Hi-resolution versus 16 bit 44.1 Khz

[Niktech]

Is anyone really noticing a benefit to listening to 24 bit 48 KHz and above to listening to 16 bit 44.1khz?

Some tracks I’ve listened too at 16 bit 44.1 KHz sound way better than the the hi -res. Some hi-res has so much air and treble it’s unlistenable  to my ears- it’s like nails down a chalkboard .

 

It would seem what is more important than the resolution is the quality of the Master, i.e. un-butchered, and of course your equipment. I guess this in someways ties in with the desktop / streaming thread. What are forumites preferences here?

 

 I did do a search to see if we’ve discussed/debated this previously on SNA, but I’ve had a few vodka beverages now and may have overlooked it. I did see one dated 2012. 
 

Cheers
 

 

 

Edited January 17, 2023 by Niktech

[muon*]

My sig says it all for me, and my personal preference.

[Niktech]

5 minutes ago, muon* said:

My sig says it all for me, and my personal preference.

LOL, I can’t see it, mate. I view the forum on my

phone, so I can’t see anyone’s cards. However, IIRC your are a Red Book man 

Edited January 17, 2023 by Niktech

[muon*]

1 minute ago, Niktech said:

LOL, I can’t see it, mate. I view the forum on my

phone, so I can’t see anyone’s cards

Marantz CD60 modded, with discrete regs, TDA1541A S1 Single Crown running in non over sampling mode ect'.

 

I played with high res on and off a few years back and found over all, I prefer well recorded red book done well to be my preference.

[Niktech]

1 minute ago, muon* said:

Marantz CD60 modded, with discrete regs, TDA1541A S1 Single Crown running in non over sampling mode ect'.

 

I played with high res on and off a few years back and found over all, I prefer well recorded red book done well to be my preference.

Cool, mate. Just edited my post as I do remember reading you are Red Book man

[Niktech]

Actually, about 20 years ago I had a CD63. The lazer had failed and I disposed of the player. I still kick myself to this day for doing so instead of getting it fixed

 

Edited January 17, 2023 by Niktech

[Dingbat]

No scientific comparison here, but I have a few albums in higher resolution than red book format, and mostly they seem to sound quite good. Most of these I don't have a version in red book format to compare. I did notice though that Sonic Youth's album Goo just sounds particularly harsh, despite the so-called Hi-Res format. It's somewhat painful to listen to. I was thinking about whether to get the CD to work out if it's an artifact of the Hi-Res, but can't justify spending twice for the same album.

[aechmea]

You will find that there are a lot of people on SN using 16/44. 

 

I use PCM digital out from my source devices to an external DAC.

 

I have tried 24/96, DVD-Audio and SACD but can't be bothered with them any more.  I even went to the trouble of extracting the hidden 24/96 stereo tracks from my handful of DVD-Audio disks.  [Hi rez is not supposed to be available as a digital output unless encrypted or ... but that's another story.]  

 

I think that it is the mastering that matters.  There are some masterings on CD that I don't much like, but that's life.

 

Technically,

extra bits means that the available dynamic range is greater, but no-one even uses 16 bits = 96dB

higher sample rate means that higher frequencies can be carried by the digital data, but who can hear > 20KHz anyway.

That's it; there is nothing magic about more bits and higher sample rates even though marketing would have you believe otherwise.

[andyr]

5 minutes ago, aechmea said:

Technically ... higher sample rate means that higher frequencies can be carried by the digital data, but who can hear > 20KHz anyway.

 

Then again, if you have tweeters like yours - which are rated to 40kHz ... why do you want to limit them by using CD quality (44.1k) or 48k?

 

That's why I run my system at 96kHz.

 

[bob_m_54]

9 hours ago, andyr said:

 

Then again, if you have tweeters like yours - which are rated to 40kHz ... why do you want to limit them by using CD quality (44.1k) or 48k?

 

That's why I run my system at 96kHz.

 

So you don't scare the neighbourhood dogs... LOL

[bob_m_54]

10 hours ago, Niktech said:

It would seem what is more important than the resolution is the quality of the Master, i.e. un-butchered, and of course your equipment.

That's the crux of it..

[Nada]

Its important to not get confused between 44.1/16 and "high definition" file effects verses different mastering.

 

Many HD versions are remastered. Sometimes with higher volume which makes them immediately sound better initially. But also with dynamic compression to allow the higher dBFS which with time can make them seem less "alive" sounding.

 

If you just compare RB vs HD you dont know if any perceived change is

1. placebo - HD should sound better right?

2. mastering

3. sample rate/bit depth

 

A little experimentation avoids a lot of spurious speculation. Foobar is like a lab. A free lab with great tools. Say you have two versions  of  a track:

 

A. RB (Red book 44.1/16)

B. HD (eg 192/24)

 

In Foobar

1. do ABX to compare the two tracks and see if you can hear a difference

2. Run replay gain on both note the numbers and apply replay gain

3. do ABX again to see what happens with volume matching

4. run Dynamic Range out of interest then find out if the sample rates are audible

5. Set up  DSP resample SOX SINC highest quality with resample to 44.1 and save the HD as RB 44.1/16 named as eg HD.2

6. ABX HD vs HD.RB - can you hear any difference with the same track at native HD versus downsampled to RB?

7. DSP upsample   RB to  192/24

8. ABX RB vs RB.2

 

This can tell you:

1. if the remastered HD sounds better independent of the sample rate

2. the volume and dynamic compression of the HD file vs RB

3. if you can hear any difference if the RB is upsampled

4. if you can hear if the HD is down sampled

 

Many who have invested in HD tracks will find this sobering.

[BugPowderDust]

1 hour ago, Nada said:

Many who have invested in HD tracks will find this sobering.

It doesn't help that there is some persistent belief in the Audiophile community that 16/44 was always insufficient for task with respect to regeneration of a digital file into the analog equivalent.

 

In fact the bit depth and sample rate were both selected based upon existing sampling theory (Nyquist and Shannon) and convenience with respect to discrete integer factors and simplicity. It's a great example of engineering theory designing products to fit the task at hand.  

 

The real reason early CDs sounded like cack was that mastering technologies had not developed alongside the media and it took some time for tools and also mastering engineers to work out how to leverage the most out of the raw media.  

 

A well mastered red book CD on my DAC still sounds superb. I listen to high bitrate stuff, but often to get access to remasters that bring something new to an old mix / master.

 

[davewantsmoore]

14 hours ago, Niktech said:

Is anyone really noticing a benefit to listening to 24 bit 48 KHz and above to listening to 16 bit 44.1khz?

 

Not if the audio content is the same.    ie.  if you get a source file (of any "resolution") and resample it to 16/44 (or whatever) you won't be able to tell the difference.

 

There are a few issues that can make this more murky.... and mislead people into thinking that 16/44 is "insufficient".

 

• For files you buy or stream, the content is not always the same.... so if you compare a "high res" download to the CD, then it isn't always the same content.  ie. it's a different mastering or mix of the audio  (so of course it will sound different).    This is especially true for older stuff that has been re-released, re-mixed, re-mastered
• Some DACs might benefit from an input which is in a higher sampling rate.   But this isn't because the higher sampling rate is "better quality audio" ..... so it doesn't matter if you use a "high res" source you download or stream, or if you resample 16/44 to the higher rate yourself.
• Poor quality resampling (both up and down) can hurt the fidelity of the audio.   I think people will come to realise in the future that the effect of this has been understated.    If audio is created in a certain sampling rate, I think there is a very good case of leaving it at this rate, and not resampling it.

 

14 hours ago, Niktech said:

Some tracks I’ve listened too at 16 bit 44.1 KHz sound way better than the the hi -res

 

There is zero reason to expect that higher sampling rates than 44 will offer any improvement.   It just comes down to what the artists put in the container.

 

14 hours ago, Niktech said:

. Some hi-res has so much air and treble it’s unlistenable  to my ears- it’s like nails down a chalkboard .

It's nothing to do with the bitrate or sampling rate.... it's simply what the artist created.

 

If you resampled it to 16/44, it would still sound as "air and treble it’s unlistenable"

 

14 hours ago, Niktech said:

 I did do a search to see if we’ve discussed/debated this previously on SNA, but I’ve had a few vodka beverages now and may have overlooked it. I did see one dated 2012. 

The internet is just full to overflowing with misinformation and bad-takes about how digital audio works.... surprisingly (perhaps UN-surprisingly) from people who should know better.

 

• If you increase the sampling rate you can include higher frequencies.
• If you increase the bitrate you can include quieter sounds.

^^^

Precisely this and nothing more.

 

If you increase the sampling rate it does not allow you store the lower frequencies with any "higher precision" .... and if you increaser the bitrate it does not allow you store the louder sounds with any "higher precision".

 

For the portion of the spectrum that is covered by 16/48 .....  24/192 covers that same portion of the spectrum with the same precision.

[stereo coffee]

16 bit plain is acceptable , to pretty good  particularly as @muon* refers to using the TDA 1541,  in fact  with a simple addition can be better than any hi-res specification *.   But where power supplies are based on 78x and 79x type regulators in hi res players - you are much better off with 16 bit ,as the nasty regulator really is designed for washing machines or equally mundane devices like computer mother boards . As what you hear as hi res + dreadful regulation  which is truly horrible.

 

There is answer in choosing players as hi res where their power supplies are designed for audio purpose, and then such players are much better than 16 bit. You would look for voltage references and discrete jfet circuitry with ripple reduction via capacitance multipliers. 

 

* There is also opportunity with 16 bit  to out-do even the best hi res players , by using a real time compander like a DBX 150x where the digital noise floor for a 16 bit player is then absent. The detail is on Page 11   https://dbxpro.com/en/product_documents/150x-owners-manualpdf

Edited January 18, 2023 by stereo coffee

[davewantsmoore]

13 hours ago, andyr said:

Then again, if you have tweeters like yours - which are rated to 40kHz ... why do you want to limit them by using CD quality (44.1k) or 48k?

Because:

 

Sounds above 20khz are not audible.

 

The response of your tweeters are absolutely garbage above 20khz..... and so even if it was audible (it's not) you'd probably want to turn it off.

 

It would be tempting at this point to trot out all the "studies" showing that sounds above 20khz might be kinda audible somehow.   They are garbage studies which show nothing.

 

Disclaimer:   I spent many years trying to prove to myself that higher frequencies than 20khz (and hence  "high res" audio) WAS audible..... because it c/would be  a "no brainer" way to improve fidelity.     The rub is that even if you have a reasonably linear transducer, and even if you amplify the high frequencies by HEAPs, they are still not audible..... BUT you have to be careful, as doing this can create distortion in the (actually) audible band( below 20khz), which can sound "different" (and make you think > 20khz is audible).

 

FWIW, after all that..... I came to think there's not even much relevant content above 10 kHz (!!!!!)

 

You also do want to be careful that the phase between 5 and 20khz is linear .... and if you have some sort of sharp rolloff at 20khz ish, then this can be a problem (audible).   That is simple to correct though (if you have digital filters, anyways).

 

[davewantsmoore]

6 minutes ago, stereo coffee said:

16 bit plain is acceptable

Yes.

It's obvious as to why 24bit is used in a production environment, or for "archival" or whatever  .... but in a playback system it is completely unnecessary.

[acg]

2 hours ago, davewantsmoore said:

FWIW, after all that..... I came to think there's not even much relevant content above 10 kHz (!!!!!)

 

Me too.

 

Redbook is more than adequate for high end playback.  Up/oversample if necessary to suit your dac but there is zero technical penalty for starting with redbook media. 

[stereo coffee]

2 hours ago, davewantsmoore said:

Because:

 

Sounds above 20khz are not audible.

 

The response of your tweeters are absolutely garbage above 20khz..... and so even if it was audible (it's not) you'd probably want to turn it off.

 

It would be tempting at this point to trot out all the "studies" showing that sounds above 20khz might be kinda audible somehow.   They are garbage studies which show nothing.

 

Disclaimer:   I spent many years trying to prove to myself that higher frequencies than 20khz (and hence  "high res" audio) WAS audible..... because it c/would be  a "no brainer" way to improve fidelity.     The rub is that even if you have a reasonably linear transducer, and even if you amplify the high frequencies by HEAPs, they are still not audible..... BUT you have to be careful, as doing this can create distortion in the (actually) audible band( below 20khz), which can sound "different" (and make you think > 20khz is audible).

 

FWIW, after all that..... I came to think there's not even much relevant content above 10 kHz (!!!!!)

 

You also do want to be careful that the phase between 5 and 20khz is linear .... and if you have some sort of sharp rolloff at 20khz ish, then this can be a problem (audible).   That is simple to correct though (if you have digital filters, anyways).

 

But limiting such frequencies has good opinion , that such limiting irreversibly changes,  the frequencies we do hear, much for the worse. .

https://worldradiohistory.com/hd2/IDX-Audio/Archive-Studio-Sound-IDX/IDX/90s/Studio-Sound-1999-01-OCR-Page-0092.pdf. 

 

 

[Dingbat]

I should also add there's nothing wrong with CD bit-rates - most of my material is in this format, and a lot of it sounds very good. I agree generally that how it is recorded and mastered is more important. Something interesting, a lot of the Indie stuff I get in digital form which would never be listened to by a big audience sounds really excellent on my system, which means the source recording and mastering must have been done very well, even though I suspect these artists don't have access to a lot of resources.

 

I think the idea behind sampling rates higher than 20kHz is not that it's beyond human hearing, it's the interpolation in the digital signal. It might give some intangible benefit, but I can't say that I'm qualified to quantify this. The idea is probably a bit like anti-aliasing in graphics rendering.

[stereo coffee]

27 minutes ago, acg said:

 

Me too.

 

Redbook is more than adequate for high end playback.  Up/oversample if necessary to suit your dac but there is zero technical penalty for starting with redbook media. 

Certainly starting, but comparison between a hi res player with good power supplies and a good plain 16 bit player shows up where 16 bit leaves off and the hi res player continues on retrieving more from the recording. Drawing similarity to many areas of electronics, where we have available headroom, digital replay likes having no boundaries around it.    

 

But given the ability to remove a 16 bit noise floor as alerted to with a DBX 150x, 16 bit certainly proves itself.  Whether you wish to spend on sky is the limit resolution, or divert to use equally as good companding as suggested, there is  certainly more, if you wish to find it.

[stereo coffee]

27 minutes ago, Dingbat said:

I should also add there's nothing wrong with CD bit-rates - most of my material is in this format, and a lot of it sounds very good. I agree generally that how it is recorded and mastered is more important. Something interesting, a lot of the Indie stuff I get in digital form which would never be listened to by a big audience sounds really excellent on my system, which means the source recording and mastering must have been done very well, even though I suspect these artists don't have access to a lot of resources.

 

I think the idea behind sampling rates higher than 20kHz is not that it's beyond human hearing, it's the interpolation in the digital signal. It might give some intangible benefit, but I can't say that I'm qualified to quantify this. The idea is probably a bit like anti-aliasing in graphics rendering.

Yes the comparison between image resolution and audio resolution is a good one, Almost the same language is used for each.

https://www.photometrics.com/learn/camera-basics/bit-depth

 

The limitation of what technically is Nyquist sampling , to 20Khz was a economics decision , Sony simply asked Herbert Von Karajan what was the maximum expected length of a classical recording to which he replied 74 minutes. From there Sony in tandem with Philips using the foundation work of Alec Reeves from 1932, worked out that the disc could be quite small which was then easy to accommodate with hardware sizing, and likely would be popular with consumers too. 

 

If audio minded thought had gone into the medium then a higher sampling rate such as 66.2 Khz would have resulted in frequency response to 30Khz , with close to 60 mins on each disc. As such digital would have matched analog cartridge ability, and given a much better audio experience to us all.  That said Pioneer explored extending frequency response with their Legato Link players, which get good reviews. 

 

Whilst I bemoan higher bandwidth not being available, instead where instead dynamic range and noise floor  is catered for above 16 bit, CD remains a good medium. It's what is done between the player and your power amp that makes the most difference to maintain its linearity.   

[muon*]

6 hours ago, stereo coffee said:

16 bit plain is acceptable , to pretty good  particularly as @muon* refers to using the TDA 1541,  in fact  with a simple addition can be better than any hi-res specification *.   But where power supplies are based on 78x and 79x type regulators in hi res players - you are much better off with 16 bit ,as the nasty regulator really is designed for washing machines or equally mundane devices like computer mother boards . As what you hear as hi res + dreadful regulation  which is truly horrible.

 

I replaced all my 78xx and 79xx regs with Sparko discrete regs, was a very nice improvement.

[acg]

2 hours ago, stereo coffee said:

Certainly starting, but comparison between a hi res player with good power supplies and a good plain 16 bit player shows up where 16 bit leaves off and the hi res player continues on retrieving more from the recording

 

Examples?

 

2 hours ago, stereo coffee said:

But given the ability to remove a 16 bit noise floor as alerted to with a DBX 150x

 

You are talking about a unit using compression to remove tape hiss in 1993...we've come a long way since then.

[acg]

Just now, muon* said:

I replaced all my 78xx and 79xx regs

 

As far as I know there are none of these regs in any of my gear, but I have used them myself to provide services voltages to run fans and relays in preamps and amplifiers.