How does a Digital Volume Control Work?

[boyblue]

I'm not sure where to ask this question, it is related to the interaction of hardware and software, but it also pertains to computer streamed music.

Many programs such as Windows Media Player or similar have their own volume control. But exactly what is the volume control doing?

One possibility is that it is simply digitally controlling the level of the analog output of the sound card. Digital controlling analog.

But what about when you are streaming music via optical, coaxial, or USB to an external DAC. The volume control on the software player still seems to affect the output levels. But since the output is a digital stream, it is not the voltage level of the stream that is changing. It has to be the numbers contained in the stream that have been altered.

So, if the volume control is at 50%, does the software divide all the digital voltage number by TWO?

In other words, during the recording of music, the Analog-to-Digital converter made this sequence of value capturers - 10, 20, 30. So with the volume control at half, those values would be reduced to 5, 10, 15?

I'm not saying that is how it works. I'm asking, how does it work??????

How can you change the volume in a pure digital music stream???

Steve/bluewizard

[Platypus]

Your analysis is correct. Within the digital domain, the amplitude of a waveform is simply defined by numbers. To reduce the amplitude (volume) the values of the numbers in the data stream are reduced proportionally, the reverse to increase volume.

The relationship isn't linear (halving the numbers doesn't halve the volume) because human hearing perceives volume along a logarithmic curve, but the principle is the same.

That means as the amplitude reduces, the digital resolution decreases also, as there is a lower total number of digital values defining the waveform.

Then, as you have observed, once the waveform passes through the DAC, digitally controlled analog volume control is possible, typically using an electronically switched resistive ladder.

[boyblue]

Thanks for replying and confirming my suspicions.

One additional point, though I'm not sure if this can be known.

While the data is pure digital, if you turn the volume control UP, will it amplify the numbers?

In other words, if I'm using an external DAC and Stereo amp, I probably want the data to enter the DAC at standard encoded levels, neither amplified or multiplied, nor attenuated or divided.

If the computer volume controls are set to 100%, can I be sure the data coming out, has the exact value of the data encoded onto the disc or into the file?

CONTEXT:

Occasionally I get questions in audio forums about connecting a computer to a stereo. Using the analog output, I usually tell people to set the volume control on the amp to a known listening level, then bring the computer volume controls up until what they hear is consistent with the stereo's volume control setting.

That many not be best for accuracy of the data, but it keeps input level well with in the normal range of the inputs of the Stereo, and keeps those inputs from being overdriven.

Digital, between the computer and the DAC/Stereo, is different. In this case, it would seem best for the computer to send accurate information. That is, values in the data stream match the values from the source. Then adjust the volume exclusively with the Stereo volume control.

However, what is the setting of the computer volume, and the application volume control, that gives a one-to-one ratio between the data source and the data stream?

Thanks.

Steve/bluewizard

[Capstan]

While the data is pure digital, if you turn the volume control UP, will it amplify the numbers?

In other words, if I'm using an external DAC and Stereo amp, I probably want the data to enter the DAC at standard encoded levels, neither amplified or multiplied, nor attenuated or divided.

If the computer volume controls are set to 100%, can I be sure the data coming out, has the exact value of the data encoded onto the disc or into the file?

Steve/bluewizard

In the digital domain only attenuation is possible. Recordings take benefit of the full dynamic range (so, in case of 16 but the associated numbers could swing between 0 and 65535) so very little headroom is usually left. 'amplification' by multiplying each sample could lead to overflow max allowed value and clipping.

In theory to keep the signal 'untouched' you should left digital level controls to 100% and use the analogue pot of the amplifier to adjust volume, anyway the different processes in the digital chain and the operating system in itself can alter at some extent the signal.
Audiophiles have created a so called 'bit perfect' theory, where you can use dedicated hardware and software to get the maximum quality in digital sound reproduction.
Windows has implemented a WASAPI signal processing solution that is supported by some software (e.g. XBMC) supporting bitperfect compliant sound processing.
Google for bitperfect and I am sure you will found suitable resources on the Internet.

Hope this helps.
Ale.

[photolady]

Capstan, though your post may help others please check dates of threads as this one is three years old.