Upsampling

[Irri]

I was just wondering if there have been any experiments at Upsampling 44.1KHz audio to a higher frequency. Maybe this is already done by a DS.

If we have an image sequence, say 1 second of 24fps footage, we can make that fit in a 2 second time period in a number of ways:

We could repeat the previous frame
A A B B C C D D
We could blend the frames:
A AB B BC C CD D DE
Or we can interpolate the difference between two frames to make a new frame:
A ABx B BCx C CDx D DEx

My examples are used to retrospectively slow an image sequence, but could be used to increase the frequency rather than time.

Would anything like that work with audio?

[grahamd Linn Staff]

Nearly all digital audio devices upsample, including the DS. It is part of the process of recreating an analogue waveform that is as close as possible to the original.

Graham

[Irri]

Thanks.

What process is used to fill in the gaps, or is that "secret sauce"?

[grahamd Linn Staff]

Well, our up-sampling is better than that normally found in digital audio hardware, and I don't think we will be giving the exact details away, but the broad principles are explained in Discrete-Time Signal Processing by Oppenheim and Schafer.

Graham

[Irri]

Thanks.

[Love_The_Music]

The book at 1124 pages is a bit too up-sampled for me unless I dust off my maths degree and turn the music off and concentrate!