HOLO Audio MAY DAC

[louisxiawei]

I see people starting discussing the MAY DAC on Spring thread and ASR forum has given some detailed measurements of MAY. As an early MAY user, I feel it‘s necessary to open a new thread to have more discussions for this exciting DAC.

 

I just received it a few days ago and it is still warming up. So bear with me with the listening impressions as I haven’t got too much to say yet.

 

I purchased MAY mainly for the purpose of using its NOS feature with HQplayer (So I don't care other OS features). I used to have T+A DAC8 DSD and guess I might not help myself comparising these two DACs later on.

 

Persoanlly, I'm confident that my system is quite transparent and it's good enough to demonstrate differences among sources. Nevertheless, I have my own listening preference and may not suit your taste. If you want to pull the trigger on this DAC, don't judge too much of my comment as I'm not that type of person who is enjoying typing lengthy paragarph of subjective listening impression.

 

I will also share some of my practicial findings when using MAY. Sometimes these little trivial things can be helpful or can even expose some bugs, small issues so that product can be improved later on accordingly.

 

Ok, no more jibber jabber. Hope you find this thread and MAY interesting.

 

 

[louisxiawei]

I will now start with more objective operational stuff and later on towards a bit subjective things.

 

For connectivity, it is a VERY HQplayer friendly DAC. The manual will tell you to install windows USB driver in order to support DSD1024 & PCM 1.536M playback, but it is not necessary. I strongly recommend you use to burn HQplayer NAA ISO image on the USB and plug onto your NAA, then HQplayer will recognize the MAY's end. I believe the ISO image is a customized linux kernal by Miska and allows your NAA to playback native DSD1024 on MAY. As I'm sick of windows OS for audio output, I can't be more happier for this.

 

However, I do find that there is a tiny pop during DSD manual track switch on MAY (no pop during gapless playback track switch) and Jeff told me that there should be no pop during switch track on Windows OS using the offered driver. For linux, he has little idea and cannot guarantee. I'm aware that this pop noise is quite a common caveat for DSD playback, but the latest custom T+A DAC8 DSD firmware has solved this issue after years of struggling and effort of Amanero. If only for the sake of perfection, it would be better if MAY can make this pop disappear. 😉

[louisxiawei]

1 hour ago, Hyde27 said:

No pop on my Holo May. iMac HQPlayer4Desktop

As I said, it’s more a Linux issue. Good to know MAC doesn’t have that hassle.

[louisxiawei]

Another thing I like about MAY is its usb connection. Since it has a quite good USB galvanic isolation internally, I tried not using the USB dongle ISOregen and found the the sound "pure, clean" enough to be satisfied.
 

I’m sure someone can observe some differences/improvement if you plug these ISOregen-ish things back. But for me, MAY itself suffices. I enjoy more of the trouble-free operation. I’m a bit fed up with plugging dongle to the dac and powering the dongle with another LPS and then worrying about their disconnection from time to time. 😂
 

In my opinion, all DAC nowadays shall consider implementing galvanic isolation for its USB port or consider them outdated.

[louisxiawei]

28 minutes ago, Miska said:

 

Spring 2 is on my loudspeaker system, but the preamp has headphone socket as well. Spring 1 is on my headphone system.

 

If you mean start/stop pops/clicks, absolutely nothing on either. I cannot check again right now.

 

I also don’t have start/stop pops. The pop I meant only occurres when switching to another track manually. Dac8 used to have this kind of pops and it might still have it on Windows OS.

[louisxiawei]

Some subjective listening impressions:

 

To better check any difference MAY could possibly bring compared to my previous DAC8, I applied the same HQplayer settings on MAY as on DAC8 DSD, so only DSD512 part. I will check the R2R or DSD1024 bit later on. MAY just has more fun to explore, needs take some time I'm afraid.

 

In general, I find MAY's sounding is very neutral, transparent and resolving. By comparison, I think DAC8 is obviously warmer while MAY is obviously calmer. This "clam" is not related to "not musical" or "thin" sounding, maybe I should say more neutral. 

 

- I'm quite picky about solo piano tone and will always test the piano track first on any new dac. So I played some good DSD piano recording from Challenge Classics label and I think MAY offers equally satisfactions as DAC8. I personally judge the performance of piano playback as an indication of sound tuned to be authentic. 

 

- You might be worried about this "calmness" of MAY might have some trade off like the mid range will not be soothing or musical. But it is not the case. I picked some tracks from some female singers who have edgy or airy voice like Kate Melua, Stacey Kent, Kat Edmonson. The vocal part is musical and pleasing with good density yet the instrumental part is being more neutral and transparent. 

 

- The thing I'm struggling/uncertain to judge now is about the bass. At first listen, I think DAC8 has more punchy and bouncy bass but then I notice that I prefer some other tracks' bass on MAY. For example, I prefer DAC8's more juicy bass when playing Coldplay's album. When playing some Daft Punk's tracks, then I think the bass on MAY is just about right and I can hear more bass details. I'm not sure if it's appropriate to conclude MAY offers better quality bass rather than quantity. I'm also not sure whether the warm/clam tune may have an effect on bass or not. Need more time to check.

 

As you can see in my system, I'm also a Spectral Audio fan. So you might somehow speculate my taste as natural, neutral, transparency, fast. If you want to ask me which one of these two DACs possesses more similar character to Spectral gear. I'm quite sure it's MAY. 

 

 

[louisxiawei]

7 minutes ago, Miska said:

 

That is not very clear case, because what happens depends on the tracks being played. If they are same sample rate it is pretty much gapless operation. If they are different sample rate, then it depends if output rate family is being switched or not. If output rate is being switched between tracks, then the DAC is stopped, reconfigured and started again. If the output rate/format doesn't change, then it is pretty much gapless operation from the DAC's point of view.

 

 

Ok, let me make a simpler example:

 

When I dragged the same track time bar (e.g. from 1:20 to 2:45), there is a pop occurred on MAY (or DAC8 on windows). I remember I sent an email to T+A regarding this and they confirmed this pop noise and told me it's normal.

 

Yes, the buffer time was set to be default. Maybe you have a different firmware? 

[louisxiawei]

4 hours ago, mevdinc said:

Am I correct in assuming that MAY DAC has no volume control?

Yes, MAY has no volume control in either OS or NOS mode.

[louisxiawei]

On 1/7/2020 at 6:13 AM, MhtLion said:

Thanks for great review! The rest of your system looks fantastic! Hope you will post your experience with R2R listening as well.  Thanks.

 

For R2R part, I applied the same filter as SDM: poly-sinc-xtr-lp, new dither: LNS15 and upsampled everything at 1.536M. The sound is very transparent, nothing warm or blur as many said R2R used to be. I believe it's a good and rare thing for a R2R dac? 

 

Frankly speaking, I personally think R2R part sound is a bit dull compared to SDM part when using HQplayer. Less holographic and organic. I don't have other R2R dacs to compare with it at the moment because I used to avoid purchasing ladder type dac.

 

I would say DSD512 gives obviously more 3D airy sounding while DSD256 gives smoother, more density, more analogue sounding but less airy (This apply on both DAC8 and MAY) . From my observation of many "hiend" product marketing slogan or reviews. This "air" thingy is quite a critical index for a DAC to be valued. 

 

Unlike DAC8 gives "clean" and “wide" LPF options, I was informed by Jeff that MAY applies 45K LPF. So I believe 7th order modulator should fit MAY very well technically. At least it sounds good to me.

 

I prefer ASDM7EC DSD256 over AMSDM7 512+fs DSD512 a lot when listening to vocal. I was a bit shocked when I hear a much better presence of vocal detail like spittle, lip smacking and saliva these kind of things. For orchestral, DSD512 is preferred. 

 

I also did my own trick/compromise and this may not suitable for others. For using Spectral gear, MIT/Spectral interconnects are compulsory and my MIT Oracle Matrix HD90.2 speaker panel has a 2C3D switch - when it's on, giving a more 3D holographic image. To be honest, I don't consider it as a serious function and prefer it to be off as it has its own limit and sounds a bit fake in some music genre. By comparison, DSD512's 3D effect is more natural than MIT's artificial 2C3D.

 

But the 2C3D works well with ASDM7EC DSD256, offering smoother, more analogue yet providing 3D airy sounding similar to DSD512 until there are hardware can fulfill EC modulator at DSD512. 

[louisxiawei]

On 2/8/2020 at 3:23 AM, clipper said:

Hans Beekhuyzen - Holo Audio May Dac (Level 2):

 

 

I sense him discouraging the oversampling apporach. Quite a pity he has missed so much fun. 😏

[louisxiawei]

 I think it's worthy to share my findings regarding the DSD1024 on my MAY. The holographic effect from DSD1024 is significant compared to DSD256 and DSD512, not subtle at all.  My PC can do either EC 7th order modulator at DSD512 or non-EC modulator at DSD512. I compared AMSDM7EC+512 at DSD512 with AMSDM7+512 at DSD1024 several times back and forth. DSD1024 is really something else even with non-EC modulator. I provide this information is because Miska once said several times that the difference shall not be that big, and sampling rate at DSD256 onwards is tecnically sufficient.😉 In my case, acoustically, DSD256 is not sufficient, DSD512 is starting to be good, DSD1024 is the starting point to be shined.

 

I don't want my post to be subjective or try to make a big fuss out of something trivial. But everyone who is able to reproduce DSD1024 on the fly really need to give it a try. I can go for the anology like this: the difference is just like a 1080p resoltion vs 720p resolution.

[louisxiawei]

16 minutes ago, Quadman said:

I've been telling everyone here for months DSD1024 is where the magic happens.  You are correct IMO, when you say 1024 with no EC is greater than 512 with EC.  My next build I hope to test 2048, which Holo says is theoretically possible.  So many people are hung up on DSD256 w EC modulators because Jussi said technically this is excellent.  I can't listen to it anymore it is clearly inferior. 1024 and occasionally 512 w EC modulators is where I live and play now.

I searched the key word "1024" in this topic after my last post to see whoever talked about dsd1024. Seems to be you are the only few. 😂

 

I tried to avoid using subjective or fancy wordings these days in any audiophile forums as we aduiophiles do make a fussy impression from time to time, the more frenquent audiophiles exaggerate the SQ, the less credbility a product will have accordingly.

 

We are still the minorities to spot the DSD1024. DSD 1024 is not that just simple as "bigger" stage, or more transpranecy or more delight overtone. I would say that I sense a strong "air" across the whole frequency when the sampling rate is higher. At DSD256, I sense no "air" at all. 

 

If we are talking about soundstage, then a multi-instrument scene or orchestral pieces will be a good example to discuss. But it not just about sound stage.

 

For example, I played a few tracks of solo singer with just piano accompaniment or no instrument at all, a solo singer with or without a piano cannot demonstrate a complex soundstage and allow us to disucss the stage is bigger or wide. But still in DSD1024, the difference is clear, I can sense the air around every syllable, that vibe is real. 

 

Same speaker, same setup, just sampling rate change. I think it's pretty amazing but shamed that not many tried yet.

 

The link below is from soulution audio, which also mentioned about "the air". I think the great minds think alike and maybe some great mind have not realised them yet.

 

https://soulution-audio.com/series7/soulution-760-dac/

 

 

 

[louisxiawei]

BTW, one of my friends uses T+A DAC 200, he also upsamples everything to DSD1024. Since then, he told me he cannot go back to DSD512 anymore. Same as me, he is also hyped about the "air" vibe. 5 years ago, he bought the DAC8 DSD and loathed the upsampling until I told him there is a thing called HQplayer. 😏

 

So, HOLOY MAY dac is not the only case here.  😁

[louisxiawei]

28 minutes ago, Quadman said:

My listening hours per month once I hit 1024 have gone up over 30%.

Same here.

 

I used to stick to the DSD256 with EC modulator as my CPU is not heavily overclocked. I do appreciate more refining sounding from the EC modulator, but I'm not in a hurry to get a new CPU just for the EC DSD1024. I used to be in the competition to get the fastest PC hardware for the HQplayer upsampling. Eventually, I got my diminishing return. CPU evolves through time, so does HQplayer's optimisation. 

 

After another round of search, I've also found that people are quite skeptical about such DSD1024 magic as the measurement differs (HOLO audio dac at DSD256 seems to be the best in measurement). Then talking about your TUBE distortion sound preference, ultrasonic noise handling at the amp. Luckily, I do not like tube sounding and my amps are Spectral that can handle ultrasonice noise well. 

 

I will have a deep search to see whether there is a higher levels of distortion with DSD rates above 256 in holo audio DAC presented in this forum. If so, I might bring the measurment and ask Jeff for his opinion. I prefer to be scientific yet combined with collective subjective opinions.

[louisxiawei]

38 minutes ago, Quadman said:

Very few Wyetech 211 SE amps live in the world unfortunately, and with the designers passing earlier this year that will remain true.  I have never heard a tube amp do what this amp can do.  It is a 211 SE amp with SS balls, fast and dynamic.  It blew my pass labs XA amp out of the water.  This relatively big amp, 22 inches long, is mostly power supply with 2/3rds of the chassis being devoted to the PSU section.  Roger knew how to get the best sound from a 211.  I am lucky that I have skills and can keep my amp running should something fail as I will never part with this wonderful amplifier.  You know what they say "you don't know what you don't know, until you do"  That is so true for 1024 and especially the wyetech 211 SE amp.

I have little experience in TUBE gear and only a few limited experience in some mediocre tubes. The place I live in now has very limited resources to fetch any good old tubes. That's why I said I don't like the cliche euphonic tube sounding. I'm sure yours must be good.

 

 

43 minutes ago, barrows said:

Certain types of jitter artifacts can be perceived as euphonic, considering some listeners preferences for sound: for example, a widening of the skirt on the plot is often associated with an "expansive" soundstage, though we should recognize this as an artifact, and not indicative of higher fidelity conversion.

@barrowsThis is intersting. I do like dialectic thinking and having a corrective approach to guide our listeners. 

 

Like Quadman said, we shall wait for the measurement as I did not find any DSD256/512/1024 measurment of MAY. 

 

In the meanwhile, I will go to T+A thread, trying to get what Lothar thinks about DSD 1024 as T+A also encourages such DSD 1024 sampling rate. I can understand people might think the scale of Holo audio company is not that big and imperfection exists. On the other hand, I believe T+A has more solid R&D capabilities and more preliminary tests and measurments before launching a product. 

 

As I said before, my friend used T+A DAC 200 at DSD1024 through HQplayer, he is feeling the same "magic airy vibe".  I don't consider these identical findings from two different DACs as concidence. These can be explained and measured for sure.

 

Either HOLO and T+A are having the same issues at such high rate and we audiences need to be properly guided instead of being dazzled by the magic or such high rate does offer some advantages. 

[louisxiawei]

3 hours ago, Miska said:

But you need to ditch USB and move over to Ethernet or other better technologies. Main limiting factor at the moment is USB Audio Class. Especially for multichannel, since the available bandwidth drops as function of number of channels

I feel disappointed about this. There is still no standard set for the ethernet audio class yet in this industry. Only a few have their own non-standard protocols (merging and playback) but these protocols do not fit the network definition but point to point. If no mistaken, T+A flagship SDV3100 has an ethernet port for NAA, but it's also a kinda USB-ish connection on their circuit board. 

 

HIFI audio remains at a small scale and people who in this small circle want higher bandwidth for upsampling purpose are the minority of the minority. 

 

Asking for over 1GHz or several hundred Mbps bandwidth just for audio data transmission seems to be ridiculous for the majority. 

 

Hopeless.

[louisxiawei]

35 minutes ago, The Computer Audiophile said:

AES67 is a standard. Ravenna builds upon that standard and can work point to point, over LAN, or even over WAN. 

Yes, I got it wrong. Merging is using AES67. 

 

I remember there is a Network-capable DACs thread and the disucssion just ended there at AES67. Miska has comment several times that AES67 is far from satisfication for a real network practice and I understand his concern.

 

Just check his comment in 2017, AES67 is more of a "go with flow", a real network is "push and pull".  I strongly echo that. If Ethernet protocol is designed to behave like point to point, then it is still an USB-ish protocol disguised by the ethernet port.

 

Anyway, AES67's bandwidth cannot meet the GHz data transimission, lol.

[louisxiawei]

48 minutes ago, Miska said:

And it is relatively easy for manufacturers to include support for various custom protocols as we can see from products like iFi NEO Stream etc.

 

One reason I created a custom protocol for HQPlayer is that there was no standard that would be sufficient for HQPlayer requirements. Likely similar thing was for Roon too.

 

The reason I feel disappointed is that I consider ethernet has the potential to be a good candidate to replace the audio USB as it is galvanic isolated in nature while having proper bandwidth yet very few elaborates. I assume things may not be that easy to implement in practice.

 

Audio USB port is not audiophile at all yet universal. What I meant to say is that we audiophiles need a unverisal port at the back of any DACs that designed for the purpose of high fidelity playback at sufficent bandwidth while fitting for all sorts of custom networking protocols. USB port is definitely not the best choice, it's for mice, keyboard and printers (even my BROTHER printer transmit data via ethernet cable).

[louisxiawei]

1 hour ago, barrows said:

Ultimately, ideally, I would prefer to see a DAC have a well implemented optical Ethernet receiver built in, but except for one or two cases, we are not there yet, and it is a bit expensive to implement a really good Ethernet receiver inside the DAC chassis.  But for now, USB is usually the best choice, as many DAC developers have figured out how to implement USB without compromises.

I'm a plain computer audiophile and I have a dream:

 

For example, if there is a unversial ethernet protocal available on the market for DACs, I want my dac to connect to a well-desgiend switch by a solid enterprise like Hewlett-Packard or NETGEAR not a self-called "audiophile" small company that DIY/mod swtiches and call it "audiophile" switches. Because I consider an "audiophile" mark on any IT product really mess things up (things like switch, router, ethernet cables, PCs) and I know people would argue. I do not think audiophile companies have more experience and talents in designing switch, routers than dedicated IT service cooperates. That's why I think the ethernet can potentially be a good opportunity to standardisze IT gadget for the audio playback purpose and leave little-to-no chance for those opportunists eyeing on the audiophile market. We can have a powerful PC with full of high frequency noise, polluted PSU, doing the HQplayer upsampling without worrying "should I mod my PC and my USB port for audiophile purpose?"  

 

I know DACs manufactures are more capable of handling USB now than 5 years ago as I used to have to use products like Uptone regen, Intona, microrendu and "audiophile" USB cable to have a proper SQ from PCs to DAC. Even though, no matter how well the DAC is designed in USB, it will still sound better at different degrees by connecting the dac to another audiophile grade PC, renderer or a USB-regen dongle. That's why I consider USB is flawed. I do appreciate these USB pheripherials and they do bring the sonic benefit but for me it is just painful and not a minimalistic solution. If things can be minimal, why bothering the daisy chain?

 

I know I'm being ideal now but some products are necessailry to be audiophile, some don't if they are so well-designed at the first place. The marketing game don't look that way, but the industry standard set by the talents with know-how can make things a little better.

 

 

 

[louisxiawei]

4 minutes ago, Quadman said:

So if one only plays 2 channel with no convolution can USB handle sample rates up to DSD2048 or are we pretty limited to 1024?

2 channel 2048 will be at speed of around 180 Mbps. USB 2.0 is around 480 Mbps in theory.  So should be ok. But I would worry other things like whether the clock in the dac can handle DSD 2048 properly.

[louisxiawei]

38 minutes ago, barrows said:

For perfectionist audiophiles, everything is going to matter to some degree, power cables, Ethernet cables, USB cables, SPDIF cables, DC power cables, etc...

We audiophiles sometimes need to find the silver line between fussy OCD and perfection. Many of us can't help and consider the OCD as a perfection instead of a waste. I had my lesson and got my diminishing return.

 

38 minutes ago, barrows said:

The problem with commercial computer peripherals, is that they are not designed to produce best sound quality for music, such is just not a concern in their design briefs, and cost is a huge factor in engineering these types of products.  Engineers of such are tasked with producing the Switch which meets the design goals, at the absolute minimum cost, this is what they take pride in, and the engineer who can meet the design goals at the least cost wins.  When an engineer is contracted as a "hired gun" for an Audiophile company, they are advised to not be so cost conscious, if the switch ends up costing three times as much in order to reduce noise, that is totally OK.

Why would you think the enterprise grade switch is inferior to the "audiophile" switch? Just because audiophile company offers a holy contract that the final goal is a cost-no-object project to reduce noise? Even the goal is not for audio playback, the quality of its function is there. EEE for lower transmit power for shorter cables, flow control for smooth data transimission....I don't think these features were invented by an audiophile mindset yet it brings benefit to the audio playback.

 

Why not the other way around? if there is an audio playback goal set for those big enterprise like NETGEAR and HP, they are more capable to launch a product at the absolute minimum cost than companies soley with audiophile mindset that are struggling surviving on this niche market.

 

Anyway, ted has mentioned we are off the topic. So I will stop this disucssion. 

[louisxiawei]

There should be a category in this forum pinned at front page with a catchy notice like "something you must know before ramblings" and what miska said about upsampling/oversampling must be put there. Lol.

[louisxiawei]

7 hours ago, Extreme_Boky said:

 

You posted on forum that is constantly  patrolled by people who have vested interest in promoting upsampling/oversampling DAC's and software. They'll do anything to sell their product, and would refrain from nothing to make you feel stupid. 

 

If you are really interested in DSD VS PCM, or up sampling / oversampling... do your own research.... read a lot. I have attached something, if you are willing to read.

 

It doesn't really matter what anyone says. The only thing that matters is what sounds best to you. But, be open to options and do not discard any.... after all, the options are made by others to satisfy YOUR needs - as a consumer. 

 

sacd.pdf 202.63 kB · 13 downloads

 

Why so cyncial and assume things are conspiratorial?

 

Upsampling and DSP are based on a terrible precondition that music industry does not intend to release everything at mastertape quality to the public. In stead, 44.1/16, MQA these lossy files are still the majority for the audiences. Compared to the audio circle, visual technology on the other hand is far more progressive. For example, are you satisified with 720p resolution when there is a 1080p or 4k resoulation?

 

Let's just follow your logic and assume people like Miska has their hidden agenda and want to promote his upsampling software/DACs as much as possible to make a fortune and make other people against upsampling look stupid. Then where is the logic people here repeatedly emphaize that DSP and upsampling would be totally unnecessary if the files are hi-res enough at the first place.

 

Common sense were established years ago here and audiophiles are just trying what they can at the current status quo, which is the whole audio industry is not progressive at all.

 

People believe what they believe in most cases and do pick a side to suit their needs. It's all about preconscious and subconscious. Your attached file is quite a good sign of you are not reading a lot.