1 kHz missing?

[Mxsmanic]

I was looking at the spectrum of a song and saw that there's nothing between
about 10 kHz and 11 kHz, as if that had been chopped out. I assume it's
deliberate. Why would anyone remove a 1 kHz band from the recording?

[Ron Capik]

How did it sound ?

Later...
Ron Capik
--

[Mike Rivers]

On 2/8/2012 8:39 PM, Mxsmanic wrote:

Your spectrum analyzer probably doesn't have enough
resolution to show that. What's there above 10 kHz anyway?

--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[Carey Carlan]

Mxsmanic <mxsm...@gmail.com> wrote in
news:cs86j7ppo1uo5radl...@4ax.com:

To remove a whine (what at 60 Hz would be called a hum) in the recording.

[Mark]

what song?
Mark

[Frank Stearns]

You weren't looking at an MP3, were you? All sorts of strange holes come and go in
the HF, depending on the bit rate and the codec. Were you seeing information between
16K and 20K, or was it gone too?

Frank
Mobile Audio
--
.

[Mxsmanic]

Ron Capik writes:

> How did it sound ?

It was a crummy mono recording of a song that I found somewhere, but I didn't
notice anything unusual about it other than the generally poor quality. I do
not have trained ears, though.

[Mxsmanic]

Mike Rivers writes:

> Your spectrum analyzer probably doesn't have enough
> resolution to show that. What's there above 10 kHz anyway?

It was a very obvious blank space between two frequencies, and I don't see
that on other recordings, so I don't think it was the software. It looked like
someone had removed it.

[Mxsmanic]

Carey Carlan writes:

> To remove a whine (what at 60 Hz would be called a hum) in the recording.

What sorts of things would cause a 10 kHz whine?

[Mxsmanic]

Mark writes:

> what song?

I think it was "Street Life," at least that was the refrain in the lyrics. I
looked it up and apparently it is by The Crusaders.

The recording also showed evidence of the "loudness wars" effect that I've
been reading about (which is what I was originally looking into).

I've heard of bizarre copy-protection schemes that worked by removing part of
the sound spectrum from a recording, and I thought it might be something like
that, although I don't see how removing part of the sound would protect a
recording. I remember some negative buzz about the technique many years ago, I
don't know if it was ever widely used.

[Mxsmanic]

Frank Stearns writes:

> You weren't looking at an MP3, were you? All sorts of strange holes come and go in
> the HF, depending on the bit rate and the codec. Were you seeing information between
> 16K and 20K, or was it gone too?

Well, I stored it into a MP3 before looking at it, but I did that with some
other songs before looking at them as well, and I didn't see this hole.

A similar hole starts at 21 kHz, but I think I've seen that on all the samples
I looked at.

I extended the scale on the spectrum and the hole is actually a valley (or
abyss) that is very abruptly 50 dB lower than the rest of the spectrum.

[Meindert Sprang]

"Mxsmanic" <mxsm...@gmail.com> wrote in message
news:gd47j71osdicgl8pl...@4ax.com...

Mmm..... what is the sample rate of that file? Could be some sort of
aliasing.

Meindert

[anahata]

On Thu, 09 Feb 2012 10:21:43 +0100, Mxsmanic wrote:

> It was a crummy mono recording of a song that I found somewhere, but I
> didn't notice anything unusual about it other than the generally poor
> quality. I do not have trained ears, though.

Copied from a tape cassette played on a machine with misaligned head,
peahaps? HF nulls at various frequencies are typical when that happens.

--
Anahata
ana...@treewind.co.uk --/-- http://www.treewind.co.uk
+44 (0)1638 720444

[William Sommerwerck]

Back in the 90s (I think), I believe there was a proposal to "notch"
recordings in this way, as a form of copy protection.

[Les Cargill]

1kHz was frequently used as the inband "advance" signal for slide
projectors. Might have been something like that; could even have been
a transcription error ( somebody forgot to turn the filter off ).

--
Les Cargill

[Peter Larsen]

Best guess: an alias tone from digitizing due to a bias leak from a
misaligned taperecorder notched out.

Kind regards

Peter Larsen

[Mxsmanic]

anahata writes:

> Copied from a tape cassette played on a machine with misaligned head,
> peahaps? HF nulls at various frequencies are typical when that happens.

Cool ... I didn't know that.

[Mxsmanic]

Meindert Sprang writes:

> Mmm..... what is the sample rate of that file? Could be some sort of
> aliasing.

It's a MP3 file, so 48 kHz I guess.

[Mxsmanic]

William Sommerwerck writes:

> Back in the 90s (I think), I believe there was a proposal to "notch"
> recordings in this way, as a form of copy protection.

So I'm not imagining that. I've forgotten the details though. I think the
playback equipment was supposed to look for the notch, or something. I was
adamantly opposed to it, as I am opposed to anything that diminishes sound or
image quality. I don't use digital watermarks for the same reason.

[Mike Rivers]

On 2/9/2012 4:28 AM, Mxsmanic wrote:

> I've heard of bizarre copy-protection schemes that worked by removing part of
> the sound spectrum from a recording, and I thought it might be something like
> that, although I don't see how removing part of the sound would protect a
> recording.

CopyCode was proposed, in about 1986, to be required by law
in the US as a means of identifying a copyright recording.
It worked by removing frequencies that presumably would be
masked anyway, and detecting the lack of those frequencies.
Congress tasked the (then named) National Bureau of
Standards to test the process to determine if it was
actually transparent to listeners. I participated in that
test myself. I could identify the difference with and
without CopyCode about 80% of the time, as could many others
in my test group (we were from the local AES chapter, not
that it means we were critical listeners). Congress
eventually nixed it, and it was abandoned.

You surely didn't hear a recording with this means of copy
protection. It just plain didn't work.

[Mxsmanic]

Mike Rivers writes:

> CopyCode was proposed, in about 1986, to be required by law
> in the US as a means of identifying a copyright recording.
> It worked by removing frequencies that presumably would be
> masked anyway, and detecting the lack of those frequencies.
> Congress tasked the (then named) National Bureau of
> Standards to test the process to determine if it was
> actually transparent to listeners. I participated in that
> test myself. I could identify the difference with and
> without CopyCode about 80% of the time, as could many others
> in my test group (we were from the local AES chapter, not
> that it means we were critical listeners). Congress
> eventually nixed it, and it was abandoned.

So you saved us!

> You surely didn't hear a recording with this means of copy
> protection. It just plain didn't work.

That's fine with me. I'm opposed to any copy-protection system that corrupts
the data it's supposed to be protecting.

[Mxsmanic]

Peter Larsen writes:

> Best guess: an alias tone from digitizing due to a bias leak from a
> misaligned taperecorder notched out.

That seems plausible, insofar as I can understand it.

[harry hall]

This wasn't a 1kHz notch but a 1kHz tone on a second track on the
tape. There were a number of systems using various tones to allow
multiple projector use and complex visual effects.

Harry

[Steve Maki]

I have the CD (1st issue from 1984) and it shows no notch. Incidently it (at least the
version by the Crusaders) is full stereo and is NOT crunched at all. It was recorded in
1979 so probably started it's life as an analog recording, and a decent sounding one. So
what you have has been butchered beyond belief for unknown reasons.

-Steve Maki

[Les Cargill]

Yes. I am saying that perhaps somebody had equipment set up to notch
that out. I don't remember if there was a second track or if it was
simply inband ( and it probably varied from system to system ).

--
Les Cargill

[Mxsmanic]

Steve Maki writes:

> I have the CD (1st issue from 1984) and it shows no notch. Incidently it (at least the
> version by the Crusaders) is full stereo and is NOT crunched at all. It was recorded in
> 1979 so probably started it's life as an analog recording, and a decent sounding one. So
> what you have has been butchered beyond belief for unknown reasons.

I think it was off YouTube, and there are about ten million different videos
of that song.

[Mxsmanic]

Steve Maki writes:

> I have the CD (1st issue from 1984) and it shows no notch. Incidently it (at least the
> version by the Crusaders) is full stereo and is NOT crunched at all. It was recorded in
> 1979 so probably started it's life as an analog recording, and a decent sounding one. So
> what you have has been butchered beyond belief for unknown reasons.

I'd like to get the CD, although now, after what I've read of loudness wars,
I'm afraid to buy a CD that may turn out to be terrible. And $20 is a lot to
pay for one song.

[harry hall]

On Fri, 10 Feb 2012 17:43:56 -0600, Les Cargill

The original poster said that a 1kHz "notch" appeared between 10kHz
and 11kHz, therefore removing at 1kHz tone bleeding from a second
track could not cause this.

[Mxsmanic]

harry hall writes:

> The original poster said that a 1kHz "notch" appeared between 10kHz
> and 11kHz, therefore removing at 1kHz tone bleeding from a second
> track could not cause this.

Yes. Everything betwene 10 kHz and 11 kHz was dozens of dB below the rest.

[Scott Dorsey]

harry hall <spar...@talk.net> wrote:
>
>The original poster said that a 1kHz "notch" appeared between 10kHz
>and 11kHz, therefore removing at 1kHz tone bleeding from a second
>track could not cause this.

Yes, but then the original poster said it had been downloaded through
Youtube, so it's been through at least one generation of the most horrible
perceptual encoding that can be imagined.

The amazing thing isn't that some band is missing but that there is
any left.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Mxsmanic]

Scott Dorsey writes:

> Yes, but then the original poster said it had been downloaded through
> Youtube, so it's been through at least one generation of the most horrible
> perceptual encoding that can be imagined.

Granted, but that doesn't explain why YouTube would remove only one band of
frequencies. And other videos on YouTube aren't like that.

[Scott Dorsey]

Okay, here's an analogy.

You put your head under a piledriver. It comes down, pieces of your head
are torn off.

Now, someone else does the same thing, in more or less the same position,
but different pieces come off.

That's how the encoding works. It's putting your music under a pile driver
and ripping bloody gobbets of it off. However, when gets removed in one
instance isn't the same as what gets removed in another instance. And, it
will vary from moment to moment in the same recording.

Sometimes you can look at see several bands totally eliminated. Sometimes
you can see a band eliminated and replaced with a single narrowband signal,
especially on the low end and high end.

Depends a lot on how the encoder is set up and what parameters were picked
for encoding.

[Jeff Henig]

Scott Dorsey wrote:

> That's how the encoding works. It's putting your music under a pile driver
> and ripping bloody gobbets of it off. However, when gets removed in one
> instance isn't the same as what gets removed in another instance. And, it
> will vary from moment to moment in the same recording.

Scott, am I detecting a little bit of frustration with violent intent?

What do you see when I show you this ink blot? ;^)

---Jeff

[Peter Larsen]

Depends on the perceptual encoding, below the notch you have what remains of
the music and above it you have - in my opinion, not something I can claim
to be a A Fact -some white noise that is inserted to represent the treble
they didn't have spare bits for.

Kind regards

Peter Larsen

[Norbert Hahn]

Mxsmanic <mxsm...@gmail.com> wrote:

>Carey Carlan writes:
>
>> To remove a whine (what at 60 Hz would be called a hum) in the recording.
>
>What sorts of things would cause a 10 kHz whine?

AM broadcast with two channels interfering.

Norbert