging is difficult to get right with NOS. Usually pretty claustrophobic. You're lucky when you get good width. It's layered depth and height that seem trickiest for NOS. These elements and plankton are where NOS suffers the most. I try not to make a secret of this.

OTOH, I'll keep repeating that for all the depth and layering OS provides, every little element on each layer 2D and flat to me (think those shooting games at fairs with various layers front to back but with 2D popups you shoot), whereas NOS makes individual elements sound more 3D in and of themselves...at the expense of crowding each other out and making the overall stage sound very forward and flat. You could also call this more "body" on the instruments, vocals, and farts. I know some or most will disagree.

This is just with headphones. I am still developing my critical listening skills with speakers and find that more difficult (crappy rooms don't help).

 

Holo Audio - Spring DAC - Level 3 - "Kitsune Tuned Edition" - Impressions (Limited)

One thing I would like to first make clear is that the Holo is a very good DAC. I feel that this needs to be said to put any criticisms of it in the proper context. What we do here is nitpick, especially for the better sounding products. When something sounds bad, it’s much easier to succinctly say “this unacceptably sucks in X and Y ways” and then move on.

The Holo was fed from two CD transports: A modified Marantz 5004 and a Theta Data III. I do not use USB for serious listening. For less serious listening, I use a Sony ZX2 and my Andromeda IEMs. Any digital audio that goes through USB exhibits varying layers of suck, at least when compared to CD transport or on the computer end, non-USB to SPDIF solutions such as the Rednet (Ethernet to SPDIF) or Lynx (PCI-E to SPDIF). These CD transports are still the best sounding sources, this is why I used them. DAC sources will account for between 25%-45% of a DAC’s sound. Although difficult enough to believe, those who have screwed enough with this know this to be true: digital transports or computer sources will have a sound that carries forth across DACs.

The reference DAC used in the comparison was the Gungnir MB. I’m too lazy to haul the Yggdrasil around and worry about power cycling / warm up which can take up to a week for Yggdrasil. The Gungnit MB warms up in three days or less. In any event, if the Gungnir Multibit was using the Gen 3 USB interface of the Yggdrasil, it should essentially sound the same from 100Hz up, with the Yggdrasil being a bit more punchy, weighty, and faster down low. But haha, we’re not using USB. We are using SPDIF, so essentially the same applies (the Yggdrasil has AES input, which does sound a little bit better than SPDIF via coax). The Data III transport has AES out, and the Holo was used with AES from the Data III. The Marantz only has a coax SPDIF output. The thing that needs to be said is that the Gungnir MB sounds quite different from a good CD transport compared to USB. The sound tends to be more fluid and more focused at the same time. In other words, better, not different as I usually hear it with the so-called USB de-fuckifiers or converters in the market today.

Only RedBook 16/44 content was used for the test. I own a few DSDs and have some ripped SACDs, but 99% of my music exists in 16/44 and the ancient “hires” format known as vinyl. I don't have the patience to dick around with DSD or HDCD. Besides these are resistor ladder or resistor string DACs. They don't do native DSD. Two systems were used for the evaluation. A headphone based system using KISS modified Sennheiser HD650s fed balanced from the Schiit Jotunheim, which was in turn fed balanced from both DACs. SPDIF out from the Marantz CDP was used to feed the DACs. The second system centered around a custom paralleled 45 tube amp (with Monolith ISTs and Lundahl double C-core OPTs. The laserdisc based Theta Data III transport was used to feed both DACs. In the case of the Holo which has an AES input, the AES output from the Data III was used. The amp was used to drive the HD650s and Moth Cicada 8” wide-banders. A modified Hsu STF series subwoofer was used to augment the 8” wide-banders. Speaker level inputs were used and crossover frequency was set to 38Hz for the best blending with the mains.

The DACs were left on for four days. In the case of the Holo, it was almost always flipping bits, with the CDT on repeat playing music into it. The majority of the changes in both DACs were heard in the first day or two. Both DACs seemed to reach stability after two or three days of being left on. Feel free to ask me more detailed questions about this process in the thread because I don’t want to waste space covering this topic here. Finally, the Holo was evaluated in only NOS mode.

I’ll start by saying the things that the Holo does well. First of all, it’s an R2R DAC. That means no sigma-delta hashy garbage, treble glare, or stridency. This is totally welcome news: more R2R type DACs in the sea of delta-sigma DACs. The treble timbre is more natural than delta-sigma DACs. What’s most striking about the Holo is the “blackground”. The background canvas upon which the sounds are painted is jet black. Perhaps the best I have heard since the PWDmk2, maybe even better. There is super clarity across with board with fine incisive lines straddled with space between crisp transients. This is good thing: we are not smothered by a blurry wall-o-sound. I mentioned this in my early impressions, but there is an enticing sweetness to the treble rendering, a pristine clarity that reminds me of the 45 tube. The Holo is most certainly not a thick, warm, syrupy sounding DAC like the PCM1704 based stuff. Those seeking this kind of sound should go elsewhere.

The macrodynamic capability was not unexpected, but the detail, focus, and speed make this unlike any other NOS DAC I have heard before. (I have not heard the Metrum Pavane). Once full warmed up for a few days, the Holo calms down and any stridency issues go away. A similar thing also happens with the Gungnir MB. When not fully warmed up, the Schiit DACs exhibit in the treble a sandy "ch" sound and the Holo a sharp "s" sound. These timbral differences still exist between these two DACs even when they are warmed up, but they are nothing annoying in the sense of delta sigma hash or glare. Where the Holo handily beats the Gungnir MB is with its “blacks” and clarity”. The Gungnir Multibit’s slight greyness or haze has always been one of its greatest weaknesses (I am surprised to this day that few other people have pointed this out. This is actually easier to hear on headphones than speakers too.)

A few people have mentioned that the Holo seemed like a bright sounding DAC. My wife said this too, at least in relation to the Gungnir Multibit. I feel it’s a bit more complex than that. The more incisive straighter lines do contribute to this perception. But at times, the Holo can be rather warm sounding with a strong mid-bass punch. But what it really comes down to is insufficient extension down in the lowest two octaves. This is beyond measurements. On the Gungnir Multibit, there is a strong sense of foundation to sounds. The Gungnir Multibit has more sustain, heft, and explosiveness to how the very low end is rendered. On speakers with sufficient low end extension, this is very noticeable, especially with low frequency sounds that straddle the border of audibility and palpability. The Holo seems to just give up here. The energy seems missing or muted. On headphones, the low end from the Holo is a bit more decipherable (probably because of lower distortion compared to speakers), but it is rendered in a kind of mushy and ill-defined way. The start-stop bass on Daft Punk’s Giorgio Moroder sounds like a series of lightweight farts. I guess the good thing is that it's rendered softly so we can’t really hear it.

The last few areas where the Holo lags behind the Gungnir Multibit are microdetail, microdynamics, and soundstage. This is not unexpected. To be fair, the Holo carries with it a NOS penalty (along with NOS bonus points). I think @Hands does a great job in describing the pros and cons of NOS DACs here: http://www.superbestaudiofriends.or...-impressions-reviews.3172/page-11#post-101411

So in that sense, the Holo really does an exceptional job given this penalty. Nevertheless, I will elaborate. As I mentioned above, in terms of macrodynamics and transient response, the Holo does not lack liveliness or speed. However, where it fails to keep up with the Gungnir Multibit is with microdynamics: the rendering of minute up and down modulations, the waverings, the tremblings with sounds during vocal and piano sustains and decay. The Holo seems to simplify sustains or decays, without the minute volume modulations and nuance of the Gungnir Multibit. If anything, there was tendency to compress and even sounds upward in loudness, which I think can make it sound more exciting for a lot people. This was easily noticeable in vocals (Alicia Keys and Adam Levine) and piano on Alicia Key’s Unplugged. In essence, the Holo lacks the immediacy or as my wife puts it, the “more real”-ness of the Gungnir Multibit. To be fair, the Holo is still no slouch in this regard, but it just doesn’t quite illicit the goose-bump effect of the Gungnir Multibit. My wife adds "flat" or "muted". Closely related to microdynamics is “plankton” or microdetail. The Gungnir Multibit presents much more ambient information, both timbral and temporal around rendered sounds, guitar plucks, percussion hits, anything. With a capable recording, the entire soundscape gets flooded with musical information from the Gungnir Multibit. Switching back to Holo makes me strain to hear this missing information for a while, until at least I get acclimated.

Finally there is headstage and soundstage. Nothing unexpected from NOS. I couldn't tell jack on the HD650 and Jotunheim. Both these components don't portray depth well (however, the HD650 + good tube p + good vinyl does.) Essentially, from the headphone rig (using the Marantz CDT, which is not as good as the Theta Data III), the headstage was the same more or less between these two DACs: upfront with little depth. Keep in mind that I am not sensitive to headstage.

On speakers and my custom amp with the 45 DHTs with the HF heaters, the difference is very noticeable. This amp is all about depth and space. I've mentioned all this before: The Gungnir Multibit stage is upfront. You are in row 5. The stage is 15 feet in front of you. The soundstage is fairly wide with ambient information coming from beyond the left and right side of the speakers. There is a great sense of depth. (I strongly disagree with people who say the Yggdrasil / Gungnir Multibit lacks stage depth, but I can see this to be the case with headphone listeners). Combined with up down left right ambient cues, you can get a really good sense of the venue of the recording. With the Holo (NOS), you are in row 1. The soundstage is well between and about half a foot in front of the speakers. There is also no ambient information beyond this limited space. No up, down, way back wall, ceiling, etc. Once the Holo is warmed up for three days, the depth of the soundstage is not as 2D flat as in the first two days, but it only extends merely a few feet back at best. Because of this, front and behind layering is limited. Although the audience here is mostly IEM and headphone users, I figure I'd include my speaker observations because I know they translate for those folks who are able to decipher soundstage cues in headstages.

That pretty much sums it up. Again, this is really good DAC. If I had nothing to better to compare it to, it would sound absolutely great. Also, I would hope that folks don’t get hung up on the comparative weaknesses. This is most certainly a case where the sum of the attributes is greater than the individual parts – think Beatles (but not Pink Floyd, ugh). Besides, it’s got that NOS charm, which I know many of you are big fans.


P.S.

I will come back in a bit and draw the localization of instruments and vocalists to best explain soundstage differences. Pictures are worth a thousand words. See this: http://www.superbestaudiofriends.or...-impressions-reviews.3172/page-13#post-102176

 

@Marvey: okay, that was a hugely dick move making the review sound super positive when you clearly don't like this thing and you think the Gungnir multibit is a vastly superior DAC. You're super contradictory in this review and I find it really disappointing.

1. You leave out the HUGE price differential
2. You only mention a couple of positives:
   1. there's no treble garbage (then you go on to say this is true for all decent R2R in your experience) and the treble timbre is excellent
   2. it has an amazingly black background with great clarity
3. Then you list of some pretty damning problems: poor bass extension, no microdetail, weak microdynamics, and absolutely garbage staging

 

Stage was compact for me in both width and depth, but I am primarily a speaker listener and don't notice headstage differences. For that, I would defer to others.

Soundstage diagram pending...

I will address Ravi's concerns separately at a later time.

 

@gbeast Read all my comments about OS vs. NOS. For NOS DAC impressions you have to speak more relative to other NOS DACs than OS DACs due to inherently different presentations. However, my general OS vs NOS thoughts, both in this thread and others, should explain where I'm coming from if you are comparing the two types.

Otherwise, I listen almost entirely on headphones. Maybe my thoughts will change as I work on my speaker setup more down the road. There are many possible reasons why I might have heard, or at least thought I did, otherwise. Too many to list.