IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Loud 16kHz tone in analog recordings, Where does it come from?
Gecko
post Jan 6 2004, 13:43
Post #1





Group: Members
Posts: 937
Joined: 15-December 01
From: Germany
Member No.: 662



I recently accquired the soundtrack to the movie "American Beauty" (score version). Wonderful music! However on almost every track there is a very loud constant tone at about 16kHz. It's volume varies according to what instruments are playing. For example there's only a piano playing and when the string instruments start playing the tone gets louder. I guess the instruments were recorded individually and then mixed together. Often there is also a tone at ca. 18.75kHz, but either my ears have degenerated or it is masked.

On one part they're playing a flute from the sampler at a lower note than it was recorded, and the original 16kHz tone moves down to almost 14kHz.

The 16kHz tone is seriously annoying and I just wanted to know where it comes from. I've heard it (and also seen it in spectral view of Cool Edit) on several (usually classic) recordings as well. Don't they hear or (if their hearing isn't as good) see this during mastering?

On a side note: the CD isn't mastered very well. While usually being a remarkably soft soundtrack, some loud bass notes are heavily clipped (and later scaled down again, probably to match the overall volume of the CD)
Go to the top of the page
+Quote Post
2Bdecided
post Jan 6 2004, 14:33
Post #2


ReplayGain developer


Group: Developer
Posts: 5101
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



Are you sure it's not around 15.5kHz - 15.625kHz or 15.734kHz?

These are the line frequencies of PAL and NTSC TV systems, and this signal can usually be heard coming from the back of a TV set. Any TV monitor in the recording room can radiate this sound, which is easily picked up by microphones.

I always found the sound annoying when I was a kid, but can't hear it now. Likewise the recording engineer was probably unaware of the problem.

Many acoustic/orchestral recordings have a slight line visible at 15.5kHz in the spectral view for this reason. Sometimes it's much louder than others, and will annoy more people.

You can easily filter it out using Cool Edit Pro - set up a notch filter using the FFT filter. It will probably sound much better to you then.

Hope this helps.

Cheers,
David.
Go to the top of the page
+Quote Post
smack
post Jan 6 2004, 15:07
Post #3





Group: Members
Posts: 187
Joined: 16-January 02
Member No.: 1046



The high-frequency tones may result from EMI (Electromagnetic Interference) during analog recording. This would match your observations about the bad mastering of the CD, by the way. The fact that the tone is at about 16 kHz seems to encourage this idea: the "line frequency" (a.k.a. horizontal frequency) of a TV signal is about 15.6 kHz.

I know that from my own experiences. Several years ago I wrote a digital oscilloscope program for my Amiga computer (R.I.P. - sigh!). When I digitized the audio signal from my TV set's analog output there was always a tone at 15.6 kHz clearly visible. Even when I looked at the analog signal from my CD player I could see the 15.6 kHz tone - but only while the TV set was on! Obviously the cables I used were poorly shielded. rolleyes.gif

edit: David was faster...

This post has been edited by smack: Jan 6 2004, 15:09
Go to the top of the page
+Quote Post
Gecko
post Jan 6 2004, 16:13
Post #4





Group: Members
Posts: 937
Joined: 15-December 01
From: Germany
Member No.: 662



You guys seem to be right. I knew about the pilot tone of TVs being in that range, but I didn't think this would be recorded during a professional recording session.

Here is a picture of that range (FFT 8192 bands and blackmann-harris window). Interestingly there seem to be two signals oscilating synchronously in frequency at about 0.8Hz.

I get good results using a 5th order Butterworth bandstop filter in the range of 15.5kHz - 15.775kHz.

Do you have any info on the other tone? It is centered at 18440Hz.
Go to the top of the page
+Quote Post
smack
post Jan 6 2004, 16:29
Post #5





Group: Members
Posts: 187
Joined: 16-January 02
Member No.: 1046



QUOTE (Gecko @ Jan 6 2004, 04:13 PM)
Here is a picture of that range (FFT 8192 bands and blackmann-harris window). Interestingly there seem to be two signals oscilating synchronously in frequency at about 0.8Hz.

So it was an NTSC TV signal (or maybe two) that interfered here!

QUOTE
Do you have any info on the other tone? It is centered at 18440Hz.

Sorry, I have no idea. Perhaps it's just the TV tone played at a higher rate?
Go to the top of the page
+Quote Post
KikeG
post Jan 6 2004, 16:57
Post #6


WinABX developer


Group: Developer
Posts: 1578
Joined: 1-October 01
Member No.: 137



Maybe it could also be a computer monitor set to whatever resolution and vertical refresh rate. I've also seen it in some cds, but at inaudible levels, just seen on a FFT analysis.
Go to the top of the page
+Quote Post
smack
post Jan 6 2004, 17:48
Post #7





Group: Members
Posts: 187
Joined: 16-January 02
Member No.: 1046



QUOTE (KikeG @ Jan 6 2004, 04:57 PM)
Maybe it could also be a computer monitor set to whatever resolution and vertical refresh rate.

That's unlikely, if you ask me. Look at this:
NTSC: 525 lines * 29.97 fps => 15.734 kHz
PAL: 625 lines * 25 fps => 15.625 kHz

The 18.44 kHz tone is too low for computer monitors. They usually operate at much higher horizontal frequencies, i.e. 31 to over 100 kHz - depending on resolution and refresh rate.
Go to the top of the page
+Quote Post
MugFunky
post Jan 6 2004, 17:53
Post #8





Group: Members
Posts: 493
Joined: 28-December 03
From: Melbourne, Aus
Member No.: 10767



that 18.75k tone is a bit mysterious... i can't seem to find a monitor refresh rate and resolution that would produce it (not like tvs for example where it's simple to determine by dividing the freq by the framerate, like PAL = 15625/25 = 625 lines or NTSC = 15734.25/29.97 = 525 lines).

filter it out anyway though - it's wasted bits in whatever encoder you use (possibly even mp3 at high rates)

it really annoys me that producers/sound guys assume these sounds don't exist just because they can't hear them. prime example is the A/V lab at my uni - there's a whopping great trinitron monitor in there that is so incredibly loud that even an 80 year old could probably hear it.

producers should know better IMHO. grrr.

This post has been edited by MugFunky: Jan 6 2004, 17:54
Go to the top of the page
+Quote Post
Lev
post Jan 6 2004, 18:05
Post #9





Group: Members
Posts: 524
Joined: 7-November 02
From: Gloucester, UK
Member No.: 3716



The unfortunate thing is that most of Thomas Newman's soundtracks that I have found exhibit this - possibly a lot of other stuff I listen to has, but that stuff is generally louder, so I'd be hard pressed to notice.

Some of the bass notes are incredibly clipped, having watched the DVD again recently, it seems to sound better than the released CD soundtrack (although maybe I was just in a good mood as my favourite film was on) wink.gif


--------------------
http://www.megalev.co.uk
Go to the top of the page
+Quote Post
KikeG
post Jan 6 2004, 18:19
Post #10


WinABX developer


Group: Developer
Posts: 1578
Joined: 1-October 01
Member No.: 137



QUOTE (smack @ Jan 6 2004, 05:48 PM)
QUOTE (KikeG @ Jan 6 2004, 04:57 PM)
Maybe it could also be a computer monitor set to whatever resolution and vertical refresh rate.

That's unlikely, if you ask me. Look at this:
...
The 18.44 kHz tone is too low for computer monitors. They usually operate at much higher horizontal frequencies, i.e. 31 to over 100 kHz - depending on resolution and refresh rate.

Yes, it seems so, I didn't think much about actual values.

This post has been edited by KikeG: Jan 6 2004, 18:30
Go to the top of the page
+Quote Post
MugFunky
post Jan 6 2004, 18:22
Post #11





Group: Members
Posts: 493
Joined: 28-December 03
From: Melbourne, Aus
Member No.: 10767



i've encountered a large number of very high tones when recording from FM radio with a (circa 25 year old) tuner. possibly there's some very strange RF action going on somewhere in the recording process. i'm clutching at straws here, but i do love a mystery.
Go to the top of the page
+Quote Post
Gecko
post Jan 6 2004, 19:59
Post #12





Group: Members
Posts: 937
Joined: 15-December 01
From: Germany
Member No.: 662



I'm going out on a limb, but perhaps the instruments which exhibit the 18.44kHz tone were originally recorded in "A" key and then digitally repitched to play back at "C" key.

A3 = 440 Hz
C3 = 523.251 Hz

So we have 18440 * 440 / 523.251 = 15506,13 which comes close (when taking into account the accuracy of my frequency readout).

I don't have the ear to hear what key the pieces are actually played in.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 27th August 2014 - 19:25