Some VST3 Plugin causes short artifact at beginning of track (buffer problem?)

I think I have potentially the same issue as here:

Loud ‘thud’ noise when using Waves Abbey Road Plugin - Origin - Audirvana

When I use some VST3 plugins (tested at the moment with CanOpener, but I think I also had the problem with some EQ plugins), then I get a short artifact when I start the playback of a new track after I have previously played some other track. It does not happen with the very first track played and it also does not happen when a new track in a playlist gets played automatically.

I also think (but would have to retest) that the artifact is shorter or longer depending on the VST in use.

All of the above looks to me as if there is something in a buffer (potentially in the plugin, not Audirvana) that is not cleaned out when starting a new track, could that be? Is there any setting I could try to prevent this?

This happens on Windows with a RME ADI-2 Pro FS connected via USB, but I remember having had the same issue on macOS as well.

Welcome…
Please paste your debug information report here, so folks can help the best possible… It is found here: Settings–> My Account–> Help–> Debug info

Available System RAM and gain-structure management and sample-rate changes can play into this.

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

That would be this:

Audirvana Studio 2.9.3 (20903)

Windows 11 (22631) x86_64 with 32GB physical RAM

Connected account of : Stefan XXXX

NETWORK
Status: available
Available network interfaces:
Ethernet ({3e8c1c4f-d0fa-467a-81b5-XXXXX}) is private
Windows Defender Firewall status for this instance of Audirvana Studio
Active profile types: all
Private profile:
Firewall: enabled
Inbound: allowed
Outbound: allowed
Notifications: enabled
Public profile:
Firewall: enabled
Inbound: blocked
Outbound: allowed
Notifications: enabled

SIGNAL PROCESSING:

Polarity Inversion:
	Globally: OFF
	Per track: ON
Effects plugins ACTIVE in realtime mode
	VST3 plugin #0: C:\Program Files\Common Files\VST3\Ghz CanOpener Studio 3.vst3
		ClassID: 5653544353335867687A2063616E6F70
	VST3 plugin #1: None
	VST3 plugin #2: None
	VST3 plugin #3: None

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: \diskstation\Media\Lossless Audio
Library database path: C:\Users\stefan\AppData\Local\Packages\Audirvana.Audirvana-4118-9684-d80dbb7827cd_q3nymrkmej12j\LocalCache\Local\Audirvana\Audirvana\AudirvanaDatabase.sqlite

Local audio files fingerprinting
Tracks with no MBID: 318

Remote Control server:
Listening on 2a02:1210:50e0:8500:9425:6039:XXXX:XXXX on port 51012

ACTIVE STREAMING SERVICES
TIDAL: Connected as PREMIUM

APPEARANCE SETTINGS:
UI theme: light
Font size: large
Language: System language
Show album covers in tracks list: yes
Source list sorted:
My Music
Radios
Podcasts
Streaming
Local
Startup view: My Music: Albums
Show local extended in source list: no
Use media keys: yes
Use media keys for volume control: yes
Use legacy Bonjour protocol: no
Number of paired remotes: 1
Remote pairing code required: yes
Screen saver disabled: no

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

Active method: Local

Max. memory for audio buffers: 640MB

Local Audio Engine: ASIO 2
Driver version 1
Use max I/O buffer size: ON

Preferred device:
ASIO MADIface USB
Model UID:ASIO MADIface USB
UID:ASIO MADIface USB

Currently playing in Integer Mode:
Device: 2ch 32bits Integer, 8 bytes per frame 44.1kHz

Active Sample Rate: 44.1kHz

Bridge settings:
Sample rate limitation: none
Sample rate switching latency: 1s
Limit bitdepth to 24bit: OFF
Mute during sample rate change: ON

Selected device:ASIO MADIface USB
Manufacturer:
Model name: ASIO MADIface USB
Model UID: ASIO MADIface USB
UID: ASIO MADIface USB

13 available sample rates up to 11289600Hz
44100
48000
88200
96000
176400
192000
352800
384000
705600
768000
2822400
5644800
11289600

Volume Control
Physical: No
Virtual: No
Max volume alert: Enabled

MQA capability
Auto-detect MQA devices: Yes
Not automatically detected, user set to not MQA

DSD capability
Raw DSD (msb)

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

Audio channels in use
Number of channels: 2
Use as stereo device only: No
Simple stereo device: Yes

1 output streams:
Number of active channels: 2, in 1 stream(s)
Channel #0 :Stream 0 channel 0
Channel #1 :Stream 0 channel 1
2 ch Integer PCM 32bit little endian 44.1kHz
2 ch Integer PCM 32bit little endian 48kHz
2 ch Integer PCM 32bit little endian 88.2kHz
2 ch Integer PCM 32bit little endian 96kHz
2 ch Integer PCM 32bit little endian 176.4kHz
2 ch Integer PCM 32bit little endian 192kHz
2 ch Integer PCM 32bit little endian 352.8kHz
2 ch Integer PCM 32bit little endian 384kHz
2 ch Integer PCM 32bit little endian 705.6kHz
2 ch Integer PCM 32bit little endian 768kHz
2 ch DSD 8bit big endian in 8bit chunk 2822.4kHz
2 ch DSD 8bit big endian in 8bit chunk 5644.8kHz
2 ch DSD 8bit big endian in 8bit chunk 11289.6kHz

Local devices found : 1
Device #0: ASIO MADIface USB
Manufacturer:
Model UID: ASIO MADIface USB
UID: ASIO MADIface USB
Model name: ASIO MADIface USB

UPnP

UPnP devices found : 0

Chromecast

Chromecast devices found : 1
Device #0: XXXXX
ID: DnsSd#XXXXX-0b45f06e687cfc5f85c91177xxxxxxxx._googlecast._tcp.local#0
Model name: xxxxxx

So… the MADIface USB supports 24bit PCM at a maximum sample-rate of 192kHz… You will need to adjust your ‘bridge’ settings to reflect the capabilities of the MADIface device… I see you have your latency set to 1s… Have you tried longer latency times?

Something is not correct with this debug report that is showing the MADIface USB as the ‘Preferred Device’ and capable of PCM sample-rates to 32bit 768kHz and DSD to 11.2MHz, which seems to reflect the capabilities of the ADI-2 Pro DAC.

What happens when you connect the ADI-2 Pro DAC directly to your computer via USB without the MADIface USB?

Are you running TotalMix?
:notes: :eye: :headphones: :eye: :notes:

This is ridiculously low…

The ADI-2 Pro FS is just using the MADIface driver, but there’s no actual MADIface hardware involved here. That’s why the report is probably a bit confusing. No TotalMix either.

Oh ok. I was just using default settings. The system has 32 GB of RAM so I could easily increase this - what would be a reasonable value?

Try 15000

So I’ve tried first with around 4 GB and then with around 16 GB, with no effect on the problem unfortunately (but also no negative effects otherwise, so I will probably keep it at a higher value)

If you can operate the RME DAC without the MADIface driver and use a generic ASIO driver, what happens?

*A playback pre-load memory buffer of 8GB will suit your needs well…

If you are switching from DSD to PCM you may occasionally get zero-crossing ‘pops’ when the DAC switches formats…

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

With generic ASIO driver, do you mean something like ASIO4ALL? I just tried that, but to no avail.

However, I found one setting in the plugin I’m using (CanOpener) which does seem to make a difference, it’s the HQ setting

CanOpener Studio by Goodhertz

From what I can tell, this introduces some additional latency of 512 samples (not sure if one mode is using linear phase filters and the other minimum phase).

The problem only appears with HQ mode on. Or it is so miniscule with HQ mode off that I didn’t hear it so far. Also those 512 samples could roughly be the length of the artifact.

Thinking about that, I think the other plugin where I experience those artificats was also a linear phase EQ. I’ll see if I can do some further investigation and reproduce this.

Thanks in any case for any input so far!

PS: Only tried PCM files so far, no DSD

No DSP possible on 1-bit PDM (DSD) files… Only when they are converted to DXD typically for digital volume control… If the RME box is using the ESS DAC chipsets… all signals (PCM and DSD) are converted to multi-bit PCM for output.

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

1 Like

The ADI-2 Pro FS R I have is using AKM AK4493 I believe. But as far as I know, the same applies to those, it’s also a delta-sigma multibit DAC.

In the case of DSD, the AKM chipset has a pure 1-bit signal path to a simple low-pass filtered output of both native 1-bit PDM and PDM signals extracted from the DoP carrier, if so employed in the RME topology (most likely employed)… However, “no-can-do” DSP on 1-bit DSD files… Because of this fact, the plug-in module is bypassed when playing DSD files. :wink:

*Note: You still need to manage your signal gain structure from the input of the plug-in, through to the plug-in output and to the next plug-in architecture I/O and/or to the player audio-engine I/O architectures.

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

1 Like

Some signal processing algorithms (read software) perform better or worse depending on the hardware platform (read CPU architecture and I/O bandwidth) it is running on…

A doctor once told me: “if it hurts when you do that, don’t do that”…

This may be a case where this plugin may run fine on a high scale DAW with the HQ setting used for rendering audio output for mixing/mastering and not so much for Audirvana in a more or less realtime streaming context…

My experience is that I have some plugins that will run in HQ/Pristine linear phase oversampling mode just fine on a water-cooled de-lid 5GHz windows DAW… And cause thermal speed step down on a MacBook Pro with Audrivana…

1 Like

I agree that the platform architecture/CPU/Memory bandwidth topology and System RAM will have implications on the integrity of the playback performance… However, I am running 112dB’s ‘Redline Monitor’ with 24/352kHz AIFF files, with no artifacts or playback anomalies as described by the OP… I have run Waves “Abbey Road Studio 3” virtualization without these anomalies as described… Up until very recently, these plug-ins were running on my 2016 Touch-bar MacBook Pro, quad-core i7, 2.7GHz with 16GB of System RAM… Most recently things are quite a different story, where my new platform is a M2 Max, Mac Studio with 64GB of RAM, however, the only aspect that I find improved, is with the track-to-track ‘select and play’ experience, which seem flawless, with about 15GB allocated for playback pre-load playback memory, however I am going to drop this down to 8GB, which seems more-than-adequate for DSD128 files and 352.8kHz DXD files being modulated after the Redline Monitor plug-in DSP, to DSD128 via r8Brain (No DSP on binaural DSD128)… I think my SSD on the USB 3.1 bus is probably the biggest bottleneck for preloading these files…

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

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