Holo Audio Red Streamer

[bodiebill2]

I have to try it as DDC some day. I can easily move my server machine next to the setup and connect it to the Red with USB. I don’t own any isolators anymore though so I’m not sure if connecting the server machine directly to the Red is a smart move. That’s why streaming is so convenient since ethernet is quite well isolated as is.

Then there’s the matter of USB cables.. which unfortunately make a difference to sound quality in my setup. I’ve used Intona Premium and nothing has beaten it so far. I’d imagine i2s cable also matters.

In my experience it is indeed important that the source is as clean as possible, and a USB isolator helps. This one does wonders for me:
https://nl.aliexpress.com/item/1005...t_main.32.3c761802HEVYmH&gatewayAdapt=glo2nld

Actually there is a third way to use the Red (and some other streamers): install an OS, such as moOde ore Volumio, that can play local files, attach an SSD drive, preferably powered with a good LPS, and start a playlist. Here this already sounds better than streaming the audio over the network. And it even gets better when the network cable is removed during Play. This way the Red is used as a standalone player. I use a smart switch so that I can remotely turn off and on the EtherRegen/network.

 

[Patu]

In my experience it is indeed important that the source is as clean as possible, and a USB isolator helps. This one does wonders for me:
https://nl.aliexpress.com/item/1005...t_main.32.3c761802HEVYmH&gatewayAdapt=glo2nld

Actually there is a third way to use the Red (and some other streamers): install an OS, such as moOde ore Volumio, that can play local files, attach an SSD drive, preferably powered with a good LPS, and start a playlist. Here this already sounds better than streaming the audio over the network. And it even gets better when the network cable is removed during Play. This way the Red is used as a standalone player. I use a smart switch so that I can remotely turn off and on the EtherRegen/network.

I need to be able to use HQPlayer so the SSD-method wouldn’t work for me. Red needs to be able to see my server which runs Roon and HQP.

Fibre is also something that I haven’t tried yet.

 

[doitttt]

holo red is a good stream but too advanced for me
then bought an EVERSOLO DMP-A6 streamer 900 dollars
the best user-friendly. compare streamer cost 4000 dollars, EVERSOLO is better.
you can put sdd m2 2280 in at 2 terabytes.
the sound absolutely fantastic
the best thing about sdd, you can shuttle files.
very little jitter

 

[GoldenSound]

Maybe @GoldenOne can confirm how the measurements were taken?

Anyway, how would you describe Red’s sound quality as a streamer, compared to your Diretta chain? Diretta protocol is an interesting approach but I have no idea could it be used with HQPlayer.

I measured both in DDC mode and streamer mode. Performance is identical in both. As expected given as there's no synchronous input in either case.

 

[Patu]

I measured both in DDC mode and streamer mode. Performance is identical in both. As expected given as there's no synchronous input in either case.

Thank you. Well there goes the theory that RPi4 would somehow contaminate the output while streaming with Red.

 

[Dandoudou]

I measured both in DDC mode and streamer mode. Performance is identical in both. As expected given as there's no synchronous input in either case.

Hi @GoldenOne ,

If I understand correctly, you measured the Red in its 'DDC mode', in other words with the SD card of Red OS inserted, and the OS running on the Rpi.

I'm not using it this way. When I say that I use it as a DDC, it's as a pure DDC, without any SD card inserted, and no OS running on the Rpi. (The led on the front of the device is not green, it's red.) And the Red works this way as a DDC (without an OS on which the Rpi would boot), like any pure DDC does: AGD DI-20, Gustard U18, and so on.

I think that this is the reason for the confusion.

 

[GoldenSound]

Hi @GoldenOne ,

If I understand correctly, you measured the Red in its 'DDC mode', in other words with the SD card of Red OS inserted, and the OS running on the Rpi.

I'm not using it this way. When I say that I use it as a DDC, it's as a pure DDC, without any SD card inserted, and no OS running on the Rpi. (The led on the front of the device is not green, it's red.) And the Red works this way as a DDC (without an OS on which the Rpi would boot), like any pure DDC does: AGD DI-20, Gustard U18, and so on.

I think that this is the reason for the confusion.

So I just tested this to see if performance was different and it's again behaving identically to when the pi is running.

The pi itself is quite a low noise device, and also the RED has some excellent filtering so that's not surprising.
I have been able to show slightly higher noise if connected to a very noisy PC (big desktop running furmark + prime95 at the same time), but the increase is very small.

When running in streaming mode, DDC mode with the pi running or not, the performance is identical. The only thing I've found that affects performance is if you have an excessively noisy source device.

 

[Dandoudou]

When I run Red OS, the device is hot like a tube power amp.
When I run it without any OS, like I do with my DI-20HE, the Red runs at the temperature of the room.

In both cases, I feed it with a very clean stream.
As a streamer, it is optically isolated. As a DDC, it receives fiber USB.
The difference in sound quality is very important without an OS running on it.

 

[GoldenSound]

When I run Red OS, the device is hot like a tube poweramp.
When I run it without any OS, like I do with my DI-20HE, the Red runs at the temperature of the room.

In both cases, I feed it with a very clean stream.
As a streamer, it is optically isolated. As a DDC, it receives fiber USB.
The difference in sound quality is very important if no OS runs on it.

Heat dissipation doesn't mean that the device is performing any differently in terms of noise or jitter. Just means it's consuming more power. So long as the filtering is effective enough (which it is), the performance impact could be basically nothing.

If you feel it sounds better in a particular way of using it, then there's no reason not to continue using it that way. But in terms of the actual jitter and noise performance, they don't change at all

 

[bodiebill2]

The pi itself is quite a low noise device, and also the RED has some excellent filtering so that's not surprising.
I have been able to show slightly higher noise if connected to a very noisy PC (big desktop running furmark + prime95 at the same time), but the increase is very small.

So what comes before the Red still has some effect. What @Dandoudou 's and my setup have in common, is that we precede the Red in DDC mode with another streamer resp. stand alone player. So this may explain the difference in sound quality that we perceive.

Supposing that jitter and electrical noise are the only parameters influencing the SQ (but are they?), your hypothesis would probably mean that these for instance (supposing that two Reds are available) would sound the same:
(1) ethernet > Red as streamer > USB out > USB isolator > USB in > Red as DDC > AES > DAC
(2) ethernet > Red as streamer > AES > DAC
Or am I rambling?

 

[GoldenSound]

So what comes before the Red still has some effect. What @Dandoudou 's and my setup have in common, is that we precede the Red in DDC mode with another streamer resp. stand alone player. So this may explain the difference in sound quality that we perceive.

Only in extreme situations.

The RED's USB input is not fully galvanically isolated like the Holo DACs for instance, but it still has extremely effective filtering.
Connecting it to my laptop or my desktop showed no change in performance at all. It was only when I literally maxed out the PC (running furmark and prime95 which are two of the most intense stress tests available for your CPU/GPU) that a small increase in noise getting through was observed.

But if you're running >1000W of compute tasks then you've got bigger problems than a 1-2mV increase in noise getting through and optimising for that degree of improvement when you've got fans running full blast next to you seems a little silly.

Supposing that jitter and electrical noise are the only parameters influencing the SQ (but are they?)

The only other factor for a digital connection would be data integrity, some DDCs or streamers may alter the audio data in some way. Be it compression, DSP of various types, some may need to perform resampling.
The RED however (and indeed the vast majority of DDCs) is bitperfect so that isn't a concern.

Besides that, yes, for a digital source device the only two other factors are how much noise it may be conducting to the connected DAC, and the timing accuracy of the clock signal it provides to the DAC.
There are many different types of distortion, alteration and ways in which analog devices can behave differently. But digital devices whilst it's definitely not as simple as 'bits are bits' as some say, there are also only a couple factors that can be of influence.

your hypothesis would probably mean that these for instance (supposing that two Reds are available) would sound the same:
(1) ethernet > Red as streamer > USB out > USB isolator > USB in > Red as DDC > AES > DAC
(2) ethernet > Red as streamer > AES > DAC
Or am I rambling?

Barring any unusual grounding issues that the first instance could potentially cause (though unlikely to be the case), yes.
You're using exactly the same clock source, and with the red being fed by USB in either case (the pi communicates to the RED's digital section via USB internally) you get absolutely identical performance.

Some DDCs will behave slightly differently if they have an asynchronous input (SPDIF input on a DI20HE or MScaler for instance) vs on the USB input. But that's not the case on the RED

 

[Dandoudou]

If the measurements show that the Red produces the same quality with Red OS, and without an OS at all, it should produce, logically, the same sound quality with any OS.
It would mean that the Red as an endpoint performs the same with: Red OS = NAA = PicorePlayer = Diretta Target on GentooPlayer = Volumio…

 

[GoldenSound]

If the measurements show that the Red produces the same quality with Red OS, and without an OS at all, it should produce, logically, the same sound quality with any OS.
It would mean that the Red as an endpoint performs the same with: Red OS = NAA = PicorePlayer = Diretta Target on GentooPlayer = Volumio…

Correct, you'll just lose certain configuration options when using something other than RED OS.

 

[Dandoudou]

Correct, you'll just lose certain configuration options when using something other than RED OS.

Did you test the Red as an endpoint for a server on the LAN or as a standalone device, connected to the DAC?

 

[GoldenSound]

Did you test the Red as an endpoint for a server on the LAN or as a standalone device, connected to the DAC?

Im not entirely sure what you're meaning.

Connecting ethernet didn't alter noise output if that's what you mean