Lab12 - Greek hifi manufacturer

Lab12 dac1 Reference - review

https://www.lab12.gr/product/dac1-reference/

Ever heard of Lab12? Apparently it is a Greek hi-fi manufacturer established in 2012.
The first time I ever heard of them was just a couple months ago, when an audiophile I've been chatting with for a while mentioned the manufacturer and the dac in question. Supposedly it was much better than Soekris dac2541, which of course immediately sparked my interest. Lab dac1 reference's MSRP here is 2950€, which places it pretty close to other known qualities like Yggdrasil, Burl Bomber, DM Convert-2 etc.

Full disclosure: this was a brand new device that I got through a local dealer. I was able to test it at home without any commitments. The unit then went on to an another audio enthusiast, who happens to be a good friend of mine. I will share impressions from both our sets, but the impressions are mainly just from myself.



The dac1 reference is based on eight Philips TDA1543 multi-bit chips, which are apparently older multi-bit chips meant for lower cost devices. This was actually the first time I got to listen a dac based on this particular dac chip, so I didn't know at all what to expect. Other related buzzwords are NOS and tube output. Actually, tube output in a dac was also something that I hadn't experienced before. Exciting times ahead!

When I received the dac, I couldn't help noticing the Finnish importer (!=dealer) of the device happens also to carry Trinnov. I had a "run-in" with the Trinnov Amethyst a few years back. I got one for home demo straight from the importer in question. And I had no choice but to shat all over it on a local hi-fi community. During unboxing the Lab12 I chuckled to myself and thought "I hope this isn't a premonition of anything".

The overall design of the dac1 reference is pretty basic: it has exactly two buttons and pretty straightforward aesthetics.
I wouldn't call it beautiful, but the VU meters certainly add something into the mix and I must admit, I do like them. The overall build quality feels good. The inputs are: one USB, two coaxial S/PDIFs and one toslink. I personally prefer AES/EBU over those for serious listening, but I guess they've decided at Lab12 it brings nothing more to the table.

I started by connecting the dac1 reference to my speaker setup, which at the time consisted of:
- Buchardt S400 Mk2 standmounts
- Ayre AX-7e integrated amplifier
- Bluesound Node 2i streamer
- Soekris dac2541 (for comparison)

Music used in this and following comparisons:
- Eva Cassidy - Nightbird
- JMSN - Heals me
- JD Allen - Love stone
- Norah Jones - Come Away With Me

The Bluesound streamer was connected with a coaxial cable. I tested the toslink input by connecting it to my Panasonic TV and playing some music from Youtube (lol, I know). I mostly played music via Qobuz and the first thing I listened to was Eva Cassidy's Nightbird. Both dacs were connected with balanced XLR cables to the amplifier.

Very initial impressions after the device had been powered on approximately for 30 minutes:
- Wow, there's actually more depth to the soundstage
- Also, front-to-back layering seems to be a tad better
- maybe a tad more resolution too?
- Not as lean sounding as dac2541, tonally darker in the midrange
- also less dense than dac2541
- I kinda expected this to be a warm sounding device, but I wouldn't describe it as such
- Not really that tubey sounding either

Usually the first impressions are quite telling if I end up liking the device or not. This was a brand new (tube) device though, so I let it stay on continuously for three days. I did listen to music right from the start though and I didn't detect any MAJOR changes in the general sound. Though I must say even if the dealer told me no long break-in period is required, the manufacturer recommends a break-in period of 200 hours.

When I got myself to analyse Lab12's sound more closely, I found out that the tonal balance was slightly tilted up at the very top. Both compared to Soekris dac2541 and what I'd consider as neutral or reference type of tonal balance. The treble wasn't completely smooth either, as it became apparent after half an hour or so after the initial start up. I did initially put this on the unit being brand new. If initially it would've been fair to say the treble was definitely grating, my opinion of it during the home demo period drifted only slightly. After a couple of days I settled calling it coarse - definitely not smooth, but I guess how insulting this ends up being depend also on your setup.

After three days of listening in my speaker setup I moved the dac to my headphone setup which consisted of:
- ZMF Verite Closed LTD Olive headphones
- Raal HSA-1b amplifier
- Pi2AES streamer + PC via USB
- Soekris dac2541 (for comparison)

Here a couple of things became immidiately apparent: The resolution advantage I originally heard against dac2541 was actually just different type of tonality. More prominent highs, more air and less dense tone made some details more highlighted. In the end I'd say dac1 reference and dac2541 are both pretty much equal in their microdetail resolving abilities. Where the dac1 reference still holds it grounds is the soundstages front-to-back separation - layering seems to be a tad better than with Soekris.

Where dac1 reference really falls apart is dynamics. It hits like a... wet noodle. It is not anything like dac2541 when it comes involvement. My foot stops tapping immediately when I switch to lab1 reference. Tapping starts again when I switch back to Soekris. The macrodynamics are not miles apart, but it's the microdynamics that really separates these two apart. Compares to Soekris, you really don't form any kind of emotional connection to the music you are listening to. Everything sounds just so flat.

After doing some more back-to-forth comparions I couldn't help noticing the difference between the fluidity between both dac's presentations. Like I said previously, dac1 reference's treble is somewhat coarse and this affects the whole presentation making it less fluid than with dac2541. It also makes dac1 reference's background greyer than dac2541's. Still, neither of the devices sound wet nor dry.

What's interesting I didn't really hear any difference between the USB and coaxial inputs. They were either equally good - or equally bad. I wasn't able to test the toslink input in my headphone setup, but it sounded OK before.

The lab12 was eventually moved to my friend's place and connected to his headphone setup, which consisted of:

- Stax SR-009 headphones
- Stax SRM-T8000 headphone amplifier
- Nuforce CDT-8 Pro cd transport
- AVM CS 5.2 integrated amplifier, which was used solely as a dac for comparison

When I got a change listen the setup, the dac had been more or less continuously on for well over 100 hours.

There was a slight difference in the nature of the presentations between the dacs, since AVM's dac is based on D/S technology. Squareness of the soundstage was certainly one of them. Lab12 isn't really warm or tubey sounding, but there's still distinct if faint timbral quality that can be associated with tubes. Midrange was darker than the reference. It didn't take much of listening to discern that the AVM's dac was already pretty much up to its task. Pretty much all technicalities favored the AVM. It had better dynamics, blacker background and more even frequency response. I guess you could also say the AVM sounded a tad "wetter" and more euphoric in the midrange, though I wouldn't say those two qualities are prominent in isolation. As a whole AVM's dac could be thought of as a Chord Hugo 2 with smoother&less bright treble and maybe slightly worse detail retrieval. In this company, the dac1 reference didn't have a change.

We also tried what happens to the sound if we used RCA outputs instead of the XLR outputs.
The overall presentation stayed the same, but the soundstage lost some width and maybe also some depth.
The difference was noticeable enough for me recommend using the XLR outputs with this dac.

When it comes to system synergy, my friend's Stax setup and dac1 reference don't really play well together. SR-009's are already quite bright and when you add bright and coarse sounding dac, you get something that is uninvolving and slightly offending at the same time.

Conclusion

Adding eight low grade multi-bit converters with a tube output stage do not equal to magic.
Lab12 dac1 reference is overengineered, technically mediocre and worst of all, bright sounding and uninvolving dac. Price-to-performance ratio is poor.
3000€ buys you an Yggdrasil or a Burl B2 Bomber, which are both miles ahead of this in every single category.
Even the Soekris dac2541 is a much better dac as a whole. I'd also choose a Bifrost 2 over this in any given day.

I might prefer this over a Modi 3, but only if the rest of the chain is dark sounding enough.

I really kinda wished this was a better product, but oh well. Still decided to write a full review, since I haven't really written anything here in a long time.

 

Just a technical note, SPDIF over RCA is the same data and protocol as AES over BNC. Just a lower voltage that makes no difference in sound, just will not drive as long a cable as AES.

 

Gratuitous VU meters!

 

The TDA1543 is a garbage DAC chip, but it is also an absolute delight. It was always a low end part, and it's been discontinued since the late 90's I think. Even after being discontinued, it found new life in a quite expensive NOS DAC from 47 Labs. That's the DAC that more or less started the whole NOS thing, and clones of it are still available in eBay.

Essentially, the 1543 will take an I2S signal directly, and it has a high enough current output (2.3mA) that one can simply use a single resistor (usually around 1K) for the I to V and take the output directly from there. That provided around 2V and an output impedance of about 1K - good enough for many uses.

It was often paired with the CS8412 receiver which will take in SPDIF and output I2S. Neither of the parts need any sort of programming, or glue logic. So for about $20 in parts ($18 of which are for the 8412) one could make a crude, but surprisingly good sounding DAC on breadboard.

The sound of these DACs is controversial. They measure terribly, high frequency is rolled off, impact is minimal, etc. But yet, there is a sort of magic in the midrange that anyone who likes full range speakers will appreciate. They are the opposite of modern hifi. They also made it easy for lots of DIYers to learn how to build a DAC.

One of the ways they were used was in parallel. Essentially it is a statistical argument were paralleling increasing numbers of chips will lead to there being more transcription errors, but each error becomes less costly. At some point the tradeoff is worth it. I think most experimenters found that two chips sounds worse than one, four is simply different, and by eight things start to sound if not better, at least more hifi. People using 32, or 64 was not uncommon (these were available for less than $1 each). Additionally, as more are used, the output current is increased which means a lower value I to V resistor, which means a lower Zout, and more current. For eight chips, the IV (and thus the Zout) would be 125 Ohms. I remember seeing instances of enough chips to drive some headphones directly from the DAC output.

At any rate, none of this bears on the above reviewed dac, about which I know nothing. But the history of these chips is fascinating. And for anyone who is keen on diving into a bit of DIY, the TDA1543 is an excelling place to start. Also, FWIW, 47 Labs also made a power opamp based amplifier called the Gaincard, clones of which (the so called "Gainclone") also became a DIY cottage industry.

 

TDA1543, NOS and tube output stage? Uh-oh.

But then again, there have been, although rarely, gears that shatter misconceptions and perform surprisingly well. The Matrix X-SABRE Pro was well received here, even though no one is a fan of ESS SABRE sound. The latest LCD-X was also met with praise, despite being Fazor.

 

looking to get some input/advice.

i recently sold my Holo May KTE for a Meitner Audio MA3. I have a Rockna Wavelight, and am taking delivery of a Rockna Wavedream Signature XLR this evening. I am thinking not much point in having two Rocknas, and have the opportunity to get the Lab12 Dac1 Ref to replace the Wavelight.

I had the Lab12 Dac1 Ref in my house for several days and quite enjoyed it. the top end is pleasant and the mid range is lush. didnt find anything glaringly wrong with the low end or the overall presentation tbh. It was connected to a Primaluna Evo 400 pre and the Lab12 Suara power amp / NAD M23 power amp, with Ascend Acoustics Sierra 2EX bookshelves + Rel T7i sub. (listening with 2 channel setup is quite rare for me, as i'm mostly on headphones).

essentially, my question is, am i going to be making a mistake swapping out the Wavelight for the Dac1 Ref? Not really "needing a change" but probably more of an itch... price wise it isnt that massive of an outlay relative to my other dacs.