Head-Fi.org › Forums › Equipment Forums › Sound Science › Frequency Response Question
New Posts  All Forums:Forum Nav:

Frequency Response Question

post #1 of 38
Thread Starter 

So I'm in the middle of adding music to my iTunes library and I noticed the "Sample Rate" is 44.100 on most of my songs. Does this have any relation to the frequency response/range that headphones have? I'd just like to be educated in this topic, so that I can have optimal sound quality. Thanks everyone.

post #2 of 38
Quote:
Originally Posted by Digital DJ View Post

So I'm in the middle of adding music to my iTunes library and I noticed the "Sample Rate" is 44.100 on most of my songs. Does this have any relation to the frequency response/range that headphones have? I'd just like to be educated in this topic, so that I can have optimal sound quality. Thanks everyone.

No,  you're fine.  If you're adding music from CDs 44.1 is perfect as it's the original sample rate of all CDs.  If you are buying music from the iTunes Store you'll see the files are mostly 44.1 also.  The figure to watch out for is the "Stereo Bit Rate" which should be at least 256kbps or higher for either AAC or mp3 files for the best quality.  


Edited by jaddie - 5/19/13 at 10:24pm
post #3 of 38

Well 44.1 will give flat response to or very near 20 khz. 

 

48 khz might give flat response to a bit more than that.  Likely your hearing is limited to 20 khz or something lower.  So well done 44.1 is enough more is overkill.  Chances are you won't receive any benefit with higher sample rates.
 

post #4 of 38
Thread Starter 
Quote:
Originally Posted by jaddie View Post

No,  you're fine.  If you're adding music from CDs 44.1 is perfect as it's the original sample rate of all CDs.  If you are buying music from the iTunes Store you'll see the files are mostly 44.1 also.  The figure to watch out for is the "Stereo Bit Rate" which should be at least 256kbps or higher for either AAC or mp3 files for the best quality.  


The Settings for my iPod Nano that I'm using right now are set to convert 256bit songs to 192bit to save space, but if I have enough left over space on the hard drive on my Nano after all the music is on, I'm gonna revert it back to 256bit.

post #5 of 38

Sample rate of 44.1 kHz means that the highest frequency you can in theory reproduce is just below 22.05 kHz. However, there are practical considerations regarding the reproduction of the digital samples that affect the 22.05 kHz region. 44.1 kHz sampling rate is not bad IMO, but 96 kHz might be a little better (allows for an easier and better behaving reconstruction filter and stuff.)

 

Anyhow, most headphones should go as high as 22.05 kHz. The problem with headphones though is that they tend to emphasize and de-emphasize  certain frequencies relative to others in the audio (i.e. they are not ruler flat.) They also sometimes introduce frequencies that were not originally in the recording (non-linear distortion) though hopefully a lesser degree... and that is why (among other things) different headphones may sound and perform different.

 

Music, at any sampling rate, is also severely affected by the recording and mastering process...

post #6 of 38
Quote:
Originally Posted by Digital DJ View Post


The Settings for my iPod Nano that I'm using right now are set to convert 256bit songs to 192bit to save space, but if I have enough left over space on the hard drive on my Nano after all the music is on, I'm gonna revert it back to 256bit.

Good idea.  The audible difference between 256kbps and 192kbps is small, but important.  If you had to eliminate a few songs you barely listen to in order to use the higher bit rate and get all your favorites to fit, I think you'd be happier in the long run.  AAC would be the best choice in this case, avoid .mp3, it's not as efficient at higher bit rates.  

post #7 of 38
Thread Starter 
Quote:
Originally Posted by jaddie View Post

Good idea.  The audible difference between 256kbps and 192kbps is small, but important.  If you had to eliminate a few songs you barely listen to in order to use the higher bit rate and get all your favorites to fit, I think you'd be happier in the long run.  AAC would be the best choice in this case, avoid .mp3, it's not as efficient at higher bit rates.  


Yeah, I've got my settings on AAC. I'm hoping I'll have enough space to revert all the songs back to 256bit, but we'll see. I'll have about 1,500 songs on my Nano by the time its finished, my Nano has 16GB of storage. But about 1.5 of that is dedicated memory for the Nano, so I actually have about 14.5 GB of usable memory. I've got almost 500 songs on it as of now, and about 9.50 GB left over so far. I'm thinking if I have about 3GB left over after all is said and done, that might be enough space to revert all my songs back to 256bit. But if not, I'll be fine listening to 192bit, it's better than 128 after all hahaha.

post #8 of 38
Quote:
Originally Posted by esldude View Post

Well 44.1 will give flat response to or very near 20 khz. 

 

48 khz might give flat response to a bit more than that.  Likely your hearing is limited to 20 khz or something lower.  So well done 44.1 is enough more is overkill.  Chances are you won't receive any benefit with higher sample rates.
 

This would only be true of original recordings.  Once on CD, it's 44.1, so 22KHz max, and even if those files are converted to 48KHz, there will be no additional high end response as it's already been limited by the record anti-alieas filter..

post #9 of 38

As already written, a 44.1kHz file is able to capture frequencies up to 22 kHz (half the sample rate). The human hearing is limited to 20 kHz anyway, sometimes a little higher (young people with excellent hearing), but often lower. So 44.1 kHz is definitely fine, especially for an iTune library.

 

Take the online test here : http://www.audiocheck.net/audiotests_frequencycheckhigh.php

 

You can download a 44.1 kHz .wav file, by clicking the down arrow. The file will prove two things: 1/ that a 44.1 kHz file is able to play a 22 kHz tone, and 2/ you probably won't hear the tone, because it is above your hearing limit.

post #10 of 38
Quote:
Originally Posted by audiosampling View Post

As already written, a 44.1kHz file is able to capture frequencies up to 22 kHz (half the sample rate).


thats with 16 bit audio i believe.

 

anyone know whats the case with 24 bit? i know theres no audible difference, im just interested

post #11 of 38
Quote:
Originally Posted by adamlr View Post

 

thats with 16 bit audio i believe.

anyone know whats the case with 24 bit? i know theres no audible difference, im just interested

 

It makes no difference regarding the maximum frequency that can be reproduced. It is 22.05 kHz (sample rate / 2) with 8, 16, or 24 bit samples.

post #12 of 38
Quote:
Originally Posted by adamlr View Post

thats with 16 bit audio i believe.

 

anyone know whats the case with 24 bit? i know theres no audible difference, im just interested

 

Bit depth is for dynamics (such as here : http://www.audiocheck.net/audiotests_dynamiccheck.php) - not frequency response. Visit the link and check out if you can hear the lowest dynamics (it's a 16-bit test). If you can, then you may need 24-bit resolution...

post #13 of 38
Quote:
Originally Posted by stv014 View Post

 

It makes no difference regarding the maximum frequency that can be reproduced. It is 22.05 kHz (sample rate / 2) with 8, 16, or 24 bit samples.

thank you for explaining that, i was wondering.

Quote:
Originally Posted by audiosampling View Post

 

Bit depth is for dynamics (such as here : http://www.audiocheck.net/audiotests_dynamiccheck.php) - not frequency response. Visit the link and check out if you can hear the lowest dynamics (it's a 16-bit test). If you can, then you may need 24-bit resolution...


naaa, he got to -30 and i could barely hear him. im not surprised, ive never claimed to have super hearing or anything, i just gave it a try because you provided the link. mind you, i live in the town center, and i have a fan working, so my room is pretty noisey.

 

so bit depth is to provide for greater dynamics? huh, no wonder ~10 is enough... (i say enough based on ethan wieners famous bit degradation video, where ~10 bit was where i still felt no quality difference, and lower than that i did)

post #14 of 38
Quote:
Originally Posted by adamlr View Post


thats with 16 bit audio i believe.

 

anyone know whats the case with 24 bit? i know theres no audible difference, im just interested

...and as with sample frequency, converting from a lower to higher bit depth doesn't do anything audible, you can't increase dynamic range and push the noise floor of the original down.  If you made an original recording, possibly, but still highly unlikely because the dynamic range of any kind of music in a recording studio is barely the equivalent of 16 bits.  The main advantage to 24  bits originals is the convenience of post processing in production with more "room" to work in.

post #15 of 38
Quote:
Originally Posted by ultrabike View Post

Sample rate of 44.1 kHz means that the highest frequency you can in theory reproduce is just below 22.05 kHz. However, there are practical considerations regarding the reproduction of the digital samples that affect the 22.05 kHz region. 44.1 kHz sampling rate is not bad IMO, but 96 kHz might be a little better (allows for an easier and better behaving reconstruction filter and stuff.)

 

Well, the problems with reconstruction filters were true years back.  Practically speaking they aren't an issue anymore.  Nearly all modern DACs are some form of sigma-delta converter.  They upsample internally to much higher rates and use simple well behaved filters.  Most quote phase at 20 khz being +/- .5 degrees or maybe a couple degrees.  If you go to 96 khz, all you do usually is reduce the upsampling multiplier.  Lets say your DAC upsamples 48 khz by 128x, if you go to 96 khz many will then upsample 64x.

 

And as others have pointed out, if your original is 44 at 16 bits you cannot create additional information.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Frequency Response Question