1

I am listening to Spotify/MP3 on my laptop with external sound card and headphones.

Is there something like dedicated track that will help me evaluate the level of internal noise of my system? I found a track called digital silence but the system probably didn't output anything at all because I could not hear any noise, no matter the volume, but with some silence inside other tracks I was able to hear the noise difference between the laptop's internal sound card and the external one.

EDIT:

There is this song on Spotify named Digital Silence but when I play it on max volume from my laptop (over headphones) I can't hear ANYTHING. While in case of other songs on max volume I think I can hear noise in them that gets improved when I use external DAC. My assumption is that the laptop will not output any signal at all in case of the Digital Silence track.

migrated from avp.stackexchange.com Jan 27 '14 at 15:05

This question came from our site for engineers, producers, editors, and enthusiasts spanning the fields of video, and media creation.

2

It may be that your sound card has a 'gate', which practically disconnects it's circuit when not in use (that is to say, a digital silence - meaning a stream of 0s will not activate your sound card, thus it may appear silent because of this).

If you want to get technical you could use a sine wave generator (I'm sure you can find one for Audacity) and either use your ears. Do you hear any noise? The more scientific way would be to connect an oscilloscope if you happen to have one, and take a look at the sine wave itself, and preferably a frequency response charting of the signal. A sine wave should have one clear peak. If you see other peaks, that means you have harmonic distortion.

The bit resolution of your source and sound card will also affect the noise level you're going to experience. You get -6dB of noise per bit you add.

16 bits = 16 * -6 = -96dB 24 bits = 24 * -6 = -144dB This simply mean that any audio below these levels will not be audible, because the noise that is inherent because of the sample width is louder.

What I'm getting at is this: Spotify is not a reliable audio source for determining the system signal to noise ration as the tracks you get off Spotify has noise introduced because of sample rate and sample width.

  • 1
    At least with proper noise-shaped dither you can actually discern signals even well below that quantisation-noise level. – leftaroundabout Dec 12 '13 at 15:22
  • I am not sure I understand your math with bits - I thought the more bits the better sound. Also the Spotify streaming should not be a bottleneck, at least according to this answer: avp.stackexchange.com/questions/9586/… – daniel.sedlacek Dec 13 '13 at 11:01
  • The math is explained here: en.wikipedia.org/wiki/Audio_bit_depth and I'm not saying spotify is a bottle neck, simply that I would not use Spotify as my audio source if I was trying to figure out the minimum noise in my system. – AlexanderBrevig Dec 13 '13 at 12:32
2

We need to disambiguate a couple of things here.

"Internal noise level" of a soundcard is an analogue phenomenon, coming from the DACs and the headphone amp. Unless you have an analogue volume pot on your external sound card, this noise level will be completely independent of what you're doing. Certainly independent of software, so you can as well measure it with no file playing at all!

As a matter of fact, that level may well be so low it's simply inaudible. That's certainly no big deal to achieve in a professional audio interface, after all we don't need to amplify any particularly low signal on a high-impedance channel (the only reason noise remains an issue for microphone preamps).

Sure, laptop sound cards are notoriously noisy, but that mainly is a problem in the inputs and when you connect to other devices with a ground difference that makes power-supply bursts audible.

If you do hear a sound difference between internal and external sound card, but only between certain tracks (and be aware that this may always be just a placebo effect, or due to different total volume levels!) then it's probably not the internal noise but some specific kind of artifact. In particular, if the DAC is not designed well it will cause distortions on the whole dynamic range that is particularly obvious on low-level, high-frequency sounds. So if there is some noise on the recording itself, the DAC might make this worse and it sounds like it's coming from the sound card itself. You might say this is extra noise from the sound card itself, but I'd rather keep to saying it's distortion of the file's audio signal.

  • Thanks for the explanation. What I was refereing to is that when I max the volume and connect headphones direclty to laptop I can hear lot of high level "noise" in a quite part of one particular track, but when I do the same via the FiiO E7 there is clearly less noise, while the heights are not cut off nor suppressed so I suppose the external DAC does a better job eliminating the "noise". – daniel.sedlacek Dec 13 '13 at 14:05
1

It would be simple to create a track that consisted of digital samples of zero. Using Audacity, start a new session, add a stereo track at the sample rate and depth you desire, then select Silence for the duration you want. Save as an uncompressed file type (WAV, for example) and listen.

You can of course save as any type you like, to see if the codec has any effect, and vary the sample rate etc to experiment.

  • There is this song on Spotify named Digital Silence (play.spotify.com/album/0ukIPSztyTel8F3gtMbv2d) but when I play it on max volume from my laptop (over headphones) I can't hear ANYTHING. While in case of other songs on max volume I think I can hear noise in them that gets improved when I use external DAC. My assumption is that the laptop will not output any signal at all in case of the Digital Silence track. – daniel.sedlacek Dec 11 '13 at 10:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.