Can you hear the difference between 16bit and 24bit audio files?

[gmgraves]

Musicophile's system: "iMac late 2009 8GB > Audirvana+/Amarra > Audioquest Forest USB > BelCanto mlink > Atlas Ascent BNC > Exposure 2010S2 Dac > Audioquest King Cobra > Exposure 3010S2 Integrated > Chord Carnival Classic > B&W CM8"

Hardly "upper mid-fi.' You are too modest. my friend. These days a good Integrated is nothing to sneeze at. My own Harman-Kardon HK990 is a good example. It made class 'B' in the Stereophile ratings two years in a row, and that's no slouch. There are some mighty expensive and good sounding separates in the same class. the 990 sounds better than my former VTL twin 150's/Audio Research SP11 setup as well as a pair of very expensive Krell separates I had on loan from my publisher for over a year while I made up my mind what to replace my tube gear with after I sold the VTLs (I'll never sell my SP11, even though I no longer use it). So don't denigrate your system.

[gmgraves]

There's a whole world of difference between the recording / mixing stage of a recording and the final playback.

Capturing an uncompressed instrument sound at 24 bit allows for a lot more headroom and better signal to noise ratio.

 

Once a whole record, of which each song is usually made of 100+ individual tracks is mixed and mastered, and dithering has been properly applied, everything changes.

 

So we need to separate recording from playback. Again, two completely different worlds.

 

Starting this thread, I just reported that there's an interesting crowdsourced test by archimago that demonstrates you can't hear differences between 24 bit and 16 bit versions of the same audio file

 

 

+1

[sandyk]

So we need to separate recording from playback. Again, two completely different worlds.

 

Not if the recording and mastering engineers are people like Barry Diament ,Cookie Marenco and several others though,

who don't go through all those damaging steps.

[tne]

DEMAND IGNORED. I am not a pupil in your classroom/lecture room.

 

Of course not. But when you make such a sweeping statement, I would expect you to be able and willing to justify it. It seems to me that there is a basic philosophical paradox to the statement you made, and that is what makes the statement so interesting. Furthermore it was hardly a demand - I said "please".

 

To address the comments of Paul and George, many experimental psychology experiments involving perception and complex cognitive phenomena utilize DBTs as their basis. Regardless, why should the fact that it involves perception/memory override the basic premise of DBTs?

 

Keeping mind that DBTs can me incorrectly performed, for several experimental reasons - sample size and selection, poor methodology, etc., - as well as the complexity of evaluating SQ for all the reasons that we all know, I would still state that they can be used, if properly designed and executed, to a positive end for evaluating audio.

[Jud]

I did read that paper and it seems like the authors might have messed up their down-sampling of the files, based on some of their odd results.

 

"Furthermore, our findings show that listeners were

more sensitive to differences between files recorded

at 88.2 kHz and their 44.1 kHz down-sampled

version, than to differences between files recorded at

different sample rates. As we down-sampled the files

through a single software program, further

investigation of down-sampling algorithms is

required to draw conclusions regarding the impact of

down sampling vs. recording at 44.1 kHz. "

 

That must be like all the other peer reviewed academic literature I've cited to "objective" friends - all these papers, reviewed by experts, turn out to have flaws easily spotted by amateurs who frequent audio forums!

 

Truth be told, I tend to look, if not for flaws, reasons for inapplicability of papers whose conclusions I don't like, too. Pretty nearly all of us do this, I think. Human nature and all that.

[Paul R]

Of course not. But when you make such a sweeping statement, I would expect you to be able and willing to justify it. It seems to me that there is a basic philosophical paradox to the statement you made, and that is what makes the statement so interesting. Furthermore it was hardly a demand - I said "please".

 

To address the comments of Paul and George, many experimental psychology experiments involving perception and complex cognitive phenomena utilize DBTs as their basis. Regardless, why should the fact that it involves perception/memory override the basic premise of DBTs?

 

i was specifically addressing DBX and rapid, multiple switching between sources. DBTs can be incredibly useful.

[Jud]

And that would only matter if human hearing had infinite bandwidth.

 

Whoa, that's certainly a misreading of the real world effects of the fact that we don't have the idealized conditions under which the Shannon-Nyquist theorem is proven. Those real world effects mean we don't so far have filters that can get around certain mathematical problems. Audible effects of these problems are a topic for discussion, but infinite bandwidth hearing certainly would not be necessary to detect them.

[gmgraves]

Of course not. But when you make such a sweeping statement, I would expect you to be able and willing to justify it. It seems to me that there is a basic philosophical paradox to the statement you made, and that is what makes the statement so interesting. Furthermore it was hardly a demand - I said "please".

 

To address the comments of Paul and George, many experimental psychology experiments involving perception and complex cognitive phenomena utilize DBTs as their basis. Regardless, why should the fact that it involves perception/memory override the basic premise of DBTs?

 

Keeping mind that DBTs can me incorrectly performed, for several experimental reasons - sample size and selection, poor methodology, etc., - as well as the complexity of evaluating SQ for all the reasons that we all know, I would still state that they can be used, if properly designed and executed, to a positive end for evaluating audio.

 

I have been party to many DBTs and all of them have been scrupulously set-op and carried-out. So, while what you say is probably true, I have no experience with such poorly executed tests. The closest I came to that was a DBT held at a hi-fi shop by a famous DAC manufacturer. Their intent was clearly to gain commercial advantage, but having said that, there also was nothing fishy about their set-up or execution of the test. Their product marginally came out on top (according to them) but other than that, I found noting wrong with their methodology or execution of the test.

 

I think that things that are simple and constant (like cable sound) can be accurately tested with a DBT, but complex devices such as amps, preamps, disc players, and DACs are a different story and my experience seems to back that opinion up to my satisfaction.

[Leonardo Piras]

Not if the recording and mastering engineers are people like Barry Diament ,Cookie Marenco and several others though,

who don't go through all those damaging steps.

 

Those people record directly to DSD. There are some others who record directly to vinyl.

While those people are great, those recording techniques imply no overdubs, no editing and very little EQ.

As such, they are suitable for classical,jazz or any kind of production that's strictly about recording performing musicians.

 

Contemporary recordings, where effects, layering and sonic "mangling" of all sorts are used for creative purposes, are currently simply impossible to produce with such techniques, due to the complexity of the process involved.

[sandyk]

Those people record directly to DSD.

 

Actually, Barry records directly to .aiff, and these days at 24/192, with a new album soon to be recorded.

[Musicophile]

Musicophile's system: "iMac late 2009 8GB > Audirvana+/Amarra > Audioquest Forest USB > BelCanto mlink > Atlas Ascent BNC > Exposure 2010S2 Dac > Audioquest King Cobra > Exposure 3010S2 Integrated > Chord Carnival Classic > B&W CM8"

Hardly "upper mid-fi.' You are too modest. my friend. These days a good Integrated is nothing to sneeze at. My own Harman-Kardon HK990 is a good example. It made class 'B' in the Stereophile ratings two years in a row, and that's no slouch. There are some mighty expensive and good sounding separates in the same class. the 990 sounds better than my former VTL twin 150's/Audio Research SP11 setup as well as a pair of very expensive Krell separates I had on loan from my publisher for over a year while I made up my mind what to replace my tube gear with after I sold the VTLs (I'll never sell my SP11, even though I no longer use it). So don't denigrate your system.

Thanks and sorry for the OT, I'm actually very happy with my Exposure integrated, it recently won a group test in two independent UK magazines (HifiNews and HifiChoice) against serious competitors like the Creek Destiny 2, NAD 390DD and Naim Nait XS2, and was very positively reviewed in some German magazines (e.g. Stereo), so I'm pretty sure I have one of the best things money can buy in this price range. Overall, I'm pretty sure in terms of bang (or rather musical enjoyment ) for the buck I'm a very happy person. I've heard many expensive systems that are giving me much less musical pleasure than mine.

 

But still, I also know that a let's say $20K and above systems can and often are more revealing than my system, so the relatively minor (to my ears) improvements from high-res over redbook may be more dramatic there.

 

I'll one day check out one of my local dealers and ask them if they are willing to participate in a double blind on a $30K and above system, out of curiosity.

[bdiament]

Those people record directly to DSD. There are some others who record directly to vinyl.

While those people are great, those recording techniques imply no overdubs, no editing and very little EQ.

As such, they are suitable for classical,jazz or any kind of production that's strictly about recording performing musicians.

 

Contemporary recordings, where effects, layering and sonic "mangling" of all sorts are used for creative purposes, are currently simply impossible to produce with such techniques, due to the complexity of the process involved.

 

Hi Leonardo,

 

As Alex pointed out, I record to 24/192 .aif (and am no fan of DSD).

 

With regard to overdubs, while I don't use the technique for my own label, I have developed a variation that does indeed include overdubs and allows for multitracking - yet keeps the mic pickup stereo and in focus throughout. This has been used by folks I know, for recordings containing several dozens of tracks.

 

In all cases that I've heard, 24-bit makes a difference, not just for recording and production but for listening too.

Too many folks seem to confuse 16-bit's signal-to-noise ratio with its dynamic range. Low level signals in 16-bit can still be quite far above the noise floor and yet still have their harmonic structure decimated by the effectively diminishing resolution that goes with diminishing level.

 

Best regards,

Barry

Soundkeeper Recordings

The Soundkeeper | Audio, Music, Recording, Playback

Barry Diament Audio

[GrahamJohnMiles]

Does it have to matter? Can anyone tell the difference between a diamond and a piece of glass from a few feet away; or a Pearl and a plastic bead. Do we say why buy a gemstone when a piece of colored glass looks just as good? Some people want the best even if others cannot discern the value or exclusivity. I have nothing but WAV files in my audio system, because I want WAV files. I don't care if someone tells me there is no difference between WAV and FLAC. If I could replace all my music with 24 bit, I probably would, because I like the idea of having what came off the master-tape before it was mixed for CD. There is nothing wrong with that and if I want to pay a little extra for the privilage, that's my business, its my money.

[mwheelerk]

Does it have to matter? Can anyone tell the difference between a diamond and a piece of glass from a few feet away; or a Pearl and a plastic bead. Do we say why buy a gemstone when a piece of colored glass looks just as good? Some people want the best even if others cannot discern the value or exclusivity. I have nothing but WAV files in my audio system, because I want WAV files. I don't care if someone tells me there is no difference between WAV and FLAC. If I could replace all my music with 24 bit, I probably would, because I like the idea of having what came off the master-tape before it was mixed for CD. There is nothing wrong with that and if I want to pay a little extra for the privilage, that's my business, its my money.

I'm a little confused by a couple of things you said and I may be simply misunderstanding your points. You state "if I could replace all my music with 24 bit, I probably would, because I like the idea of having what came off the master-tape before it was mixed for CD." Much of the music that is available and being reissued in various forms is sourced from an analogue tape. There is no 24 bit on that tape if that is the source. If it is captured and converted via ADC then you have 24 bit. I believe most tapes considered master are already mixed and the process for conversion to another medium is mastering but it's possible that someone might go back to the premixed tapes, remix and then remaster. Some 24 bit simply is t what it all cracked up to be. Just yesterday I replaced an HD Track 24/96 download of Steely Dan Gaucho which always sounded harsh to me with a ripped copy of the MFSL Gold CD and to my ears that version though 16 bit is superior but on base that on the quality of the mastering not the bit level.

 

At at this point after chasing hi res downloads for a few years I am of the opinion that what matters first is the mastering period. Secondly for myself I have decided that unless an album was recorded in 24 bit I will pass from now on on all analog conversions to hi res as they seem to me too much a mixed bag of results.

[Leonardo Piras]

In all cases that I've heard, 24-bit makes a difference, not just for recording and production but for listening too.

Too many folks seem to confuse 16-bit's signal-to-noise ratio with its dynamic range. Low level signals in 16-bit can still be quite far above the noise floor and yet still have their harmonic structure decimated by the effectively diminishing resolution that goes with diminishing level.

 

When tracking, we experimented lots of times with different bit depth and sample rates.

When I listen to the single track I recorded I can tell the difference btw 16 and 24 bits.

I've always tracked to .aiff

But later, when all the instruments add up and we mix and master the final thing, with dithering and so on, it's very, very hard to tell which is the 16/44 master and which is the 24/96.

Never had any experience with DSD.

[bdiament]

Does it have to matter? Can anyone tell the difference between a diamond and a piece of glass from a few feet away; or a Pearl and a plastic bead. Do we say why buy a gemstone when a piece of colored glass looks just as good? Some people want the best even if others cannot discern the value or exclusivity. I have nothing but WAV files in my audio system, because I want WAV files. I don't care if someone tells me there is no difference between WAV and FLAC. If I could replace all my music with 24 bit, I probably would, because I like the idea of having what came off the master-tape before it was mixed for CD. There is nothing wrong with that and if I want to pay a little extra for the privilage, that's my business, its my money.

 

Hi GrahamJohnMiles,

 

Well said.

Agreed 100%.

 

Best regards,

Barry

Soundkeeper Recordings

The Soundkeeper | Audio, Music, Recording, Playback

Barry Diament Audio

[bdiament]

When tracking, we experimented lots of times with different bit depth and sample rates.

When I listen to the single track I recorded I can tell the difference btw 16 and 24 bits.

I've always tracked to .aiff

But later, when all the instruments add up and we mix and master the final thing, with dithering and so on, it's very, very hard to tell which is the 16/44 master and which is the 24/96.

Never had any experience with DSD.

 

Hi Leonardo,

 

I understand you find it hard to tell the 16/44 from the 24/96.

My experience is different. I find them to be quite readily differentiated - at least with the recordings I've tried on the monitoring in my listening room/studio.

(I find the difference between 24/96 and 24/192 to be even greater.)

 

It could be the gear, the setup and/or the listener... or something that is occurring during your mix process. Different folks do seem to have different sensitivities to different aspects of sound. I could say the same about different systems (and techniques) too. I'm glad we can all choose the paths we wish to take. ;-}

 

Best regards,

Barry

Soundkeeper Recordings

The Soundkeeper | Audio, Music, Recording, Playback

Barry Diament Audio

[Leonardo Piras]

I'm glad we can all choose the paths we wish to take. ;-}

 

Agreed. May I ask why you're no fan of DSD?

[bdiament]

Agreed. May I ask why you're no fan of DSD?

 

Hi Leonardo,

 

I don't know why but I find the treble discomforting -- to a similar degree (but not a similar direction) that I find the treble on CD discomforting.

It bothers me to the extend that I personally don't consider DSD a high resolution format--aside perhaps from in the bass.

 

Many folks I know, including a number of cyberfriends on this forum seem to really like the format. It just isn't my cup of sonic tea.

 

Best regards,

Barry

Soundkeeper Recordings

The Soundkeeper | Audio, Music, Recording, Playback

Barry Diament Audio

[gmgraves]

Those people record directly to DSD. There are some others who record directly to vinyl.

While those people are great, those recording techniques imply no overdubs, no editing and very little EQ.

As such, they are suitable for classical,jazz or any kind of production that's strictly about recording performing musicians.

 

Contemporary recordings, where effects, layering and sonic "mangling" of all sorts are used for creative purposes, are currently simply impossible to produce with such techniques, due to the complexity of the process involved.

 

 

DSD certainly can be edited and overdubbed and electronically "mangled" just like LPCM or analog tape. But it's not as easy as it is with LPCM.

 

You're right, however about direct-to-disc, though. What you hear is pretty much what was captured as no editing is possible. In fact, if somebody goofs, the entire side must be re-recorded from scratch. Not a big deal in the days of 78's where the longest side was about 6-minutes, IIRC, in LP production, mistakes are expensive!

[gmgraves]

Hi Leonardo,

 

I don't know why but I find the treble discomforting -- to a similar degree (but not a similar direction) that I find the treble on CD discomforting.

It bothers me to the extend that I personally don't consider DSD a high resolution format--aside perhaps from in the bass.

 

Many folks I know, including a number of cyberfriends on this forum seem to really like the format. It just isn't my cup of sonic tea.

 

Best regards,

Barry

Soundkeeper Recordings

The Soundkeeper | Audio, Music, Recording, Playback

Barry Diament Audio

 

 

Perhaps it's your equipment. I've never noticed anything "discomforting" about the DSD high-end. In fact, just the opposite. I find that DSD's treble is the most natural and life-like that I've ever heard from a recording. I hear much less high-frequency distortion than I do at 24/96, for instance.

[kumakuma]

DSD certainly can be edited and overdubbed and electronically "mangled" just like LPCM or analog tape. But it's not as easy as it is with LPCM.

 

Much as I like the sound of DSD, it is my understanding that DSD cannot be edited directly and requires conversion to PCM for editing and then conversion back to DSD within the DAW. This would seem to defeat most of the benefits of recording in DSD for the vast majority of albums.

[bdiament]

After I responded to another member's question about *my* take on DSD, what would lead anyone to believe I've only heard DSD on my own equipment, as opposed to say, a good variety of systems?

 

Barry

Soundkeeper Recordings

The Soundkeeper | Audio, Music, Recording, Playback

Barry Diament Audio

[gmgraves]

Much as I like the sound of DSD, it is my understanding that DSD cannot be edited directly and requires conversion to PCM for editing and then conversion back to DSD within the DAW. This would seem to defeat most of the benefits of recording in DSD for the vast majority of albums.

 

Well, sort of. It needs to be converted to DXD in order to edit it. It's a kind of PCM designed to interchange freely with DSD.

 

Digital eXtreme Definition - Wikipedia, the free encyclopedia

[Le Concombre Masqué]

unbelievable

When tracking, we experimented lots of times with different bit depth and sample rates.

When I listen to the single track I recorded I can tell the difference btw 16 and 24 bits.

I've always tracked to .aiff

But later, when all the instruments add up and we mix and master the final thing, with dithering and so on, it's very, very hard to tell which is the 16/44 master and which is the 24/96.

Never had any experience with DSD.