Auralic Vega: my review

[Wang Xuanqian]

My laptop almost never recognizes Vega on reboots, same problem I had with iLink and HiFace2, all XMOS based. What works for me is putting DAC in sleep mode and turning ON after computer starts. When I forget doing that before reboot, cycling in and out of sleep mode usually works as well, last resort is powering it down, trying to avoid that as it takes an hour for femto clock to warm up to use EXACT setting.

 

It seems to be a common problem with XMOS and we have been reporting this since January this year but never get resolved. I doubt if there are some bug inside their chip but who knows...

What I do is just switch to other input channel and switch back again, usually it resolve the problem.

 

Will try out the new SU1 series XMOS chip next month, hope they have fixed this bug in the new hardware.

[ejn1]

ASIO with DoP correct?

 

yes, ASIO and then have DSD over DoP setting.

[jtwrace]

Will try out the new SU1 series XMOS chip next month, hope they have fixed this bug in the new hardware.

 

What's new about it? Anything sonically or features?

[Vitor]

 

It seems to be a common problem with XMOS and we have been reporting this since January this year but never get resolved. I doubt if there are some bug inside their chip but who knows...

What I do is just switch to other input channel and switch back again, usually it resolve the problem.

 

Will try out the new SU1 series XMOS chip next month, hope they have fixed this bug in the new hardware.

 

On a diferent note Mr. Wang, and since you are here, can you elaborate about the fact that whenever the EXACT mode is choosed in the clock setting i get dropouts?

However, they don't seem consistent, sometimes they "appear", sometimes not.

It's a shame, because it only makes sense to use the EXACT mode, which is clearly the best clock setting.

To contextualize, i'm using the Vega with a MacBook Pro using Audirvana+.

Thanks in advance.

[jtwrace]

On a diferent note Mr. Wang, and since you are here, can you elaborate about the fact that whenever the EXACT mode is choosed in the clock setting i get dropouts?

However, they don't seem consistent, sometimes they "appear", sometimes not.

It's a shame, because it only makes sense to use the EXACT mode, which is clearly the best clock setting.

To contextualize, i'm using the Vega with a MacBook Pro using Audirvana+.

Thanks in advance.

This is the email response I got back from AURALic (Dewei Kong) when I asked that question.

 

Firstly, I have to claim that both EXACT and FINE are experiential options that designed only for testing, AURALIC do not ensure you can user both of them without drop out or hiss noise during daily use.

 

To see both options, you need to keep the unit power up uninterrupted for one hour then the options will show up in the menu. Put the unit into sleep status will not effect the show up of two options.

[Wang Xuanqian]

What's new about it? Anything sonically or features?

 

Nothing related to sonic I believe. XMOS just put a lot of around circuit such as USB PHY, reset circuit inside the chip inside the SU1 chip so it will request less components on the board.

[Wang Xuanqian]

On a diferent note Mr. Wang, and since you are here, can you elaborate about the fact that whenever the EXACT mode is choosed in the clock setting i get dropouts?

However, they don't seem consistent, sometimes they "appear", sometimes not.

It's a shame, because it only makes sense to use the EXACT mode, which is clearly the best clock setting.

To contextualize, i'm using the Vega with a MacBook Pro using Audirvana+.

Thanks in advance.

 

EXACT and FINE are for fun. We never promote this function as we can not ensure everyone can use it. The value of VEGA is not just a funny function but the entire design, that's the reason we got multiple awards. We asked all reviewers to judge the sound of VEGA by clock setting at AUTO and they I can see them all following the guideline.

 

By the way, OS X never works in our lab or show room. The only machine works without any problem of EXACT is C.A.P.S V3. If you are seeking for ultimate sound quality, you should probably goes to a customized Windows or even a customized distribution of Linux since both are far superior to Apple. OS X is just easier to use for most people.

[Andrew_WOT]

Was always wondering how far AUTO mode goes, the manual says

AUTO (default)

VEGA will use the best possible internal clock precision to lock on any inputs.

This function provides perfect balance between sound quality and locking

stability.

Can it go into FINE or EXACT, is there chance to get some indication on actual effective clock for AUTO in the future firmware upgrade?

 

And another question on ActiveUSB, how exactly that "up to 2 seconds buffer" mentioned here works, I definitely do not hear any delay when starting playback. Had Chord DAC64 before and on that one the internal buffer was adding significant latency on playback startup.

The USB input implementation here is special. Data enters the Vega via a firmware-tweaked XMOS receiver chip, is passed to a Sanctuary processor chip where it is buffered for up to two seconds – AURALiC call this ActiveUSB technology – and then up-sampled to 32bit/1.5MHz before being sent along the chain to DAC silicon (Sabre ESS 9018) and analogue output stage (the aforementioned ORFEO Class-A module).

[Wang Xuanqian]

Was always wondering how far AUTO mode goes, the manual says

 

Can it go into FINE or EXACT, is there chance to get some indication on actual effective clock for AUTO in the future firmware upgrade?

 

And another question on ActiveUSB, how exactly that "up to 2 seconds buffer" mentioned here works, I definitely do not hear any delay when starting playback. Had Chord DAC64 before and on that one the internal buffer was adding significant latency on playback startup.

 

AUTO is dynamic adjustment but the stability is 1st priority. Our Sanctuary processor is a dual core ARM926 and we have programmed a FIFO that can utilize the external SDRAM, which is 64MB, to buffer the data. However we do not use it to maximum as the delay will be unacceptable. Most of the times, the latency are pretty small as the program will detect the input signal's stability, the maximum condition I have experienced is only 100ms.

 

We will introduce a new quad core Cortex-A9 computing solution in our network streamer coming next year. The new CPU have the capability of 12,000MIPS compare to Sanctuary's 1,000MIPS. It will enable us to stream and output DSD/DSD128 and DXD though USB to any DoP compatible DAC and also possible for high quality DSD upsampling and high precision room EQ at digital domain before DAC.

[Vitor]

EXACT and FINE are for fun. We never promote this function as we can not ensure everyone can use it. The value of VEGA is not just a funny function but the entire design, that's the reason we got multiple awards. We asked all reviewers to judge the sound of VEGA by clock setting at AUTO and they I can see them all following the guideline.

 

By the way, OS X never works in our lab or show room. The only machine works without any problem of EXACT is C.A.P.S V3. If you are seeking for ultimate sound quality, you should probably goes to a customized Windows or even a customized distribution of Linux since both are far superior to Apple. OS X is just easier to use for most people.

 

Despite the fact that you didn't "promote" them doesn't mean they shouldn't work the correct way in all platforms (in cd I have no problems with it, only via USB).

Saying that it works well in Windows/Linux environment and not iOS isn't, by all means, an answer I expected.

Why is even that feature available if it's not fully tested in all environments, that's the question. If it was me, I would've preferred not to include them at all.

This is just my opinion, besides that all I have to say is that the Vega is a wonderful DAC for the price.

Cheers!

[Andrew_WOT]

Despite the fact that you didn't "promote" them doesn't mean they shouldn't work the correct way in all platforms (in cd I have no problems with it, only via USB).

Saying that it works well in Windows/Linux environment and not iOS isn't, by all means, an answer I expected.

Why is even that feature available if it's not fully tested in all environments, that's the question. If it was me, I would've preferred not to include them at all.

This is just my opinion, besides that all I have to say is that the Vega is a wonderful DAC for the price.

Cheers!

I guess that's the problem is with Apple and their USB 2.0 Audio driver implementation, nothing Auralic can help with, I am afraid.

FINE and EXACT as manual says should be reserved for purposely build for playback computes "For USB input, you will experience drop off on any computer without special optimization for music playback".

The fact that many here are getting away by running it w/o any drop outs on generic purpose Windows on Linux computer is just a nice bonus. It does require some tinkering to make it stable like adjusting ASIO buffer size, disabling USB Power saving, bumping up playback thread priority, etc.

[dr.sah]

AUTO is dynamic adjustment but the stability is 1st priority. Our Sanctuary processor is a dual core ARM926 and we have programmed a FIFO that can utilize the external SDRAM, which is 64MB, to buffer the data. However we do not use it to maximum as the delay will be unacceptable. Most of the times, the latency are pretty small as the program will detect the input signal's stability, the maximum condition I have experienced is only 100ms.

 

We will introduce a new quad core Cortex-A9 computing solution in our network streamer coming next year. The new CPU have the capability of 12,000MIPS compare to Sanctuary's 1,000MIPS. It will enable us to stream and output DSD/DSD128 and DXD though USB to any DoP compatible DAC and also possible for high quality DSD upsampling and high precision room EQ at digital domain before DAC.

 

 

Any more details on new streamer?

[phaeton]

Hi everyone,

 

I'm a Vega owner since a couple of months, from the very beginning my Vega started to have some weird but not so serious problems, but today a very serious problem showed up....

 

Problem number 1: sometimes when I turn on form sleep mode, the display will not show the frequency, even if music is playing. That's not so much annoying, it just needs to switch freq and it will restart showing, I'm reporting this just as a little bug. This is a problem that showed up from the beginning.

 

Problem number 2: sometimes, even when playing the same album (same freq) but not during playback of the same file, the filter mode 4 starts to cut high freq in a very audible way, and also starts to distort very low freq. To get back to normality a switch of the filter mode is needed. Unfortunately I have not tested with other filter mode, since this is something that happens randomly and usually I only use filter 4. This is somewhat annoying, and I'm reporting it as a functional bug. This is a problem that showed up after a couple of days of so of use.

 

Problem number 3: whenever I play a dxd file (352.8k) with filter 1 I hear a lot of distorsion. This is something very strange, because I can hear a lot of "scratches" (like when you have a deeply damaged vinyl) even with music stopped and even with the usb cable unplugged. This is happening only with filter 1 and 352.8k freq, every other combination of filter/freq is working just fine, including dsd64/128. This is very annoying, since even if I do not have so much dxd file, I'm very worried that this is just the start of a more serious broke of the Vega. This is a problem that I discovered today but I don't know when it really started since I usually use filter 4, and I consider this a serious hw problem. This problem will show up with the Vega just switched on (no sleep mode), or going up from sleep mode, and it is completely repeatable (while the previous 2 are completely random event).

 

Would like to ask to other Vega owners if they are experiencing the same problems, or a subset, or maybe my unit is definitely going broken, and before it breaks completely I have to send for assistance.

 

My Vega is a HW version 1.2, SW version 1.5, S/N 1586

 

Can someone help, or do I have to send the unit for assistance?

 

rgrds

mike

[Andrew_WOT]

Hi everyone,

 

I'm a Vega owner since a couple of months, from the very beginning my Vega started to have some weird but not so serious problems, but today a very serious problem showed up....

 

Problem number 1: sometimes when I turn on form sleep mode, the display will not show the frequency, even if music is playing. That's not so much annoying, it just needs to switch freq and it will restart showing, I'm reporting this just as a little bug. This is a problem that showed up from the beginning.

 

Problem number 2: sometimes, even when playing the same album (same freq) but not during playback of the same file, the filter mode 4 starts to cut high freq in a very audible way, and also starts to distort very low freq. To get back to normality a switch of the filter mode is needed. Unfortunately I have not tested with other filter mode, since this is something that happens randomly and usually I only use filter 4. This is somewhat annoying, and I'm reporting it as a functional bug. This is a problem that showed up after a couple of days of so of use.

 

Problem number 3: whenever I play a dxd file (352.8k) with filter 1 I hear a lot of distorsion. This is something very strange, because I can hear a lot of "scratches" (like when you have a deeply damaged vinyl) even with music stopped and even with the usb cable unplugged. This is happening only with filter 1 and 352.8k freq, every other combination of filter/freq is working just fine, including dsd64/128. This is very annoying, since even if I do not have so much dxd file, I'm very worried that this is just the start of a more serious broke of the Vega. This is a problem that I discovered today but I don't know when it really started since I usually use filter 4, and I consider this a serious hw problem. This problem will show up with the Vega just switched on (no sleep mode), or going up from sleep mode, and it is completely repeatable (while the previous 2 are completely random event).

 

Would like to ask to other Vega owners if they are experiencing the same problems, or a subset, or maybe my unit is definitely going broken, and before it breaks completely I have to send for assistance.

 

My Vega is a HW version 1.2, SW version 1.5, S/N 1586

 

Can someone help, or do I have to send the unit for assistance?

 

rgrds

mike

Hi Mike, I recommend to email Auralic, most likely it's the same problem I have. I'll pm you on this.

[phaeton]

Hi Mike, I recommend to email Auralic, most likely it's the same problem I have. I'll pm you on this.

 

ok, thanks.

[tiff_needle]

Hi Mike, I recommend to email Auralic, most likely it's the same problem I have. I'll pm you on this.

 

Keep the rest of us posted, please.

 

The vega is on my short list and I would like to be aware of any bugs other users are experiencing...

[Andrew_WOT]

Keep the rest of us posted, please.

 

The vega is on my short list and I would like to be aware of any bugs other users are experiencing...

I'd rather let Xuanqian handle it, my understanding that this is isolated issue affecting very small number of units.

[phaeton]

Mr Wang has already responded to my mail!!!!

 

Kudos to Auralic for their customer service, really first class. I'm really happy to have bought Auralic products, since I can always count on a first class support!

[Andrew_WOT]

Mr Wang has already responded to my mail!!!!

 

Kudos to Auralic for their customer service, really first class. I'm really happy to have bought Auralic products, since I can always count on a first class support!

Second that, they quickly identified the problem and provided solution. All is good now. Two thumbs up support.

[Andrew_WOT]

Sometimes I am getting this white noise on start of DSD128 tracks when switching back and forth between PCM, DSD64 and DSD128.

Restarting playback usually fixes it. This is foobar2000 with foo_dsd_asio.

 

I'm at the 500 hour mark.

 

Stunning DAC.

 

Some notes:

 

Got DAC warmed up and into Exact mode.

 

Native sample rate for PCM and DSD - seemed like DAC would start to get stuck with dropouts also noted white noise zing at start of some DSD - like switching between DSD and PCM was too muc for it.

 

Tested jplay ASIO, vega WASAPI, and vega ASIO

 

WASAPI limits to 24bit

 

Raising the JRMC19 buffer to 100ms has cured the dropout/white noise issue but I'm also up sampling all PCM to 384 (wanted to see if DAC played smoother i.e. Less dropouts without having to switch sample rates)

 

Exact mode using Adnaco S3B is fine with vega ASIO - USB mode min. Latency and ASIO buffer in vega control panel at 2048

 

Completely powering down the vega seemed to help reset it so I have had no more dropouts after doing a bunch of pcm2dsd testing which seemed to through the DAC into an unstable state

 

I think possibly since the DAC has an ARM chip, maybe this is possible. Mr wang did tell me to reboot the DAC actually.

 

Jrmc19 does not support 2xdsd real time DSD upsampling over DoP so you have to use jplay or foo ASIO proxy

 

Running direct to tube amp with 2.5v input so 60-70 really loud but doable. Building a light speed with Tortuga board. Seems like best bang for buck and leaving resistor attenuators behind will be fun, and it's cheap.

[hifitubes]

Sometimes I am getting this white noise on start of DSD128 tracks when switching back and forth between PCM, DSD64 and DSD128.

Restarting playback usually fixes it. This is foobar2000 with foo_dsd_asio.

 

My experience tells this is possibly software related as tweaking (restarting JRMC) or trying different drivers (KS w/ JPLAY or Vega ASIO) fixes the issue. Sometimes I restart entire Windows system and put Vega to sleep, then wake up.

 

FWIW, I have tested a PPA Studio USB card and Adanco fiber-over-USB system and both are working nicely with Vega in Exact mode, using JRMC PCM2DSD or native DSD 128.

[Engvet]

Yeah, its the one I have and it gives you the optionality to go in either direction...

 

 

ejn,

How do you find the Vega through the Taurus, do you think it improves the presentation and how does the Taurus sound purely as a pre? I have a turntable so need an analogue input so am thinking Vega plus Taurus into my TrigonTRE50s and Audio Physic Scorpio 25s. Vega would use the balanced analogue inputs and the phono would use the single ended inputs, and then 5m balanced cables to power amps next to speakers withnd short speaker cables. This is what I have already and the Vega plus Taurus would be replacing a Perreaux DP32.

Where did you get your system ejn? very similar to mine and see you are in uk, your system has an Audio Emotion look to it :-).

Anyone out there with any other ideas about how to incorporate a Vega and turntable into the system? Music First Copper pre? Goldpoint stepped attenuator (all balanced or all single ended so maybe not)?

Best regards,

Keith.

 

By the way, have listened to Vega in my system against Perreaux and SotM sdp1000.

[phaeton]

Sometimes I am getting this white noise on start of DSD128 tracks when switching back and forth between PCM, DSD64 and DSD128.

Restarting playback usually fixes it. This is foobar2000 with foo_dsd_asio.

 

I got the same when I was experimenting with ASIO settings and foobar buffer.

 

completely solved setting ASIO to safe and 8192 samples, and foobar buffer to 30000ms, but it works well with no hiss whatsoever even with smaller values.

the smallest I found working in my system was min latency / 512 and 200ms, but of course this is highly dependent on the pc used as player.

 

btw, I found no difference in SQ going from the min values to the max values.

[Andrew_WOT]

Thanks Mike,

I am actually using Extra Safe and 8192 now but foobar2000 buffer is set to 1090 (default), will play with that as well.

But the issue is so minor and easily fixable to lose any sleep over it.

I got the same when I was experimenting with ASIO settings and foobar buffer.

 

completely solved setting ASIO to safe and 8192 samples, and foobar buffer to 30000ms, but it works well with no hiss whatsoever even with smaller values.

the smallest I found working in my system was min latency / 512 and 200ms, but of course this is highly dependent on the pc used as player.

 

btw, I found no difference in SQ going from the min values to the max values.

[phaeton]

.........

But the issue is so minor and easily fixable to lose any sleep over it.

 

fully agree, moreover considering how well the Vega sounds :-)