I'm a home theater moron - What are DAC's?

[scasarano]

Just wondering - what are DAC's?

 

[MDS]

No reason to call yourself a moron. It's a valid question.

DAC = Digital to Analog Converter. Conversely, ADC = Analog to Digital Converter.

Receivers and amps amplify an analog signal, although there are newer receivers/amps that use digital amplification. So incoming digital signals must be converted to analog before amplification. When you play, for example, a CD and the cd/dvd player is connected to the receiver with a digital cable, the receiver must convert the digital signal to analog before it can be amplified and that is the function of the dacs. That's the high level overview.

I can elaborate further if you want, but maybe it would be better if you just ask any further questions you have.

 

[nibhaz]

Digital Analog Converter

To play a digital-audio signal, it must be converted back into analog form. After some error correction, the digital signal is sent to a digital-to-analog converter (DAC). If it's a stereo signal, it is first demultiplexed to separate the right and left channels.



I must have started to write just before MDS posted....

 

[MDS]

I must have started to write just before MDS posted....

It happens to me alot. That's ok, the more replies the better. Everyone has a slightly different way of explaining things.

 

[scasarano]

Wow - sounds really technical to me and I 'm an engineer. The reason I am asking is that the reciever I bought has 192 kHz/24-Bit DAC's which is apparently better than the 502 which has 96 kHz/24-Bit x 6. I was just wondering what this does for me. I'm sure it's something I'll never notice.

 

[nibhaz]

Digital

Check out this article about digital audio. It's a bit intense at times, but being an engineer it should be easy reading for you

 

[scasarano]

Still pretty technical for me (I'm a structural/civil engineer) but thanks.

 

[MDS]

That was a good article. From the conclusions, this is the most important point: To wrap up, oversampling is not used to create or restore lost information. It is used to simplify the analog reconstruction filter.

When you play a 16/44.1 CD, the DACS will upsample the signal to 24/96. Going from 16 bits to 24 bits is irrelevant but increasing the sample rate pushes the Nyquist frequency up to 48 kHz so that the reconstruction filter can be much gentler and any distortion products will now be much higher than our ears can hear.

 

[avnetguy]

Going from 16 bits to 24 bits is irrelevant but increasing the sample rate pushes the Nyquist frequency up to 48 kHz so that the reconstruction filter can be much gentler and any distortion products will now be much higher than our ears can hear.

Wouldn't the Nyquist frequency only be important on the original ADC conversion? Just thinking this as the upsampling interpolation shouldn't introduce any additional frequencies higher than the original. Also, a 22050 Hz (-1) Nyquist for 44100 Hz CD sampling is already well beyond our normal hearing range.

I'm pretty certain that the upsampling is there just to simplify the output stage ( or done for bragging rights ) by going to the highest common digital denominator, like 96kHz.

Steve

 

[racquetman]

Also, a 22050 Hz (-1) Nyquist for 44100 Hz CD sampling is already well beyond our normal hearing range.
Steve

Well, 22 khz isn't exactly "well beyond" our hearing range.

There seems to be an argument that frequencies above what we can hear still affect the overall sound. You can read a little on the subject here:

http://mixonline.com/mag/audio_world_above/

 

[MDS]

That's the point from the article that I was trying to reinforce. The upsampling is to improve the output stage. The Nyquist frequency is important both ways. On ADC, it defines what sampling frequency would be required to accurately reproduce the analog signal being sampled, but on DAC it also applies because there must be an analog reconstruction filter that limits the output to the Nyquist frequency.

Using the CD analogy, sure 22.05 kHz is above our range of hearing, but the reconstruction filter cannot cut it off exactly at that frequency. To do so, it would have to be a so-called 'brick wall' filter; in other words the theoretical ideal. Some frequencies below 22.05 kHz will get attenuated somewhat and they would be in our hearing range. So the theory goes, if we upsample, the Nyquist frequency is much higher and the reconstruction filter can be simpler (less steep cutoff slope). If the frequency is 96 kHz and gets cut to 48 khz, some of the frequencies at or below 48 khz will still get attenuated but the ones most affected will still be well above our range of hearing. The frequencies near 20 kHz will (should be. ) relatively unaffected.

 

[avnetguy]

Well, 22 khz isn't exactly "well beyond" our hearing range.

There seems to be an argument that frequencies above what we can hear still affect the overall sound. You can read a little on the subject here:

http://mixonline.com/mag/audio_world_above/

OK, how about ... its generally accepted that 22 khz is well beyond our hearing range.

Even that article states human hearing is down about -120db at 20khz so its not something I'm going to worry about.

Steve

 

[avnetguy]

Yes the upsampling definitely makes the filtering method simplier for the harmonics created by the DAC analog reconstruction. Probably good to point out that the source digital data does not contain these harmonics but rather is created in the conversion to the analog domain.

One point on the ADC side though, audio frequencies near or at Nyquist are not accurately reproduced. The frequency will be retained but accurate reproduction of the analog signal is approximately 10x lower (depending on who you speak to) than the sampling rate, 4.4kHz in the case of CD audio.

Steve