Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Signal above digital full scale (Read 4106 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Signal above digital full scale

Here is a picture of a signal where the true "analogue" peaks lie between samples...

Cheers,
David.

Signal above digital full scale

Reply #1 –
Yes, and if you use an application such as Adobe Audition to increase the sampling rate there will be samples above 0dB in a 32 bit file, or distorted peaks in a 16 bit file. I actually dealt with this in some detail in the old Syntrillium forum a year or two ago.

This has implications for the designers of CD players. One hopes they are aware the effect and allow the necessary headroom. What about oversampling players? Do they actually create the necessary samples above 0dB?

Cheers,
Alan
Cheers,
Alan

Signal above digital full scale

Reply #2 –
Quote
Here is a picture of a signal where the true "analogue" peaks lie between samples...

Nice picture. Is this in reference to a particular thread or query? If so, could you post a link.
Quote
This has implications for the designers of CD players. One hopes they are aware the effect and allow the necessary headroom. What about oversampling players? Do they actually create the necessary samples above 0dB?

I would certainly hope so. If the engineer knew what they were doing then they should be aware that there can be peaks in the analogue signal higher than digital 0dB.
Most consumer CD players seem to lowpass at 20kHz. As this effect drops off extremely quickly as you move away from the Nyquist frequency it shouldn't make a big difference to the analogue stage.

Signal above digital full scale

Reply #4 –
Quote
btw, the effect doesn't have to reduce as you go much below nyquist - the image I posted shows a sine wave at half nyquist, i.e. fs/4.

Cheers,
David.
[a href="index.php?act=findpost&pid=227791"][{POST_SNAPBACK}][/a]

Oh, right, I suppose it doesn't. It does drop off very quickly after the nyquist limit, but then goes up again near (Fs/2^n). I have attached a graph showing exactly what effect this has, if anybody is interested. What the graph shows is the ratio of max sample to signal peak versus sampling frequency. The wave is a 1024Hz sinusoid and the Nyquist limit is at 512Hz.