Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Default Visualization: Spectrum (Read 4190 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Default Visualization: Spectrum

HI!
Simple question:
I have the default Spectrum there on the top.
What does it mean?

1. What does the height of the bars mean?
2. What does it mean to show bars on the left side,what on the center, and what on the right side?
3. Does it have to do with Stereo, Joint Stereo?

Default Visualization: Spectrum

Reply #1 –
The X-Axis represents the frequency and the Y-Axis the sound pressure level.
Just google "audio spectrum", there are even a lot of YouTube videos which explain exactly what it all means.

Default Visualization: Spectrum

Reply #2 –
left to right is increasing frequency, bar height is volume.

Try some minimalist music, with for example a well-defined bass kicks.

Default Visualization: Spectrum

Reply #3 –
A OK.
So on the left are the lower frequencies and on the right the higher.
Now, does some particular aspects of a song belong to left or right?
For example, do the vocals appear to the higher frequencies and the bass to the lower?

Default Visualization: Spectrum

Reply #4 –
I can't tell if your question is serious.

I'm sorry if I'm being incredulous, but I've never seen someone ask whether bass is made of low frequencies.

Edit
Double sorry,  but if I am to explain things, I have to ask: is this your first time wondering about frequencies and music and things like that?

Default Visualization: Spectrum

Reply #5 –
Quote from: dhromed on
I can't tell if your question is serious.

I'm sorry if I'm being incredulous, but I've never seen someone ask whether bass is made of low frequencies.

Edit
Double sorry,  but if I am to explain things, I have to ask: is this your first time wondering about frequencies and music and things like that?


Yes.

Default Visualization: Spectrum

Reply #6 –
Ok!

What we normally call bass is roughly the range of 20-500 Hz, which is the leftmost part of the spectrum visualizer. Speech is around 2000-6000Hz, which is to the left of the middle. Instruments like violins and pianos are mostly in the same range as speech/vocals. To the right you have all the high frequencies, which you'll mostly hear in the crashes and cymbals of drums, and the noisy parts of electric guitar.

It may seem like most of the sound is bunched up to the left. That's because humans can hear clearly from about 30Hz all the way to about 19000Hz (if you have perfect, young ears). That's a huge range to cram into such a tiny space, so it uses a logarithmic distribution, and as such most things we hear fall on the left of the graph.

In foobar, you can make some pure tones. Create a new playlist, go to Add Location (Ctrl+O) and enter this command:

tone://300,5

The first number is the frequency of the tone, and after the comma you put the number of seconds you want the tone to last. So this command gives you a 300Hz tone for 5 seconds. You can change the numbers to anything you want.

There's also:
sweep://500-10000,10

This one is essentially the same as tone:, but instead creates a transition from 500Hz to 10000Hz over 10 seconds.

Be warned! These tones are fullscale, which means they are LOUD! Make sure to turn down your volume when you stay playing around. Try opening various visualizers while you play the tones.

There's lots of stuff to read about sound and signals on Wikipedia, but this should give you a basic feel for it. If you're hungry for more, grab a copy of Audacity and do... things.

Default Visualization: Spectrum

Reply #7 –
One correction: speech is from 300Hz to 8000Hz, at least this is how "full speech band" is defined in telecommunication. Upper value may be lowered to 3400Hz when we speak about minimum band that allows proper understanding of speech. Fundamental frequencies of human voice are around 100-250Hz: a typical adult male will have a fundamental frequency of from 85 to 155 Hz, and that of a typical adult female from 165 to 255 Hz. Children and babies have even higher fundamental frequencies. Infants show a range of 250 to 650 Hz, and in some cases go over 1000 Hz.  A 10-year-old boy or girl might have a fundamental frequency around 400 Hz. Fundamental frequency is responsible of how we percept pitch of human voice. So in fact even telecommunication does not allow us to properly perceive pitch of human voice. But music will and this is what we will find while using foobar. Thus "bass frequencies" are strongly mixed with "speech frequencies" - especially those most important of how we perceive someone's voice.

Sorry, I couldn't left it without such clarifications, despite this is far beyond foobar's scope.

Additionally I will lower the range of bass to 200Hz, maybe 250Hz. Look at given fundamental frequencies of voice and see how we percept human voice with fundamental frequencies of females voice... Additionally 300Hz tone sounds so "beepy" that I think no one will classify this as bass... But this may be my subjective opinion.

Default Visualization: Spectrum

Reply #8 –
This may help as well:

When a single note is sung or played on a typical instrument, you are not hearing a pure tone... you are hearing a partial: a combination of pure tones... a whole lot of them, usually! The lowest-frequency tone is the fundamental, and it is usually among the strongest, although you may notice the higher-frequency overtones more. Some of the overtones are called harmonics; they are at frequencies that are whole-number multiples of the fundamental.

When you hear noise, like white noise (hiss), or the dominant sound in the sound of snare drums, hi-hats and cymbals, it is the most complex combination of all: essentially all frequencies together...infinite, random overtones, rather than the orderly set that make tonal instruments like the trumpet sound as they do.

The spectrum visualization is showing you a kind of movie in which each frame is a graph of the tones in a split-second of sound. foobar2000 is analyzing that segment of the playing audio, determining what tones it contains and how intense they are, and plotting them on the graph. So as you watch, you are seeing the graph update many times per second.

To see what I am talking about, go get this trumpet sample. It was posted here a while back because it's difficult for some MP3 encoders to handle without mangling the sound, and that's the kind of thing we like to talk about around here. It also makes a good demo of partials. Play it over and over, and just watch the spectrum in full-screen mode as it plays. Right-click anywhere on the spectrum display and change Bands to 160 so you see the frequencies with more precision.

You will see that each note the trumpet plays is actually a simultaneous set of tones, and you will see some interesting patterns, with the louder harmonics looking like the tines of a lopsided fork sticking up, and with much smaller, quieter tines in between. As different notes are played, the general pattern is the same, just shifted up or down the scale...but there are some slight differences in the patterns as well, since it's a real trumpet being finessed by a live performer:



You may find the Spectrogram visualization is also helpful. It is like the spectrum, but turned sideways (high frequencies are up top, low ones at the bottom). The horizontal axis is time, with the most-recently played audio on the right. It's not as good at showing you the intensity, but it's a good way to see the harmonics:



(I have my spectrogram [a href='index.php?showtopic=102273']configured to show color[/a])

Once you understand this stuff, you can learn what MP3 & AAC encoders do. It's more complicated than this, but basically they do this same kind of analysis, producing a whole bunch of snapshots of overlapped split-second segments of sound. In each of these snapshots, they get rid of the tones that you can't hear (usually because they're overpowered by nearby, louder tones). The encoder stores the remaining frequencies with reduced precision: "close enough" values, say. The audio player converts these snapshots back into a complex waveform...and if the encoder did a good job, the resulting sound will be pretty close, if not indistinguishable from the original.

You may also enjoy foo_musical_spectrum.

Default Visualization: Spectrum

Reply #11 –
Quote from: TuNk77 on
Please share the RGB code for the spectrogram

From top to bottom:
0 0 0
4 1 71
48 0 115
86 32 136
176 0 91
255 0 0
255 147 0
255 247 0
255 255 255
255 255 255

Default Visualization: Spectrum

Reply #12 –
OT: epic colours  very nice modification mjb2006 - i'm going to "fork" my foobar config and use your colour set for some time.
and once again thank you dhromed for your original instructions and colour set

Default Visualization: Spectrum

Reply #13 –
Quote from: EpicForever on
Additionally 300Hz tone sounds so "beepy" that I think no one will classify this as bass...


Thats the only thing that I failed to understand.
This:
Code: [Select]
tone://300,5
, is 300Hz. And you said its lower frequencies. Still, the sound that comes when I play it, is like way too high. All the music I have, did not have such high sound.
But then, is it because you said the keyword, "fullscale", like perhaps meaning, that all the other music is put in logarithmic, or something like that?

Yeah. Thanks a lot ppl! I got the idea and all the information I needed. Now I can see the Spectrum and seems more logical.
Also, I havent finished reading...

Default Visualization: Spectrum

Reply #14 –
By "beepy" I meant that it sounds like all those "beeeep" sounds from buzzers. And this is far from bass.

You said your music don't have such high sound as 300Hz... of course it has, but this is the moment when thing become more complicated and bit tricky - psychoacoustics - and all those differences how we perceive pure tones, harmonic sounds consisting of many pure tones, non harmonic sounds, noises etc...

Default Visualization: Spectrum

Reply #15 –
Grande Valse Brillante In E-Flat Major, Op.18: Vivo // Arthur Rubinstein
"Fantaisie-Impromptu" In C-Sharp Minor, Op.Posth.66: Allegro Agitato // Arthur Rubinstein

View->Visualizations->Spectrum

Double-Click the window

Right-Click->Bands->160

This is ABSOLUTELY fascinating...
Thanks a lot guys!!!

Default Visualization: Spectrum

Reply #16 –
Quote
This is ABSOLUTELY fascinating...


You've caught the bug. 

Default Visualization: Spectrum

Reply #17 –
I would like to commit 2 questions too.
They have to do with female vocals.
I listen to one popular singer.
Not classical soprano or stuff like that, just folk music lets say.
I investigate the Spectrum, and focus only on the "fundamental" bands, that "capture" her voice.

1. In PIANO I noticed clearly, the "main" tone. But in female vocals, they are split in 3 or 4 neighbor areas. Strange...

2. When, in some songs, she reached "high" tone, the "fundamental" band, which was clearly visible, went very high in the spectrum.
Like around -10dB.
But, before in this post, it was mentioned, that the height of the bands, indicate the "volume", or someone else said, the "sound pressure."
What do I see here? What does the height show?
What I thought, is that, when the female singer, reaches a high tone, that I would see some bands on the right of the spectrum.
Instead, I see a band, on the middle, as usual in the whole song, but it just goes high above on the spectrum.

PS: I will investigate further, with actual sopranos some time later. Perhaps, Maria Callas...

Default Visualization: Spectrum

Reply #18 –
Quote
went very high in the spectrum.

That's normal. It means she sang REALLY REALLY LOUD. Because of the way our throats work, singing or shouting louder usually also means singing higher.

The height of the band vibrates a little because the volume and tone vary a little, expecially with a lot of vibrato.

You'll never hear a voice, no matter how high, on the right side of the spectrum, since no human being can create tones of 10KHz or higher. Some people have very high voices, some have very low (Barry White etc), but this is still only a small part of what your ears can hear.

Try Visualisations > Spectrogram as well. For the best detail, Right-click on it > FFT Size > 4069.
You'll notice the fundamentals and overtones much more easily.

Default Visualization: Spectrum

Reply #19 –
Quote from: dhromed on
Try Visualisations > Spectrogram as well. For the best detail, Right-click on it > FFT Size > 4069.
You'll notice the fundamentals and overtones much more easily.


Sorry as this was a post for the spectrum.
For the Spectrogram, I don't understand what I am seeing.
Only maybe that the horizontal axis is the time.
Sorry.

Default Visualization: Spectrum

Reply #20 –
In the spectrogram, the vertical axis is frequency, the brightness (and color, if so configured) is intensity/loudness, and the horizontal axis is time (past to present). It's like a diagram of the last however-many seconds of sound that just played.

Default Visualization: Spectrum

Reply #21 –
Quote from: CC-GoE on
For the Spectrogram, I don't understand what I am seeing.


The spectrogram is a very intuitive way of looking at music. Time goes from left to right, and tones go up and down.


Default Visualization: Spectrum

Reply #22 –
Quote from: mjb2006 on
In the spectrogram, the vertical axis is frequency, the brightness (and color, if so configured) is intensity/loudness, and the horizontal axis is time (past to present). It's like a diagram of the last however-many seconds of sound that just played.


ALLRIGHT.
EEE.
Why are there 2 stripes going on?

About the vertical axis: So when we go higher, it means higher frequencies?...

The brightness: Which is the lowest setting and which the highest? I guess, black is lowest and full white highest?

CHEEERS

PS: Just a gift...

http://grooveshark.com/#!/search/song?...uinn+The+Eskimo
(Select the first one from the compilation: The Essential Bob Dylan)
(!)