Newbie question, mp3 quality
 

In message , Don Pearce
writes
On Sun, 23 Oct 2005 17:17:41 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 16:27:22 +0200, hwh wrote:

"David Jones" schreef in bericht
...
The mp3 I used was Pink Floyd, Speak to me. Stats according to AVICodec =
Audio : 1.68 MB, 192 Kbps, 44100 Hz, 2 channels, 0x55 = Mpeg-1 audio
Layer 3 (MP3), Supported

It is normal that a 192 kbps MP3 does not sound as good as a CD on
reasonable equipment.

gr, hwh

It should be very hard to tell the difference between a 192kbps MP3 and a
CD. There will probably only be a very points during a complete CD when the
differences will become audible, and they will be very brief, during
complex noise-like bits - particularly protracted sibilants - where the
perceptual coding fails. For most music you should hear no differnce.

d

If only! If Stewart were to offer £1000 to blind detect 192k MP3 and
original CDs I'd take him up on this one like a shot!
(If I could use my HD580s to listen to the recordings though, not
speakers).

Interesting. What is your blind discernment rate? And what do you find to
be the audible shortcomings?

d

I've never tried to deliberately do a controlled comparison test. I
noticed it when I first coded some tracks I already have on SP Minidisc,
initially at 160kbps (thinking it would be OK). As this proved to be
very poor, I re-recorded them at 192k which was a noticeable
improvement, but still poorer than the Minidisc. In reality I find 192
to be adequate for purpose when played back using the MP3 portable
player. It's just that the standard ATRAC Minidisc codec in my Sony
MDS-JB940 seems to be so good. (What's the bit rate for ATRAC in SP mode
I wonder?).

Now of course it may not be the fault of MP3 at all, but the freeware
codec on my PC. I'm using Freerip, and have nothing else to compare it
with.
--
Chris Morriss

Newbie question, mp3 quality
 

On Sun, 23 Oct 2005 18:21:43 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 17:17:41 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 16:27:22 +0200, hwh wrote:

"David Jones" schreef in bericht
...
The mp3 I used was Pink Floyd, Speak to me. Stats according to AVICodec =
Audio : 1.68 MB, 192 Kbps, 44100 Hz, 2 channels, 0x55 = Mpeg-1 audio
Layer 3 (MP3), Supported

It is normal that a 192 kbps MP3 does not sound as good as a CD on
reasonable equipment.

gr, hwh

It should be very hard to tell the difference between a 192kbps MP3 and a
CD. There will probably only be a very points during a complete CD when the
differences will become audible, and they will be very brief, during
complex noise-like bits - particularly protracted sibilants - where the
perceptual coding fails. For most music you should hear no differnce.

d

If only! If Stewart were to offer £1000 to blind detect 192k MP3 and
original CDs I'd take him up on this one like a shot!
(If I could use my HD580s to listen to the recordings though, not
speakers).

Interesting. What is your blind discernment rate? And what do you find to
be the audible shortcomings?

d

I've never tried to deliberately do a controlled comparison test. I
noticed it when I first coded some tracks I already have on SP Minidisc,
initially at 160kbps (thinking it would be OK). As this proved to be
very poor, I re-recorded them at 192k which was a noticeable
improvement, but still poorer than the Minidisc. In reality I find 192
to be adequate for purpose when played back using the MP3 portable
player. It's just that the standard ATRAC Minidisc codec in my Sony
MDS-JB940 seems to be so good. (What's the bit rate for ATRAC in SP mode
I wonder?).

Now of course it may not be the fault of MP3 at all, but the freeware
codec on my PC. I'm using Freerip, and have nothing else to compare it
with.

I tried an experiment once with both Lame and Fraunhofer at 192. Listening
to both, I was convinced I could hear nasty bits just occasionally. I noted
them down carefully and could identify them every time - MP3 artifacts for
sure. Then I listened to the original CDs, and damn it - they were still
there.

Maybe Freerip does have shortcomings; I'm not familiar with it, but
although I believe I have a good ear, I really can't hear a difference
between a CD and an MP3 at 192. 128 is a totally differnt matter, though,
and for many records I can hear the difference quite easily. Not all
though.

d

Newbie question, mp3 quality
 

In message , Don Pearce
writes
On Sun, 23 Oct 2005 18:21:43 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 17:17:41 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 16:27:22 +0200, hwh wrote:

"David Jones" schreef in bericht
...
The mp3 I used was Pink Floyd, Speak to me. Stats according to
AVICodec =
Audio : 1.68 MB, 192 Kbps, 44100 Hz, 2 channels, 0x55 =
Mpeg-1 audio
Layer 3 (MP3), Supported

It is normal that a 192 kbps MP3 does not sound as good as a CD on
reasonable equipment.

gr, hwh

It should be very hard to tell the difference between a 192kbps MP3 and a
CD. There will probably only be a very points during a complete CD when the
differences will become audible, and they will be very brief, during
complex noise-like bits - particularly protracted sibilants - where the
perceptual coding fails. For most music you should hear no differnce.

d

If only! If Stewart were to offer £1000 to blind detect 192k MP3 and
original CDs I'd take him up on this one like a shot!
(If I could use my HD580s to listen to the recordings though, not
speakers).

Interesting. What is your blind discernment rate? And what do you find to
be the audible shortcomings?

d

I've never tried to deliberately do a controlled comparison test. I
noticed it when I first coded some tracks I already have on SP Minidisc,
initially at 160kbps (thinking it would be OK). As this proved to be
very poor, I re-recorded them at 192k which was a noticeable
improvement, but still poorer than the Minidisc. In reality I find 192
to be adequate for purpose when played back using the MP3 portable
player. It's just that the standard ATRAC Minidisc codec in my Sony
MDS-JB940 seems to be so good. (What's the bit rate for ATRAC in SP mode
I wonder?).

Now of course it may not be the fault of MP3 at all, but the freeware
codec on my PC. I'm using Freerip, and have nothing else to compare it
with.

I tried an experiment once with both Lame and Fraunhofer at 192. Listening
to both, I was convinced I could hear nasty bits just occasionally. I noted
them down carefully and could identify them every time - MP3 artifacts for
sure. Then I listened to the original CDs, and damn it - they were still
there.

Maybe Freerip does have shortcomings; I'm not familiar with it, but
although I believe I have a good ear, I really can't hear a difference
between a CD and an MP3 at 192. 128 is a totally differnt matter, though,
and for many records I can hear the difference quite easily. Not all
though.

d

It's not the HF that's the problem, although at 53 years old my HF is
totally useless, perhaps even worse than most 53 year olds, with my
effective limit at about 10kHz.

The places where I hear the significant differences are where there's a
single clearly recorded female solo vocal which also has a
high-amplitude low-bass note at the same time. The vocal gets very
degraded when the bass note is present, although listening on the
Sennheiser HD580s to the original CD, there is no degradation of the
vocal during the bass note. It's almost like an intermodulation effect.
It's not perceivable (to me)) in complex music or in orchestral music,
probably due to the significant masking that this provides, the
small-scale female folk-ish material I like seems to be the easiest to
notice the effect on.
--
Chris Morriss

Newbie question, mp3 quality
 

"Chris Morriss" schreef in bericht
...
What bitrates do others use for MP3?

I use VBR with the high quality setting. Lame encodes the tracks at about
170 kbps for old 1960's tracks up to about 260 kbps for more difficult to
encode tracks from Oasis for example.
This method reduces the number of tracks on my 6 GB. iPod to about 680, but
I think this is enough.
By the way, if you have the possibility, try to use AAC, as that can sound
good at lower bitrates than MP3.
(I'd say slightly more than half the bitrate in comparison).

gr, hwh

Newbie question, mp3 quality
 

On Sun, 23 Oct 2005 19:43:31 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 18:21:43 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 17:17:41 +0100, Chris Morriss wrote:

In message , Don Pearce
writes
On Sun, 23 Oct 2005 16:27:22 +0200, hwh wrote:

"David Jones" schreef in bericht
...
The mp3 I used was Pink Floyd, Speak to me. Stats according to
AVICodec =
Audio : 1.68 MB, 192 Kbps, 44100 Hz, 2 channels, 0x55 =
Mpeg-1 audio
Layer 3 (MP3), Supported

It is normal that a 192 kbps MP3 does not sound as good as a CD on
reasonable equipment.

gr, hwh

It should be very hard to tell the difference between a 192kbps MP3 and a
CD. There will probably only be a very points during a complete CD when the
differences will become audible, and they will be very brief, during
complex noise-like bits - particularly protracted sibilants - where the
perceptual coding fails. For most music you should hear no differnce.

d

If only! If Stewart were to offer £1000 to blind detect 192k MP3 and
original CDs I'd take him up on this one like a shot!
(If I could use my HD580s to listen to the recordings though, not
speakers).

Interesting. What is your blind discernment rate? And what do you find to
be the audible shortcomings?

d

I've never tried to deliberately do a controlled comparison test. I
noticed it when I first coded some tracks I already have on SP Minidisc,
initially at 160kbps (thinking it would be OK). As this proved to be
very poor, I re-recorded them at 192k which was a noticeable
improvement, but still poorer than the Minidisc. In reality I find 192
to be adequate for purpose when played back using the MP3 portable
player. It's just that the standard ATRAC Minidisc codec in my Sony
MDS-JB940 seems to be so good. (What's the bit rate for ATRAC in SP mode
I wonder?).

Now of course it may not be the fault of MP3 at all, but the freeware
codec on my PC. I'm using Freerip, and have nothing else to compare it
with.

I tried an experiment once with both Lame and Fraunhofer at 192. Listening
to both, I was convinced I could hear nasty bits just occasionally. I noted
them down carefully and could identify them every time - MP3 artifacts for
sure. Then I listened to the original CDs, and damn it - they were still
there.

Maybe Freerip does have shortcomings; I'm not familiar with it, but
although I believe I have a good ear, I really can't hear a difference
between a CD and an MP3 at 192. 128 is a totally differnt matter, though,
and for many records I can hear the difference quite easily. Not all
though.

d

It's not the HF that's the problem, although at 53 years old my HF is
totally useless, perhaps even worse than most 53 year olds, with my
effective limit at about 10kHz.

The places where I hear the significant differences are where there's a
single clearly recorded female solo vocal which also has a
high-amplitude low-bass note at the same time. The vocal gets very
degraded when the bass note is present, although listening on the
Sennheiser HD580s to the original CD, there is no degradation of the
vocal during the bass note. It's almost like an intermodulation effect.
It's not perceivable (to me)) in complex music or in orchestral music,
probably due to the significant masking that this provides, the
small-scale female folk-ish material I like seems to be the easiest to
notice the effect on.

Well, that most certainly isn't a normal MP3 artifact - it just doesn't do
that. So I don't know exactly what is going on for you, but I would suggest
you find yourself an alternative MP3 codec; there are plenty out there.

As for hearing, at 55 I can still hear comfortably to about 16kHz, which
may be better than average for my age - I will have to check.

d

Newbie question, mp3 quality
 

In message , Don Pearce
writes
Well, that most certainly isn't a normal MP3 artifact - it just doesn't do
that. So I don't know exactly what is going on for you, but I would suggest
you find yourself an alternative MP3 codec; there are plenty out there.

As for hearing, at 55 I can still hear comfortably to about 16kHz, which
may be better than average for my age - I will have to check.

d

I doubt if I could hear 16kHz even as a child, but I had various ENT
problems at around the age of 5 so I guess that didn't help.

I'll try a few other codecs and also check carefully by comparing the
sound card output of the pc with the MP3 player, although I don't think
it's the player's analogue output because the effect got much less when
I went from 160 to 192kbps.
--
Chris Morriss

Newbie question, mp3 quality
 

The places where I hear the significant differences are where there's a
single clearly recorded female solo vocal which also has a
high-amplitude low-bass note at the same time. The vocal gets very
degraded when the bass note is present, although listening on the
Sennheiser HD580s to the original CD, there is no degradation of the
vocal during the bass note. It's almost like an intermodulation effect.
It's not perceivable (to me)) in complex music or in orchestral music,
probably due to the significant masking that this provides, the
small-scale female folk-ish material I like seems to be the easiest to
notice the effect on.

I can't see, or hear!, how any system that throws info away can be
totally transparent unless it can code the original info in the same
space....
--
Tony Sayer

Newbie question, mp3 quality
 

On Mon, 24 Oct 2005 09:15:48 +0100, tony sayer wrote:

The places where I hear the significant differences are where there's a
single clearly recorded female solo vocal which also has a
high-amplitude low-bass note at the same time. The vocal gets very
degraded when the bass note is present, although listening on the
Sennheiser HD580s to the original CD, there is no degradation of the
vocal during the bass note. It's almost like an intermodulation effect.
It's not perceivable (to me)) in complex music or in orchestral music,
probably due to the significant masking that this provides, the
small-scale female folk-ish material I like seems to be the easiest to
notice the effect on.

I can't see, or hear!, how any system that throws info away can be
totally transparent unless it can code the original info in the same
space....

Then you probably need to do some reading.

d

Newbie question, mp3 quality
 

"Don Pearce" wrote in message

On Sun, 23 Oct 2005 19:43:31 +0100, Chris Morriss wrote:


The places where I hear the significant differences are
where there's a single clearly recorded female solo
vocal which also has a high-amplitude low-bass note at
the same time. The vocal gets very degraded when the
bass note is present, although listening on the
Sennheiser HD580s to the original CD, there is no
degradation of the vocal during the bass note. It's
almost like an intermodulation effect. It's not
perceivable (to me)) in complex music or in orchestral
music, probably due to the significant masking that this
provides, the small-scale female folk-ish material I
like seems to be the easiest to notice the effect on.

Well, that most certainly isn't a normal MP3 artifact -
it just doesn't do that. So I don't know exactly what is
going on for you, but I would suggest you find yourself
an alternative MP3 codec; there are plenty out there.


Sounds like clipping in the headphone amp - tubed? ;-)

Newbie question, mp3 quality
 

"Don Pearce" wrote in message

On Mon, 24 Oct 2005 09:15:48 +0100, tony sayer wrote:

The places where I hear the significant differences are
where there's a single clearly recorded female solo
vocal which also has a high-amplitude low-bass note at
the same time. The vocal gets very degraded when the
bass note is present, although listening on the
Sennheiser HD580s to the original CD, there is no
degradation of the vocal during the bass note. It's
almost like an intermodulation effect. It's not
perceivable (to me)) in complex music or in orchestral
music, probably due to the significant masking that
this provides, the small-scale female folk-ish material
I like seems to be the easiest to notice the effect on.

I can't see, or hear!, how any system that throws info
away can be totally transparent unless it can code the
original info in the same space....

Then you probably need to do some reading.

Agreed. The fact that MP3 works at all is proof that masking
is a very powerful effect.

The irony of audiophools doing so much listening via
perceptual coders should not be lost on us of the more
rational persuasion.

They seem to still claim to hear that which can not be
measured, but now they are not hearing the loss of that
which can be clearly measured.

Note:

http://www.dedicatedaudio.com/inc/sdetail/4345?noc=true