Audio Confusion
May 7, 2013 at 12:51 PM Thread Starter Post #1 of 53

fihidelity

100+ Head-Fier
Joined
Nov 26, 2011
Posts
237
Likes
16
I have quite a few basic questions and am a little confused so I thought it might warrant a thread.
 
  1. Am I right in thinking that CDs rip to 41kHz at 16bit, and If that was the case then what is the point in DACs supporting playback at 96kHz and 24bit or 32bit? Can you get higher than CD rip quality elsewhere and what is the sample frequency of .wav files?
  2. I was reading up on MP3 compression and it seemed like a lot of work went into finding ways of cutting out parts of a track that were inaudible and that's the only real loss. I don't know how many people can hear the difference between MP3 and FLAC or ALAC say, I can but I really have to look out for it andI couldn't tell if I was just listening to a track passively. Are the differences that striking to people and if not then why go to the extra effort of using lossless compression?
  3. Can anyone hear the difference between 41 and 96kHz say? (I've not done a test myself)
 
I'll post more as I think of them, apologies in advance 
rolleyes.gif

 
May 7, 2013 at 1:26 PM Post #2 of 53

xnor

Headphoneus Supremus
Joined
May 28, 2009
Posts
4,092
Likes
221
Audio CDs are sampled at 44.1 kHz (= 44100 Hz), two channels, with a bit depth of 16 bits per sample.
 
DACs that support higher bit depth give you the possibility to use digital attenuation without quality loss within certain limits (which depend a lot on the performance of the DAC).
 
You can get higher sample rates by buying such tracks or ripping stuff other than CDs. The main point is that you can play such sample rates natively, i.e. without conversion.
 
Lossless is a 1:1 copy of the CD or whatever you're copying. You can use that to create lossy files to save space on your portable player. You can do that every time a new lossy encoder version is released without the need to re-rip. Depending on the bitrate of the lossy file, it can be easy to virtually impossible to distinguish it from the lossless one.
 
Read: http://www.xiph.org/~xiphmont/demo/neil-young.html
 
May 7, 2013 at 1:29 PM Post #3 of 53

chewy4

Headphoneus Supremus
Joined
Jun 27, 2011
Posts
1,591
Likes
76
Quote:
I have quite a few basic questions and am a little confused so I thought it might warrant a thread.
 
  1. Am I right in thinking that CDs rip to 41kHz at 16bit, and If that was the case then what is the point in DACs supporting playback at 96kHz and 24bit or 32bit? Can you get higher than CD rip quality elsewhere and what is the sample frequency of .wav files?

There are plenty of formats that can go higher. Digital downloads have no restrictions for example, and I've seen them being sold at 24bit 192kHz.
 
 
  1. I was reading up on MP3 compression and it seemed like a lot of work went into finding ways of cutting out parts of a track that were inaudible and that's the only real loss. I don't know how many people can hear the difference between MP3 and FLAC or ALAC say, I can but I really have to look out for it andI couldn't tell if I was just listening to a track passively. Are the differences that striking to people and if not then why go to the extra effort of using lossless compression?

This is debated a lot. I've seen very few people post positive ABX logs between 320kbps LAME mp3 and FLAC. Try an ABX test yourself to see if you notice the differences.
 
A lot of the people(like myself) that use FLAC just do it for the peace of mind.
 

 
  1. Can anyone hear the difference between 41 and 96kHz say? (I've not done a test myself)

No humans can, and especially not with normal music. Although maybe a baby could. All that higher sampling rate gives you is the ability to accurately produce more frequencies. The highest frequency that can be produced correctly is whatever half the sampling rate is, and 44.1kHz brings you over 20kHz, which is the upper limit of human hearing.

That being said, a lot of these high resolution downloads are completely different masters than their CD resolution counterparts, so they will sound different.
 
May 7, 2013 at 2:46 PM Post #4 of 53

fihidelity

100+ Head-Fier
Joined
Nov 26, 2011
Posts
237
Likes
16
Quote:
Audio CDs are sampled at 44.1 kHz (= 44100 Hz), two channels, with a bit depth of 16 bits per sample.
 
DACs that support higher bit depth give you the possibility to use digital attenuation without quality loss within certain limits (which depend a lot on the performance of the DAC).
 
You can get higher sample rates by buying such tracks or ripping stuff other than CDs. The main point is that you can play such sample rates natively, i.e. without conversion.
 
Lossless is a 1:1 copy of the CD or whatever you're copying. You can use that to create lossy files to save space on your portable player. You can do that every time a new lossy encoder version is released without the need to re-rip. Depending on the bitrate of the lossy file, it can be easy to virtually impossible to distinguish it from the lossless one.
 
Read: http://www.xiph.org/~xiphmont/demo/neil-young.html

Thanks for that, that article was a good read too!
Quote:
There are plenty of formats that can go higher. Digital downloads have no restrictions for example, and I've seen them being sold at 24bit 192kHz.
 
 
This is debated a lot. I've seen very few people post positive ABX logs between 320kbps LAME mp3 and FLAC. Try an ABX test yourself to see if you notice the differences.
 
A lot of the people(like myself) that use FLAC just do it for the peace of mind.
 

 
No humans can, and especially not with normal music. Although maybe a baby could. All that higher sampling rate gives you is the ability to accurately produce more frequencies. The highest frequency that can be produced correctly is whatever half the sampling rate is, and 44.1kHz brings you over 20kHz, which is the upper limit of human hearing.

That being said, a lot of these high resolution downloads are completely different masters than their CD resolution counterparts, so they will sound different.

I'll have a go at an ABX test sometime, it'd be interesting to see how I do. I see what you mean about peace of mind, can FLAC be poorly encoded or is that just a limitation of lossy codecs? 
 
May 7, 2013 at 2:56 PM Post #5 of 53

mikeaj

Headphoneus Supremus
Joined
Jan 10, 2010
Posts
1,639
Likes
109
Every compliant lossless encoder (including for FLAC) should produce a result that can be decoded by any compliant decoder into the exact original.  A better lossless encoder might be faster or produce a slightly smaller file size.  That's it.
 
May 7, 2013 at 3:09 PM Post #6 of 53

xnor

Headphoneus Supremus
Joined
May 28, 2009
Posts
4,092
Likes
221
Poorly encoded with lossless encoders means little to no compression (but the decoded output has to be equal to the input in order to be lossless), with lossy encoders it means plainly audible artifacts.
 
FLAC for example has a compression ratio of about 56% which is good. The FLAC encoder offers several compression levels from 0 to 8. Higher compression level gets you a bit lower compression ratio (the resulting .flac file will be a bit smaller) but uses more CPU time, but it's lossless in any case.
 
May 7, 2013 at 4:09 PM Post #7 of 53

fihidelity

100+ Head-Fier
Joined
Nov 26, 2011
Posts
237
Likes
16
More questions:
 
  1. What is headroom? I've seen people talking about how a higher sample rate gives better headroom.
  2. Do different DACs convert the digital signal to an analog differently and is that why they can sound different to one another? What would say a Lampizator have over an ODAC? I still don't really understand what benefit higher bit depth has.
 
May 7, 2013 at 4:29 PM Post #8 of 53

xnor

Headphoneus Supremus
Joined
May 28, 2009
Posts
4,092
Likes
221
Assuming we're talking about DACs that abide by the sampling theorem there of course are measurable differences but not necessarily audible ones. Different noise floor, distortion, anti-aliasing filters ...
If the measurable differences are not big then whether there are audible differences or not ideally should be tested in a double blind test. The main reason for reviewers hearing differences is that the comparisons are often sighted and not properly level matched, which is very important since some DACs output 1.5 V and others over 2.5 V instead of the quasi standard 2.0 V.
 
Lampizator? Do they even provide any specs let alone basic measurements?
 
Basically, given a certain bit depth there is a fixed noise floor. With digital attenuation you push the signal into the noise floor. If the performance and bit depth of your DAC exceeds that of your music the noise floor is a couple of dB lower. You can therefore digitally attenuate by about the same amount without quality loss.
 
 
edit: I don't see how headroom is related to sampling rate.
 
May 7, 2013 at 5:21 PM Post #9 of 53

mikeaj

Headphoneus Supremus
Joined
Jan 10, 2010
Posts
1,639
Likes
109
Quote:
  1. Do different DACs convert the digital signal to an analog differently and is that why they can sound different to one another? 


 
 
The circuits, low-level implementation, etc. can definitely be different.  But hopefully, the results are similar regardless, up to a certain point.  I mean, all analog circuits have a certain amount of imperfections, and these are hopefully small enough for the task at hand.  You could think of the (digital) input information as a code: it's technical audio-speak for what the output should be.  Unless different DACs are deviating from the script, they should land somewhere close to what they're "supposed" to produce.  As xnor said, most of the differences perceived (which are not necessarily directly related to the differences in signals produced) are probably just due to different output volume levels.  Some audiophile DACs kind of deviate from the script though.
 
 
Quote:
edit: I don't see how headroom is related to sampling rate.

 
Headroom in terms of implementation (complexity, steepness, etc.) of the antialiasing filter?  *shrugs*

 
May 11, 2013 at 4:55 PM Post #10 of 53

fihidelity

100+ Head-Fier
Joined
Nov 26, 2011
Posts
237
Likes
16
When I have no music playing (and sometimes when I do), I can hear what sound like sine sweeps and today I could even hear people talking and music but really quietly in the background. This happens when I'm using my Schiit Magni with DT770 Pros. What could be causing this, is it the amp or maybe the long headphone cable? It's very annoying!
 
May 11, 2013 at 6:14 PM Post #11 of 53

xnor

Headphoneus Supremus
Joined
May 28, 2009
Posts
4,092
Likes
221
Radio reception is not a bug, it's a feature of Schiit amps!
 
May 12, 2013 at 3:45 AM Post #12 of 53

fihidelity

100+ Head-Fier
Joined
Nov 26, 2011
Posts
237
Likes
16
Quote:
Radio reception is not a bug, it's a feature of Schiit amps!

Funny you should say that, I was emailing Jason yesterday and I said something along the lines of 'Whilst it's nice that the amp has multi functionality, if I wanted a radio I probably would have bought one' 
wink.gif

 
Short of taking the lead off of the roof and lining the amp, is there any way of stopping it? When I move the headphone cable about the signal changes, is the cable basically functioning as an antennae?
 
May 12, 2013 at 3:33 PM Post #14 of 53

fihidelity

100+ Head-Fier
Joined
Nov 26, 2011
Posts
237
Likes
16
Quote:
A cheater plug(3 prong to 2 prong) will probably fix it. Or at the very least make sure its in the same power strip as the device you're playing music from.
 
Mine had that same problem at my parents house.

Is that possible in the UK?
 

Users who are viewing this thread

Top