24/48 Flac files instead of 24/96?

[guitarrerodiabolico]

Hi everybody, this is my first post. I tried to check if there was something like my inquire somewhere in the forums and I couln´t find any.

So, my question is: if a flac file at 24/96 (music) shows on the spectrum analysis that there is nothing above let´s say...20Khz, why not convert it to 24/48? Am I missing something? And even if there is something going up to 40khz or 50khz, what about our ears? Are we really able to listen beyond 22khz or 24khz?

I´m asking this because is half space speaking in megabytes. Of course if somehow goes against sound quality i should forget it, but maybe somebody ou there has an aswer or expanation for this. I´m new at hi-fi files, so if what I´m asking is a complete nonsense please forgive me, but still, please explain it to me.

Many thanks to all.

 

[Crocodile]

96kHz is the sampling frequency, not the audio frequency.

 

[John Duncan]

Yes, and the sampling frequency needs to be double the audio frequency to recreate it accurately, which was why 44k was chosen for CD.

I'm not a big believer in high res audio tbh, but then I'm 47 so probably a bit deaf. You should try taking the 24/96 files and downsample to a copy in 16/44 and ABX them, then decide 🙂

 

[TimothyRias]

For that experiment I would downsample to 48 kHz rather than 44 kHz. Downsampling to 44 kHz requires a lot more interpolation, and consequently and percieved difference could be due to artifacts of the downsampling method. (Rather than inherent to the lower sampling rate.)

 

[John Duncan]

For that experiment I would downsample to 48 kHz rather than 44 kHz. Downsampling to 44 kHz requires a lot more interpolation, and consequently and percieved difference could be due to artifacts of the downsampling method. (Rather than inherent to the lower sampling rate.)

Interesting, thanks.

 

[davedotco]

For that experiment I would downsample to 48 kHz rather than 44 kHz. Downsampling to 44 kHz requires a lot more interpolation, and consequently and percieved difference could be due to artifacts of the downsampling method. (Rather than inherent to the lower sampling rate.)

Spot on..... :clap:

Once you get above CD standard resolution the are few if any advantages to be had, all other things being equal.

Mucking about with the bit depth or sampling rate can alter the sound audibly though rarely does if done with care.

For example, digital volume controls effectively throw away bits, reducing resolution but given that most digital processors are 24 bit or greater this is not usually a problem. Even with a 24 bit system it is unlikely that the effects of a quite substantial level reduction will be audible. It is worth remembering that even the best CD players or dacs will struggle to resolve better than 17-18 bits.

Similarly Nyquist shows that a sampling rate of twice the highest frequency is sufficient to perfectly reproduce the original signal, so reduction of sampling rate from 96 to 48 khz should be inaudible. Halving the sampling rate is simple arithmetically, but resampling from 96 to 44.1 khz is more complex, so the performance of the resampling mechanism comes into play.

As always in audio, it is often simpler to hear the difference than it is to work out what is actually causing the difference.

 

[guitarrerodiabolico]

Ok, I did the experiment converting a few 24/192 files to 24/96 and then to 24/48.

This files are 1) a Coltrane´s tune downloaded from HDtracks, 2) a Pat Metheny´s vinyl rip, and 3) a Rickie lee Jones from a SACD iso taken from a PS3.

I analyzed these 3 files with audacity. On #2 an #3 there was signal above 45Khz, on #1 the max. was 21Khz.

The truth is I didn´t notice any notorious change ´till converting 16/44. When i say notorious means that maybe you can say there is "something" missing but not for sure. (On #2 and #3)

I compared the sound with the vinyls I have and there was no difference. This means this flac files (down to 24/48) have the same quality as using the Lp (according to my ears)

My audio equipment: Amp.: Yamaha Cr-1020, Blueray player Sony S-580 (a cheap one but very good), Tannoy 15" Monitor Gold speakers. turntable: Pioneer PL-680 with Stanton 880.

 

[davedotco]

Ok, I did the experiment converting a few 24/192 files to 24/96 and then to 24/48.

This files are 1) a Coltrane´s tune downloaded from HDtracks, 2) a Pat Metheny´s vinyl rip, and 3) a Rickie lee Jones from a SACD iso taken from a PS3.

I analyzed these 3 files with audacity. On #2 an #3 there was signal above 45Khz, on #1 the max. was 21Khz.

The truth is I didn´t notice any notorious change ´till converting 16/44. When i say notorious means that maybe you can say there is "something" missing but not for sure. (On #2 and #3)

I compared the sound with the vinyls I have and there was no difference. This means this flac files (down to 24/48) have the same quality as using the Lp (according to my ears)

My audio equipment: Amp.: Yamaha Cr-1020, Blueray player Sony S-580 (a cheap one but very good), Tannoy 15" Monitor Gold speakers. turntable: Pioneer PL-680 with Stanton 880.

The reality is that modern digital recording, even at CD standard is pretty damn good.

If it isn't, then this is due either to incompedence on behalf of the the recording and mastering crew or commercial decisions made by the record companies.

Mostly it is the latter, with 'loudness wars' and deliberately compromising CD standard recordings being the most obvious.

Bastids.

 

[andyjm]

Hi everybody, this is my first post. I tried to check if there was something like my inquire somewhere in the forums and I couln´t find any.

So, my question is: if a flac file at 24/96 (music) shows on the spectrum analysis that there is nothing above let´s say...20Khz, why not convert it to 24/48? Am I missing something? And even if there is something going up to 40khz or 50khz, what about our ears? Are we really able to listen beyond 22khz or 24khz?

I´m asking this because is half space speaking in megabytes. Of course if somehow goes against sound quality i should forget it, but maybe somebody ou there has an aswer or expanation for this. I´m new at hi-fi files, so if what I´m asking is a complete nonsense please forgive me, but still, please explain it to me.

Many thanks to all.

The engineers back at Philips in the late 70's were not idiots. The redbook standard (16/44.1) is perfectly adequate for audio reproduction. Increasing the bit depth is a total waste of time, particularly given the tendency for current record producers to compress the hell out of modern CD recordings. 12 bits would be more than adequate for the rubbish produced at the moment.... There are engineering reasons why increasing the sample rate may be helpful as it makes life easier for the anti - aliasing filters on the output of the DAC. A well implemented over sampling approach can get over this issue however.
I will leave it for the interested student to explain why Philips used 44.1KHz and not 44KHz for the redbook sample rate.

Edit: I should have added that 24/96 is marketing, not engineering.