Songs cut off before end

Software cuts songs off about 8-10 seconds before the song is finished.

Provide more informations, such as your OS, your DAC, if you’re using USB, UPnP, if you’re playing local tracks or streaming, if you’ve enabled upsampling etc.

Why all these? I’m tired of all these discussions. Since my beginning, in March with A 3.5 and now with AS I had this problem (USB Dac or Realtek). If within a year (I subscribed to AS for a year) this bug will not be fixed I will say bye bye Audirvana forever. I already canceled my subscription. However, for everybody with this problem I suggest to use WASAPI, large WASAPI buffer, no upsampling and very low buffer (I use 512 MB right now).

I’ve been using WASAPI since day 1. Using the default settings of AS, and I’ve never had this issue. Not with AS or with A3.5. (Windows 10 Home).

In the past I’ve used a Terratec Aureon 7.1 USB soundcard hooked up analog. Nowadays I’ve hooked up a Audiotrak Prodigy Cube Black Edition USB soundcard digitally through toslink.

Same thing with macOS. Streams without issues through Bluetooth to a set of monitor speakers.

Because it’s important to have informations to understand where the problem can be. You can’t expect a fix if you don’t provide any infos.
For me, it happened in the past on macOS using Audirvana 3.5 when the software wasn’t rebooted since a week ; rebooting fixed things. Some old topics show, for example, that disabling upsampling can fix the issue for some users. And of course, USB, UPnP, macOS, Windows are all very different configurations. WASAPI ? No idea what it is.

1 Like

In Windows you have 3 options in AS: WASAPI, ASIO, Kernel Streaming. I agree with you but I didn’t see a professional answer like: to avoid this problem you have to do this (from Damien or other developers). Anyway, with my settings the problem is less frequently, especially with streaming. I know there are users without this problem but I know also that there are many users with good equipement that have this problem only with AS or A.

1 Like

What makes it difficult is the wide application of hardware, drivers and conditions.

From a hardware manufacturer with roots in pro audio the ASIO drivers will give the most stability.
Older DACs from a niche audiophile market will work well with KernelStreaming.
And for almost all newer DACs, WASAPI will work fine.

But this is not a fact and Audirvana does not have the capacity to test this. It is an opportunity as a community to create an overview with which OS, which DAC and which driver you achieve the best result.

But this also does not give any guarantee. An OS update or installing a tool like fidelizer can cause problems again.

Apart from that, Audirvana has some issues that I haven’t seen with other software.

1 Like

And I saw on this forum that Mac version is better. Damien loves Macs, raised with Macs and Audirvana was at origins a Mac app. Some fanatics will say: “of course, Macs are better”. Maybe. But a good PC with Windows is also stable and good, I use Windows since 1999 and I know what I say. The main problem is with AS. And also I agree with @Jacob.

1 Like

To me, Audirvana is a music player. Tag management and searching for information can also be done with other tools.

The core should be rockstable. If Damien has a certain brand and type of PC and DAC where he guarantees stability, I would like to use it.

When I turn on the music, I want to relax for a while. Don’t annoy me with hitches or skipping tracks. Also on Mac Audirvana has its challenges.


This was true in the past.
AS has many problems on Mac too.
And A3.5 is slower on a Mac than a snail. If you load to it a lot of music, you see the beach ball all the time.

1 Like

I also saw this “load of” time (Windows), but I thought was my slow CPU fault. But if I remember well the CPU didn’t have high usage, A 3.5 was less responsive but other apps worked well in parallel.

So, I have a Mac Mini running Audirvana. The signal is sent to a Audiolab MDac plus. I bought this software because I AM NOT A TECH! My relationship is with the music, not the software version, not the hardware version or operating system.

I was running Jriver prior to this and it is wonderful and stable. I only swi to Audirvana because I was told by a friend that the SQ was better… But at what cost?

The most common reason for this is upsampling, just disable it a check again.

To put it into perspective.
There are also plenty of satisfied users. And I can also listen to music without any hickup much more often than there are problems.

Sometimes it takes a while to find the right stable setup.
But it remains a great Musicplayer that sounds very good.


I totally understand that development takes time and money but, this is a revision of an existing product and therefore some of these issues should have been foreseen.

The other iss that I have not brought up is skipping during playback - like 40 years ago with a scratched record. Last night I had friends over to listen to music and they were impressed with my system and the SQ but were confused about paying for software that skips and cuts off music.


That’s the trade off you pay for the sound quality. It interacts more directly with the hardware and you can have issues with some devices.

You should be able to fix this by disabling upsampling. If that still doesn’t help, reduce the pre-load cache size.

1 Like

I agree but it’s not ok. Because usually with upsampling the SQ is better. This software must do a right job with all the features activated with most DACs.


I agree, I spent good money on a Dac with capabilities and I do not want to have it limited by the software I am paying for.

I am a patient guy, but these issues need to be resolved if Audirvana has the expectation to be the industry leader.


But with a small buffer and without upsampling, does it work without any problems?

1 Like

That’s irrelevant. Most of us are probably using Delta-Sigma DACs. These chips only decode DSD natively. It makes more sense to up-sample to DSD for this type of DAC to transfer the burden of conversion from PCM to the computer’s CPU. If the feature is there, it should work. Period.