Frequenc Spectrum of CDs vs. LPs (1 Viewer)

[Mike Co]

Just a little input from an old analog guy. Human hearing is generally regarded as sensitive to about 20KHz, depending on age, etc. However, sounds in nature occur significantly above that and CD's were brickwall filtered at 20K on this basis (sampling at 44.1 also because of this).
Human hearing is most sensitive in the

 

[RobertR]

That's a very general statement which doesn't prove that the actual sound of musical instruments as

a. captured by a real world microphone at a real world distance at a real world recording studio and

b. recorded at a significant level on a real world record master on a real world commercial music label

contains much of anything above 20 kHz. Recording bats and other such "sounds in nature" doesn't count. Here's a link to a web page which shows the frequency range (including overtones) of many musical instruments:

http://www.hgpaudio.com/downloads/musicalsignals.pdf

 

[Mark All]

Here's some interesting reading about why the frequency response potential for vinyl is superior to that of CD:

Tannoy Research Paper

I think the findings here were what Matthias was referring to. Of the available formats now, I'd suggest that DVD-Audio has the widest potential usable frequency response if record labels would use its potential.

 

[Chu Gai]

Well you can only use what's there on the recording and if there's no significant audible output at say 17K, it's a moot point. As was mentioned earlier, consider the microphones for one. Hearing tests on young adult males, for example, indicate that while they might have normal hearing sensitivity at 90 dB SPL @ 15 kHz, they have NO sensation at 18 or 20 kHz. What do you think it is at say 30 kHZ? Any idea how loud you'd have to play something to have it register?
The blame for poor sounding CD's is the same as poor sounding LP's or SACDs or DVDA's - it's the recording. Time I think we stopped giving these bastards a free pass.

 

[Mike Co]

When you talk about the frequency of an instrument, remember there are lots of harmonics--don't even ask how many db down at, say, 20 KHz.
They keep going way beyond human hearing and the point is the more you capture, the more true to the recorded sound it is. Why? I don't know, but I do know that SACD sounds a lot better than any CD known to man; it may have something to do with wider bandwidth, but it's still conjecture. It makes sense to me since a recorded performance never seems to sound as good as the original as long as the original isn't amped or otherwise electronically processed. There are some very good mics, amps, pre's etc. that have very wide bandwidth and fewer speakers, processors (EQ, DSB's, etc) that go way over audible. The need hasn't been great because the main playback media has been CD for the past 20 years.
With the new digi media (SACD & DVD-A) the need for more is here.
Blah, blah, blah...

 

[RobertR]

You've compared EVERY SACD and CD "known to man"? That's quite a feat. To my knowledge, no one has EVER done a comparison of SACD and CD using a proper double blind protocol where the ONLY difference was the use of DSD vs. 16/44 PCM. Without such control of the variables (different recording methods, mixing, ancillary equipment, microphones, etc.) involved, such comparisons don't tell us much.

 

[Wayne A. Pflughaupt]

Mike,

Kudos for a rational and thoughtful presentation. I’d like to note a couple of things:

How much more? If the link Mattias provided is accurate, vinyl’s limitation is about 25kHz. That may sound a lot higher than 20kHz, but it isn’t. The difference between 20 and 25kHz is virtually nothing, only 1/4 octave – that’s only three half-steps on a musical instrument, the distance between C and D#.

I submit that if you take two identical audiophile recordings on vinyl and CD (i.e., the cream-of-the-crop for each medium), if one is perceived to sound better than the other, the reason will be factors other than extended frequency range. Especially when you consider that the mics used most likely didn’t even get response beyond 20kHz.

Regards,
Wayne A. Pflughaupt

 

[Mark All]

Sennheiser has had a studio microphone on the market now for a few years, the MKH800, that has a flat frequency response to 50kHz and costs around $2500. Still, most studios seem unlikely to have microphones that extend beyond 20kHz. Why? That's the limit of CD quality playback, the dominant format for popular music for the last 20 years, and ultrasonic extension is widely perceived, arguably incorrectly, as all that's necessary. The limit set on CDs was based on scientific research in the late 1970s and earlier. CDs are great, but they do not present a faithful reproduction of natural sound in music.

The Japanese study report cited by Tannoy is linked below.

Oohashi

The study involved subjecting volunteers to music recorded using a system capable of extension to 100kHz and looking inside their brains to see the effects. Of course, the system did not have to rely on the limits of either CD or LP technology.

 

[RobertR]

The more important question is whether they used ANY of the real world music that people commonly listen to when looking inside people’s brains. If they didn’t, the study has no relevance to what people actually listen to with respect to music, REGARDLESS of format. What the study really needed to do was a proper double blind comparison between an ultra wide bandwidth system and 16/44 CD using music that you or I would listen to. If people have NO awareness of a difference (regardless of "brain changes"), then there is no practical difference.

 

[LanceJ]

IIRC, somewhere on Steve Hoffman's forum he was discussing old mixing & EQ consoles from the 1950s & 1960s and he mentioned seeing one made for EQ purposes by RCA (RCA equipment in the studio is highly regarded). It had a total of three knobs, the last one labeled for 10kHz.

10kHz.

I'll let the reader come to his own conclusions concerning this........

LJ

 

[Chu Gai]

Is that what they call sizzle?

 

[Mark All]

More reading material on why professional sound engineers now consider the limited frequency range of CD to be inadequate...

Meridian's White Paper


It's not surprising for equipment designed 50 years ago to have a equalization channel centered around 10kHz since the frequency range of recorded music in the 1950s was largely limited to less than 15kHz until "full frequency range" recordings became available. I'm sure some people during the transition to full frequency range recordings in the 1960s doubted whether it would be of any significant benefit to them.

 

[RobertR]

He mentioned equipment made in the 50s AND 60s. It's reasonable to think that the full range of a typical LP was used by then.

 

[Allan Jayne]

Higher frequencies can be recorded in the first few tracks (near the outer edge) of an LP compared with near the center.

I am led to believe that some of the complaints of harsh sound on CD has to do with poor filtering of the original source to eliminate content above about 22 KHz and also poor final D/A conversion. (The brickwall filter mentioned earlier is what is needed.) If content above 22 KHz finds its way into the digial processing, it will show up in the final output as spurious frequencies below 22 KHz which may amount to sour notes.

If I remember correctly, I have heard 18 KHz reproduced from an LP with test tones years ago. But now I can't hear over about 12 KHz from any source.

This may be of interest:
http://members.aol.com/ajaynejr/nyqam.htm

 

[Chu Gai]

Are you familiar with the pcabx.com website Allan? There are various wav files recorded that would let you examine that premise.