No output sound when using 192kHz

After solving the issue with being stuck to 48kHz, I can successfully configure JACK and Ardour to the sample rate that I want. Specifically when using 192kHz though, there is no output sound.

Note, that I’m more than happy with lower sample rates, 96kHz is more than good enough. No need to get into a conversation on why do you need it. I can be quite technical and I want to understand why it does not work, know that I have set up my workstation without issues, test things for myself, and have it working for the case that I will need it.

Here is my testing:

  • On fresh 192kHz session with just one track for testing
  • I can record successfully, see the correct waveform in Ardour
  • When I play the recorded track, there is no sound
  • In Mix view, everything seems ok
  • I open JACK mixer and route the sound from Ardour’s master out to JACK mixer input
  • JACK mixer visually shows audio is received

So, I though it may not be Ardour, but:

  • I switch to another session with the only difference being the sample rate being other than 192kHz e.g. 96kHz
  • Playback works as expected
  • I export the 96kHz file to FLAC, can open and play successfully with Audacity
  • I switch back to the 192kHz and export the file to FLAC
  • I open the file in Audacity and it is empty i.e. flat line waveform

I use Focusrite Scarlett 4i4 that supports 192kHz and successfully records on 192kHz, as mentioned above. Throughout all tests there’s no changes done that relate to the sound card.

Admittedly no answer for your direct question… But there is really no basis for using either 96 or 192kHz in real world situations with human hearing…

Yes I know, and thank you for your response. I just want to know what is wrong. Also, if it’s something wrong with Ardour, report it so that it gets fixed.

1 Like

Do you have a way to verify that it actually successfully recorded at 192kHz?

How are you starting jackd? There are times that jackd will not actually open the device at the requested rate, but will pick the closest rate it can get to work. Capturing the output startup messages from jackd will help confirm that the jackd server really starts at 192k as requested.

Have you tried the ALSA backend in Ardour rather than JACK backend?

Guitar Modeling plugins sound more realistic when given a 96khz input. When used at 44.1kHz or 48kHz, they tend to sound a little ‘wooly’ for lack of a better term. When 96kHz is used, they sound more like an amp in a room, and there is a nice ‘clarity and definition’ that comes through. I am guessing that it is the inaudible harmonics of the guitar being captured that add to the audible range after going through the sim. I tried 96kHz vs 192kHz, and the difference was negligible, but audible; not worth the extra CPU needed. Recording at 44.1kHz with oversampling in the plug-in does not help 44.1kHz sim amps sound like 96kHz input.

Guitars through amp sims sound GREAT at 96kHz, otherwise, they sound like my Boss GT-8 :frowning:

If you use amp modeling software, definitely give it a try! Let me know if you do.

EDIT: It has been my experience that Amp modelers cannot replace a good mic’d amp. They seem to fail in the dynamics department. Also, running a REAL distortion/overdrive pedal into the audio interface adds a nice touch to help boost the dynamics and touch of the simulated amp sims.

Some plugins do upsample internally for various technical reasons but there is no reason to sample a guitar with more than 48kHz, even assuming you have transducers that can cleanly reproduce high frequency signals.

An electric guitar does not even produce frequencies above 10kHz, not to mention those classic 4x12" speaker cabinets at best reproduce signals up to 5kHz (at 10kHz the signal is already attenuated by more than -60dB).

It may well be that things sound different when you use higher sample-rate, but what you hear are likely artifacts caused by cheap transducers that won’t be present with proper high-end equipment.
It could also be placebo effect. There’s no shame in that. It is easy to fool the human brain.

1 Like

Have you compared the 44.1kHz to the 96kHz input in an amp sim with high end equipment and found no difference ?

It is not placebo effect. It is likely my hardware.

FYI, I am using a Native Instruments Audio Kontrol 1. Native Instruments Audio Kontrol 1

I can agree guitars don’t produce freq over 10kHz. I was trying to rationalize what is going on. So that means there are no upper harmonics being adding to the signal. Ok.

It works this way every audio interface I have ever had.

All the guitars are made in USA or JAPAN. Mogami cables.

Not only my audio interface, but outboard guitar modeling gear that samples at 44.1kHz has the same ‘wooly’ tone. 96kHz outboard modeling gear sounds similar to my amp sims ran in 96kHz. I will never buy digital modeling guitar gear that uses a 44.1kHz sample rate. Maybe they have improved in 10 years, but at over $600, the sound did not justify the price to me.

The concept is also talked about in one of these books : Best Selling Books - Bobby Owsinski

Would there be a newer USB audio interface around $300 on the market that you would highly suggest ?

If you play guitar and have some high end gear, give it a go some time and tell me what you think. 44.1kHz input vs 96kHz input amp sim :slight_smile:

Do you have a way to verify that it actually successfully recorded at 192kHz?

I can see the recorded waveform in the Edit view, and it follows the expected volume envelope and rhythmic pattern.

How are you starting jackd? There are times that jackd will not actually open the device at the requested rate, but will pick the closest rate it can get to work. Capturing the output startup messages from jackd will help confirm that the jackd server really starts at 192k as requested.

I start the jackd server via qJackCtl which displays in it UI the sample rate and it correctly displays 192kHz. I then open Ardour, which also correctly displays the session rate as 192kHz.

Have you tried the ALSA backend in Ardour rather than JACK backend?

I will try that and see if it works, report back here.,_executive_summary

I had to customize this file to get my Audio Kontrol 1 to use 96kHz sample rate.

It may be possible that ALSA is not currently configured for 192kHz. I dont know for certain. It may be something to check into.

ALSA needs to be configured properly for JACk to work.

I currently have problems with the JACK backend, but I was able to record and playback successfully through a Scarlett Solo at 192k using the ALSA backend.

FYI ALSA MIDI works better than JACK MIDI. My experience is that JACK MIDI has a much higher latency.

My copy of Ardour 7.5 was crashing when trying to use JACK at 192k, but a recent git snapshot records and plays back with no problems using the JACK backend on a Scarlett Solo running at 192k.

This is incorrect. And in fact, if you use JACK for app-to-app MIDI routing, it has zero latency.

1 Like

Thank you for the correction. If JACK is not the cause, I would guess it was a2jmidid. Would a2jmidid cause about 40 to 100 ms of MIDI latency?

There are older versions of a2jmidid that suffer from jitter, not latency (that is, they have variable amounts of latency, not excessively high latency).

I do not believe that the current versions suffer from this. I use JACK1 (because I helped write it) which has a builtin version of a2jmidid, and that adds zero latency with zero jitter.

I managed to export audio file from Ardour in 192kHz that plays back in VLC/Audacity when JACK is stopped; I don’t know why that time I got a flat line waveform. Moreover, I tried again to put a couple of plugins and connected them to Ardour master output e.g. jaaa analyzer and JACK mixer and they both receive audio.

I don’t think it’s Ardour, it seems like JACK. I’ll look for support there.

Both JACK and Ardour take over full control of the chosen audio interface while they run.

Non-JACK applications cannot play audio via the same interface while JACK is running.

No other applications can play audio via the same interface while Ardour is running.

The situation is slightly different with PipeWire,

1 Like

Do a blind test and report back when you’ve successfully made a 10/10 score.

This forum really should have the same rule as the HydrogenAudio one : All members that put forth a statement concerning subjective sound quality, must - to the best of their ability - provide objective support for their claims. Acceptable means of support are double blind listening tests (ABX or ABC/HR) demonstrating that the member can discern a difference perceptually, together with a test sample to allow others to reproduce their findings.


I’ve tried with Ubuntu Studio on live USB, same.
I’ve tried with AVLinux-MXE on live USB that does not use PipeWire, same.

All other sample rates work fine. At this point I give up, I don’t need such high sample rate anyway.

This topic was automatically closed 91 days after the last reply. New replies are no longer allowed.