"If it measures good and sounds bad, it is bad.
If it sounds good and measures bad, you've measured the wrong thing."


Building a reference system - Part 1: Metrum Octave MKII NOS DAC

Posted by Amine Slimani on October 10, 2014 at 5:10 PM

Many audiophiles believe that the source is the most important piece of equipment in an audio chain. Indeed, if information is missing from the source, there is nothing that even great preamplifiers, amplifiers, speakers or headphones can do to recover it and add it back to the chain. Conversely, if there is something inherently wrong with the source (distortion or coloration), there is nothing the rest of the chain can to totally cover it; admittedly, it is possible to carefully choose equipment downstream in order to minimize whatever flaw is present upstream, but by doing so, one is only counterbalancing the flaw partially and is perhaps losing other qualities in the process of doing so.

That is why I personally have always paid a lot of attention to the source in my system, and I would also try to do as much research as possible before getting a source component as it seems to be, in my opinion, the fastest evolving type of audio components (be it USB to SPDIF converters or DACs). Given that I was very satisfied with the performance of my Aqvox powered Audiophilleo 2 USB to SPDIF converter, the first item I decided to upgrade in my system was the DAC: I wanted to have a clean base on which to build upon the rest of the system.

But before getting to the Metrum NOS MKII DAC, I think that it might be relevant to share my experience with DACs in order to allow the reader to better put in perspective where I am coming from and what qualities I am looking for in a DAC.

The first DAC that amazed me with its superior sound quality was the discontinued Audio-gd DAC19MK3. It was an all discrete PCM1704UK based DAC using the great sounding Pacific Mircrosonics PMD100 HDCD digital filter. The DAC19MK3 was not only as much detailed or even more detailed than most reasonably priced DACs and CD Players (i.e. under $3000), it also made many of them sound broken. The DAC19MK3 was the first DAC in my experience that sounded “analog” in a good way. The analog sounding part did not come from false warmth or haziness but rather came from the lack of obvious digital colorations. Non-amplified instruments such as violins and pianos started to sound real for the first time. Since then, I had the opportunity to listen to many DACs, and the R2R PCM1704UK based ones always sounded better, to my ears, than the sigma-delta based competitors. However, given that I haven’t sampled every DAC out there, I am sure that there are excellent sigma-delta based DACs out there.

While trying to improve the sound my audio system, I made the “mistake” of selling the PMD100 based DAC19MK3 and replace it with the newer DAC19DSP which replaced the PMD100 (or DF1704) digital filter with a custom DSP filter from Audio-gd. It took me a while to realize my “mistake” (it can perhaps best be qualified as regret instead of mistake since the DAC19DSP is also an excellent sounding DAC, for its price). While the new DAC was better sounding by objective standards (bandwidth extension, details, soundstage… ), I felt, nonetheless, that after a (very) long term period of listening that something was lost in the subjective realm. Indeed, after months of listening, I started noticing that even though the DAC19DSP was better sounding than many sources I had listened to, it didn’t sound as pleasing as the PMD100 on unamplified instruments such as pianos and violins. How did I know that it was coming from the DAC and not from the rest of my chain? The answer is pretty easy: once I experimented with different upsampling methods, I found that some of the faults that I had noticed, would lessen when upsampling to 96K (Upsampling would have probably yielded better results at higher sample rates, as was the case with Kingrex UD384 DAC I reviewed that unsurprisingly tops at 384K), and that by “relaxing” the customizable DSP filter (to -50db passband instead of -130db passband), the DAC surprisingly sounded better.

Experimenting with upsampling on both the DAC19DSP and Kingrex UD384 made me realize that software based upsamplers (such as JRiver built in upsampler or Foobar’s SoX plug-in) sounded better to my ears than straight feeding 44.1K files into those (oversampling) DACs. The (oversimplified) explanation of that finding is pretty simple: upsampling makes us listen more to the selected software upsampler and less to the stock hardware oversampler built in the DAC chips or their hardware digital filter. Those who think that upsampling cannot have any positive impact on sound and can only degrade data should consider this: most, if not all, sigma-delta based DACs already have oversampling filters which are doing … integer upsampling (8x, 16x… ). So there is already upsampling (called oversampling) and digital filtering going on the DACs whether we like or not. Also, one has to keep in mind that the digital filters that are most difficult to construct are those made for the 44.1K (in comparison with 88.2K and 96K+ frequencies) CD/ Redbook format. Programmers have to arbitrate between frequency domain and time domain performance; that is why we see different types of filters: slow roll-off, fast roll-off, minimum phase, intermediate phase…

As a result, it is not straightforward and easy to construct very good sounding digital filters. Some companies, such as Ayre or Meridian to name a few, have developed some interesting sophisticated digital filters and their products are acclaimed by many audio critics. However, those DACs are relatively expensive and are still based on sigma-delta with built-in oversampling.

After reading a lot about different approaches to building modern DACs, I came to the conclusion that I wanted a DAC that was R2R based (that is the architecture my ears seem to favor) and that was flexible in terms of digital filtering.

Ironically, in my case and according to the criteria I had set for myself, the state of the art DAC for me had to be a NOS (Non-oversampling) R2R based DAC – those technologies were already around in the 1980s. There were quite a few choices out there, from the affordable (often Chinese made) TDA1543 based DACs to some very impressive discrete based DACs (TotalDac for instance). I settled on the Metrum Octave MKII because it seemed like a competently built and reasonably priced DAC. Also, it seemed to be very well reviewed by very different people (although that is no guarantee for sound in one’s system).

Given that the Metrum Octave MKII has been extensively reviewed elsewhere, I won’t go into much detail about its technical specifications. I will jump right ahead to the rest of the review.


Review system:

System 1 – Headphone Based (optimized)

Computer source: HP Pavillon DV6 (Core i7, 8GB), Windows 7 64bits SP1, Fidelizer, JRiver Media 17

Transports: Audiophilleo 2, Kingrex UD384, Jkeny’s modified Hiface MK1, Audio-gd Digital Interface (w/ Tentlabs upgrade clock), Teralink X2, Musiland Monitor 01 USD, EMU 0404 USB

DACs: Metrum Octave MKII, Audio-gd DAC-19 DSP with DSP1 V5 upgrade, Kingrex UD384

Amplifiers: Glenn’s OTL headphone amplifier, Audio-gd C2, Little Dot MKIII

Headphones: Sennheiser HD800 with Wiplash Audio TWau Reference cable, Beyerdynamic T1 with ALO Chain Mail cable, Sennheiser HD650 with Artisan Silver Dream upgrade cable

Digital interconnects: Hifi Cables Sobek BNC, OyaideDB-510 BNC, Wireworld Ultraviolet USB, Artisan Twinline Pure Silver USB cables

Analog Interconnects: Artisan Ultimate Silver Dream RCA, Norse Audio custom 8 conductor UP-OCC ACSS, Deep Sounds SPS ACSS, Kimber PBJ RCA

 Power filtration: Bada LB-5600 Filter, Essential Audio Tools Noise Eater, Essential Audio Tools Pulse Protector, Supra Mains distributor

Power cords: Hifi Cables & Cie PowertransPlus (x2), Hifi Cable & Cie SimpleTrans, Olflex power cords

Vibration Control: Aktyna ARIS decoupling feet, Maple and Acrylic Platforms, E&T rack, Stabren Damping pads, Sandbox, Brass cones, Vibrapod, Yamamoto Ebony footers and Various Herbie’s Audio Labs tweaks


System 2 – Speaker Based (not fully optimized)

Yamaha CD 600/ Audio-gd DAC19DSP --> Kimber PBJ --> Yamaha AS500 / Populse T150 --> Hifi Cables & Cie Maxitrans II --> Triangle Antal EX floorstanding speakers


Reference tracks used for the review:

CD Quality:

Mahler - Symphony n 5- Decca - 16/44

Sol Gabetta - Schostakowitsch Cellokonzert Nr. 2/Cello - 16/44

Vivaldi - Concertofor 2 violins - Carmignola/Mullova - 16/44

Natalie Dessay - Italian Opera Arias - Emi Classics -16/44

Puccini - La Boheme - Decca - 16/44

Glenn Gould - The Goldberg Variations (1981) - 16/44

The Essential James Bond - City of Prague Philharmonic orchestra - 16/44

The Dave Brubeck Quartet - Take Five - HDCD - 16/44

Diana Krall - Live in Paris - 16/44

Norah Jones - Come Away With Me - 16/44

Patricia Barber -Companion - 16/44

Johnny Cash – The Essential - 16/44

Soundrama - "The Pulse" Test CD - 16/44


High Resolution quality:

Rachmaninoff Dances -HD Tracks - 24/96

Mozart Violin Concertos - Marianne Thorsen - 2L - 24/96

Dunedin Consort -Messiah - Linn Records - 24/88

Glenn Gould – The Goldberg Variations (1955) – HDTracks - 24/96

Keith Jarrett - Paris/ London - Testament - 24/96

Jazz at the Pawnshop- HD Tracks - 24/88

Alison Krauss and Union Station - Paper Airplane - 24/96

Ella Fitzgerald /Louis Armstrong - Ella & Louis - 24/96

Diana Krall – From this Moment on - 24/88

Diana Krall – The Girl in the Other Room - 24/96

The World's Greatest Audiophile Vocal Recordings - Chesky - 24/96

The Kinks - One for the Road Live - 24/96

The Eagles - Hotel California - HD Tracks - 24/96

Head-fi/Chesky Sampler - Open Your Ears - 24/96



The set-up:

Burn-in and warm-up:

Regarding burn-in, the Octave MKII sounded very good out of the box. Its sound improved a little bit after a few hundred hours of burn-in but the change was not as dramatic as I had previously experienced with other audio components. As for warm-up, once it is fully burnt-in, it still benefits from being left on for a couple of hours. Given how little heat it generates, one can just leave always on and forget about it to get optimum results – at least that is what I do.



As Metrum offers to get their DAC with or without the USB module, I think that many buyers will wonder whether it is worth to get the USB module or not when they are faced with their buying decision.

I personally got the USB version mainly out of curiosity, in order to see how their module performs against stand alone USB converters. The end result is that while USB built-in modules have come a long ways since the early days of horrific sounding and 16/48 limited USB inputs, they are still a little bit away from competently built standalone USB converters such as the Audiophilleo 2 (AP2).

When running the Octave MKII through its USB input, the sound gets more upfront and in-your face than when using the AP2 as a USB transport. I had the (false) impression of greater details for the first few minutes of listening until I realized that the USB input on the Octave MKII was acting like the dynamic contrast you would find on older TV sets with primitive electronics. By turning up the contrast ratio (vs. the Audiophilleo), some of the finer details and nuances are lost in the process. The difference in sound quality between the two inputs is not huge but is pretty noticeable in a transparent system. Would I have noticed the deficiencies of the USB input without comparing it to a better transport? I probably would have not been able to do so.

So when does it make sense to get the USB version of the Octave MKII? If one does not have a source of the caliber of the AP2, and is not ready to invest in an expensive digital cable, it makes sense to stick with the built-in USB input, as it will do (most of) the job just fine. One has always the possibility to upgrade later, with probably better alternatives than the AP2, as the technology is moving pretty fast in the sector of USB transports. Also, the USB Octave MKII is probably going to beat most variations of a USB transport – digital cable – DAC combination.

If you have a transport of similar performance level or better than that of the AP2, you can skip on the USB module. If not, the built-in USB will probably do a fine job.



After a lot of experimentation, I ended up settling on SoX upsampling (on Foobar) with the following parameters:

Upsampling to: 176.4K (for 44.1K files only)

Quality: Best

Passband: 90%

Allow aliasing: checked

Phase: 25% (i.e. intermediate phase)

In many ways, SoX upsampling (with the parameters chosen above) is the closest thing to no-upsampling. Yet, its contribution to the sound is very recording dependant. Sometimes, SoX can sound more relaxed, bigger and portray a deeper soundstage in particular recordings, especially on heavily processed ones. Other times, it seems to reduce the depth of the soundstage on very well recorded classical music pieces. Overall, in the grand scheme of things, those differences are rather tiny. I can definitely live with either configuration.

One has to note, though, that I have chosen the most “relaxed” setting for SoX. Theoretically, lowering the passband to 90% and allowing aliasing should minimize as much as possible (pre and post) ringing. Indeed, it would be counterproductive to use an upsampler with heavy ringing (because it uses a “brickwall-type” of filter), on a NOS DAC that is supposed to have a perfect impulse response (vs. regular oversampling DACs). As for the phase, I settled on 25% as it sounds somewhere in between linear phase and minimum phase. Keep in mind regardless of the digital filter (or lack of filtering) that is applied in the playback chain, the ringing (and other faults) recorded on the digital file cannot be undone, except perhaps with the help of some sophisticated apodizing filters (as claimed by some manufacturers).

An interesting alternative to Foobar’s SoX upsampler is JRiver Media Center built-in upsampler. It gives the most pleasing, musical and analog like representation I have experienced from a PC based media player. And though it is less detailed than SoX upsampling (in Foobar), it remains the closest thing to the analog flavor of the PMD100 digital filter I have come across. I will probably go into more details on the comparison between media players in a separate article as it is not directly related to the Metrum DAC.


Other tweaks:

Although I have read elsewhere (I think it was hifi critic, from memory, but I did not want to go back to other reviews before finishing mine to give people as fresh of a perspective as possible) that the Metrum Octave (MKI) was very sensitive to external factors such as vibration control, I have to admit that I did not spend much time experimenting around with various tweaks and cables. When I received the Metrum Octave MKII, I did a simple drop-in replacement of the DAC19DSP I was using before. Given that I was very satisfied with the sound of the Octave from the start, I did not bother experimenting with specific tweaks.

As a side note, I used the Metrum Octave MKII with the Hifi Cables & Cie PowertransPlus power cord (my reference for more than five years), the Aktyna ARIS 2 aftermarket feet (my reference for more than 4 years) as well as some maple platforms (my reference for more than 5 years). As for interconnects, I exclusively used my also long trusted Artisan Ultimate Silver Dream RCA cables. Although I am an audiophile that likes tweaking his system, I have kept my surrounding “accessories” inventory relatively stable for the past few years as I am both familiar and satisfied with their effect in my system.


The sound:

Timber and tonal balance:

One of the first things you would notice with Octave MKII is how beautifully balanced it is from top to bottom. One of my fears going the NOS route was to encounter a rolled off or heavily distorted high register. Actually, it turned out the Octave was much more energetic than the darker sounding DAC19DSP in the higher frequencies. Meanwhile, the top end is as pure and diversified as I have ever listened to in a DAC, combining the analog-ness of the PMD100 DAC19MK3, the energy of the DAC19DSP and the sophistication of the Kingrex UD384 (when playing high resolution files).

When directly comparing the Octave MKII against the DAC19DSP and the UD384, one might think at first that the DAC19DSP has the more powerful bass, followed by the Octave MKII and then trailed last by the UD384. However, after some more listening to various kind of recordings (high resolution and 16/44 sample rates, well recorded albums as well as poorly recorded ones…;), it became apparent that the DAC19DSP had a small emphasis on the mid bass that made it appear as more powerful sounding than the Octave MKII and UD384. The difference between the Octave MKII and the UD384 is that the Metrum has some powerful low bass when the recording calls upon it, that is greater that of the DAC19DSP, while the UD384 does not have that same impact on the lowest frequencies.

Nonetheless, dissecting the tonal balance piece by piece does not fully do justice to the Octave MKII as its biggest strength is the way it reproduces voices and instruments, as a whole, from a single cloth, in a realistic manner. This does not mean that everything is blended together, without differentiation between sources of sound playing at the same time. It actually means that every single instrument and voice exists as a single entity and not as discontinued compilation of bass, mid and high frequencies.

Granted the DAC19DSP, and many other DACs, already do that seamless integration of the different frequencies, but the Octave MKII just goes a step further. Given the right conditions (i.e. proper source upstream, as well as transparent amplifiers and headphones downstream), the Octave MKII’s depiction of timber of instruments and voices is absolutely stunning. If you like to listen to classical music or non-heavily processed music, you might find the Octave MKII the right match for your system.

In my personal experience, I have found that when a digital source can render properly and convincingly both pianos and violins, there is a good chance that it can render everything else in a satisfying manner.

So, one might ask, how did the violins get rendered? Did the Octave MKII focus on the wooden resonance of the body or the sheen of the strings? The Octave MKII did both and neither. The violins sounded different from one recording to the other, and if different type of violins were used in the same recording, the Octave MKII lets you hear every nuance of it.

Although, I had already heard differences on the DAC19DSP between the Stradivarius and Guadanigni in the nicely recorded Vivaldi - Concerto for 2 violins - Carmignola/Mullova, the Octave MKII just makes it easier to spot differences between the two superbly rendered violins. Meanwhile, despite that greater ability of separation, everything seems to be more enjoyable to listen to.

It seems that instead of focusing on the flaws of a recording, as many analytical DACs tend to do, the Octave MKII seems to be focusing on revealing the details that are most relevant to enjoying the music.

So, is there any weakness or limitation in the timber or tonal balance of the Octave MKII? I suspect that it is in possible to make the bass even more powerful (with a bigger power supply perhaps), but other than that, and without having access to a truly exceptional reference DAC, I cannot find any objectionable fault with its tonal balance and timber rendering of various instruments and voices.

During my almost one year experience with the Octave MKII, I did notice on occasions that the sound was not perfect, that it sounded sometimes harsh or congested in the midrange or simply that some instruments just sounded off. However, the culprit was usually elsewhere in the chain.

Also, the more transparent my components became downstream, the more I started appreciating how beautiful and realistic the tone and the balance of the Octave MKII was.


Soundstage & Imaging:

I was very impressed from the start with the soundstaging capabilities of the Octave MKII. If I had to make a direct comparison in size, I would say that its soundstage size is noticeably bigger than that of the DAC19DSP and slightly bigger than that of the Kingrex UD384. However, it remains a simplistic way of comparing soundstage sizes as it varies from one recording to another and from one sampling rate to the other. That it is just to tell you that it soundstage is as big as I have listened to in my system but I have not had much experience with (very competent) sigma-delta based DACs, other than the UD384, that are supposed to be champions in soundstaging.

However, what struck me when listening to the Octave MKII was not its ability to portray a big soundstage but, rather, its ability to transport me to the recording venue. On many recordings, you can get a pretty good idea of the size of the recording room not by trying to calculate the size of the room that is being painted in front of you but by measuring the size of the room that you are actually occupying.

While this was not very obvious at first, but still noticeable, when using the audio-gd C2 amplifier and the ALO recabled Beyer T1 combination, it was much more obvious and apparent when ending up with Glenn’s OTL amplifier and the HD800s (with Wiplash TWau Reference cable).

While the soundstage size of the HD800s is not clearly as big as that of the Triangle Antal EX speakers (in a 750 sq. feet room), for instance, it is however more realistic and immersive.

Hence, I can imagine other DACs (and most possibly the Metrum Hex) having a bigger soundstage size than that of the Octave MKII. But on its own, and fed from the Aqvox powered Audiophilleo 2, the Octave MKII gratified me with the best soundstage I have heard from a digital source.

As for the layering and imaging, all I have to say is that the Octave MKII combines image specificity and image body beautifully. While the Octave MKII makes it very easy to pinpoint the location of various performers and instruments in the soundstage, it also retains enough “body” to the sound to make it lifelike.

Actually, the imaging specificity is better than what you get in real life. Or to put it in other words, I have never been in a live event that approached the 3D performance and image specificity of the Octave MKII. There are maybe a few possible explanations. Either we do not listen the same way at home and at various concerts. Another explanation is that using multiple microphones has an impact on the 3D perception, but it has the same surprising and counterintuitive result.



The Dynamics of the Octave MKII are as good as I have listened to in a DAC. In comparison with the DAC19DSP which seems to have a better (at least bigger) power supply section and a dedicated all-discrete output section (vs. direct feed from the DAC on the Octave MKII), the Metrum DAC held up surprising well and had overall the edge in dynamics in my system.

More specifically, and even in comparison with live music, the micro-dynamics are as good as I have ever heard. Every little dynamic variation of a voice or instrument is tracked exceptionally well and makes the listening sessions very engaging.

As for the macro-dynamics, and conversely to the soundstaging and imaging department, where I found that the Octave MKII (as well as other DACs) exceed live concerts, I find that macro-dynamics are not as good as what you might experience in live concerts.

Listening to my “ultimate” tracks such as the Dark Knight OST or Gladiator OST, the Octave MKII can make you go through some terrifying moments. But on other tracks, I felt that perhaps there is still room for improvement. Or maybe, it is just that many recordings are too much (dynamically) compressed for their own sake.

In any case, I am just nitpicking, and in comparison with most DACs and CD players I have listened to, the Octave MKII is able to deliver a very lively and dynamic representation of the music.


Transparency & resolution:

The Metrum Octave MKII has a very high level of transparency to the recording. One of my fears when getting a NOS DAC was to have a “romantic” sounding component that would impose a certain character on every single recording (I had the same fear when getting and OTL tube amplifier and my fear also turned out to be unfounded). However the Octave MKII is the most transparent DAC I have listened to in my system (or elsewhere). On Messiah - Dunedin Consort album, for instance, I was able to hear the performers move as if they were just in front of me.

The nice thing about the Octave MKII is that its added level of transparency does not come at the expense of false or excessive brightness. In comparison with previous sources (either CD players or DACs) I have listened to, the Metrum DAC seems to take the dust of all the recordings I have while making them better to listen to at the same time. So, is the Octave MKII cheating somehow? Maybe it is the case, but there is no way to be sure.

The level of details that this little DAC can render is unbelievable. On a Chesky classical music recording, I was able to hear the traffic noise outside the recording venue. On my reference Jazz at the Pawnshop recording, I was able to hear, better than ever, the background chatting and silverware noises while the music was playing at a high volume. Although, there might still be information to extract from the various recordings I have, I do not believe it is necessary to musical enjoyment.




I believed that we have reached a very interesting age in the Digital reproduction of music. Today, a €1000 DAC buys you equipment that can reveal flaws, or at least the very specific character, of most recordings out there.

So why bother spending this much money or even more on a DAC? The answer is in the subtlety. Even though the Metrum Octave MKII NOS DAC can show flaws on many recordings, it has rewarded me with the best sound I have heard in my system. Contrary to what you will find on analytical DACs that focus on laying bare the recording piece by piece regardless of the music as a whole, the Metrum Octave MKII is able to put all that information into context, in a very musical way. Getting a truly good DAC can have the combined effect of being both more revealing and more enjoyable to listen to at the same time.

The main weakness of the Octave MKII DAC, in my personal opinion, is it high sensitivity to the quality of the source (jitter… ). Its USB input is good but can be bettered by the Audiophilleo 2. Meanwhile, its SPDIF input is sensitive to every little change made at the source (media player, usb transport… ). Whether that is a sign of transparency (of its analog section) or a possible improvement in the digital section, I believe that the truth lies somewhere in the middle as even the mega-buck high performing DACs still seem to benefit from better transports (be it CD players or USB converters).

However, given its price performance ratio, I cannot recommend highly enough the Octave MKII to anyone looking for a single ended DAC in this price range.

This review is the first of a series of articles that I will be posting in the coming weeks. As I have elected the Octave MKII as my reference DAC, the readers can get a better sense on how it sounds in my system by reading the next reviews.


Categories: Source

Post a Comment


Oops, you forgot something.


The words you entered did not match the given text. Please try again.

You must be a member to comment on this page. Sign In or Register

1 Comment

Reply Amine Slimani
6:54 AM on October 10, 2014 
Please post the comments here: