Yes but those other locations could still have been using wireless mice and keyboard is the catch. So I agree it is a worthwhile test.
Seablade
How do you define well above this? As mentioned above, when I listen to it it sure sounds like it is close to the noise floor. Of course part of the problem is that the recording I am finding is from Vocaroo which I am not familiar with but seems to be a web based recording solution, which means it likely compresses audio losslessly before transmission.
It would be good to get a lossless file (WAV, FLAC, etc.) demonstrating this out of Ardour or Mixbus along with typical recordings etc. so we can know for certain where the noise floor sits in this, and how it compared to a typical recording.
Seablade
1 Like
I guess we would need a description of what processing was done when exporting that audio file (for example abnormally high gain), but the noise when played from that vocaroo.com link was nowhere near -93 dB FS. I currently have the system gain on my computer set for comfortable background listening, meaning commercial recordings are probably at around 70 dB SPL give or take, and when playing that file from vocaroo I can hear the noise easily without having to turn up the volume. At this system volume level setting I cannot hear the system noise floor at all, and it is just a 16 bit device. Unless you are in a pretty extreme environment (e.g. high quality studio) you would usually need a very large gain to hear the quantization noise floor of even a 16 bit system.
I think you are misunderstanding ‘noise floor’ as I am referring to it. I am not referring to the theoretical electronic noise floor, but rather the noise floor of the recording. I can hear significant noise that is close to the levels of the beeps in that recording and not in fantastic environments. I suspect there is a significant gain happening here, thus my requests above.
Seablade
I was responding specifically to John’s comment (which I quoted in my post) about whether dither could help (and if the denormal handling setting was related to dither).
In that specific context I think it is obvious (unless something very weird was done) that the interference is well above the quantization noise floor, which is where dither settings are relevant.
@Seablade @Chris the reason dither came to mind was because each time I listen to deperito’s sample it never sounds quite the same. This made me wonder if the ‘beep’ had in fact been some kind of psychological effect - i.e. the brain trying to make sense of what it was hearing. But in fact your theory is much more likely - i.e. the signal coming from vocaroo.com isn’t lossless (so there’s no guarantee we’re even hearing the same thing!!) These are all things that deperito will need to think about if he’s back here next month.
My favourite theory (so far) is that deperito’s guitar lead’s picking up a wireless signal from somewhere - but for me at least, the jury’s still out on whether it’s a remote signal or something much more local.
Yes. I download the file and put into Ardour and the noise floor is really high. Just looking at the waveform shows it’s way above a normal noise floor:
Running it through LSP Spectrum Analyser shows the peaks at around -47dB and a strong “colour” to the noise, which seems to be centred at around 3 kHz.
The “beep” seems to be at around 5400 Hz and is actually lower than the noise floor, by around 10dB.
As stated, we really need a raw, unprocessed version of this file, preferably a wav file, to understand it fully.
Given the strong corellation around 3kHz, it may be possible to deal with it with some careful EQ, although finding and fixing the root-cause would be better.
Cheers,
Keith
Se also the time analysis above: Weird beep when using audio interface - #8 by x42 There is a fainter periodic signal (around 26 impulses per second).
1 Like
I think you are misunderstanding ‘noise floor’ as I am referring to it. I am not referring to the theoretical electronic noise floor, but rather the noise floor of the recording. I can hear significant noise that is close to the levels of the beeps in that recording and not in fantastic environments. I suspect there is a significant gain happening here, thus my requests above.
I think it could be deeply related to gain actually, since - as I said in my previous answers - I just cranked up my volume to be able to hear the beep more clearly and assess its presence or absence.
As soon as I’ll be able to use my home studio again I will get you a recording of me playing the guitar at the gain I usually play when not trying to see if the beep is still there, I can assure you that even in that case the beep is rather clearly audible even when playing some phrases on top. As I was saying, I noticed that the beep becomes clearer as you introduce harder distortions/increase the gain, in fact I noticed it the very first time while using a metal distortion.
It would be good to get a lossless file (WAV, FLAC, etc.) demonstrating this out of Ardour or Mixbus along with typical recordings etc. so we can know for certain where the noise floor sits in this, and how it compared to a typical recording.
In order to bring you the beep I just recorded it using Ardour while I was playing guitar using Guitarix as effect processor (I/O schema: Guitar → Audio Interface → Guitarix → Ardour) without noise gate to be able to hear it better, exported it (but I just can’t recall the audio format… Sorry) and then uploaded it to vocaroo.com just because was the simplest service for audio file hosting I found.
Of course I can export for you a WAV file of me playing the guitar as I usually do, but could you suggest me how can I get it to you all? WeTransfer maybe?
I was doing a research on how to get the DI BOXes you suggested and it seems that Amazon doesn’t ship many of them to Italy… May I ask you if there’s a parameter in the specifications which I can also check in other devices in order to get a good DI BOX that is suitable for my problem?
Thanks again
1 Like
Do you have Google Drive, or Dropbox, or similar?
Also, where are you applying the metal distortion? If it’s getting amplified with the distortion, it sounds like it’s before whatever that device is (guitar amp?).
Cheers,
Keith
Yes, I have Google Drive for instance.
The distortion is applied at Guitarix level, following the signal chain I wrote before: Guitar Input (jack cable connected to audio interface) → Guitarix application for distortion → Ardour
Then I simply listen to Ardour’s output signal from my speakers or headphones
Ah right, so Guitarix is providing the distortion.
Any amplification (including gain applied to guitars) will amplify the noise as well as the guitar sound.
For the audio file, you could put the raw file into Google Drive and share a link to it using the sharing feature on Google Drive.
Cheers,
Keith
Hi there, just replying in the topic so it doesn’t close.
I’ll be back to my studio this weekend, make the tests you suggested and let you know.
Cheers
Ok, here I am with some updates.
First of all, @John_E unluckily removing batteries from my didn’t have any effect and the beep is still there…
I also did some recordings (regions exported by Ardour in WAV format) while playing at a “normal” gain. You can hear the signal is humming, in fact - since I often use strong distortions - I usually use some Guitarix noise gate to shut the noise down when not playing, anyway I disabled it for these recordings in order to be able to hear the beep more clearly.
As you can hear in the recording the beep is still there but - as expected - it’s VERY quieter. Anyway, even though when playing it’s well covered by the guitar signal and practically inaudible, it’s still there and can still be heard when guitar signal is quieter (e.g. when I’m not playing).
I did two recordings: a shorter one while I’m not playing and the beep can be heard (just listen carefully at 0:04 approximately) and a (very 
) longer one of me going through a bunch of exercises. I tried to listen carefully to the last one but didn’t manage to hear the beep, as I said it’s probably covered by the guitar signal.
-
This is the recording of me NOT playing in which you can hear the (very quiet) beep (second 4 approximately). The maximum digital peak of the track is -28.3 dBFS and the maximum digital peak of the Master track is -14.2 dBFS (K14/RMS meter)
-
This is the longer recording of me going through the exercises. The maximum digital peak of the track is -14.3 dBFS and the maximum digital peak of the Master track is -0.2 dBFS (K14/RMS meter)
Please let me know what if you think and if you have other useful advices
For the recording where you’re NOT playing, I hear a weird sound at about 1 sec and the same sound repeated (but quieter) at about 4 secs. No idea what it is though… 
I’m sorry, at about 1 sec I lightly brush against the strings so that’s what you’re hearing 
… I forgot to mention, sorry
So it is a somewhat tough one to an extent.
Given that the beep is about 5.4k at -80dBFS or so… I don’t think I would spend much time worrying about it personally. It would be relatively easy to have a sharp notch drop it another 20dB and you would never hear it most likely. As it is that low you are unlikely to ever hear it compared to useful signal as it is.
As to the exact cause, that is a great question. I honestly don’t feel like this is an external cause, but rather internal electronics, obviously some of what you described above would mean it should be external, but it just doesn’t feel like an external source. Internal would mean something at the electrical circuit level that I don’t know you will have a lot of control over.
Seablade
His very first posting seemed to indicate that it happens at locations close to his apartment but not when recording in a different city. Which leads me to believe it must be an external cause though of course I might be wrong. Weird stuff going on.
I tested the setup also in another apartment within the same building, in an apartment within another building in the same city (just about 50m away from my home) and in an apartment within a building located in another city about 33km away from mine, finding out that I can hear the beep in the other apartment within my same building and in the apartment 50m away from me; while I don’t hear the beep in the apartment in the other city
Oh I understand, but I am also wondering how much could be coincidence is the problem.
At this point I don’t have much to base it off of other than my gut having listened to all sorts of RF noise over my career of all types, but definitely more than possible I am wrong.
Seablade