Home Vinyl Asylum

Welcome Licorice Pizza (LP) lovers! Setup guides and Vinyl FAQ.

That signal does not exist. This is a strawman argument.

Instead of explaining the math here, I will simply link to a helpful image posted by a HydrogenAudio regular. It is a picture of an oscilloscope trace of a 14khz sine wave, played back at 44.1khz sampling rate, on a PC motherboard's AC97 sound output. I will also link to the original thread this came from. (Thanks uart!)

This image shows that the operation of digital-to-analog converters, as shown by the chart you link, simply does not occur with even the cheapest DACs available today. As long as you trust how this reading was taken - and I assure you, it should take very little faith - then sample & hold doesn't exist, and that chart is a lie.

Please follow the link and compare it to the samples that are actually stored on disk. The look quite a bit like the image you're posting. There are a couple differences - Audacity uses linear interpolation to draw lines between samples, instead of sample&hold as what you're describing - and the frequencies are a bit different. But the fundamental similarity is the same. What you see on a computer monitor is not what actually happens on the wire.

Of course this chart does match reality for NOS DACs with no output filters, but those are defective to begin with. Razz them out all you want, and I'll agree with you. They have no bearing on how DACs are supposed to be designed, nor on what even the cheapest DACs actually do.

If you're willing to keep supporting this argument (that this chart has some basis in reality), you should actually provide some physical evidence that DAC outputs behave like this. As in, wire up a scope, and plot it. Otherwise, you are sprouting nothing but hot air cooked up by marketing types and non-engineers.


This post is made possible by the generous support of people like you and our sponsors:
  Sonic Craft  


Follow Ups Full Thread
Follow Ups


You can not post to an archived thread.