Allo Digione Signature

[linuxfox]

1 hour ago, nbpf said:

Thus, my temporary conclusion is that either the S/PDIF interface of the Naim DAC is not as good as its USB interface or that the DigiOne Signature is not as good as I hoped.

Today I have received and used the 2 x 18650 batteries on the clean side. I'm quite happy with it. It is close to the usb interface of the NDAC, the closest S/PDIF I have tried so far. The usb interface has still many advantages, but the only annoying difference (for me) is a little bit of bass authority missing.I have to try for a better PS for the dirty side (at the moment I'm using an Amazon Basics smps). Tomorrow I'll try to use batteries on the dirty side as well (an usb powerbank).

In my tests the difference between real-time kernel vs. standard kernel has always been too small: setting the cpu governor on "performace" is usually enough.

In terms of sound quality my favourite player is mpd (alone or with UpMpd), squeezelite is second.

 

[LowMidHigh]

2 hours ago, nbpf said:

Thus, my temporary conclusion is that either the S/PDIF interface of the Naim DAC is not as good as its USB interface or that the DigiOne Signature is not as good as I hoped.

 

The DS gives out notebly clean signal, so I hazzard to guess the coax implementation on the Naim isn't as good as USB.

When I feed my Mutec MC3+USB spdif coax, the sound is nothing less than amazing. Switching to USB, including over SOtM 200 Neon, is big step down--the reversed situation than yours.

[LowMidHigh]

2 hours ago, nbpf said:

In other words, the differences between Ghent Audio cables and  Uptone Audio cables, transcoding and non transcoding, real-time vs. standard kernel are negligible in comparison with the difference between replay from USB and replay from the DigiOne Signature. 

 

I usually buy cables from Dyson: quality components, immaculate construction, down-to-earth approach with zero snake oil.

I asked for a DC cable, and his recommendation included basic components for half the price. I asked if it would perform as well as Ghnet. His reply:

"That's just the way I would build it. This isn't the kind of cable that benefits from esoteric materials."

So here you have it.

[wanta911]

11 hours ago, nbpf said:

 

Thus, my temporary conclusion is that either the S/PDIF interface of the Naim DAC is not as good as its USB interface or that the DigiOne Signature is not as good as I hoped.

 

Thanks for this post. It's a reminder that after the euphoria of getting something new that all might not be as it seems. As I've hinted in a previous post/s, the SQ of the DS sometimes seems too warm and hazy (mid bass) in my setup and it was bugging me.

 

It motivated me to put the Aries back in and do some more comparisons which I've been doing all afternoon using the Lumin App for the DS and Auralic's Lightning App for the Aries....both drawing from Minimserver on my NAS. Well the difference in the bass is quite obvious and I'm not sure why it wasn't this obvious when I initially did the comparison? I run the Aries with an SBooster with the Ultra add-on after ditching the Auralic LPS very early on, which I would describe as mushy. The SBooster makes the Aries bass really tight and resolving but it can tend to sound thin on poor recordings. In comparison the DS bass is warmer, which can make it more musical but can also make some material sound muddy. For me though, the DS still has a better soundstage and more ability to unearth small details.

 

One of the advantages of using room correction is that I can make separate target curves for each unit, so that's what I'll be doing later to do some more comparisons ?

[LowMidHigh]

2 hours ago, wanta911 said:

 

Thanks for this post. It's a reminder that after the euphoria of getting something new that all might not be as it seems. As I've hinted in a previous post/s, the SQ of the DS sometimes seems too warm and hazy (mid bass) in my setup and it was bugging me.

 

It motivated me to put the Aries back in and do some more comparisons which I've been doing all afternoon using the Lumin App for the DS and Auralic's Lightning App for the Aries....both drawing from Minimserver on my NAS. Well the difference in the bass is quite obvious and I'm not sure why it wasn't this obvious when I initially did the comparison? I run the Aries with an SBooster with the Ultra add-on after ditching the Auralic LPS very early on, which I would describe as mushy. The SBooster makes the Aries bass really tight and resolving but it can tend to sound thin on poor recordings. In comparison the DS bass is warmer, which can make it more musical but can also make some material sound muddy. For me though, the DS still has a better soundstage and more ability to unearth small details.

 

One of the advantages of using room correction is that I can make separate target curves for each unit, so that's what I'll be doing later to do some more comparisons ?

 

 

The difference is sound can only be attributed to jitter.

The more jitter, the muddier.

 

In theory, the Arie’s jitter is greater than DS, so I suppose that’s what behind your perception.

 

If you want bring jitter levels to bare minimum, you have to take the single through a re-clocker, or a chain of them, and even tether it/them to a 10MZ reference clock.

[wanta911]

7 minutes ago, LowMidHigh said:

 

 

The difference is sound can only be attributed to jitter.

The more jitter, the muddier.

 

In theory, the Arie’s jitter is greater than DS, so I suppose that’s what behind your perception.

 

If you want bring jitter levels to bare minimum, you have to take the single through a re-clocker, or a chain of them, and even tether it/them to a 10MZ reference clock.

 

Agreed....but the DS is muddier than the Aries, which suggests the DS has more jitter?

[LowMidHigh]

23 minutes ago, wanta911 said:

 

Agreed....but the DS is muddier than the Aries, which suggests the DS has more jitter?

 

There's no other explanation, which is interesting since Allo brags how much they put into reducing the jitter (0.4fs, if I recall). 

 

Of course, the DS has to be powered with clean DC, lest jitter is introduced.

Also 75ohm coax cables are prone to pick up interference. Keep them short (3 meter is really pushing it).

Lastly, the acrylic case can be an issue if RFI and EMI are abound. A good 3mm metal case would go a lot way to shield the boards (you hear that, Allo?)

 

A bit off tangent, if you stream, set the DS as server. Don't stream to it. In my system the difference is painful. 

[RX8R3ROD]

And, important though it undoubtedly is, there is a little bit more to it than jitter alone, ... such as EMI and RFI, as you say.

 

Just changing capacitors in the power supply can make a significant difference to how muddy / tight / boomy / etc the DS sounds.

 

Whilst the capacitors themselves may affect jitter (something beyond my ability to measure reliably given my current tools) they certainly affect the ‘speed’ of the power supply, and some combinations will work rather better with some systems and recordings than others.

 

For example, in my system using Implulse TD712Z’s with just 4” single drivers, there’s tons of speed but relatively limited bass.

 

One type of supercap can add a beautiful, meaty ‘whack’ to (for example) the drums in Terence Trent Darby’s Wishing Well so that that they time (to my ears)  phenomenonally well, but muddle the interplay between bass and piano in, for example, Temptation by Diana Krall. Other supercaps reverse the effect, and the ESR values in the different caps are quite different, which is thought, at least in part, to explain this effect.

 

Perhaps something to bear in mind, and possibly why it can be such a struggle to get a system that sounds consistently good across many recordings and musical styles.

[nbpf]

20 hours ago, linuxfox said:

Today I have received and used the 2 x 18650 batteries on the clean side. I'm quite happy with it. It is close to the usb interface of the NDAC, the closest S/PDIF I have tried so far. The usb interface has still many advantages, but the only annoying difference (for me) is a little bit of bass authority missing.I have to try for a better PS for the dirty side (at the moment I'm using an Amazon Basics smps). Tomorrow I'll try to use batteries on the dirty side as well (an usb powerbank).

In my tests the difference between real-time kernel vs. standard kernel has always been too small: setting the cpu governor on "performace" is usually enough.

In terms of sound quality my favourite player is mpd (alone or with UpMpd), squeezelite is second.

 

For me too, the DigiOne Signature is the best S/PDIF source that I have had in my system so far. I have never tried the "canonical" NDX, NDX2, etc.

 

As I reported, I would possibly have difficulties taking apart the Signature S/PDIF feed from direct replay from USB sticks on instrumental music. But when listening to vocal music I can hear an obvious difference. In my perception, it is not a matter of more or less glare, resolution or bass authority but rather of something that I would describe like a tonal shift. Listening to track 1 of 

 

  https://www.discogs.com/Johann-Sebastian-Bach-Ophélie-Gaillard-Pulcinella-Bach-Arias/release/9874603

 

in 24bits/88.2kHz, the voice of Sandrine Piau sounds obviously more natural (to me) when replay is from the USB stick. Which leads me to the question of MPD and MinimServer settings. Here are my MPD audio output, buffering and resources settings: 

 

audio_output {
  type                "alsa"
  name                "Allo DigiOne"
  device              "hw:sndallodigione"
  mixer_type          "none"
  replay_gain_handler "none"
  dsd_usb             "yes"
  auto_resample       "no"
  auto_channels       "no"
  auto_format         "no"  
  use_mmap            "yes"
}

 

audio_buffer_size               "8192"  # 2018-10-28, from Volumio
buffer_before_play               "10%"  # 2018-10-28, from Volumio

max_connections                   "20"
max_playlist_length            "81920"
max_command_list_size          "81920"
max_output_buffer_size         "81920"

And here are my MinimServer settings:

+*props
aliasTags=director:conductor
.autoUpdate=true
contentDir=/media/sd/share/audio/data
displayName=MinimServer[rpi6]
http.port=9790
indexTags=composer:Komponist, work:Werk, conductor:Dirigent, ensemble:Ensemble, All Artists:Künstler, Genre, Form, period:Epoche, Date:Aufnahmedatum, -incdate:Seit, Gramophone's top 10 violin concertos, Gramophone's top 10 symphonies, Gramophone's top 10 piano sonatas, Gramophone's top 10 piano concertos
itemTags=Comment
.logFile=minimserver.log
.logLevel=info
mergeFolderAlbums=true
ohnet.debug=Default
ohnet.port=9791
showExtras=true
startupScan=true
stream.transcode=flac:wav
tagCustom=AlbumArtist.displayRole={artist}, Composer.displayRole={artist}
tagFormat=Title.displayFormat={$title$composer^^^ | }, Artist.displayFormat={$artist$ensemble$conductor}, Comment.displayFormat={$*infoFileURI^<a href="^">Digital booklet</a>$comment}
tagOptions=Album.sortTags={Album, Artist}
.updateReminder=1
+:status=0

Is there anything that strikes you as being different in your MPD settings? When you use MPD in conjunction with upmpdcli, what is the UPnP server? Thanks, nbpf

[nbpf]

21 hours ago, LowMidHigh said:

 

The DS gives out notebly clean signal, so I hazzard to guess the coax implementation on the Naim isn't as good as USB.

When I feed my Mutec MC3+USB spdif coax, the sound is nothing less than amazing. Switching to USB, including over SOtM 200 Neon, is big step down--the reversed situation than yours.

I am not talking about differences between S/PDIF and USB streaming. The Naim DAC does not support USB streaming. It has two USB ports for direct replay from a USB drive. It replays the files in the order in which they appear in the file system of the connected USB drive and the only possible user interactions are to skip to the next and to the previous files and to stop replay.Thus, USB replay is not very practical. But it sounds better than any source that I have so far attached to the DAC!

[nbpf]

11 hours ago, wanta911 said:

 

Thanks for this post. It's a reminder that after the euphoria of getting something new that all might not be as it seems. As I've hinted in a previous post/s, the SQ of the DS sometimes seems too warm and hazy (mid bass) in my setup and it was bugging me.

 

It motivated me to put the Aries back in and do some more comparisons which I've been doing all afternoon using the Lumin App for the DS and Auralic's Lightning App for the Aries....both drawing from Minimserver on my NAS. Well the difference in the bass is quite obvious and I'm not sure why it wasn't this obvious when I initially did the comparison? I run the Aries with an SBooster with the Ultra add-on after ditching the Auralic LPS very early on, which I would describe as mushy. The SBooster makes the Aries bass really tight and resolving but it can tend to sound thin on poor recordings. In comparison the DS bass is warmer, which can make it more musical but can also make some material sound muddy. For me though, the DS still has a better soundstage and more ability to unearth small details.

 

One of the advantages of using room correction is that I can make separate target curves for each unit, so that's what I'll be doing later to do some more comparisons ?

Thanks for the report, please let us know your findings with room correction!

[clipper]

On 12/20/2018 at 10:13 PM, LowMidHigh said:

I'm in this game from streaming, particularly SoundCloud. When LMS streams directly, via the plugin, the SQ is amazing; simply gripping. On the other hand, when streaming off my Android device or MacBook Pro (Airplay), the yields is markedly less stellar. I guess the initial processing by the none-audio-grade gear, compounded by the additional network traffic, inject the signal with high levels of jitter. 

 

Unfortunately, LMS plugins leave a lot to be desired—really a lot. Therefore, my first proposal is Chromecast built-in capabilities. Spotify, Tidal, Deezer, YouTube, SoundCloud and many others would be then streamed directly by LMS, with the corresponding SQ. The user would still enjoy the sleek and rich UI provided by the vendors. Truly, the best of both worlds. 

I just noticed that the LMS ickstream plugin (see http://www.ickstream.com) has SoundCloud available to add as a service (along with Tidal, Qobuz, Calm Radio, Deezer, radio.net, Radionomy, Spotify and UberStations).

 

Accessing SoundCloud with the ickstream plugin looks a little different than it does with the SoundCloud plugin...

 

See http://wiki.ickstream.com/index.php/Open_Beta_Squeezebox_Installation for installation instructions for ickstream.

[LowMidHigh]

4 hours ago, nbpf said:

I am not talking about differences between S/PDIF and USB streaming. The Naim DAC does not support USB streaming. It has two USB ports for direct replay from a USB drive. It replays the files in the order in which they appear in the file system of the connected USB drive and the only possible user interactions are to skip to the next and to the previous files and to stop replay.Thus, USB replay is not very practical. But it sounds better than any source that I have so far attached to the DAC!

In terms of SQ, the principle is the same though: the implementation that suffers from least jitter will sound best. 

[LowMidHigh]

6 hours ago, RX8R3ROD said:

And, important though it undoubtedly is, there is a little bit more to it than jitter alone, ... such as EMI and RFI, as you say.

 

Just changing capacitors in the power supply can make a significant difference to how muddy / tight / boomy / etc the DS sounds.

 

Whilst the capacitors themselves may affect jitter (something beyond my ability to measure reliably given my current tools) they certainly affect the ‘speed’ of the power supply, and some combinations will work rather better with some systems and recordings than others.

 

For example, in my system using Implulse TD712Z’s with just 4” single drivers, there’s tons of speed but relatively limited bass.

 

One type of supercap can add a beautiful, meaty ‘whack’ to (for example) the drums in Terence Trent Darby’s Wishing Well so that that they time (to my ears)  phenomenonally well, but muddle the interplay between bass and piano in, for example, Temptation by Diana Krall. Other supercaps reverse the effect, and the ESR values in the different caps are quite different, which is thought, at least in part, to explain this effect.

 

Perhaps something to bear in mind, and possibly why it can be such a struggle to get a system that sounds consistently good across many recordings and musical styles.

Tweaking with the power supply would affect the jitter and jitter characteristics. Nice of you to push the envelope. 

 

Still, I don't think PS can affect the timing per se. That is the exclusive responsibility of the clock. 

[nbpf]

9 hours ago, LowMidHigh said:

In terms of SQ, the principle is the same though: the implementation that suffers from least jitter will sound best. 

I am not sure: the Naim DAC reclocks the data prior entering the DAC stage, see page 4 of  https://www.naimaudio.com/sites/default/files/products/downloads/files/naim_dac_august_2009.pdf. Thus, I would expect jitter figures to be the same for both inputs.

[LowMidHigh]

50 minutes ago, nbpf said:

I am not sure: the Naim DAC reclocks the data prior entering the DAC stage, see page 4 of  https://www.naimaudio.com/sites/default/files/products/downloads/files/naim_dac_august_2009.pdf. Thus, I would expect jitter figures to be the same for both inputs.

 

The concept of buffering the incoming stream and he re-clocking it is as obvious as it is elusive.

 

I heard the CEO of a high-end re-clocker assert they do just that. “So why does optical and coax sound different with your product?” I retorted. “Why your USB is not as resolving as your AES?”

 

He huffed that he never got the time to listen to his own products so I couldn’t agree with me, and that was it. 

 

In fact, you hear the difference yourself, do you not? If the process were bullet-proof, unaffected by the incoming clock, the sound would be source-neutral.  But your own experience refutes that.

[LowMidHigh]

1 hour ago, nbpf said:

I am not sure: the Naim DAC reclocks the data prior entering the DAC stage, see page 4 of  https://www.naimaudio.com/sites/default/files/products/downloads/files/naim_dac_august_2009.pdf. Thus, I would expect jitter figures to be the same for both inputs.

 

BTW, my theory—which is worth very little—is that the jitter enters the receiver with the signal, and although the data is stripped and re-dressed with the new clock, it comes back to mutilate the signal along its output path with vengeance. 

[linuxfox]

18 hours ago, nbpf said:

Is there anything that strikes you as being different in your MPD settings? When you use MPD in conjunction with upmpdcli, what is the UPnP server? Thanks, nbpf

 

My mpd configuration is very similar, the only real differnce is the dsd part:

audio_output {
        type            "alsa"
        name            "DigiOne"
        device          "hw:0,0"
        mixer_type      "none"
        auto_resample   "no"
        use_mmap        "yes"
        dop             "yes"

}

NDAC doesn't support direct DSD, but it supports dop (DSD over PCM) so "dop  yes" should be the right choice.

As UPnP server I use logitech media server, but I never noticed differencies between direct mpd and upmpdcli+media servers.

In addiction this is the rilevant part of my config.txt :

gpu_mem=16
dtoverlay=pi3-disable-wifi
dtoverlay=pi3-disable-bt
dtoverlay=i2s-mmap
dtoverlay=allo-digione

 

And I use cpufrequtils with this  /etc/default/cpufrequtils:

GOVERNOR="performance"

 

I use ferrite chokes on all the ethernet cables

[DancingSea]

Does the Signature, original digione or the USBridge support DSD?  Does it pass DSD to the DAC?

[agladstone]

1 hour ago, DancingSea said:

Does the Signature, original digione or the USBridge support DSD?  Does it pass DSD to the DAC?

It’s the best streamer ever ! 

[quanghuy147]

I have just tried Volumio, however, i just learned that Signature can be run on Moode and Rune.

 

Which software do you think is best for natural and organic sound?

 

Thank you.

[burmar]

1 hour ago, quanghuy147 said:

Which software do you think is best for natural and organic sound?

 

I think the best software is picoreplayer as squeezeplayer and a separate LMS server on anything you want..

I'm using a little fanless pc as server.
The sound quality is, in my test (tried all kinds of mpd players) the best

[LowMidHigh]

4 hours ago, quanghuy147 said:

I have just tried Volumio, however, i just learned that Signature can be run on Moode and Rune.

 

Which software do you think is best for natural and organic sound?

 

Thank you.

 

I don't think thd OS makes a difference, rather the music server. LMS is probably best supported--for whatever it's worth--so I'd stick with it. 

 

The most important decision is where the rendering takes place. You'll get best SQ if LMS renders directly, but the user interface leaves a lot to be desired, and glitches are common with plug-ins that can be finicky. 

 

Still, I opted for that option, hoping Allo would implement Chromecast Built-in one day. The advantages are overwhelming: server based quality streaming, while the user can interact with the service via a native, cutting-edge phone app, sleek and packed with features. Best of both worlds. 

 

The merit hasn't been lost on others: Both Cambridge Audio and Naim have incorporated CC built-in in their new network players. 

 

I've floated that idea to Allo. Hope you guys would partition them too. 

[Flyman]

Well guys a bit Confused...

 

with method is the best...bnc to bnc or coaxial to coaxial? Also a vfm cable suggestion would be a help.

 

What about coaxial to coaxial vs bnc to coaxial?

 

second thing the power.the dirty side with allo psu.ok that's ok.

ypu say for the clean side 7v...5v....but the 18650 as i know is 3,7v for the battery pack.also if there is a 7v battery....did it damage the card?for all that thing that i read there isn't reason for so much voltage.

 

maybe a newbie question...i have an ifi ipower 9v...can i use it in the dirty side or not?

 

Ps...ihave for now a raspberry pi3 b+ And i use with a powerbank with a very good results in sound.better than my linear power supply.so why in the dirty side you can't use a powerbank?

[LowMidHigh]

7 minutes ago, Flyman said:

Well guys a bit Confused...

 

with method is the best...bnc to bnc or coaxial to coaxial? Also a vfm cable suggestion would be a help.

 

What about coaxial to coaxial vs bnc to coaxial?

 

second thing the power.the dirty side with allo psu.ok that's ok.

ypu say for the clean side 7v...5v....but the 18650 as i know is 3,7v for the battery pack.also if there is a 7v battery....did it damage the card?for all that thing that i read there isn't reason for so much voltage.

 

maybe a newbie question...i have an ifi ipower 9v...can i use it in the dirty side or not?

 

Ps...ihave for now a raspberry pi3 b+ And i use with a powerbank with a very good results in sound.better than my linear power supply.so why in the dirty side you can't use a powerbank?

 

I'll just address the interface.

 

Both BNC and RCA are audio coaxial. BNC adheres to strict impedance, 75ohm in the Signature case (which is the AES specs). RCA can be all over the place.

 

As you may have guessed BNC is the way to go, given the receiving end adheres to 75ohm as well. Mix is match is not recommended.

 

I'm cable skeptic, especially digital ones, I'd go with anything that is well constructed and meets the specs. Blue Jeans is a good option, but I like Dyson Cables best.  Zero snake oil, 100% value.