Audirvana Studio

[Dandou]

The USB input is noisy in Chord DACs. Toslink and Coaxial inputs are better indeed.

Did you ever try the BNC inputs with SRC-DX? 

[stefano_mbp]

On 1/19/2022 at 9:30 PM, Dandou said:

There are plenty of threads on Audirvana forum that prove it.

I can say that Audirvana do transcode pcm files (flac, aiff, Alac …) to wav before sending them to the network device, bit perfect or not depends on upsampling rules defined in Audirvana

[jos]

10 minutes ago, Dandou said:

The USB input is noisy in Chord DACs. Toslink and Coaxial inputs are better indeed.

Did you ever try the BNC inputs with SRC-DX? 

No, and I don’t now anything about SRC-DX too, sorry 😞. Some time ago, because of my old ears, I asked some young guys visiting my place to watch Live Aid, to compare the sound of Toslink and Coax on my Chord Qutest, and both of them said: Toslink is a tiny little better.

[Dandou]

Yes, and it converts the bit-depth of the tracks to the native bit-depth of the DAC. 

[Dandou]

12 minutes ago, jos said:

No, and I don’t now anything about SRC-DX too

It's a device that is produced by a Canadian maker, and it allows capturing the USB output of a streamer (or a computer), and to input it to Chord DACs through their double BNC entry.

There are Chord users that say that they get the best sound with the double BNC entry. One of them is a friend of mine who uses SRC-DX with a Qutest. 

[Dandou]

20 minutes ago, jos said:

I don’t now anything about SRC-DX

SRC-DX allows also to connect between the streamer and the DAC another device that is called "Opto-DX". This device converts the electrical stream of the sound to optical. In this way, it removes all the noise that is produced by the server, the LAN, and even the streamer. A second converter converts the optical stream to an electrical stream that is input to the Chord DAC through its double BNC connectors.

Opto-DX improves greatly the sound quality, and is made for Chord DACs. They can take advantage of it, because they have a double BNC entry. 

 

Sorry for the off-topic. 

[Jud]

35 minutes ago, Dandou said:

Yes, and it converts the bit-depth of the tracks to the native bit-depth of the DAC. 

“Convert” is perhaps less accurate than “zero padding” for the way this usually works.

 

Most DACs internally use 32-bit floating point or 24-bit integer. If sending a16-bit file, or a 24-bit file to a DAC utilizing 32-bit, AS, at least in my limited understanding, simply adds zeroes for the least significant bits. So 1.00000000 versus 1, no real difference or conversion involved.

[Dandou]

10 minutes ago, Jud said:

“Convert” is perhaps less accurate than “zero padding”

You are right, and I don't use always the most accurate words in English. But I did not mean in any way that this "zero padling" does not keep the sound bit-perfect. 

[Jud]

36 minutes ago, Dandou said:

You are right, and I don't use always the most accurate words in English. But I did not mean in any way that this "zero padling" does not keep the sound bit-perfect. 

Sure, and your English is just fine. 🙂 I mentioned it only because the question has come up previously as to what exactly happens and whether it is some form of DSP that affects the sound.

[jos]

1 hour ago, Dandou said:

SRC-DX allows also to connect between the streamer and the DAC another device that is called "Opto-DX". This device converts the electrical stream of the sound to optical. In this way, it removes all the noise that is produced by the server, the LAN, and even the streamer. A second converter converts the optical stream to an electrical stream that is input to the Chord DAC through its double BNC connectors.

Opto-DX improves greatly the sound quality, and is made for Chord DACs. They can take advantage of it, because they have a double BNC entry. 

 

Sorry for the off-topic. 

Thanks a lot, I will look-in to it. 👍

[jos]

On 1/19/2022 at 5:55 PM, Jud said:

We may be referring to the same thing in different terms. When I mentioned the “network,” I was talking about UPnP/NAA and all the hardware past the PC/Mac.  I’m guessing your reference to the “interface” is to either UPnP or the Yamaha, so it would be included in what I called the “network.”

 

In any case, checking with HQPlayer (trial version) if you can will give you a good idea whether it’s the Yamaha hardware or its particular implementation of UPnP that is the problem.

Thanks Jud, I agree: everything in the network = the network, but as an old project manager, we also did make a difference between ‘the network (servers/routers/switches’ and so called ‘clients in the network’ (PC/MAC/printers/Scanners and so on). Sorry for being off topic too. With interface I mean: UPNP tied to the Yamaha, both communicating with AS.

[rodrigaj]

Has anyone here streamed their music with AS to the LAN port of a minidsp SHD? Here is my setup:

NAS > Macmini running AS > iPad running Aremote > LAN minidsp SHD

 

I'm running a trial and don't want to waste time experimenting. Currently, I have everything set at 24/96, nothing else.

 

If you specifically run AS to the LAN port of a SHD, would you mind sharing your thoughts on what has worked out best for you. PM is fine, if this is too specific for the intended purpose of this thread.

 

(I did ask at Audirvana forums and got no response. The minidsp shd forum  pretty much all use Volumio)

 

 

[jos]

Limetree bridge installed without a hitch and so far no stuttering with 192khz/24bit files in native Limetree bridge mode (upscaling done trough r8brain). I didn’t compare sound quality yet, but AS seems to work fine trough with this bridge. 😀😊😇

[jos]

Still no problems using the Limetree bridge to my DAC. 🤗 I use it in native mode, so no upsampling is done by this bridge. In upsampling mode I can upsample. I can also upsample trough AS (r8brain), but I also could leave it to my Chord  Qutest DAC. Is there any experience out there…what to do best?

[Dandou]

15 minutes ago, jos said:

Is there any experience out there…what to do best?

For upsampling, you'll get an excellent result only with HQPlayer. 

[Jud]

4 hours ago, Dandou said:

For upsampling, you'll get an excellent result only with HQPlayer. 

I’ve been using both for years, and my advice would be to try upsampling with each of Audirvana Studio and HQPlayer to see which you personally like best. Not surprisingly, people’s tastes in these things vary. As my signature suggests, my own preference over these many years has turned out to be for Audirvana, but HQPlayer is also excellent, and my choice when I’m using Linux.

 

For the same reason, also try the Chord. While my preference is for upsampling in software, that is not the case for everyone.

[jos]

57 minutes ago, Jud said:

I’ve been using both for years, and my advice would be to try upsampling with each of Audirvana Studio and HQPlayer to see which you personally like best. Not surprisingly, people’s tastes in these things vary. As my signature suggests, my own preference over these many years has turned out to be for Audirvana, but HQPlayer is also excellent, and my choice when I’m using Linux.

 

For the same reason, also try the Chord. While my preference is for upsampling in software, that is not the case for everyone.

Unfortunately, I can’t try HQPlayer with my current Catalina OS-version, so AS is the way to go. On the web they say in general: ‘don’t upsample to a Chord DAC’, because Chord DAC’s can do that better. My new Bridge can also upsample, but I leave it in non-oversampling (native) mode, just as AS with r8brain mode active. My Chord DAC can’t be set to oversampling, or not, because it’s ‘build in’ to do so.

[Dandou]

58 minutes ago, jos said:

On the web they say in general: ‘don’t upsample to a Chord DAC’, because Chord DAC’s can do that better.

This is untrue. 

 

As upsampling is a necessary step for Delta-Sigma DACs in the chain of the conversion from digital to analog, the sound has to be brought to high frequencies of DSD before its conversion to analog. DSD512 in the case of your Qutest. The problem is that high quality upsampling to this rate requires a lot of computing power, and the tiny chip of the DAC is far for being powerful enough. So it operates a low quality upsampling. To upsample to the highest quality level, a powerful computer is necessary. 

This is true for all Delta-Sigma DACs. If Chord DACs were an exception, Chord would not sell for them the M Scaler: a very expensive device that has for only purpose the upsampling of the sound before its input to Chord DACs (through their double BLC entry). 

[pl_svn]

14 minutes ago, Dandou said:

Chord would not sell for them the M Scaler

as far as I recall Chord DACs do upsample internally to 1.5MHz before starting the actual digital/analogue conversion

I believe the MScaler, with its 1 million taps, just does a much better job

[Dandou]

13 minutes ago, pl_svn said:

internally to 1.5MHz

They upsample to 768 KHz, and then convert to DSD512. 

M Scaler does the PCM upsampling better for sure, but it costs (with its cables) about 3-4 times the price of the Qutest. 

Still, there are Chord users who prefer to upsample with HQPlayer. 

[pl_svn]

13 minutes ago, Dandou said:

They upsample to 768 KHz, and then convert to DSD512

 

right, 768KHz, not 1.5MHz (that was the Vega): sorry

though... are you sure about also converting to DSD? Rob Watts deeply dislikes DSD (he says it is an intrinsically limited format, whilst PCM is only limited by decoding costs/engineering)

Chord DACS, Dave apart I believe, all do convert (... decimate, as RW says) DSD to PCM

[Dandou]

30 minutes ago, pl_svn said:

are you sure about also converting to DSD?

Yes, DSD is the format in which all Delta-Sigma DACs work. They need to bring the digital sound to this format, before operating the conversion to analog asound. 

For this reason, Delta-Sigma DACs sound better if you play DSD files. It's not because the sound in the DSD tracks is superior to the sound of a PCM track. It's because, the DAC receives the sound in its native format, and does not have to operate a conversion to DSD with its tiny chip (as it does when you play a PCM file). Since the chip of the DAC is not powerful enough to operate the conversion at high quality, the DAC sounds better when you feed it with DSD tracks that it does not need to convert. 

[Jud]

2 hours ago, jos said:

Unfortunately, I can’t try HQPlayer with my current Catalina OS-version, so AS is the way to go. On the web they say in general: ‘don’t upsample to a Chord DAC’, because Chord DAC’s can do that better. My new Bridge can also upsample, but I leave it in non-oversampling (native) mode, just as AS with r8brain mode active. My Chord DAC can’t be set to oversampling, or not, because it’s ‘build in’ to do so.

What you could try in that case is to upsample with Audirvana to the highest rate the DAC (or streamer) will accept as input, whether PCM or DSD, and see if  you like that better than letting the DAC handle all the upsampling.

[Dandou]

55 minutes ago, Jud said:

whether PCM or DSD

Unfortunately, there's no way to get a good sound by inputting DSD to Chord DACs. 

The reason for that is that the only port through which it is possible to input DSD to Chord DACs is their USB port. But the implementation of the USB port in Chord DACs is very poor, and can be qualified as defective. When soud is input through this port, it triggers inside the DAC an Anamero chip that is extremely noisy. For this reason, it's better to avoid the USB port of Chord DACs, and to input through other ports. But their other ports are limited to a PCM input. 

[pl_svn]

2 hours ago, Dandou said:

Yes, DSD is the format in which all Delta-Sigma DACs work

 

ok, but... Chord DACs are not delta-sigma 😉

and, btw, no Amanero USB receiver