Jump to content

Rate this topic

Recommended Posts

thanks i get 1ms latency with my config ,  i will let the latency checker running to see if there is a change my noise problem appears again. 

 

miska what HQP buffer time will be the most stable  with the dac8 in DSD512  ?

Share this post


Link to post
Share on other sites

I get 499 micro second current and 620 absolute on the checker. I am not sure what this means. It does say the system is capable of real time streaming without dropouts.

 

Share this post


Link to post
Share on other sites
Quote

I tested again. I rebooted the PC and ran the check. This time is was around 4000 micro seconds! Red all over and not enough for real-time audio streaming. I ran Fidelizer and checked again. Back to the 499 micro seconds!

 

Since 4000 µs is 4 ms and at DSD512 the DAC8 DSD's ASIO driver buffer size is around 1 ms, problems are pretty much guaranteed. Even 500 µs is just half of the buffer size and not so great, but well enough. With Linux, already 10 years ago, getting latency performance in about 5 µs range wasn't so much problem...

 

On Win10 you usually get decent figures by setting "High Performance" power profile. If you have Win10 for Workstations you also get profile called "Ultimate Performance" which gives even better figures.

 

Share this post


Link to post
Share on other sites
4 hours ago, juanitox said:

thanks i get 1ms latency with my config ,  i will let the latency checker running to see if there is a change my noise problem appears again. 

 

miska what HQP buffer time will be the most stable  with the dac8 in DSD512  ?

 

HQPlayer's buffer time setting doesn't have effect with DAC8 DSD because the driver has fixed buffer size. So just leave it at "Default" as it doesn't have any effect anyway.

 

With most ASIO drivers it is best to leave it at "Default" and set buffer size from the ASIO driver's control panel (which you can open from HQPlayer's Tools-menu). In DAC8 DSD case, there's no such thing.

 

With WASAPI on Windows and ALSA on Linux, start with 100 ms setting. With CoreAudio on Mac you cannot adjust it either, it is managed by CoreAudio itself. With NAA, leave it at "Default".

 

Share this post


Link to post
Share on other sites

 

Quote

On Win10 you usually get decent figures by setting "High Performance" power profile. If you have Win10 for Workstations you also get profile called "Ultimate Performance" which gives even better figures.

 

 

Just to be clear my testing is with "Ultimate Performance" profile.

 

I also found a LatencyMon tool that shows better statistics. It shows current between 10-30 with occasional peaks to 100+. The max is 429 which is related to NVIDIA kernel driver. This seems more reasonable than the checker.

Share this post


Link to post
Share on other sites
17 minutes ago, Miska said:

If you have Win10 for Workstations you also get profile called "Ultimate Performance" which gives even better figures.

You can enable Ultimate Performance on Windows 10 Version 1803.  Thanks for reminding me about this.  I haven't tried it yet.

https://www.tenforums.com/tutorials/107613-add-remove-ultimate-performance-power-plan-windows-10-a.html

Share this post


Link to post
Share on other sites

I am still not sure what is causing it. One thing I am certain is that Fidelizer reduces latency. Red is how it looks just after reboot.

 

DPC.thumb.PNG.29b1a3a4de475424e77715dc1b3df5d2.PNG

 

LatencyMon shows Nvidia and Directx causing high DPC and ISR. I am still not sure if this is an issue and how to deal with it. For now I will certainly continue to use Fidelizer.

 

Share this post


Link to post
Share on other sites

If the latency stays in 500 µs range, it is enough for this case and not really worth to put lot of effort the get it lower.

 

Nvidia driver causes some amount of latencies due to desktop compositor doing huge transfers over PCIe bus. And it also depends if applications do heavy graphics on the GPU. So eye candy costs quite a bit of GPU traffic.

Share this post


Link to post
Share on other sites

It stays low and LatencyMonitor is all green. Current latency in low enough and most important stable even during heavy load, so it's good. I am happy I can see and hear the difference.

 

It could be nice to know what is the root cause and what Fidelizer is doing to reduce it. Now that I know what to do to keep latency low, I can enjoy the music without problems or white noise.

Share this post


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

It stays low and LatencyMonitor is all green. Current latency in low enough and most important stable even during heavy load, so it's good. I am happy I can see and hear the difference.

 

It could be nice to know what is the root cause and what Fidelizer is doing to reduce it. Now that I know what to do to keep latency low, I can enjoy the music without problems or white noise.

The PC is not powerful enough to run so many messy windows background processes while running HQplayer at full power. That's why linux is much more hassle-free. 

 

I will not recommend using Fidelizer on PC running HQplayer because its core isolation and processor priority policy really mess up the CPU, not good for HQplayer. Turning on Fidezlier can cause my frame stutter during gaming. I don't really think the author of Fidelizer has idea how HQplayer utilizes the CPU, although he offers option of certain software optimisation including HQplayer.

 

I only use Fidelizer on my windows NAA and I'm expecting more official approach like windows 10 lean version, but seems no further news from windows.

 

 

Share this post


Link to post
Share on other sites

Currently the only choice I have is to run Fidelizer to get Latency to the right level. I don't know of anything else that can do it better. I'll be glad to get better recommendations. I don't mind about gaming and music at the same time so for me Fidelizer is the right answer when I want to use HQPlayer and it works perfectly well  much better with it than without it.

 

If  I want to do gaming I would not mind about Fidelizer. I currently run at "ultimate power" profile which is good for gaming and I have no problems with gaming or stability with or without Fidelizer.

 

Share this post


Link to post
Share on other sites

When the T+A 8 DSD came to market, a lot was written about the day and night quality difference going from DSD256 to DSD512.

 

As I am now testing native DSD support on Linux with my T+A 8 DSD, I do not experience that performance gap between DSD256 and DSD512.  Sometimes, I believe DSD512 is better, sometimes I believe DSD256 is better (more relaxed).

This can be due to a lot of things, to start with my hearing, my room, performance of my streamer, HQembedded filter/modulator configuration, ......

 

As I am sure there are a number of CA members playing DSD512 with the T+A 8 DSD dac (using Windows), I would love to request a couple (2-3) performances, that for each of you personally, proof the quality difference between these 2 DSD upsampling levels.

It would be great if software player and its filter/modulator configuration could be added.

 

Looking forward to your recommendations.

 

Dirk

 

P.S. today I can play on my Pinkfaun 2.16 streamer with Roon/HQembedded

DSD64 upsampled to DSD512 with poly-sinc-xtr/AMD7

Redbook upsampled to DSD512 with poly-sinc-xtr-2s/AMD7 (for now I have settled on this filter choice)

 

Share this post


Link to post
Share on other sites

 

Dirk, there is no night and day difference and I don't believe anyone really experienced such a difference.

 

I can assure you both are great. From my experience DSD512 feels fuller and deeper which gives a small but significant advantage over 256. I only use poly-sinc-xtr. I just tried it with a couple of tracks, redbook and hi-res.

I never used 256  before and I just did a sort experiment so everything I say here is only based on my short testing session. 

 

For me the difference is noticeable. It's seems more pronounced playing the hi-res (24/192) less with redbook. I never tried DSD to DSD conversion. I believe some may can claim it subtle but significant some may less notice it. I've met many audiophiles and their experiences are not always the same when it comes to subtle differences (some find it very significant, some not). I am on the subtle differences side. I like ones that brings higher quality so for me it's better significant enough but not by a big margin.

 

If I had no way to compare, I can assure you I could also be happy enough with only the 256 but after the comparison I'm staying with 512.

 

BTW, some people will even tell PCM sounds better. The truth is that PCM sound different. For me it's more like comparing SS to tube. DSD is definitely warmer and I like it better but this can be a matter of taste rather than comparing the same sound and quality.

Share this post


Link to post
Share on other sites

Dirk, may I suggest trying a better DAC like the Pink Faun DAC 2.32 and use HQP to serve files in their native format?

I used to do redbook --> DSD512 into a T&A DAC8 DSD before but once I had a little session at PF HQ I was smitten...and turned to the dark side...

Share this post


Link to post
Share on other sites

Anybody willing to post the DSD settings for the imbedded version of HQ Player? Switched to a Linux base server and have no clue how to set up the imbedded version for the DAC 8. Thanks for anyone’s help. 

Share this post


Link to post
Share on other sites
2 hours ago, nc42acc said:

Anybody willing to post the DSD settings for the imbedded version of HQ Player? Switched to a Linux base server and have no clue how to set up the imbedded version for the DAC 8. Thanks for anyone’s help. 

Once you have embedded version running connect to your server using browser http://ip_address:8088/.  Logon and select configuration link at the bottom of the page.  The rest of the stuff matches up with what you had with desktop version.

Share this post


Link to post
Share on other sites

How does this dac perform connected to a sotm sms200 Ultra / txUSB Ultra?

Anyone using this combo can comment?

Thanks.

Share this post


Link to post
Share on other sites

Hi Vitor

 

I am not able to respond to your question, as I do not have such a combo.

 

However, I am a little bit surprised by your question. Applying a device such as the sms200 Ultra / tx USB Ultra is usefull to improve on the SQ of the source (PC, streamer, network player, ...) , not of the DAC itself imho. 

The DAC itself is more suspectible to format of the data it is fed with, I.e. it performs better with DSD than with PCM. Not that is bad with PCM at all!

 

What is your source that you want to improve with the sms200 Ultra/ ....?

 

Dirk

Share this post


Link to post
Share on other sites
On 6/5/2018 at 11:46 AM, louisxiawei said:

Just give you guys an update:

 

I reported the "faint pop" issue to T+A  two days ago and today I got a detailed reply from Lothar Wiemann. Thumb up for their professional attitude. 

 

They were able to reproduce the problem as well and showed me a screenshot from an oscilloscope indicating the "pop" noise on the right channel. They found out that the pop occurs right at the output of the Amanero USB receiver and told me it is not a problem of the DAC circuit or the clocking circuitry / any other hardware problem.

 

I also shown them what we discussed here on CA and some @Miska technical reply explaining the potential reasons. They also think it's either the problem of Amanero windows driver or the USB receiver firmware. 

 

They've contacted the Amanero and will try to fix the problem. Will keep you updated.

 

 

On 6/30/2018 at 7:38 AM, louisxiawei said:

Some updates of T+A DAC8 DSD firmware upgrade progress:

 

With some great help from T+A (Lothar Wiemann), I've managed to upgrade the firmware via RS232 approach. Lothar is really a helpful gentleman being very supportive regarding this matter.

 

In general, my upgrade experience matches what @RS241 said a year ago.

 

Especially this:

 

Hence, my DAC8 DSD now has been re-flashed to Amanero and the PC reads it as Amanero device and it must be used with Amanero driver (Combo384 ASIO 1.03), T+A driver will not work.

 

Please note that: Currently there is no way back to the T+A firmware once it is re-flashed to Amanero.

 

Current Amanero firmware used on DAC8 DSD is:

CPLD firmware: CPLD_for_1080

CPU firmware: DSD512x48x44

 

I've been testing it for quite a few hours, so far, the "pop sound" issue from the R channel previously reported seems to be solved.

 

One might ask: does this firmware support linux DSD512?  The answer is no.

 

But good news is: Amanero firmware 2004be is the one that supports linux DSD512. So, once this manual FW upgrade was done, one can have total freedom to try any latest Amanero firmware on DAC8 DSD. I have not tried the 2004be on DAC8 DSD yet, but this might be the potential way to experience DSD512 on linux.

 

Also, be aware that Amanero windows driver does not work with 2004be at the moment (that is also why T+A cannot issue a 2004be firmware update).

 

Regarding the detail of the FW upgrade without opening the enclosure:

 

Two steps in general:

 

1. Update the DAC8 DSD firmware to V2.6 onwards

2. initiate service of DAC8 DSD and use oemtool117 to flash the firmware

 

Step two is very safe and easy, I have nothing to comment.

 

The step one can be safely done with the help of T+A MP8, and this is also the way T+A recommends. If you don't have access to MP8, then you need to use RS232 connection as S0133 service note described for the upgrade. This is the method I used and is not very recommended by T+A. It is highly risky to turn your DAC8 DSD into a brick because it has no recovery mechanism.

There are some tricks during the upgrade and it was a bit nervy indeed since I failed three times during the RS232 upgrade. 

 

Another good news is: Lothar has been writing a programme for PC to do a safe upgrade with secure recovery for DAC8 DSD. But it might take some time since he uses his own spare time doing it. For those who dare not upgrade it manually at risk or don't have a MP8, this might be the best solution for all. So again, patience is a good virtue.

 

I might open a thread about my experience of this risky upgrade if I have enough spare time or there is an increasing need for the FW upgrade. But I doubt T+A will support this since it raises chance to send broken unit to the factory and causes more trouble.

 

Since I am still encountering the same “pop” “chirping” static noise when playing back 96/192 family files upconverted to DSD512, I have revisited this thread for some updates.

 

It appears since the above posts by @louisxiawei in June, there has been updates to the DAC firmware and several iterations of the Amanero firmware which potentionally solved the problem, sort of. But, as of today, there is no official T+A “fix” for this problem because firmware is still being beta tested. Correct? So, those of us with this problem are still in wait mode?

 

And based on the past few pages, T+A considers this issue an isolated occurrence, so an official fix might not be forth coming. Which means I may need to start splicing together RS232 cables and playing Russian roulette with my DAC.

Share this post


Link to post
Share on other sites
2 hours ago, Argopo said:

 

It appears since the above posts by @louisxiawei in June, there has been updates to the DAC firmware and several iterations of the Amanero firmware which potentionally solved the problem, sort of. But, as of today, there is no official T+A “fix” for this problem because firmware is still being beta tested. Correct? So, those of us with this problem are still in wait mode?

 

 

I was told that the units after June 2018 are all installed with the latest firmware (CPU firmware: DSD512x48x44) which fix the "pop" noise, so in terms of the firmware, I don't think it's a beta test.

 

However, with users' own RS232 cable flash, we cannot use T+A official driver but driver downloaded from the Amanero website. From my experience, Amanero driver for windows 10 has glitch (producing blow sound during the track switch). But the windows 7/8/vista has no such problem on my windows 10. I'm still waiting for Amanero to fix this. Whether T+A official driver has this glitch, no idea.

 

T+A previously commented that the pop noise only happened on few batches with certain firmware. My DAC8 DSD was purchased in early 2017 in the UK and I think the firmware version was labeled on the box saying V2.0. Some of my other friends purchased their units (V2.2) from Hong Kong sometime in 2017, and they also found the pop noise.

 

Quote

And based on the past few pages, T+A considers this issue an isolated occurrence, so an official fix might not be forth coming. Which means I may need to start splicing together RS232 cables and playing Russian roulette with my DAC.

Or borrow a MP8 unit to do the update from the local dealer if they have one. Maybe contact T+A to arrange this?

Share this post


Link to post
Share on other sites
14 hours ago, Argopo said:

 

 

Since I am still encountering the same “pop” “chirping” static noise when playing back 96/192 family files upconverted to DSD512, I have revisited this thread for some updates.

 

It appears since the above posts by @louisxiawei in June, there has been updates to the DAC firmware and several iterations of the Amanero firmware which potentionally solved the problem, sort of. But, as of today, there is no official T+A “fix” for this problem because firmware is still being beta tested. Correct? So, those of us with this problem are still in wait mode?

 

And based on the past few pages, T+A considers this issue an isolated occurrence, so an official fix might not be forth coming. Which means I may need to start splicing together RS232 cables and playing Russian roulette with my DAC.

 

I think we are talking about 2 different things here:

1.) The "faint pop noise" which @louisxiawei mentioned. This occurs only with few DACs. If your DAC is affected,  this problem can be solved by a firmware update. The normal way to get this update installed on your DAC is to contact your T+A dealer or contact T+A service dept.

[Note: If you don't want to give your DAC to a dealer or send it to T+A service, you might alternatively consider option no. 2 below.]

 

2.) The other thing is DSD512 from a Linux based source. The DAC8DSD does not officially support Linux as operating system for the source. However: with a recent Linux kernel (> 4.15) the DAC8DSD will work with PCM up to 192kSps and DSD up to DSD128 (DoP mode).

For higher sample rates under Linux there is currently no official support. Amanero works on DSD512 compatibility for the USB receiver but this is still work in progress and only beta firmware versions are available directly from Amanero.

For those who want to experiment with the latest Amanero firmware on their DAC8DSD, it exists the possibility to install a firmware on the DAC8DSD which "opens" the DAC and  then permits to flash any USB receiver firmware (experimental or stable) from the Amanero repository on the DAC8DSD USB receiver chips.

After flashing the USB receiver with original Amanero firmware, the original Amanero Windows driver has to be used instead of the Windows driver from the T+A website.

Note: The Amanero Windows driver is functional identical to the T+A driver it only uses the standard VID/PID from Amanero instead of the T+A VID/PID.

 

So it is up to you. If you want to experiment (especially with DSD512 and Linux) use option 2, otherwise go the official way.

 

 

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

×