A 16 v. 8 bit listening test.

[Les Cargill[_4_]]

I can't tell any difference for the material included.

http://www.audiocheck.net/blindtests_16vs8bit_NeilYoung.php

--
Les Cargill

[geoff]

On 10/11/2015 7:32 a.m., Les Cargill wrote:
> I can't tell any difference for the material included.
>
> http://www.audiocheck.net/blindtests_16vs8bit_NeilYoung.php
>


Neil Young ? During his grunge period(s) this isn't much difference !

That's why he feels the need for this higher resolution Pogo thing
!?!@@%#&!!!

geoff

[[email protected]]

EQ one differently from the other(add
some top to the 16bit to give the illusion
of higher quality) and you'll hear a
difference. At least the bean counters
at the label will!

[Frank Stearns]

Les Cargill > writes:

>I can't tell any difference for the material included.

>http://www.audiocheck.net/blindtests_16vs8bit_NeilYoung.php

No kidding! Lots of intentional distortion in the guitar (over-driven tubes and
transducers pushed well into non-linearity; or, should I say, even more-than-usual
non-linearity). Musically, it's in-your-face junk.

Something like windchimes or orchestra bells at a -30 or even lower, would be a much
more interesting test. And, we'd sure as hell want to do the tests with original PCM
files, not MP3s, which I assume these were given how quickly they loaded.

That said, I did some experimenting with this a few years back. All things being
equal with a good conversion, same sample rate, filtering, et al, the only real
difference between various bit depths should be the noise floor. That's it. No magic
or wistful thinking about some oblique nod toward analog's "infinite resolution"
nonsense.

I still record at 24 bit and mix and edit in 32 bit float because I don't have to
worry about headroom issues as multiple serial processing takes place, but that's
it. As usual, I suppose YMMV.

Frank
Mobile Audio

--

[JackA]

On Monday, November 9, 2015 at 2:02:27 PM UTC-5, geoff wrote:
> On 10/11/2015 7:32 a.m., Les Cargill wrote:
> > I can't tell any difference for the material included.
> >
> > http://www.audiocheck.net/blindtests_16vs8bit_NeilYoung.php
> >
>
>
> Neil Young ? During his grunge period(s) this isn't much difference !
>
> That's why he feels the need for this higher resolution Pogo thing
> !?!@@%#&!!!

At least we agree on one thing! It Pono, Pogo was a music group, I think!

Jack
>
> geoff

[[email protected]]

>
> That said, I did some experimenting with this a few years back. All things being
> equal with a good conversion, same sample rate, filtering, et al, the only real
> difference between various bit depths should be the noise floor. That's it. No magic
> or wistful thinking about some oblique nod toward analog's "infinite resolution"
> nonsense.
>

If dithering has been properly applied, then this is exactly what the theory says, it's the same as an analog noise floor.

So for once theory and practice DO agree. :-)

Mark

[JackA]

On Monday, November 9, 2015 at 3:50:30 PM UTC-5, Frank Stearns wrote:
> Les Cargill > writes:
>
> >I can't tell any difference for the material included.
>
> >http://www.audiocheck.net/blindtests_16vs8bit_NeilYoung.php
>
> No kidding! Lots of intentional distortion in the guitar (over-driven tubes and
> transducers pushed well into non-linearity; or, should I say, even more-than-usual
> non-linearity). Musically, it's in-your-face junk.

You mean it's worse than my stereo mixes??!! :)

I agree, I got 6/10!

Jack
>
> Something like windchimes or orchestra bells at a -30 or even lower, would be a much
> more interesting test. And, we'd sure as hell want to do the tests with original PCM
> files, not MP3s, which I assume these were given how quickly they loaded.
>
> That said, I did some experimenting with this a few years back. All things being
> equal with a good conversion, same sample rate, filtering, et al, the only real
> difference between various bit depths should be the noise floor. That's it. No magic
> or wistful thinking about some oblique nod toward analog's "infinite resolution"
> nonsense.
>
> I still record at 24 bit and mix and edit in 32 bit float because I don't have to
> worry about headroom issues as multiple serial processing takes place, but that's
> it. As usual, I suppose YMMV.
>
> Frank
> Mobile Audio
>
> --
> .

[Mike Rivers[_2_]]

On 11/9/2015 4:06 PM, JackA wrote:
> It Pono, Pogo was a music group, I think!

Pogo was a cartoon character. A possum.

--
For a good time, visit http://mikeriversaudio.wordpress.com

[Trevor]

On 10/11/2015 7:50 AM, Frank Stearns wrote:
> Les Cargill > writes:
>
>> I can't tell any difference for the material included.
>
>> http://www.audiocheck.net/blindtests_16vs8bit_NeilYoung.php
>
> No kidding! Lots of intentional distortion in the guitar (over-driven tubes and
> transducers pushed well into non-linearity; or, should I say, even more-than-usual
> non-linearity). Musically, it's in-your-face junk.
>
> Something like windchimes or orchestra bells at a -30 or even lower, would be a much
> more interesting test.


And if you really want to prove a point, record them at -60dB or less,
then increase the playback volume of course to compensate. That's how
they "proved" early digital systems could be worse than normally
recorded analog after all. No point comparing properly with tape of
course, since you'd barely hear the windchimes for tape noise when
recorded at -60dB :-)


> That said, I did some experimenting with this a few years back. All things being
> equal with a good conversion, same sample rate, filtering, et al, the only real
> difference between various bit depths should be the noise floor. That's it. No magic
> or wistful thinking about some oblique nod toward analog's "infinite resolution"
> nonsense.
> I still record at 24 bit and mix and edit in 32 bit float because I don't have to
> worry about headroom issues as multiple serial processing takes place, but that's
> it. As usual, I suppose YMMV.

That's the whole point though, 16 bit is more than enough for
distribution despite idiots like Neil Young's blathering. And why higher
resolution consumer systems have failed spectacularly to sell in any
number. But having some extra margin when recording doesn't hurt when
the cost is minimal these days.

Trevor.

[Scott Dorsey]

Frank Stearns > wrote:
>
>That said, I did some experimenting with this a few years back. All things being
>equal with a good conversion, same sample rate, filtering, et al, the only real
>difference between various bit depths should be the noise floor. That's it. No magic
>or wistful thinking about some oblique nod toward analog's "infinite resolution"
>nonsense.

IF it is a linear encoding. If you move from 16-bit linear PCM to 8-bit
linear PCM, the only thing that changes is the noise floor. If it is properly
dithered, you should have no problem hearing plenty of sounds well below the
noise floor too.

But... if it's 8-bit u-law encoding, there are going to be some other effects
that might be noticeable. In the real world when we see 8-bit audio systems
they are almost always using a nonlinear encoding.

I assume from the sound that the website is using linear encoding but they
don't actually say anywhere.

>I still record at 24 bit and mix and edit in 32 bit float because I don't have to
>worry about headroom issues as multiple serial processing takes place, but that's
>it. As usual, I suppose YMMV.

There's no reason not to, since computer power is cheap. I still do a bunch
of recording to 16-bit systems out of a general inertia, myself.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

>
>
> And if you really want to prove a point, record them at -60dB or less,
> then increase the playback volume of course to compensate. That's how
> they "proved" early digital systems could be worse than normally
> recorded analog after all.

only becasue these early digital systems did not use proper dither.

With proper dithering, there is no quantizing distortion, only random noise, same as anlog.

Mark

[JackA]

On Monday, November 9, 2015 at 10:17:58 PM UTC-5, Mike Rivers wrote:
> On 11/9/2015 4:06 PM, JackA wrote:
> > It Pono, Pogo was a music group, I think!
>
> Pogo was a cartoon character. A possum.

Hmm, can't say I...
Maybe in the days off Krazy Kat!!!
1916!! https://www.youtube.com/watch?v=iNrL_-jVvXo

Jack


>
> --
> For a good time, visit http://mikeriversaudio.wordpress.com

[Trevor]

On 11/11/2015 12:58 AM, wrote:
>> And if you really want to prove a point, record them at -60dB or less,
>> then increase the playback volume of course to compensate. That's how
>> they "proved" early digital systems could be worse than normally
>> recorded analog after all.
>
> only becasue these early digital systems did not use proper dither.

Which is my point, and was still better than analog when *properly*
compared anyway.


> With proper dithering, there is no quantizing distortion, only random noise, same as anlog.

No not "the same as analog", FAR less! :-)

Trevor.