Playback is not gapless

Studio
21

Dear Agoldnear,

I am reading your posts, and I appreciate your knowledge and the high accuracy you use to describe each matter. I have been an audiophile with a “good” technical background for years. I used Roon for years with all combinations it gives, including HQP. I subscribed to Audirvana Studio two months back after the trial. I am listening to Qubuz through Auirvana while typing and enjoying the superb experience.
@Agoldnear

If you do not mind, I still want to benefit from your experience. Since I am still relatively new to Audrivana, I did not try many combinations. So, what is your opinion on the below (my objective is the best sound quality):

  • MacOS of Windows and what can be the difference.
  • PC USB to the DAC or UpnP to a small streamer connected to the DAC
  • Any other advice for the best sound quality?

My setup is as follows:

  • Windows PC specially configured with i7 and 16GB of RAM, powered in a way to isolate the ground USB USB. I use Kernal Streaming. The processor is configured with a fixed clock at 5.2 Ghz and water cooling (Corsair).
  • USB from the computer goes to Chord MScaler for upsampling. MScaler is powered by a battery to isolate the power
  • MScaler is connected through dual SPDIF to Chord Dave. Dave is also powered by a LAB power supply also to isolate the power.
    Then preamp (Mcintosh and bi-amping: Tube for high and Macintosh MC452 for bass)
  • B&W speakers 802 Diamond D2

An isolation transformer is used for power, and line filters also are used. The ground is completely isolated from the GRID. I made those changes gradually over the past few years.
I am pretty happy with the sound.

Your opinion or anybody is highly appreciated.
Enjoy the music.
Thanks.

22

@imadch … I appreciate your playback system amalgamation and configuration/implementation…

You will find that I am Apple-centric and a proponent of DSD, and I do not employ UPnP for any digital-audio…

My reference playback system is USB-centric, being served by a Thunderbolt 3/4 expansion chassis hosting an ELFIDELITY AXF-107 Power Filter PCI Express / PCI Express 16x card and USB 3.0 PCIe card and WireWorld USB 3.0 cable serving iGalvanic 3.0 that delivers USB 2.0 via the UpTone USPCB A>B Adapter to my DAC (TEAC UD-501)… Power and ground/earthing is managed through a complex system of filtering and harmonic conditioning independently, with great attention paid to the mitigation of RF and EMF inducted noise.

I choose the 3.0 cable for the fundamental USB-3 design protocol of the separation of power and signaling lines and data-lines, that remove inducted power/ground noise and signaling noise on the data-line and the WireWorld cable design dogma that further improves isolation.

Personally I am not a big fan of the PCM-centric DAVE and M-Scaler design dogma (One of my best friend/audiophile has the exact Chord M-Scaler + DAVE system you are employing) … My perspective is purely a PCM vs DSD one.

It is my opinion that for best network based audio distribution, optical transmission is imperative for reasons of noise mitigation and hopefully someday Ravenna protocol will prevail.

:notes: :eye: :headphones: :eye: :notes:

1 Like
23

Today, this is simpler to answer since the M-series platforms have come to market,… The answer is this: The most important differentiating element regarding high-resolution digital-audio playback on all Apple Inc. platforms vs Windows, etc platforms, is the vertical-integration of the M-series CPU and and related platform topology with the MacOS and now iOS… It has always been about the MacOS and the Apple designed hardware synergies.
:notes: :eye: :headphones: :eye: :notes:

24

Thank you for taking the time to reply to my post.

The Mac M-series (ARM) is a fantastic platform from Apple. To share some experience, I tried Mac mini with M with a known streaming server but was not impressed. Windows performed better.

Now with Audirvana, I am so happy with the performance that I did not try Mac yet. So, sure, I will try the M-series shortly and tell you. I have mini and Pro Max.

25

If you were using the Mac Mini headless, this is a problem, because the graphics sub-system is not being engaged and the CPU does all of the graphics processing of the UI, etc…

26

@Agoldnear

WOW. I enjoyed reading every word you wrote. There are multiple ways to achieve good sound, and I am keen to learn more about your different components.

Let me ask you one question regarding USB. Many use USBs with separated power and signal wires are there. However, as we know the power in the USB cable is DC, and DC gives no magnetic field like AC, especially since the power is negligible.
What is your recommendation if I add a USB PCI card to my PC? Currently using an ASUS motherboard with ASMEdia built-in USB controller. No complaints as it is good after some careful configuration.
Thx.

27

Can you please explain more about what you mean by headless, as I used it with and without a monitor? I also tried my Macbook Pro MAX with 64GB of RAM, which is supposed to be a beast. My current setup was better.

28

The USB 3 cable design protocol, eliminates packet request interrupts on the data-line and the WireWorld USB 3.0 cable design mitigates inducted noise that may be picked-up on the power/earth lines … The transmission bandwidth/speed/request latency provides an unfettered digital-audio signal path… In my system the iGalvanic transition to USB 2.0 via the UpTone USB 2.0 adapter (very short at approximately 4") is key in reducing packet request interrupt latency.

The isolation of the USB 3 interface provided by the PCIe expansion chassis being supplied with a very clean power-signal in concert with the USB 3.0 transmission path, virtually eliminates noise induced jitter on the USB 3 controller clock. I believe the PCI power filtering is improving the Thunderbolt power signal all the way back to the Thunderbolt controller…

I can’t speak to your subjective experiences with your Apple platforms… However, “headless” means using the computer without a display attached…

29

I am using USB 3.2 port now, and it seems better than 3.1 or 2.0 (All are good in my case), knowing that vendors recommend USB 2.0. I guess most DACs do not utilize still higher than USB 2.0 (please correct me). I am using a custom silver USB cable. It is good compared to many snake oil in the market. I did not try WireWorld Cables, but they seem good.

What PCIe expansion chassis are you using? If you please state the brands and models, that would be awesome.

I will share with you my new MAC testing comparison soon. As I know, with MACyou will go with ALSA as there is no Kernal or ASIO streaming.

Cheers.

30

You can as I did add an HDMI monitor dongle that simulates a screen to activate the GPU. 10USD to solve it.

1 Like
31

There is a performance advantage in keeping the high-speed USB 2.0 cable as short as possible…

In my case with the UpTone Audio USB 2.0 USPCB A>B adapter card on the output of iGalvanic 3.0, which is directly attached to my DAC like a giant dongle, the USB 2.0 packet request interrupts are extremely short in duration and any noise precipitated in the signaling process is isolated in iGalvanic… the throughput bandwidth of the iGalvanic + USB 3.0 cable (5 Gbits/s) reduces signaling latency to and from the USB 3.0 bus controller and increases the packet speed from the bus controller, to iGalvanic… The buffering and re-clocking, etc, of iGalvanic 3.0 seem to be optimized in this operational state… My feeling is the USB digital-audio signal is contextually free of noise related jitter-precipitating distortions of the leading and trailing edges of the signal waveform…

In the evolution of my USB transmission system architectures over the years from plain-vanilla USB 2.0 transmission to the USB 3.0 system I employ today, I have personally found tangibly perceptible improvements in clarity/focus and detail across the full-spectrum of contextual harmonic color and contextual macro-dynamics/micro-dynamics and contextual Left-channel + Right-channel spatial/phase relationships.

I employ a Sonnettech two-slot Thunderbolt 3/4 PCIe expansion chassis, connected with a good quality TB 3/4 cable being served by my MBP quad-core i7 with 16GB RAM, (6 GB allocated for Audirvana playback pre-load buffer) … I up-convert/modulate all PCMxxx to DSD128 using r8Brain with custom filter settings, in concert with the FIR filter in the UD-501.

:notes: :eye: :headphones: :eye: :notes:

32

Hello. Yes it is a good practice to use cables the shortest possible to reduce the resistivity. The only exception is connectors as if they are so too short, you may not be able to achieve the required resistivity like 75ohm or 110 ohm between equipemts.

I have MScaler and using it as I do not notice any colorization or collapse in the sound imaging. I did upsample a lot to DSD and PCM using HQP but not any more. Upsampleing always changes the sound. Pure PCM sound best in my case and to my likening. I usually stream and using an online upsampler is not an option for me.

Regards.

33

This is because the Chord products are biased to PCM in the fundamental architecture of DAVE and M-Scaler (Digital to Analog Veritas in Extremis) algorithms and output conversion… There is little reason for up-conversion in Audirvana in your case… However my best friend is now using PGGB for up-conversion, and M-Scaler is somewhat superfluous for him now.

34

Yes, DAVE is a PCM DAC, but it works well with DSD. there are a few Japanese pure DSD DACs out there. But most content masters are PCM.
I have tried many DACs myself with friends. All top brands except DCS I did not try. Dave was among the top compared to many, including Esoteric Dx1. Dave is so revolving, and depending on how you power it, it can be subject to a kind of digital glare. Mscaler helps here, but no day and night difference, I agree.

Upsampling is a large subject. I may need to write about it. Using even AI to increase the resolution if digital content is now one of the hottest areas (NVIDIA DLSS3, …) for extreme audio, there can be a caveat. The human listening brain is more complex than the visual system we have. But it is an area of good potential for audio upsampling.

Since I stream, I can only work with a real-time up sampler.

Best Regards.

35

The DAVE system does not have an unfettered 1-bit PDM signal-path to the D/A output circuitry… Virtually all AKM and Burr-Brown DACs provide a direct 1-bit PDM signal-path to the D/A output circuitry… The ESS DAC design is more aligned with DAVE.

It’s obvious to those audiophiles that study the bio-mechanics/physiology/neurology/perception and psychoacoustics of the human-hearing system and and the psychology of cognitive- interpretation, that in the individual, there are many influences on those interpretations of perceived sound-quality.

Outside of streaming… The investment in a PGGB perpetual license is by a wide margin, less than the price of M-Scaler,… Anecdotally, I have insight from my friend’s interpretation of his DAVE system when listening to PGGB processed music… He is emphatic about PGGB processing, and he finds the processing blows-away M-Scaler as applied to his local music library of CD rips, etc… Personally I have not yet, had opportunity to to experience this on his system and corroborate. However, you may already have some subjective experience with PGGB…(My friend also streams files via Qobuz… I do not stream files, I purchase files… His DAVE system is near identical to yours, down to the replacement of the power-supplies and a playback system architecture on a dedicated power circuit/ground plane, optical-isolation, dedicated server, etc)

In a DAC that has a direct 1-bit signal path that delivers a more ‘analog-like’ signal to the PDM D/A circuitry/filter, I find the natural sound-quality of PCMxxx modulated to DSD 5.6MHz (DSD128) by r8Brain in Audirvana, subsequently reproduced from my DAC platform which employs a dual-mono BB PCM1795 DAC chipset, etc, and output architecture, to be very satisfactory in comparison with a native or up-sampled PCMxxx signal that is modulated to PDM in the DAC architecture and presented to the D/A circuit. (Similar to the obvious reason for your employment of M-Scaler, that usurps/bypasses the algorithm in DAVE, except the DAVE platform is not capable of handling an unfettered 1-bit DSD signal path from input to output.)

Finally, the most important aspect of sound-quality produced from any high-resolution DAC platform is the slew-rate of the D/A circuitry components, and the slew-rate of platform component topology in general…

:notes: :eye: :headphones: :eye: :notes:

36

Hi,

Reviving the thread a bit. Turns out my problem isn’t solved. I still hear a tiny dropout every time a track ends and another starts, even if I bring the buffer down to 256k (it actually doesn’t seem to make a big difference what buffer I choose now). Any other idea as to how to solve this?

I see in Debug Info that “universal gapless playback” is not activated. Could that be the problem?

Audirvana Studio 2.3.4 (20304)

macOS 13.4.1 with 8GB physical RAM

Connected account of : Ota Rie

SIGNAL PROCESSING:

Polarity Inversion:
	Globally: OFF
	Per track: ON
Effects plugins NOT ACTIVE

UPSAMPLING:
r8brain not in use
r8brain filter parameters
Bandwidth = 99.5%
Stop band attenuation 218dB
Phase linear

AUDIO VOLUME:
Max allowed volume: 100
Replay Gain: None
SW volume control: OFF

LIBRARY SETTINGS:
Sync list: 1 folders
AUTO: /Users/jba/Music/Music/Media.localized/Music
iTunes/Music library synchronization: last synchronized on Fri May 5 12:15:21 2023

Library database path: /Users/jba/Library/Application Support/Audirvana/AudirvanaDatabase.sqlite

Local audio files fingerprinting
Tracks with no MBID: 153

Remote Control server:
Listening on 127.0.0.1 on port 49153

ACTIVE STREAMING SERVICES
Qobuz: Connected as Qobuz Studio

=================== AUDIO DEVICE ========================

Active method: UPnP

Preferred device: [UPnP] Brinkmann Nyquist Model UID:Brinkmann Nyquist UID:uuid:83CC11A0-CFC1-11DE-8D30-0024C5002CA2

Currently playing to UPnP device at 24 bits 44.1kHz
Selected device:Brinkmann Nyquist
Manufacturer: Brinkmann
Model name: Nyquist
Model UID: Brinkmann Nyquist
UID: uuid:83CC11A0-CFC1-11DE-8D30-0024C5002CA2
UPnP device at http://192.168.1.17:38400/description.xml
ID 0x0

6 available sample rates up to 192000Hz
44100
48000
88200
96000
176400
192000

Volume control: Yes
Max volume alert: Disabled

MQA capability
Auto-detect MQA devices: Yes
Not a MQA device, user set to not MQA
DSD capability: Unhandled

Device audio channels
Preferred stereo channels L:0 R:1
Channel bitmap: Ox3, layout:
Channel 0 mapped to 0
Channel 1 mapped to 1

UPnP set capabilities
Maximum PCM frequency set: 192000Hz
Maximum PCM bitdepth set: 24
Not native DSD capable
Avoid RAW PCM streams: No
Unwanted playback stop workaround: No

UPnP / DLNA supported protocols :
http-get::audio/mpeg:dlna.org_pn=mp3
http-get::audio/mp4:dlna.org_pn=mp4
http-get::audio/vnd.dlna.adts:dlna.org_pn=aac_adts
http-get::audio/x-ms-wma:dlna.org_pn=wmabase
http-get::audio/x-ms-wma:dlna.org_pn=wmafull
http-get::audio/x-ms-wma:dlna.org_pn=wmapro
http-get::audio/x-ms-wma:dlna.org_pn=wmalsl
http-get::audio/x-sony-oma:dlna.org_pn=atrac3plus
http-get::audio/aiff:
http-get::audio/wav:
http-get::audio/flac:
http-get::audio/aac:
http-get::audio/m4a:
http-get::audio/ac3:
http-get::audio/x-dsf:
http-get::audio/mp4:
http-get::audio/vnd.dlna.adts:
http-get::audio/x-wav:
DLNA 1.5: Yes
Native Gapless playback: Yes
Universal Gapless playback active: No
Missing events workaround: No
Can play native DSD: No
Volume Control: dB
Number of channels: 2
Use as stereo device only: No

1 output streams:
Number of active channels: 2, in 1 stream(s)
Channel #0 :Stream 0 channel 0
Channel #1 :Stream 0 channel 1

Stream ID 0x0 0 channels starting at 0
12 virtual formats:
2 ch Non-mixable linear PCM Interleaved 16 little endian Signed Integer 44.1kHz finite length
2 ch Non-mixable linear PCM Interleaved 24 little endian Signed Integer 44.1kHz finite length
2 ch Non-mixable linear PCM Interleaved 16 little endian Signed Integer 88.2kHz finite length
2 ch Non-mixable linear PCM Interleaved 24 little endian Signed Integer 88.2kHz finite length
2 ch Non-mixable linear PCM Interleaved 16 little endian Signed Integer 176.4kHz finite length
2 ch Non-mixable linear PCM Interleaved 24 little endian Signed Integer 176.4kHz finite length
2 ch Non-mixable linear PCM Interleaved 16 little endian Signed Integer 48kHz finite length
2 ch Non-mixable linear PCM Interleaved 24 little endian Signed Integer 48kHz finite length
2 ch Non-mixable linear PCM Interleaved 16 little endian Signed Integer 96kHz finite length
2 ch Non-mixable linear PCM Interleaved 24 little endian Signed Integer 96kHz finite length
2 ch Non-mixable linear PCM Interleaved 16 little endian Signed Integer 192kHz finite length
2 ch Non-mixable linear PCM Interleaved 24 little endian Signed Integer 192kHz finite length

0 physical formats

Current device transportInfo:
CurrentTransportState: PAUSED_PLAYBACK
CurrentTransportStatus: OK
CurrentSpeed: 1

Current device MediaInfo:
NrTracks: 0
MediaDuration: 0:04:31
CurrentURI: http://192.168.1.19:49155/audirvana/audio_6.wav
CurrentURIMetadata:

dc:titleConcerto for Three Violins in F Major, RV 551: I. Allegro</dc:title>
dc:creatorGiovanni Antonini</dc:creator>
upnp:artistGiovanni Antonini</upnp:artist>
upnp:albumArtURIhttp://static.qobuz.com/images/covers/62/52/0745099455262_600.jpg</upnp:albumArtURI>
upnp:albumVivaldi: Double & Triple Concertos "Il proteo"</upnp:album>
upnp:originalTrackNumber4</upnp:originalTrackNumber>
upnp:classobject.item.audioItem.musicTrack</upnp:class>
http://192.168.1.19:49155/audirvana/audio_6.wav

nextURI: http://192.168.1.19:49155/audirvana/audio_8.wav
nextURIMetadata: <DIDL-Lite xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:upnp="urn:schemas-upnp-org:metadata-1-0/upnp/" xmlns="urn:schemas-upnp-org:metadata-1-0/DIDL-Lite/" xmlns:dlna="urn:schemas-dlna-org:metadata-1-0/">
Concerto for Three Violins in F Major, RV 551: II. Andante Giovanni Antonini Giovanni Antonini http://static.qobuz.com/images/covers/62/52/0745099455262_600.jpg Vivaldi: Double & Triple Concertos "Il proteo" 5 object.item.audioItem.musicTrack http://192.168.1.19:49155/audirvana/audio_8.wav 
PlayMedium: NETWORK
RecordMedium: NOT_IMPLEMENTED
WriteStatus: NOT_IMPLEMENTED

Current transport actions:
Play,Stop,Seek,X_DLNA_SeekTime,Next,Previous

Current device AVT service description:

<?xml version="1.0"?> 1 0 urn:schemas-upnp-org:device:MediaRenderer:1 Brinkmann Nyquist Brinkmann http://www.meridian-audio.com This model is a connected audio module which supports the UPnP/DLNA standard Nyquist CDMCM-210-001 uuid:83CC11A0-CFC1-11DE-8D30-0024C5002CA2 http://www.meridian-audio.com DMR-1.50 image/png 48 48 24 /images/small.png image/png 120 120 24 /images/large.png image/jpeg 48 48 24 /images/small.jpg image/jpeg 120 120 24 /images/large.jpg urn:schemas-upnp-org:service:RenderingControl:1 urn:upnp-org:serviceId:RenderingControl /service/RenderingControl1.xml /service/RenderingControl_control /service/RenderingControl_event urn:schemas-upnp-org:service:ConnectionManager:1 urn:upnp-org:serviceId:ConnectionManager /service/ConnectionManager1.xml /service/ConnectionManager_control /service/ConnectionManager_event urn:schemas-upnp-org:service:AVTransport:1 urn:upnp-org:serviceId:AVTransport /service/AVTransport1.xml /service/AVTransport_control /service/AVTransport_event CDM21006180372 http://192.168.1.17:38400

Current device RootDevice description:

<?xml version="1.0"?> 1 0 urn:schemas-upnp-org:device:MediaRenderer:1 Brinkmann Nyquist Brinkmann http://www.meridian-audio.com This model is a connected audio module which supports the UPnP/DLNA standard Nyquist CDMCM-210-001 uuid:83CC11A0-CFC1-11DE-8D30-0024C5002CA2 http://www.meridian-audio.com DMR-1.50 image/png 48 48 24 /images/small.png image/png 120 120 24 /images/large.png image/jpeg 48 48 24 /images/small.jpg image/jpeg 120 120 24 /images/large.jpg urn:schemas-upnp-org:service:RenderingControl:1 urn:upnp-org:serviceId:RenderingControl /service/RenderingControl1.xml /service/RenderingControl_control /service/RenderingControl_event urn:schemas-upnp-org:service:ConnectionManager:1 urn:upnp-org:serviceId:ConnectionManager /service/ConnectionManager1.xml /service/ConnectionManager_control /service/ConnectionManager_event urn:schemas-upnp-org:service:AVTransport:1 urn:upnp-org:serviceId:AVTransport /service/AVTransport1.xml /service/AVTransport_control /service/AVTransport_event CDM21006180372 http://192.168.1.17:38400

UPnP devices found : 1
Device #0: ID 0x0 : Brinkmann Nyquist
UID: uuid:83CC11A0-CFC1-11DE-8D30-0024C5002CA2
Location: http://192.168.1.17:38400/description.xml
Manufacturer: Brinkmann
Model name: Nyquist

Local

Max. memory for audio buffers: 2240MB

Local Audio Engine: CoreAudio
Actual I/O buffer frame size: 0

Local devices found : 1
Device #0: ID 0x4e Haut-parleurs MacBook Air
Manufacturer: Apple Inc.
Model UID: Speaker
UID: BuiltInSpeakerDevice
Model name: Apple Inc. Haut-parleurs MacBook Air

Chromecast

Chromecast devices found : 0

37

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.