===================================

Here are my results:

File 2 sounds the best of the five.
I don't know if its the 24 bit one, but I like it the best.

Files 1 and 3 are tied for second place.
I really cannot distinguish any differences between the two.

File 4 takes third place, because as the .wav files reached its
end in playing
I could hear hiss...white noise... whatever you call it...

File 5 takes fourth (last) place, because it had even louder
hiss at the end than file 4,
therefore it disturbed me the most.

I am very curious to know how I did.

===================================

File 5 is worse (the noise can be heard all along, not only at
the end). I didn't take the time to listen to File 4, that was
bad with headphones This time, File 2 seems better (sharper)
than file 1

ABX test (without DSP) : 7/8  (yesterday, with headphones, File
2 seemed also better than file 3, but I got 5/8 only at ABX, so
I didn't report it, it was a vague feeling) Trying 16 ABX at
lower volume (because of neighbours) : 10/16 only

File 2 seems also sharper than File 3
ABX : 10/16 only

Conclusion :

Best : File 2
then File 1 and 3
then file 4
Worst : File 5

===================================

File 1 - Very clean throughout, even at tail of wav - no hiss
File 2 - Less clean than File 1 at tail of wav - ever so slight
hiss
File 3 -
File 4 - Guitar string buzz on frets more grainy than File 1
File 5 - Lots of hiss

My guess - there in the order of highest to lowest quality - you
haven't mixed them up at all.

===================================

I did the test very quickly not in the best environment (noisy
computer and the infamous sb live as soundcard). The only one i
could really hear a difference on was nr 5 which had more noise
at the end than the others. Then I _thought_ i heard some noise
in file 1 as well, but much less. I'm very unsure about this
one, but I didn't bother to abx it. The rest sounded the same to
me.

Nice test by the way!

===================================

I started by listening to all the files in order to see whether
I could pick out the two worst files, which turned out to be not
too hard.

5 was obviously the worst, with noisy grunge apparent especially
at the end.  There was no need to perform an ABX comparison
(http://pcabx.com/) because the difference from the others was
very obvious.

4 was next worse, but much better than 5.  There is slightly
more hiss than the others at the end.  In an ABX comparison
against 1, I scored 8 out of 8 correct (probability of getting
those results by chance: 0.004)

The next three are all bunched together, and I don't think I can
tell any of them apart.  On quiet music, I can usually hear the
hiss associated with 13 bits, but this sample has a lot of
signal, so I don't.

Equipment:  M-Audio Audiophile 2496 card, Grado SR325 headphones
plugged in directly to the card

I have a similar test of bit truncation/dither at
http://ff123.net/madchallenge.html, where I am able to hear
differences from a reference 16 bit file with truncation out to
14 bits, and with dither out to 15 bits (I heard the extra hiss
from the dither at 15 bits).

===================================

I downloaded the bit-test files and took a listen.  My ratings:

File 2: best all around
File 3: sounds like number 2 but with the life sucked out
File 1: somehow odd
File 4: noisey
File 5: very noisey

So I guess it would be:

File 2: dithered
File 3: truncated
File 1: "13" bits
File 4: "11" bits
File 5: "9" bits

Of course it's hard to believe that anyone would go by the
numbers rather than just using whatever sounds best to them on
their particular system for a particular recording.

===================================

Great test.  To me the advange of 24 bits is recording at lower
peak levels to avoid clipping.  When recording at 24 bit I get
more like 16 bits on dynamic range/noise, which in turn would
sound more like 8 bits if recorded at 16 bits.  That would be
closer to your +/- 40 db test. I hope I got that one right.

Anyway, my guess is:
File2 dither
File3 truncated
File1 +/- 20 db
File4 +/- 30 db
File5 +/- 40 db

Files 4 and 5 sound noisier to me.  1 and 3 sound almost
identical. 3 just feels like it has slightly better dynamic
range, but not better sounding. File 2 just sounds harmonically
different to me, but maybe the result of processing, rather than
the lack of processing. So I my have missed the boat on it.

Reminds me of the Coke vs. Pepsi tests.  I like Coke better, but
there something about tasting them back to back that makes Coke
taste like crap. Hope you're not getting buried too deep in
emails.

Thanks,

===================================

I saw your petition for your "24-bit test" on ProRec.  Here's my
ranking:

File1 - 3
File2 - 4
File3 - 5
File4 - 2
File5 - 1

It's really hard to tell the difference between some of them.  I
hope I was close!

Thanks for your contributions to ProRec and your website.  I'm
just a hobbyist musician/recordist but I really enjoy all of the
discussion at ProRec.

===================================

these are my guesses:
1: trunkated
2: -20dB
3: -40dB
4: -30dB
5: dithered

ps. I listened through a soundblaster, a really long and bad
cable to my home stereo, into a pair of AKG 240 headphones.

===================================

1 11 bit
2 13 bit
3 16 bit dithered
4 16 bit truncated
5 9 bit


Thanks, that was fun.  I've just gotten interested in the magic
of music bits.

===================================

Why are people so scared!!

My best is 2341  and a long way behind 5 (47 years old ears)
My engineer (24) goes 3214 and a long way behind 5

Do please write with how many people have done the challenge and
the correct result, a beer depends on it!!

Great fun!!

Regards

===================================

Hi, great thing that 24-bit test ;)
Yeah I don't have trained ears but anyway my guesses:
(from worst to best) file5, file4, file3, file1, file2
Did I get *any* of them right?
Thanks for reply,

===================================

I listened at work on cheap walkman style headphones and could
not tell much difference.   Files 3 and 5 seemed a little worse
than the 1 2 & 4.

If I remember to take the files home, I'll try again in the
studio.

===================================

best sounding to worse, and i labeled what science would say is
the reason for the differences let me know how right or wrong i
am...this is a real difficult test for me considering i listen
to hip hop and i don't really know how a GOOD guitar is
supposeta sound

1. file1.wav-is 16bit dithered...
2. file5.wav- 16 bit truncated
3.  file4.wav- 16 bit volum 20 db
4. file5- 16 30db
5. file3- 16 40

===================================

Interesting.  Even if I'm way off base, it was a good listening
excercise.
Here are my guesses.

File 1 - dithered
File 2 - truncated
File 3 - 11 bits
File 4 - 13 bits
File 5 - 9 bits

How'd I do?

===================================

Great site (especially liked the part were you stole those damn
signs!)

As for the 24 bit test, great idea.  I first listened to them on
my Cambridge soundworks digital speakers ($50) and then on my
Event PS6's ($500).  I am admittedly new to recording so I may
not have known what to listen for.

There were some extremely slight differences that I noticed but
so slight as to leave me thinking I was make it up in my head.

The only definite disparity was in the hiss level in File 5
which I am assuming is the 9 bit.  The answers below are based
more on a "gestalt" thing more than evidence.  If you told me
that 1-4 were the same I would believe you.

Here are my guesses:

File 1: sounded "best" 24 to 16
File 2: 13 bit
File 3: 11 bit
File 4: 16 bit
File 5: 9 bit

I am not getting rid of my DSP factory quite yet:)

Thanks for the test.

===================================

Unfortunately, I don't have the time (or a good headset at work)
to really knock this out; but File4.wav seemed to sound the best
to me.

Which was which?

===================================

Thanks for creating this challenge. I'd never directly compared
files of varying bit depth before, though I had serious doubts
that 24 bit makes much of a difference in practice. Your test
confirmed that for me - I can't hear any difference in the
files, at least over (not the greatest) headphones here at work.
When I thought I could hear something different, I went back and
the effect was not repeatable. I suspect that is because the
brain does much filling-in and interpolating of everything we
hear so it's easy to convince yourself you're hearing some
difference even A/Bing the same sample.

One thing I noticed in your test description was this statement:

But is recording at 24 bits really worth the added overhead of
half again more disk space and CPU power?

I believe that most pro audio apps (Sonar is what I use) convert
everything to 32-bit float numbers before processing, so I don't
think there's much additonal CPU overhead for 24 bit. There
certainly is significant disk overhead and storage space,
however, and that's the main reason I stick with 16/44.1K
recording.

Even though I wimped out on the guessing, could you tell me
which sample was which? :)

Thanks again, Ethan!

===================================

First let me say that I record only in 16bit. I cannot justify
running out of tracks so I can have the knowledge I recorded in
20 or 24 bit. 2nd I mean this as friendly dialog, hoping to
learn something myself, But I think your test as you stated it
is flawed to the those with the 24bit point of view. From what I
have been able to gather, one of the main reasons to record in
24 bit is for the ability to process the signal many times with
little distortion (NOT THD but artifacts, noise, waveform shape
etc). SO if I were trying to convert or prove wrong the 24bit
crowd, I would have done the test this way.

Same recording setup, but actually send the signal to a 24 bit
and 16 bit recording device at the same time. save those files.
on a copy of the two files add processes to them. comp, reverb,
chorus, cut 20db, normalize then output all the processes to a
16bit track. both dithered and truncated.

your test doesn't seem to be realworld because it ignores all
the things we tend to do to a signal once it is recorded.

Ready for the return volley, bracing for impact! :-)

===================================

I'm listening at a PC with an SB Live! on "Walkman" headphones.
I'm having trouble distinguishing differences between the files.
Main problem is that I can only listen to them one at a time, so
it's hard to make a real comparison.   At home on my high
resolution system, it might be another matter, but here and now
about the best I can do is THINK I can hear a bit more detail in
#5 than in #1!   Sorry, but I can't "rank" them!

Nice playing, BTW!

===================================

I can't tell any difference although I sometimes like wav2 and
wav5 the best and maybe 4 the worst.

so I would pick best to worst this order.

2, 5, 1, 3, 4

===================================

To tell the truth couldn't hear the difference. Wav # 1 sounded
worst, but the others all sounded the same to me.  Maybe this
points out a problem with my monitoring system...hmmm.

===================================

Thank you, thank you, thank you.

First, I haven't taken the challenge yet, since I just read the
note and I am forty-five minutes from my equipment, but I assure
you I will download and listen this afternoon.

Second, I have spoken to the brick wall many times that I simply
cannot tell the difference between 24 bit and 20 bit resolution
on acoustic recording in any of the gear I've tried. I just
can't. Maybe my golden ears have turned to brass...

As for 20 bit vs. 16 bit, I'm going to guess there should be no
audible difference, given careful recording. Below that bit
rate, I am very interested in the listening test.

===================================

Here are my almost certainly erroneous answers:

1. = -20dB and back.
2. = Truncated.
3. = Dithered.
4. = -30dB and back.
5. = -40dB and back.


4 and 5 were easy of course. Then it started to get tricky. I
know that I like #3 the best (it's sounds better pretty much all
way around), and that I find #1 to be the dullest. But I'm not
sure if "dull" just means it lacks truncation errors! Anyway, I
went with my heart.

This has been a real eye-opener for me personally. Many thanks
for this great test!

===================================

Unfortunately I haven't gotten my sound card or monitors yet, so
I am listening on the on board sound on my motherboard which is
quite bad (very noisy, with pops and clicks) and a pair of $40
headphones.  However I thought I would give it a try anyway.

I was able to distinguish just the 2 worst files

#5 is the worst - 9 bit (You can easily hear the noise in this
one.)
#4 is the second worst - 11 bit (you can still hear the noise
here as well.)

As for the rest, I am going to blame my listening environment --
which sucks =).  Seriously, I think with better sound card and
speakers I could pick out the 13 bit one.  However, your test is
not good at all for distinguishing the difference in theoretical
dynamic range, because the guitar strumming is constantly loud.
The final bits do always add detail in theory, but it really
doesn't make an audable difference unless you have lots of
dynamics in the music (especially quiet parts).

Also, THE major advantage of 24bits is in the signal processing
side of things.  You make your calculations at 24 (or higher)
bits in order to not get distortion from rounding errors.  A
better test of the 24 bit advantage would be to run a bunch of
plugins on a file that are calculating at 16 bits (rather than
the normal 32).  Of course, no modern app will do that
anymore--but you get my point.