L.K.S Audio MH-DA004 - probably sounds bad

Woop woop. We have achieved full Head-Fi, repeat, we have achieved full Head-Fi. Anything you cannot cope with is therefore your own problem.

 

I recently picked up the MH-DA004 for $750 on the used market to try it. I only have a Modi 2 Multibit and a Pro-ject Pre Box S2 to compare it to and for sound quality it beats both handily. Build quality is solid and it's doing well with the Elac Navis active speakers I just received. I'm probably going to keep it.

 

I bought a brand new LKS MH-DA004 based on some questionable but solid reviews. My initial thoughts were the build quality and, at least on paper, the specs were good. After initial burnin ( about 150 hrs) I started some critical listening tests.

My system consists of a Levinson 31 transport, a MacBook Pro, Levinson 380s preamp, Levinson 360s and Berkeley Alpha 2 Dacs, Ayre QB-9DSD, Spectral DMA-360 amp and all cabling is Transparent Ultra. I’m using VOX software for streaming on the Mac.

For starters this Dac has a very forward presentation compared to the Levinson and Berkeley. Resolution and micro details are pretty good, but somewhat smeared compared to the higher cost dacs. The one thing I noticed almost immediately, was the timbre of instruments, especially piano and violins seemed somewhat artificial for lack of a better word. The notes lacked the natural decay and room ambience heard through the both the Berkeley and Levinson. On the plus side, I though bass definition and weight were excellent and on par with both high priced Dacs. Both upper on lower midrange registers were good but not on the same level as the Berkeley and significant more grainy compared to the Levinson. Image definition and space were pretty good and horizontal soundstage very good, but vertical was nowhere near as good as either other Dac. Treble extension was decent but somewhat on the bright side of neutral. I wouldn’t say it was harsh, but not nearly as refined as the other 2 Dacs. Dynamics and pacing were very good as to be expected using ESS chips.

Overall not a bad Dac especially given the modest price, but definitely not a giant killer as some have claimed. The Ayre QB-9DSD bettered this Dac in every way playing ripped files through the Mac. I thought the LKS sounded the best playing 24/96 ripped files and filter settings on 8. Which btw, the user manual is pretty poorly written on the different filter settings. Experimentation is a must. It might make a good headphone Dac with a good tube pre but it was definitely too bright for my system.

 

Forgot to add speakers. Speakers used were Egglston Andra IIIs and Platinum Audio Solos. The Solos still reproduce some of the best female voices I’ve heard.

 

@Spectral1212, you'll probably find this thread helpful for SBAF advice on ESS chipsets:

https://www.superbestaudiofriends.org/index.php?threads/best-value-sabre-dacs.7102/

 

@bengo Thanks got the link. I’m not a huge fan of Sabre chips. I’ve heard very few Dacs that sounded “neutral” with these chips installed. I personally prefer the sound of R2R chips. Some implementations of the AKMs are decent too. The most musical chips I’ve heard to date were the Ultra Analog 20400 found in some of the better late 90s Dacs. I still have an old Parasound D/ac 2000 Ultra. It used 2 of these chips and decodes HDCD as well. Still competes with most of these newer Dacs playing HDCD files. Of course UA went under years ago which is too bad. The old Burr Brown PCM63p-k chips were awesome as well. Thanks for the heads up..​

 

@Lautreamont
@Spectral1212

Just a word of kind advice to the new owners of the LKS 004. You simply MUST MUST shield the transformers from the rest of the circuit (or is it the other way round). It's very easily and inexpensively done with a bit of copper sheet. See the LKS 004 thread at Head-Fi.* There are a couple of useful posts there on the subject, including my own (Lennym). You will be very pleased with the improvement, and I think that improvement is responsive to some of what is written here. Skeptics with the ability to measure the EMI will see it clearly. Also, the display should always be off while you listen as it is also an EMI source.

*I do not personally vouch for any other mod discussed there.

 

If that was meant for me, I can't "have" L.K.S. do anything.

A brief look at the innards of the X-Saber Pro reveals a very similar configuration, but that as to hardware, it is a lightweight next to the L.K.S.

The all-important power supply is much, much smaller on the X-Saber Pro. One ES9038PRO chip as opposed to two in the L.K.S. What appears to be a chip implemented analog section as compared to the L.K.S.'s fully discrete analog section. Quality name brand connectors on the L.K.S. That doesn't mean the X-Saber Pro doesn't sound as good, or better--I haven't heard it. But it does suggest more potential in the L.K.S.

So if the X-Saber Pro was well received here, in spite of its ES9038PRO chip*, the L.K.S. with the slight mod I suggested might do very well.

BTW, X-Saber Pro has now gone the MQA route, something I would not support.

*I understand that ESS DAC chips are not favored on this board.

 

As far as I identified, there are (at least) two issues, which I prefer to separate.

1. Is LKS MH-DA004 good dac?
2. Is 'shielding' mod not bull-shit?

#1 could be a matter of preference, taste, etc etc. I just suggest you to wait until people consider your assessments worth listening, possibly via months or years of good to great activities/contributions.

#2 is a real problem imo. Sbaf, as a group consensus, always looks down on

• irrational (i.e., not evidenced or supported well) mods,
• people who recommend to do so
• people who evaluate them biasedly and/or incomprehensively.

It's never considered 'kind advice' here -- rather 'hazardous or malicious influence'. Better say nothing.
If you're too enthusiastic to keep silent, again, wait until you become a meaningful poster, then spend all your resource in justifying it not irrational with convincing evidences.

By the way, I don't think good dac chips, large/heavy power transformer, and full discrete circuits do necessarily better. Audio-gd and Kingwa taught me a lot. Both in general and in the specific context of ESS9038. lol.

PS. that said, I am interested in hearing that dac.

 

mfw Lenny is still here after exactly two years defending this shitty DAC, this time stating that the transformers need to be shielded. what sort of shitty DAC needs a mod to shield itself from its own transformers?

 

legitimately lol'd

Also what ChaCha said about DAC mods although I'd make an exception for stuff like Soekris where people can roll their own filters out of curiosity.

Different note - why do so many Chinese mfgs seem to favor ESS chips/dacs? Why not AKM, AD, Burr-Brown (TI), an R2R dac chip, or even just coming up with their own internal (company) or domestic production chips? Nothing wrong with choosing ESS per se, just curious.

 

AKM is also frequently chosen. Specifically latest 4493/95/97/99.
Multibit chip is used only by Schiit (among in-productions; to my knowledge). R2R discrete seems to fail more likely without legendary engineers and/or a bunch of trials and errors.

I bet primary decision factors include (1) general perception (in Asia, people seem to like Sabre and AKM in my observation; no idea why), (2) per chip cost (and discount amounts in large-volume sales), (3) easiness (documentations, tech support, etc etc), and (4) perhaps non-English speaking customer supports. Just my conjecture.