Please have a look at this spectrum graph and let me know your thoughts

[Waltz Mastering]

Originally Posted by SirElliot

It is how the human eardrum is excited under the given frequencies.

So your saying that the ear hears a frequency as being more aggressive when the peaks average to be sharp in a particular region and less aggressive when they average to be flat or natural?

Where have you heard about this?

Every sharp is a flat, so how would that work out? (G# = Ab)

[Jerry Tubb]

Hey Mark, you seem like a good guy, so I'll give you some straight answers.

1. Yes your curve looks typical of many forms of modern full range "pop" music. No red flags can be ascertained from it.

2. Your curve looks like it's probably an "average" setting. Peak settings show more detailed dynamic activity as music is playing, average settings give you more objective view of a songs frequency content over it's entire length.

3. No. once again, your curve looks typical. As to the objective of spectrum analysis, it's just a visual tool for learning, a quick check, & pinpointing problems.

4. No.

5. Yes.

The Lecture in a nutshell:

Often times spectrum analysis creates more confusion than it actually helps to solve problems. Your initial post is a case in point. The danger is that one can become reliant on visuals, like a crutch, which may even stymie the development of proper listening skills. Best to keep it's use to a minimum. Careful Listening, aural analysis, and learning to identify problems (or the lack of) by EAR is much more important skill.

Most real mastering engineers don't use these visual tools. Nor does their use ensure a "nice final master".

Good Luck - JT

Originally Posted by MotifCity.Com

I have some questions:

1. Does this curve at least look like its in the ballpark for hip hop music? If not, can you please point out any red flag areas ?

2. If you use spectrum analysis, when you mix do you use peak or average graphs or both?

3. Should the average line be straight and smooth as possible from left to right (with the exception of the low and high ends). Is that the objective with spectrum analysis?

4. Can you please post graphs from your projects so I can compare mine to yours?

5. Is it normal for the graph to slope down on the right? (Seems like it always does in my mixes and if I bring up those levels any higher my mixes are too bright)

Thanks, just want to learn more about properly using spectrum analysis to get a nice final master

[UnderTow]

Originally Posted by SirElliot

Hi.
The harshness stems from 9000 kHz because it falls within the sharp region and, preferring a bit more low-end thump is due to having the peak on 47.8 Hz, which is not aggressive enough for UnderTow’s taste. If the bass had peaked around 46 Hz, it would sound more aggressive on the bass line.

Bollocks.

Alistair

[SirElliot]

Hi

Originally Posted by Waltz Mastering

So your saying that the ear hears a frequency as being more aggressive when the peaks average to be sharp in a particular region and less aggressive when they average to be flat or natural?

Where have you heard about this?

Every sharp is a flat, so how would that work out? (G# = Ab)

What I am trying to say is that there are certain frequencies at a particular dB level in which our ears become more agitated.

This will vary based on numerous factors.

1. How sensitive the loudspeaker is at the particular frequency.

2. How well a person can hear high frequencies.

3. How loud the peak is at the given moment from a dB perspective.

If the frequency is cut off abruptly (I am not referring to using a high/low pass filter), on either end of the frequency scale you have a greater chance of creating a peak.

Loudspeakers by design will gradually roll-off the high (and low) end frequencies. It is based on the design limitations of the tweeter (and woofer). If one should severely cut-off shall we say anything beyond 12 kHz, and the loudspeaker offers –3 dB @ 12 kHz, it will have a lower chance of transmitting any peaks for it is already – 3 dB @ 12 kHz.

If the loudspeaker is – 3 dB @ 20 kHz using the above scenario, it will inform the user (providing he or she is capable of hearing 20 kHz) that there are frequencies missing which can translate into an annoyance @ high SPLs.

If the track’s peaks end on a sharp note in which the fundamentals are not present (not enough dB) to dampen the offending frequencies, it will create an annoyance if there is enough SPL at that given frequency.

1 k, 2 k, 4k, & 8 k are not sharp notes. 1125, 2250, 4500 and 9000 kHz falls within the sharp region.



I you would like to continue this discussion we will need to continue on Monday for I am about to do an event.

Cheers!

[scraggs]

WTF are you talking about? i will wait patiently until monday when you can come back and tell me what on earth you mean with this 'sharp note' business.

[dcollins]

Originally Posted by UnderTow

Bollocks.

Alistair

I'm afraid so.


DC

[trmchenry]

The only science-y sounding thing I can come up with now is something vaguely related to dissonance. But you can't really control that at the mix/master level.

[dcollins]

Originally Posted by trmchenry

The only science-y sounding thing I can come up with now is something vaguely related to dissonance. But you can't really control that at the mix/master level.

If you mean cognitive dissonance, I think you might be on to something.........



DC

[Cellotron]

I'm wondering whether forums for visual artists ever get hit with posts like "please listen to this recording of audio tests signals bounced off the surface of my latest sculpture and let me know if my work looks good to you"


Best regards,
Steve Berson

[UnderTow]

Originally Posted by SirElliot

I you would like to continue this discussion we will need to continue on Monday for I am about to do an event.

I've heard of instituitions that let the patients out for the week-end but not yet the opposit...

Alistair

[trmchenry]

Originally Posted by dcollins

If you mean cognitive dissonance, I think you might be on to something.........



DC

lol

[macc]

There are some truly amazing threads round here of late

And I'm not talking about Ben/Subvertbeats' spangled jodhpurs.

[SirElliot]

Originally Posted by trmchenry

The only science-y sounding thing I can come up with now is something vaguely related to dissonance. But you can't really control that at the mix/master level.

Yes. The characteristics are somewhat similar.

Bear in mind I am not suggesting MotifCity.Com mix should be altered. I am merely translating to MotifCity.Com what is occurring from the SMAART visualisation.


This is the curvature of the high frequencies.




With a starting point of 9 kHz and ending point of 18 kHz, there is roughly a 47 dB drop (0.4 dB @ 9.53 kHz, - 47.7 dB @ 17.95 kHz). The slope is so steep there is a boost between 8.5 to 9.5 kHz, which offers a spike. With such a steep slope and, 9kHz falling within the sharp (C#9 = 8869.844) region, all the energy in the high frequencies resides around 9 kHz.

Please observe below where the frequencies that would aid 9kHz to reach 18 kHz reside:

The response is based on freezing the graph with the 9.53 kHz peaking @ 0.4 dB

9.53 kHz +0.4 dB

10.69 kHz - 4.9 dB

11.33 kHz - 11.5 dB

12 kHz – 12 dB

12.7 kHz – 14.2 dB

13.46 kHz - 14.9 dB

14.2 kHz - 17.6 dB

15.1 kHz - 22.1 dB

16 kHz - 39.5 dB

16.95 kHz - 43.8 dB

17.95 kHz - 47.7 dB




Fletcher & Munson curve

http://www.phys.unsw.edu.au/jw/graph...her-Munson.gif

It is evident the higher frequency the more SPL is needed to compensate our inability of registering higher frequencies.

If 9 kHz requires more decibels than 4.5 kHz and 16 kHz requires more decibels than 9 kHz, what frequency are we going to hear the most when 16 kHz is – 39.5 dB?

Loudspeaker Frequency Response

A loudspeaker’s frequency response is measured using two points to determine it’s output from a dB perspective. One is –3 dB and the other is – 10 dB. Loudspeakers have difficulty producing anything below –10 dB.

If the loudspeaker can deliver 97 dB @ 1 watt/1 meter it will be 94 dB @ -3 dB and, 87 dB @ - 10 dB based on the roll-off points of the frequency response.


Amplifier Output From A Wattage Perspective

A 3 dB loss from a 200-watt amplifier will offer 100 watts.

A 10 dB loss from a 200-watt amplifier will deliver 20 watts

This all means 9 kHz will be the dominant frequency in what we hear in terms of high frequencies since it offers the highest peak and, everything below 10 kHz will be not be loud enough for us to detect.



Cheers!

[UnderTow]

Originally Posted by SirElliot

Yes. The characteristics are somewhat similar.

Bear in mind I am not suggesting MotifCity.Com mix should be altered. I am merely translating to MotifCity.Com what is occurring from the SMAART visualisation.


This is the curvature of the high frequencies.




With a starting point of 9 kHz and ending point of 18 kHz, there is roughly a 47 dB drop (0.4 dB @ 9.53 kHz, - 47.7 dB @ 17.95 kHz). The slope is so steep there is a boost between 8.5 to 9.5 kHz, which offers a spike. With such a steep slope and, 9kHz falling within the sharp (C#9 = 8869.844) region, all the energy in the high frequencies resides around 9 kHz.

Please observe below where the frequencies that would aid 9kHz to reach 18 kHz reside:

The response is based on freezing the graph with the 9.53 kHz peaking @ 0.4 dB

9.53 kHz +0.4 dB

10.69 kHz - 4.9 dB

11.33 kHz - 11.5 dB

12 kHz – 12 dB

12.7 kHz – 14.2 dB

13.46 kHz - 14.9 dB

14.2 kHz - 17.6 dB

15.1 kHz - 22.1 dB

16 kHz - 39.5 dB

16.95 kHz - 43.8 dB

17.95 kHz - 47.7 dB




Fletcher & Munson curve

http://www.phys.unsw.edu.au/jw/graph...her-Munson.gif

It is evident the higher frequency the more SPL is needed to compensate our inability of registering higher frequencies.

If 9 kHz requires more decibels than 4.5 kHz and 16 kHz requires more decibels than 9 kHz, what frequency are we going to hear the most when 16 kHz is – 39.5 dB?

Loudspeaker Frequency Response

A loudspeaker’s frequency response is measured using two points to determine it’s output from a dB perspective. One is –3 dB and the other is – 10 dB. Loudspeakers have difficulty producing anything below –10 dB.

If the loudspeaker can deliver 97 dB @ 1 watt/1 meter it will be 94 dB @ -3 dB and, 87 dB @ - 10 dB based on the roll-off points of the frequency response.


Amplifier Output From A Wattage Perspective

A 3 dB loss from a 200-watt amplifier will offer 100 watts.

A 10 dB loss from a 200-watt amplifier will deliver 20 watts

This all means 9 kHz will be the dominant frequency in what we hear in terms of high frequencies since it offers the highest peak and, everything below 10 kHz will be not be loud enough for us to detect.



Cheers!

Still bollocks.

Alistair

[Nordenstam]

Originally Posted by SirElliot

This all means 9 kHz will be the dominant frequency in what we hear in terms of high frequencies since it offers the highest peak and, everything below 10 kHz will be not be loud enough for us to detect.

Funny typo.

It doesn't make sense in any case, even if the typo hadn't been there. Here's something for you to try: cut everything above 10kHz and see if you can hear the difference. If you can't, consult an ear specialist for audiometric checks!

[dcollins]

Originally Posted by Lupo

Funny typo.

It doesn't make sense in any case, even if the typo hadn't been there. Here's something for you to try: cut everything above 10kHz and see if you can hear the difference. If you can't, consult an ear specialist for audiometric checks!

I want to know more about the sharps. Or how 47.8Hz is cool but 46 isn't.


DC

[Huntley Miller]

Originally Posted by dcollins

I want to know more about the sharps. Or how 47.8Hz is cool but 46 isn't.


DC

The Davelizer can't do 47.8Hz?

Lame.

[jdg]

i think thats the key of one of my fav jimi hendrix songs