T+a dac 8 dsd

[louisxiawei]

Hello guys,

 

Just want to ask a few questions here regarding T+A DSD upsampling:

 

 

• does T+A need a powerful cpu like I7 to do DSD 512 upsampling? I see some of users have difficulty upsampling to DSD 512 with I5 cpu. Can someone share some HQplayer setting for T+A DAC8 DSD?
  
• Does T+A support DOP? My friend is now having difficulty upsample his T+A to DSD512 by ticking the DOP option doesn't help and his CPU is I5 6500.
  
• last but not the least, what's the SQ difference between HQplayer upsampling and Foobar upsampling to DSD 512?
  

 

Sorry to be a pain asking so much and any help would be appreciated.

[louisxiawei]

Thank you so much for the reply.

 

by using HQplayer, my friend and me cannot smoothly upsample the T+A to DSD 512. There is some pop sound appearing from time to time. Poly-sinc-2s doesn't help, buffer time remedies it but doesn't eliminate the pops. CPU takes up 50% + in that case.

 

With my friend i5 6500 CPU, the pop disappear and the smooth DSD sound comes only from the DSD256. CPU takes up to 25% in DSD256 case.

 

One funny thing is my friend then used foobar to upsample T+A to 512 successfully without pop sound. But I doubt the CPU might not be taken advantage of properly because I also trust HQplayer more in terms of upsampling. (Hope someone might help me out regarding this)

 

Me myself, am using Exasound e12 with HQplayer upsampling to DSD256. Fantastic!

 

But haven't compared them AB yet, anyone got idea Exasound DSD 256 vs T+A DSD 512?

 

 

 

 

I don't think you can do DSD512 over DOP. You really should be doing ASIO for the higher DSD sample rates.

You may be able to do some DSD512 upsampling with an I5 just fine if you stick to the poly*-2s filters, but YMMV. Ideally, to get the most out of all the filters, an I7 is recommended.

Don't know about Foobar....I'm an HQPlayer fan all the way

[louisxiawei]

Thanks for the reply, hifial.

 

Sorry to copy paste the message again to you, since you might be the correct person to ask. (You have experience in both exasound DAC and T+A)

 

 

by using HQplayer, my friend and me cannot smoothly upsample the T+A to DSD 512. There is some pop sound appearing from time to time. Poly-sinc-2s doesn't help, buffer time remedies it but doesn't eliminate the pops. CPU takes up 50% + in that case. It might be due to the CPU is not powerful enough.

 

With my friend i5 6500 CPU, the pop disappear and the smooth DSD sound comes only from the DSD256. CPU takes up to 25% in DSD256 case.

 

One funny thing is my friend then used foobar to upsample T+A to 512 successfully without pop sound. But I doubt the CPU might not be taken advantage of properly because I also trust HQplayer more in terms of upsampling.

 

Me myself, am using Exasound e12 with HQplayer upsampling to DSD256. Fantastic!

 

But haven't compared them AB yet, what do you think of the SQ between Exasound DSD 256 vs T+A DSD 512 through HQplayer?

 

I will chime in but I hop so will others.

 

1) Yes, you need a powerful cpu like a Skylake i7 6700K to do upsampling to DSD512. It is not that the T+A needs it but HQ Player does in order to have the power needed to do the very intense processing.

 

2) As far as I am aware you can not do DSD512 in DOP. It is my understanding it does not have the headroom in the rapper. T+A needs a driver to do DSD512 (and DSD256) and you can get it on their web page. It is only for Windows. NO Mac for DSD512.

 

3) There is a HUGE difference, IMHO, between HQP and Foobar when upsampling to DSD512. Not that Foobar is bad sounding mind you.

[louisxiawei]

Well said and what a good metaphor for me to better understand

 

Much appreciated for the reply. Guess I need to prepare to sell my exasound sometime as well.

My personal audition experience from some HIFI shop and my friend's experience agrees that T+A DSD 512's sound can compete Chord DAVE PCM's one.

 

However, I do wish T+A release a DAC only has DSD circuit part, it will make it more compact and slim yet powerful.

 

Thanks again.

 

 

 

 

 

 

 

Well the computational power that HQP needs for its algorithms to do the upsampling is much more intensive then what Foobar needs. So that is why one is able to do DSD512 with foobar with just an i5.

 

It is like asking a Formula One race car to run on ordinary gasoline and ask why it is braking down and under performing.

 

After hearing the T+A do DSD512 I sold my beloved exasound and bought my T+A DAC. And I do not make a big change lightly.

[louisxiawei]

Hey Ted,

 

I've just placed the order of T+A DAC8 DSD. Now I can compare exasound e12 DSD256+ vs T+A DAC8 DSD512+ through HQplayer.

 

Can't wait for it!

 

Off the topic: exasound e12 upsampling to 12288000 with DSD256+fs modulator sounds the best. I would recommend untick the auto rate family option for upsampling to highest rate.

[louisxiawei]

Hey guys,

 

MY DAC8 has arrived. I can confirmed that DSD512 on T+A DAC8 is a BIG step forward compared to DSD256. I did compare the DSD256+ exasound e12 at 12288000 sample rate with T+A DAC8 dsd256, both sound good.

 

In other words, if T+A DAC8 only supports up to DSD256, exasound is still going to be my choice.

 

Now I have an issue. I believe the higher the sample rate, the better the sound will be when using T+A with HQplayer.

 

Thus, I untick the "auto rate family", tick the Pipeline SDM, tick the CUDA offload, run my I7 3770 + nivida gtx780 at 24576000.

 

Oversampling: poly-sinc-2s, Modulator: ASDM7 512+fs

 

Music playback is smooth in general, but I still have some tiny pop sound from time to time.

 

Does it mean the T+A usb interface got its bottleneck not stable at 24576000 sample rate? Or my CPU is not power enough? It takes only 30% CPU, 25% GPU running this setup.

 

Any idea? Any help would be appreciated.

[louisxiawei]

Likely your CPU/GPU is on the edge, check the per-core loads in Resource Monitor. If memory bandwidth or GPU access becomes an issue, it doesn't necessarily show up in the load figures. You could as well leave "auto rate family checked", set limit to 24.576 MHz and leave the rate in main window to "auto", it lightens up the load a little...

Thanks a lot Miska. I did find the pop sound disappear when I check the auto rate family.

You are correct, load figure doesn't mean anything, running DSD512 with auto rate family checked is still not quite stable if doing some multitask like opening web browser, which can lead to the stutter.

 

Another thing is quite weird, when I used exasound dac, when I checked the SDM pipeline, it cause the music hiccup while in T+A's case, it's the opposite, if I don't pick the SDM pipeline, it will playback continuous hiccup. Any reasons for that?

 

BTW, I also run a fidelizer pro on my PC. Hope it won't pull HQplayer's leg.

 

Guess I need to build a beast to accommodate the T+A. Upgrade my CPU to i7 7700K and GPU to Titan.

 

Thanks again!

[louisxiawei]

Hey guys,

 

I find when I switch DSD tracks on T+A DAC8, there is some tiny pop/click sound between tracks. I have these click sound only with DSD files (either HQplayer's upsampling DSD or direct DSD playback)

 

I've also tried using ROON without connecting HQplayer, the click still exists. Gapless play is ok without any pop click, only happens when I manually switch songs.

 

Is it normal? Can it be fixed with HQplayer's setting or Roon's setting?

[louisxiawei]

I was playing with HQPlayer settings last night and a small change increased the sound quality quite significantly. Try the following and see if it is replicated in your system:

 

Uncheck "Auto rate family" and set DSD upsampling to 24.576 MHz.

 

Holy s**t. I've no idea why this would change anything for better. I mean using multiples of 48KHz instead of 44.1 KHz for red book music sounds wrong. But no it sounds so good. Better prat, musicality, definition, warmth, etc.

 

I should mention that I noticed this with my previous dac Exasound E22 Mk2 while upsampling to DSD 256. But with T+A DSD 512, the change is more significant.

I did the same thing. T+A'S 24M6 sounds much better than 22M6, better to use AMSDM7 512+FS, the heaviest modulator for DSD512+

 

I also own the exasound e12, also uncheck the auto rate family and let the dac upsample to DSD256+FS which is 12288000. It can also yield a better performance on exasound.

[louisxiawei]

What filter are you using? And what sort of music do you listen to?

 

I can't do what you suggest without changing to a lighter load filter.

 

 

Sent from my iPhone using Computer Audiophile

 

I'm listening to everything, Jazz vocal, Classical through Tidal. I'm also upsampling some DSD256 files to DSD512, they sound the best compared to PCM file to DSD512.

 

Try this first:

 

Filter: Poly-sinc-2s

Modulator: AMSDM7

Upsampling rate: 24576000

SDM pipeline: checked

Auto rate family: unchecked

CUDA offload: checked

 

Basically, AMSDM7 is designed for DSD512 as Miska mentioned before, but you might not like the sound as he also mentioned there is no certain modulator match for each certain filter.

 

However, DSD512+ (24576000) sampling rate does sound better than DSD512 (22579200) in my case. Much more air, more transparent, music is a little bit more focused.

 

What is your CPU or GPU? Because if they are not powerful enough like i7 7700K, it might be not going to work.

 

I'm going to build a PC with i7 7700k and Titan X. Gonna unleash T+A's full potential through HQplayer's heaviest settings.

[louisxiawei]

Thanks.

 

So what modulator you are using at the moment? DSD7 ? amsdm7? or amsdm7 DSD512+?

 

I do believe AMSDM7 DSD512+ is designed for 24576000 just like DSD7 256+fs is designed for 12288000.

 

Maybe you are right, only works for 24576000 only works on DSD files instead of 44.1-based PCM.

[louisxiawei]

I've been thinking that if T+A DAC8 DSD has an additional external DC power supply input, like the 12V one which merging nadac has, it would be such a perfect and epic DAC!

 

Then we can use Ultra-cap linear power supply like vinnie rossi's MINI PURE-DC-4EVR to power the T+A DAC8 DSD.

 

Can't imagine the synergy of that combo.

 

If anyone knows how to mod it, do please let me know.

[louisxiawei]

Well I understand both points but the DAC 8 DSD has an analog volume control and is not a digital volume control, just so we all understand.

 

Also I use my DAC 8 DSD direct to amps with very good results. I have heard my DAC 8 DSD used with a few highly regarded preamps (BESPOKE Audio, Merrill Audio Christine), in others systems, and it made a worthwhile improvement. But we are in the $12,000 range of preamps.

 

Could a more affordable preamp of like sonic quality bring worthwhile gains? Yes, if there is one to be found.

 

My suggestion is to try if you can.

 

 

Hi Hifal,

 

Off the topic, but you might be the correct person to ask.

 

I'm thinking about purchasing T+A P3000 HV preamp matched with my T+A DAC8 DSD for the ADAM active MK3 classic speaker.

 

Have you got any experience regarding T+A p3000 HV compared to DAC8 DSDS's own preamp? Will that be a good combo setup for my system? Or you think P3000 is an overkill?

[louisxiawei]

@bibo01

 

Sorry to bring this again, but any news of DAC8 DSD firmware update for playing DSD512 on linux? Thanks in advance.

 

 

[louisxiawei]

53 minutes ago, volpone said:

Not very good !

AMANERO seems working to fix native DSD512 support FW issues on LINUX, still some problems according  https://github.com/lintweaker/xmos-native-dsd/issues/12. However i hope AMANERO will persevere and succeed.

 

 

Guess there will be a bit long wait. I might consider a windows NAA as an alternative for DAC8 DSD. Like the Logic Supply CL100 recommended on the HQplayer website, but don't know if there is a lite version of windows 10 OS for it. 

[louisxiawei]

32 minutes ago, volpone said:

 

I’m really disappointed too ! 

As an temporary alternative i’m looking for a good ASIO NAA / ROONBRIDGE networked endpoint on par with Linux dedicated ALSA endpoint SQ (mainly USB EMI/RFI isolation). Any WIN device comparable to: Micro/Ultra Rendu, ALLO USBridge, SoTM SMS-200 ultra … from SQ point of view ?

I'm thinking maybe a windows based NAA like NUC + ISO regen will be something better than Microrendu and SoTM SMS-200.

 

I remember John mentioned ISO regen has better SI than microrendu, so with the help of NAA, you still have the Ethernet isolation while connecting a ISO REGEN after the NAA will still have the regen function for better signal SI. But the OS in the windows NAA must be stable and lite enough like Sonicorbiter and support Roon remote and HQplayer.

 

I'm really eyeing on Signature Rendu SE, but unfortunately won't help either for DAC8's DSD512.

 

 

[louisxiawei]

4 hours ago, bibo01 said:

and @louisxiawei

Do not give up!  

I spoke to Amanero today and he told me that he has received an email from T&A confirming that latest firmware is working fine with DSD512 (incidentally they also use HQPlayer  ).
Very soon (probably a couple of days, maybe less) Amanero will release final version of firmware - 2003 - which is further optimized (he tested it last night) and it also contains code for Windows (2002 did not).

Thanks for the good news. Finger crossed for T+A's new firmware for linux DSD512.

 

I did kinda give up and begin to build windows based NAA. 

[louisxiawei]

Meanwhile you guys are waiting for uncertain linux DSD512 support for DAC8 DSD?, I would like to report another issue of T+A DAC8 DSD that annoyed me for quite a while.

 

This issue is mainly about the "very faint pop" or "static sound" during playback of one certain rate 24.576M (DSD512 x 48), and some of other DAC8 DSD users I know can also confirm this "pop" issue existed.

 

The "pop" sound is so faint that only using headphones can detect it when you listen to some quiet, peaceful music tracks. If you use loudspeakers or listen to some loud dynamic music,  the pop sound will probably be disguised. 

 

I mainly use HQplayer to upsampling everything to DSD512 to the DAC using "Auto rate family", and found that all 48k base file can cause the "pop" sound. But situation varies:

• 48/24 to DSD512 x48 has the most frequent pop sound (using any filters such as poly-sinc, poly-sinc-xtr, or any 2s filters like poly-sinc-2s)
• 96/24 to DSD512 x48 gets better, pop occurred much less frequently, but still have a few during one track (using any filters such as poly-sinc, poly-sinc-xtr, or any 2s filters like poly-sinc-2s)
• 192/24 to DSD512 x48 has the least frequent pop sound, sometimes you can detect none pop during a very short-time track (using any filters such as poly-sinc, poly-sinc-xtr, or any 2s filters like poly-sinc-2s)

The pop sound does NOT occur when:

• any 44.1K base file to 44.1 x DSD512 (using any non-2s or 2s flters)
• any 48k base file to 48 x DSD256 (using any non-2s or 2s filters)

I've eliminated a few potential causes for this "faint pop sound":

• the pop cannot be caused by incapability of my CPU for processing HQplayer since I use 7980XE and poly-sinc-2s filter can even cause the 48K→48x512 upsampling to have a pop sound
• I've also discussed with @Miska already. With his test, it's also not a HQplayer's problem.
• I've also reinstalled my entire windows OS (windows 10 pro for workstation) just in case some of my background process or software messed around the latency for streaming music, but the problem still existed afterwards. Plus some of the DAC8 DSD users I know running HQplayer on neat windows 7 can also spot the pop sound, which indicates the issue might not caused by the operating system.

So the problem can only be caused by the DAC8 DSD itself, I was told that it is likely the Amanero USB interface of DAC8 DSD is acting up at DSD512 x 48 rate since similar issue happened from some XMOS problems or other DACs have had.

I've attached a 48/24 track, and would like to invite the DAC8 DSD users to try it out. In this track, you will probably spot a few "faint pop" sound in the first 20 seconds of the track on your "right channel".

If the problem does exist widely among DAC8 DSD users, then we can be confident enough to report this issue to T+A.

Any help would be appreciated.

 

 

03. Hedwig's Theme (From “Harry Potter And The Sorcerer's Stone”).flac

[louisxiawei]

1 hour ago, blue2 said:

 

You'd want to try the track with another DSD512 DAC (non-Amanero USB) and not hear the pops to be sure of this statement!

That's a little difficult for a normal customer.  As far as I know, there are really limited number of dacs supporting 48 x 512 rate out there , and I'm not owning a DAC shop. It's an assumption with convincing test results rather than a solid statement. The reason I did that last post is: I'm not totally sure and want to gather more information before taking further action and report it back to T+A.

 

I have ifi micro dsd BL, which supports 48 x 512 only on linux OS. I cannot test it myself since I only have windows OS installed but Miska tested it for me and reporting back that no pop during 48 x 512 playback. (Interestingly, he also did not detect any pop during 48 x DSD 512 rate playback.)

[louisxiawei]

8 minutes ago, Em2016 said:

 

Just to rule it out (for the feel good factor) did you try without any USB conditioners in the path - i.e. just a direct USB cable from PC to DAC? I see a REGEN in your signature.

 

Yes, just direct USB cable from PC to DAC. I took the ISO-REGEN out during the test.

10 minutes ago, Em2016 said:

And do you know for sure (from T + A themselves) if the DSD512x48 sample rate is officially supported by this DAC? If they say no, then that's the end of it.

I know the iDSD and Pro-Ject S2 DAC and other cheap DACs do support 512x48 so you would expect the T+A DAC to also support it, but it's possible it doesn't.

 

Yes, it does support 24.6M (DSD512 x 48). It is clearly written on its manual.

 

[louisxiawei]

4 hours ago, blue2 said:

If you accept this result, and why not, then the problem is with your hardware not specifically a T+A DAC 8 DSD problem.

I see you have difficultly understanding things. Can't you see my test results? There are three factors causing the problem: PC, HQplayer, DAC8 DSD. I've eliminated two (PC and HQplayer) by having people do the test . What is left? Think harder!

 

 I'm not alone hearing the faint pop noise. So far, I have at least 10 DAC8 DSD users with me spotting  the same problem, and they use different PC hardware, OS system with same HQplayer settings unless you keep arguing their PC are all defective, all their DAC8 DSD are faulty.

 

Quote

To narrow it down you could install Linux on a separate partition on your Windows PC and see how the ifi plays the test track.

 

I don't see the logic. T+A can only support DSD512 on WINDOWS while ifi can do DSD512 x 48 only on Linux! 

 

If Proving ifi is working fine can  suggest my PC is working well, which I already know. Do you think it will make any sense telling T+A that if IFI can stream DSD512 x 48 on LINUX without a pop, then it's your DAC8 DSD's problem to stream DSD512 x 48 on WINDOWS with a pop? 

 

Quote

unless it's an Amenero firmware problem but Miska's test seems to refute that

So far, Miska is the only one I know cannot detect the faint pop.

 

Quote

Getting others to confirm your finding may need a large number of participants if only a small percentage of T+A's are affected -

To be honest, I think even 10 people hearing the pop noise is enough convincing. I just want to get more information on CA since people here have more chance to use HQplayer to stream 48 x 512. I know many other DAC8 DSD users, they don't even bother using HQplayer with it.

 

My question for you is: Do you have a DAC8 DSD? Can you do the quick test? If the answer is no, then you are not helping at all but doing the pointless defense for DAC8 DSD.

[louisxiawei]

53 minutes ago, juanitox said:

 

i don't hear anything  head in my horn with windows 10 , 48 to DSD512  polysinc ASDM7.  the only erratic pop i have is when the sample rate is changing between 44 to 48 or 96. but nothing to start a PopGate. ?

Thanks for testing.

 

As I mentioned, speaker system is quite difficult to detect the faint pop since there is a certain distance between your listening spot and speakers. So some speaker users don’t even bother the issue.

 

It’s not like a playback stutter caused by insufficient CPU capability when you upsample non-integer rate like 44.1 -> 48 x 512. It’s more like a static noise pops infrequently hidden at the background while the music flows.

 

I have one friend put his ear right close to the right channel of the speaker and then be able to detect the faint pop.

[louisxiawei]

31 minutes ago, blue2 said:

No need to be insulting! You haven't published any test results and no mention of others finding the 'pop' prior to my post. If you really have 10 independent confirmed results granted that is significant, but you have to also state how many have found no 'pops'. 10 positives could be 0.01% failure rate on the T+A hardware.

 

If your ifi 'pops' with your test track at DSD512 x 48 under Linux on your PC then it could be your PC/USB is causing the problem.

 

Maybe you need to think harder!!

No interest discussing with a fragile person thinking I was insulting him. Have fun with your ridiculous logic.

[louisxiawei]

Just want to add another comment and hope this can also suggest a bit more regarding this "faint pop" issue. 

 

I also have Logic CL100 as windows NAA streaming music from my 7980XE HQ PC to the DAC8 DSD. That was the original HQplayer end-point device I found the "faint" pop sound. 

 

Of course one can say that my newly built HQplayer PC and Logic CL100 NAA both have issue now. What are the odds? 

 

I also tried to recall my previous experience regarding the 48 x 512 playback since I purchased the DAC8 DSD more than a year ago but found the problem till now. I've checked the track playback history of Roon, found myself barely owning or playing any 48/24 files before, and the only 48K base files sit in my previous library are some 192/24 albums which can hardly detect the faint pop as I reported previously.

 

 

[louisxiawei]

2 hours ago, Solstice380 said:

@louisxiawei  I have a Lampi Golden Atlantic with Amanero USB using the Win Combo384 ASIO driver on a NUC NAA.   I have thought occasionally that I had heard the static pops but because they are very low level and random I wasn’t sure.  I also listen to music that generally does not have long quiet passages.  

 

I listened to the Harry Potter track 5 times and heard the pops.   Totally random, and low level, but they are there. 

 

Is it the Amanero driver?

Many thanks for testing. It helps. It's very likely the Amanero's problem. 

 

The static pop does occur randomly when playing that harry porter soundtrack, but the pop is somehow predictable with a certain time window, something like I will definitely hear the first pop after 10 seconds of the track. I've listened/tested to this track for 76 times (Roon records it).

 

I also installed the old version of T+A driver, didn't make the pop disappeared. 

 

@4est Yes, although that faint pop is quite audible through some revealing headphones like HD800, it can be ignored by people using speakers or even headphone users listen to dynamic music quite often. But I believe for some certain genre of music with long quiet passages or dead silent background such as solo piano, solo vocal pieces. These static pops are a bit annoying and can be audible through speakers.

 

I agree we shall be more forgiving and give people more time to deal with the problem, but pursuing the best possible SQ without flaw is what this audiophile hobby is all about, which includes stubborn people like me looking for reasons for these insignificant faint pops.