What does a higher bit DAC do?
- Thread starter aphex
- Start date
JohnCU
Banned
- Dec 9, 2000
- 16,528
- 4
- 0
think of an MP3 with a high bitrate, it sounds better than those with low bitrates since it has more samples.
i think this is applicable, since sound is analog and MP3s are digital thus DAC.
but i hated that class, so i'm not sure if i'm speaking in correct terms, although i passed with a B. i dropped out of computer engineering the second year, yuck. EE FTW.
i think this is applicable, since sound is analog and MP3s are digital thus DAC.
but i hated that class, so i'm not sure if i'm speaking in correct terms, although i passed with a B. i dropped out of computer engineering the second year, yuck. EE FTW.
Switching from CPEG to EE exempted you from learning more about DAC/ADC? I just did a whole EE class last semester on DSPs and the theory of quantization and all that stuff.
Zenmervolt
Elite member
- Oct 22, 2000
- 24,514
- 44
- 91
Remeber that digital is a "stairstep" approximation of a wave. A higher bit DAC allows more "stairsteps" which in turn allows a closer approximation of the analog wave.
ZV
ZV
JohnCU
Banned
- Dec 9, 2000
- 16,528
- 4
- 0
you can take tech-electives if you want to learn about DSP, but i'm specializing in power so i don't have to. i took some signals classes, but nothing about this kind of stuff. i dropped out of CpE because at the time i didn't like all the programming classes i had to take.
That's kind of what I did. I started out as CPEG but then my soph. year took a programming class using a language called Scheme. It was the most awful thing I had ever seen/used, so I switched to EE as soon as that semester was over.
Rubycon
Madame President
- Aug 10, 2005
- 17,768
- 485
- 126
Surprised nobody has mentioned resolution that -
8 bit audio allows 256 discrete voltage levels per sample
16 bit audio allows 65536 discrete voltage levels per sample
24 bit audio allows 16777216 discrete voltage levels per sample
Its 2^n where n is the number of bits = number of discrete voltage levels per sample.
8 bit audio allows 256 discrete voltage levels per sample
16 bit audio allows 65536 discrete voltage levels per sample
24 bit audio allows 16777216 discrete voltage levels per sample
Its 2^n where n is the number of bits = number of discrete voltage levels per sample.
i dont think it has more samples or more "stairsteps", but rather, each sample is more precise
[/think]
Originally posted by: LS20
i dont think it has more samples or more "stairsteps", but rather, each sample is more precise
[/think]
You just destroyed my brain.
JohnCU
Banned
- Dec 9, 2000
- 16,528
- 4
- 0
Originally posted by: LS20
i dont think it has more samples or more "stairsteps", but rather, each sample is more precise
[/think]
are you sure? the analog signal that comes out of the DAC becomes more "precise" with more bits. a true analog signal has an infinite number of points, but alas a non-ideal DAC does not have infinite memory, so at best it is only an approximation, aka the stair-step approximation mentioned above. think of a cheap DC-AC inverter sinusoidal signal.
They use Sigma-Delta DACs, which operate on a very differnet principle than the higher bit ones. Other types of DAC are the the successive approximation and flash based converters. I couldn't explain to you exactly how the Sigma-Delta converters work without digging out my notes, but we did lab experiments in my DSP class last semester in which the 1-bit DAC was able to produce recognizable speech, whereas other methods using higher bitrates could not.
"accurate" replication of a signal can be improved with either higher sampling or higher bit-rate .. which are discrete things. higher bit means each each sample is higher accuracy of each sample... goes along with Ms Dawn says... for example instead of each sample being in increments of 4v (either 4 or 8 or 12 or 16, etc)... they can be stored in 1s (1,2,3,4)... ie more "discrete levels"
my understanding of it, atleast. im not ee
JohnCU
Banned
- Dec 9, 2000
- 16,528
- 4
- 0
Originally posted by: LS20
"accurate" replication of a signal can be improved with either higher sampling or higher bit-rate .. which are discrete things. higher bit means each each sample is higher accuracy of each sample... goes along with Ms Dawn says... for example instead of each sample being in increments of 4v (either 4 or 8 or 12 or 16, etc)... they can be stored in 1s (1,2,3,4)... ie more "discrete levels"
my understanding of it, atleast. im not ee
i don't know much about signals except for very basics but i'm studying through some stuff now. maybe someone else can verify for sure.
EDIT: okay i'm pretty sure you are right. the frequency is samples per second... the bit rate is bits per sample...
shurg....
a digital to analog convertor takes the digital content "bits" and converts it into analog information that can be ensd to an amplifier.
a higher quality capacity DAC will allow for digital information coming in to be more and more complex, therefore providing you with the true resolution at which the digital content was created.
a digital to analog convertor takes the digital content "bits" and converts it into analog information that can be ensd to an amplifier.
a higher quality capacity DAC will allow for digital information coming in to be more and more complex, therefore providing you with the true resolution at which the digital content was created.
Bits per sample defines the dynamic range of the file, nothing more. It will lower the noise floor and it will also allow for sounds with rapid high-amplitude transience to be played back correctly.
Therefore certain sounds such as electric guitars running through tube amplifiers will not benefit from more bits per sample, and other sounds, particularly gun-shots would benefit considerably from a 24bit or even 32bit sample in terms of initial integrity, although there are no converters/amplifiers available that could make full use of such extreme dynamic ranges, nor are there speakers that have the excursion capabilities.
Then there are the # of samples per second. These primarily improve waveform integrity through time rather than amplitude, and a higher sample frequency has another arguable benefit of extending the upper frequency response.
Keep in mind that the specifications of a DAC are far more important than the sample size it can work with. There are many 16-bit audio converters out there that can greatly outperform various 24-bit units. Primarily because most of the audio you listen to is 16-bit anyways, and would not benefit from a 24-bit DAC, but would benefit from it's quality of output.
Therefore certain sounds such as electric guitars running through tube amplifiers will not benefit from more bits per sample, and other sounds, particularly gun-shots would benefit considerably from a 24bit or even 32bit sample in terms of initial integrity, although there are no converters/amplifiers available that could make full use of such extreme dynamic ranges, nor are there speakers that have the excursion capabilities.
Then there are the # of samples per second. These primarily improve waveform integrity through time rather than amplitude, and a higher sample frequency has another arguable benefit of extending the upper frequency response.
Keep in mind that the specifications of a DAC are far more important than the sample size it can work with. There are many 16-bit audio converters out there that can greatly outperform various 24-bit units. Primarily because most of the audio you listen to is 16-bit anyways, and would not benefit from a 24-bit DAC, but would benefit from it's quality of output.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 411
Facebook
Twitter