Jump to content

bhobba

Auralic Aries And Clays DAC's

Recommended Posts

7 hours ago, bhobba said:

 

It was Joe Rasmussen's latest effort.

 

Don't read too much into it at this stage - it was a first firing up and does a lot more such as video than the other stuff, plus is a  cheaper than things like the DS or Clay's Konveter.

 

I have one myself, but I use it for HT duties that IMHO it excels at.  Like I said its a bit of a different beast.  It has an upgraded clock that should improve the feeding of the DS I use an Oppo for - but I haven't checked it yet.  Joe says he has been at comparisons where it was better than the DS so I think it deserves more time to judge.

 

Again a GTG seems more and more whats needed.

 

Thanks

Bill

Hi Bill

I  still remember when we compared the modded 105 to a few other dac's at Mike's place.

Was just making sure it was a 205 as I was reading elsewhere that he was looking for 205's to mod.

From my experience, I preferred Clay's Klein dac to the modded 105 Oppo. (It will take a lot to convince me that the Oppo is better than a DS)

I am always interested to hear about the latest developments and I trust your, Mike and Clay's ears :love. Please keep posting.

 

 

Share this post


Link to post
Share on other sites
1 hour ago, Gieseler Audio said:

but yes you have basically got that right AFAIK.

That can only be good news for consumers and DAC manufactures.

Share this post


Link to post
Share on other sites
7 hours ago, Gieseler Audio said:

Hi Dan,

Bill is the expert on that but yes you have basically got that right AFAIK.

I think Auralic was working with the MQA group towards getting their devices hardware approved when they had a falling out.

Auralic had put quite a bit of effort in the MQA unfolding process so have implemented their own version of it in their products but obviously are not MQA certified. Anyway that is my take on it so please anyone that knows the full gist of it please explain.

 

This is like I see it too. Auralic developed their own MQA decoding but have NO support from MQA. Basically they do not like MQA at all. This contribution is from Auralics FB site in April:

https://music-room.com/magazine/insight/mqa-the-lossy-codec-no-end-user-asked-for-or-needs

 

Matt

Share this post


Link to post
Share on other sites
Moderator

Thank goodness someone that sees beyond the marketing  spiel.

Share this post


Link to post
Share on other sites

This is how I see it. Anything that is going to be streamed is going to have some sort of compression. Tidal Hifi is better than Spotify and if Tidal Masters are better than Tidal Hifi so be it. You embrace technology as they come. Some are fads, some stick along for longer. I might be wrong but DSD wasn't as readily available as MQA. Correct me if I'm wrong but didn't you have to buy DSD albums/tracks individually? If MQA is the highest quality streaming option we've got right now I don't see anything wrong with embracing it. Lets be honest, down the line someone is going to come up with some other format which will claim to be better than MQA. If you can incorporate MQA into your system without too much of an outlay I don't see it as such a big risk. I haven't read about what MQA claims to be, but from the little I've read it's the best we've got. I say just give it a go until something better comes along.

Share this post


Link to post
Share on other sites
5 minutes ago, Cardiiiii said:

Correct me if I'm wrong but didn't you have to buy DSD albums/tracks individually?

No, DSD is just what SACD is and there still a lot of those out there.

Share this post


Link to post
Share on other sites
5 minutes ago, powerav said:

No, DSD is just what SACD is and there still a lot of those out there.

 

But do you still have to buy them individually as albums/tracks as opposed to being available on a subscription basis like MQA on Tidal?

Share this post


Link to post
Share on other sites
18 minutes ago, Cardiiiii said:

But do you still have to buy them individually as albums/tracks as opposed to being available on a subscription basis like MQA on Tidal?

http://www.hdtracks.com/dsd-albums yes I on't think there is a streaming option, just like Hi Res the files are too big, too much data.

Share this post


Link to post
Share on other sites
Posted (edited)

@Cardiiiii

Hi Joe,

I do not agree with your opinion about MQA, but I like what Auralic do regarding MQA. They offer their own MQA decoding to customers but are not bound to the restrictions MQA as a company impose on their "partners".

 

Matt

Edited by matth

Share this post


Link to post
Share on other sites
11 minutes ago, matth said:

I do not agree with your opinion about MQA

Yes after listening to 2 minutes of a MQA track I went back to the normal FLAC one, man the MQA is so loud it's unbearable.

Share this post


Link to post
Share on other sites
Posted (edited)
On 2017-5-19 at 6:39 AM, powerav said:

So just to be clear. For the Aries streamers you don't need a MQA Dac because the Aries version 5 will spit its own version of MQA out, is that right?

 

Yes.  Any old Dac that does at least 178 and 196 will work.

 

Just as an overview here is whats going on in simple terms.  MQA losslessly encodes 24 to 48k stuff and puts it in the last 7 bits where its simply sounds like noise - so you get 17bit resolution instead of 24 but through a trick called subtractive dither without going into it (you simply subtract it from the 24 bits - strip the last bits and put in the new 7 bits), its effectively something like 22 or 23 bits anyway so you do not really loose anything.  Now MQA allows anyone to extract the data so you get 48k - that's called the first unfolding.  However in creating the 96k PCM they use a trick - what the trick is, is say its 384khz PCM - it can be anything even up to many mhz - is you down-sample it to 96k by means of overlapping functions called Bessel functions.   The simplest is a simple triangle.  So you take 1/4 of a sample, then 1/2 of the next sample, then 3/4. then all the next, then 3/4 the next, then 1/2 the next, then 1/4 the next.  The triangles overlap.  It has the effect of keeping every 4th sample but chucking away the other three.  It also causes a slight roll-off at about 6db per octave above about 20 k that is pretty much inaudible - but they correct for it anyway, so its within 1/2 db up to 20k but then rolls of at about 6db.   The reason they do that is its not a brick-wall filter like you would have to use in normal down-sampling.  When music hits such a filter it rings which is why High Res sounds better.  However it has the problem of 'reflecting' the content above 48k into the normal audio band.   Because of the mild filter used and the fact HF content naturally drops off this reflecting is inaudible so isn't a problem - its much lower than normal noise in the audioband.  Now lets say we want to get the full 384 back.  Its can be shown, and its rather intuitive anyway - you simply linear interpolate between the  96k samples to get the best approximation to the original (this is sometimes called sparse compression).  Its isn't exact - but in that frequency range it's pretty much all noise anyway so the inaccuracy done by this is like adding extra noise so who cares.

 

The up-sampling by linear interpolation is called the second unfolding and is in fact a digital filter optimized to match the encoding trick they use.  But MQA goes further.   It analyses the music to determine the optimum encoding method even trying to correct for issues in the original analogue to digital converter.  It then uses the corresponding optimal up-sampling filter at the DAC claiming to correct for issues in the DAC as well which is why they only will give the MQA designation to DAC's not devices like the Auralic.  Thats a lot of the MQA controversy - some like me think that last part about correcting errors in the DAC is BS.  It's the falling out Auralic had with MQA and said F you to them.  The encoding method used is put in metadata in the audio stream using inaudible methods.

 

Now Aurilic will not say what they are doing with MQA but here is my guess.  They are doing the first unfold - anyone can do that.  But for the last unfold they came up with their own up-sampling method they use for all MQA files.  They rekon their method is better.  We will find out with my Explorer DAC so we can compare all the methods.

 

Whats my bet? There will be virtually nothing in it - but we will see.

 

I have to head off for lunch and will write up my impressions of the listening session Clay talked about after.

 

Thanks

Bill

Edited by bhobba

Share this post


Link to post
Share on other sites
Posted (edited)

Ok - my views of the listening session.

 

First we tried Joe's upgraded Oppo with a 196k DVD audio disk.   Not the best - a bit dull, dry and flat.   But it was brand new and is well known not to be great with disks.   We will try it later by plugging in some ripped disks to the usb input and casting from Tidal and disks plugged into a Windows machine - the Oppo can receive casts.  My phone casts Tidal and casting is built into Windows 10 - its how I listen to my system using the  Oppo.  I think the Oppo sounds best that way.  I use it to connect to my DS.

 

Then Clay spent some time getting the Aries set-up and while he was doing it Mike plugged an Ethernet cable into the Direct Stream Junior meaning my phone, which has Bubble-Upnp can transmit to it - so I did.   Instantly better, but it has, as I have noticed before, this 'thick' quality, not bad but - well at least a bit different if not an issue.  However PS Audio is doing a lot of work on the Bridge, MQA and all that so the final verdict will need to wait until that's finished - its taking longer than expected though.

 

Then Clay got the Aries LE working - WOW - instantly better.  We listened to Blue which is MQA encoded and Mike, who is a very experienced and critical audiophile knew a particularly hard track in that album so we played it.  He closed his eyes - I know that track as well and it was simply the best rendition I have heard - period.  Mike however, while admitting its likely the best he has heard picked up some slight vocal coloration.   Its cause - Mike thought the newly minted ML5's - but we don't know - further investigation is required which Mike will do as time permits.   Its also when he 'conned' me into selling him my LE.   Ok I had asked Clay to build me a supply for a uptone regen I had - Mike had a regen (he sells them with his Curios cables) so used one with Clay's supply.   Instantly better - less grunge, clearer and with more detail.  I immediately asked Mike to get me one of the new ISO regen's and he will get one for his shop.   That should make things better again.

 

That's where it all stands now - its better that any other server Clay or Mike has tried but exactly where it all stands will take some time to sort out.   It should be a lot better once Clay builds one of his supplies to run the LE.

 

I will give it a week or two for Clay to finish the LE power supply and see how it stands then.

 

Once its all sorted out ie I get a proper audio cable for the Explorer 2, Auralic firmware version 5 comes out of beta, the new DS software updates are released, new power supplies are made and the ISO Regens arrive we will have a GTG which should prove VERY interesting.   Clay may even have his Konverter without the pre there - who knows.

 

Thanks

Bill

 

 

Edited by bhobba

Share this post


Link to post
Share on other sites
Posted (edited)

Hi guys,

unfortunately it was not possible to evaluate seriously the sound quality of the new Auralic G2 series at High End Munich.

For me the most interesting question is how the upcoming Audirvana 3.1 with UPnP/DLNA output to microRendu or SMS-200 will sound in comparison to the Aries.

 

Matt

Edited by matth

Share this post


Link to post
Share on other sites
Posted (edited)

Hi Guys

 

The following sheds some interesting light on whats going on with the Aries and MQA:

https://www.computeraudiophile.com/forums/topic/33766-upsampling-mqa-files-to-original-resolution-with-sox-will-sound-like-the-original-resolution/

 

Basically all the second unfolding of MQA does is try and get an approximation to the 192k, 384k, DXD or whatever was used as source.  In doing so it introduces aliasing components etc that may or may not be filtered depending on if the output stage of the DAC does it anyway. For example Clay's Konveter and his upcoming new DAC uses a Jensen transformer as the output - so does the Direct Stream.  They filter anything above about 100K anyway so no need to do anything about the ultrasonic aliasing components.

 

But what this guy discovered, and he and I expect Auralic also discovered, is why bother?  Simply do some good usual up-sampling like minimum phase SOX.  It greatly reduces aliasing but at the cost of post ringing.   The thing is unlike pre-ringing which even small amounts sounds bad (you never hear things like that in nature) post ringing is pretty much inaudible because of the natural decay of transients like drums etc in real life.   So it looks like that's what they did for the second unfolding - just use a high quality minimal phase normal up-sampler.

 

I have heard both the Auralic and now that PS Audio has released its MQA Bridge 2 that as well.  It was not a direct comparison - but from memory it was close.

 

I think we can expect more devices like the Auralic with their own unfolding now this little secret is out, and more choices of what to use with Clay's DAC's.  For example it would be interesting to hear how Audirvana sounds - it does the first unfolding then you can upsample it to whatever you like.

 

Thanks

Bill

Edited by bhobba

Share this post


Link to post
Share on other sites
1 hour ago, bhobba said:

Hi Guys

 

The following sheds some interesting light on whats going on with the Aries adn MQA:

https://www.computeraudiophile.com/forums/topic/33766-upsampling-mqa-files-to-original-resolution-with-sox-will-sound-like-the-original-resolution/

 

Basically all the second unfolding of MQA does is try and get an approximation to the 192k, 384k, DXD or whatever was used as source.  In doing so it introduces aliasing components etc that may or may not be filtered depending on if the output stage of the DAC does it anyway. For example Clay's Konveter and his upcoming new DAC uses a Jensen transformer as the output - so does the Direct Stream.  They filter anything above about 100K anyway so no need to do anything about the ultrasonic aliasing components.

 

But what this guy discovered, and he and I expect Auralic also discovered, is why bother?  Simply do some good usual up-sampling like minimum phase SOX.  It greatly reduces aliasing but at the cost of post ringing.   The thing is unlike pre-ringing which even small amounts sounds bad (you never hear things like that in nature) post ringing is pretty much inaudible because of the natural decay of transients like drums etc in real life.   So it looks like that's what they did for the second unfolding - just use a high quality minimal phase normal up-sampler.

 

I have heard both the Auralic and now that PS Audio has released its MQA Bridge 2 that as well.  It was not a direct comparison - but from memory it was close.

 

I think we can expect more devices like the Auralic with their own unfolding now this little secret is out, and more choices of what to use with Clay's DAC's.  For example it would be interesting to hear how Audirvana sounds - it does the first unfolding tyhen you can upsample it to whatever you like.

 

Thanks

Bill

 

Bill it really depends whether you want to be an anorak with total control over output waveforms (which is what MQA ultimately intends to do) or whether you want to unpack additional content to create a better approximation of an analogue ideal. 

 

Whilst we tend to speak of analogue reconstruction in terms of signal processing in definitive terms, it's not so exactly true: MQA (as with all ultra-high-res PCM content) is capable of very high-frequency response, though that capability isn't used to create sounds we can't hear, it's used - in a digital signal processing context - to create waveforms that are difficult to replicate exactly at lower sample frequencies (for many reasons) using PCM methods. Similarly those transformers don't filter out in a brickwall context.  

 

I would doubt the second unfolding is just upsampling, but I get your drift, and it's not a bad one. 

 

There will always be a market for those wanting the last word, and I suspect authentic MQA will be there to some degree - some and in time diminishing, as DxD and other master content is less expensive by the day, MQA in a streaming context inherently means a lossy remaster - getting off that 'last word' paradigm - I suspect the hard line Meridian has taken on rights ownership in the MQA process is inconsistent with competitive product pricing for many, so as with you I suspect this isn't the last we'll hear of such devices. 

 

Now I'm just waiting for someone to code this into a Linux driver or Amanero-type device and the rest of us can get on with it :) 

Share this post


Link to post
Share on other sites
1 minute ago, rmpfyf said:

I would doubt the second unfolding is just upsampling, but I get your drift, and it's not a bad one. 

 

It isn't.  Its key to the temporal smear MQA tries to address - but it comes at a cost - aliasing components - however they are usually high enough in frequency so many DAC's like Clays naturally attenuate them anyway.  Minimum phase up-sampling doesn't care a hoot about temporal smear but greatly reduces aliasing components - its downside is post ringing.

 

Which is better - only listening and time will tell.   Of course MQA claims temporal smear is the key - the guy that wrote the link respectfully disagrees.

 

Personally I doubt his claim its audibly indistinguishable from DXD, just as I doubt the claims MQA is better than DXD.  However its real purpose is to lift the quality of streaming so that eventually it will be the main game in town for audiophiles.  Of course there will always be those that want the very best obtainable. Even I partake of that.   I have a collection of SACD and DVD Audio rips passsed through an entropy encoder I also use, but these days my main game is Tidal.

 

Thanks

Bill

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


  • Recently Browsing   0 members

    No registered users viewing this page.

×