Orchestral music - getting it right

[Thad E Ginathom]

Similarly, one can easily get convinced that a piano sounds RIGHT only in a concert hall when listened from 33.33333 ft away at and angle of ?/2 radians from the axis of the piano and the performer, where you get a reverberation time of ? seconds.

Hence the vicious audiophile punch-up for that particular seat before each concert :lol:

The point being that when Wagner's symphonies were composed - they were done after keeping in mind the concert acoustics.

Instruments developed, concert halls developed, composition and performance changed. These things are an essential part of the history of western music. Prof Wright covers this in his Musical Appreciation course, which I am ever-grateful to have found through this forum.

I, quite frankly, don't get your tirade against the high frequencies in rock music!

And from a Mahler fan, too!

Hi hats and cymbals ARE High frequency equipment - carrying sound energy till 16 kHz.
Classical music doesn't employ any such equipment and hence roll off at 4 kHz is desirable.

But it does! My current quest is to find ways to correct the sad fact that my hearing now starts to "roll off" as low as 1k :sad:. I am struggling to understand how to boost those higher frequencies, preserving both the brightness of violin and the sound of cymbals through EQ without loosing the peaks to horrible digital clipping.

But this is funny since the most vehement opponents of compressed music are the ones who listen to Classical music - which is mathematically the least complex - and thus compression doesn't lose any musical information.

compressed as in MP3, or compressed as in engineered-for-radio-and-teenagers?
I don't know about mathematical complexity, but orchestral music has such a wide variety of different sounds combined with such a huge dynamic range... doesn't it beat everything else for almost all kinds of complexity in sound?

 

[alpha1]

Will compression not compromise the natural dynamics of classical recordings? After all most compression algorithms happily normalise sound (lower loud passages and bring up low passages). Personally, I thrive on that momentary kick of a sudden loud passage. And enjoy a conductor gesticulating wildly to perfectly time that loud passage where tympani, brass section, strings, and everyone else gives it their all.

Yes but does it make real complex for mathematical encoding?
I have put a test below.

Hence the vicious audiophile punch-up for that particular seat before each concert :lol:

Actually I forgot to add that it is T10 decay reverb time of ? seconds.

I don't know about mathematical complexity, but orchestral music has such a wide variety of different sounds combined with such a huge dynamic range... doesn't it beat everything else for almost all kinds of complexity in sound?

Dynamic range - if you compare with super companded modern music - yes.

Complexity? No.
Simply because of two reasons:
1. Very less energy and information in the high frequency domain
2. Very less overtones of the fundamental tones (all the instruments are acoustic)

Compare this to electronic and rock music, where high frequencies carry a lot of information.
(randomness also adds to complexity - applause is very complex mathematically for encoding)


One way to check the complexity is to encode a Mozart CD at vbr (say V0), and then encode an electronica/rock CD again at same vbr setting.

The average bitrate for electronic and rock music will always be more than classical. This is because more bits are required to describe that kind of music compared to classical. (complexity)

 

[jls001]

Yes but does it make real complex for mathematical encoding?
I have put a test below.

Actually my question was only with respect to the dynamics. One of the algorithms used in most audio compression algorithms is minimization of the swings in the dynamics (peaks take more bits to encode by virtue of the fact that it takes more bits to describe a signal of higher amplitude. I don't know if the reverse is also true for a low level signal). Another trick is to eliminate fast transients. So one ends up losing high dynamic swings intended in the program material, as well as fast attacks and graceful decays.

I have absolutely no idea about the mathematics and physics of it, but I would hazard that an orchestral piece with 20-30 musicians or more, playing perhaps 10-13 different types of instruments, would be more complex by most metrics, than 4-5 rock musicians belting out on an electric guitar, an electric bass, a drum kit, and perhaps a synthesizer (and of course vocals).

Assuming it is the harmonics of instruments that contribute to the mathematical complexity of a sound, we have equivalents in both genres. Let's match them note for note:

1) A wall of massed strings sounds as complex as an electric guitar on Overdrive effects (or Chorus or Flanger), be it solo or rhythm. The brass section (of trumpets, clarinets, piccolos, flutes, etc) in full blast is harmonically equally complex - if not more so - than any instrument that a rock ensemble can throw at it.

2) 4-6 cellos playing in harmony is way more complicated harmonically than an electric bass. In fact most electric bass sound flat and 2D compared to the hair-raising timbre of cellos (or violas, for that matter). Try Musica Nuda's Eleanor Rigby to see what I mean. One cello, one female voice = hair raising experience.

3) the tympanis can surely match even two kick drums as favoured by some rockers (think Alex van Halen) for SPL and tonality?

My contention: classical is more complex than rock. Or pop. Or jazz. But perhaps is the equal of some polyrhythmic music favoured by certain African bands and Brazilian ensembles.

Also, any specific reasons why some people feel that rock music is bright compared to classical? I have never had an occasion to feel this way:indifferent14:

 

[jls001]

2. Very less overtones of the fundamental tones (all the instruments are acoustic)

Violin, saxaphone, bamboo flute, cello, piano - all acoustic instruments. I don't know what to choose between sax and bamboo flute for the complexity of the overtones they emit. Cello comes next. Piano and violins are also capable of producing very complex overtones.

Compare this to electronic and rock music, where high frequencies carry a lot of information.
(randomness also adds to complexity - applause is very complex mathematically for encoding)

Rock or pop does not carry any more high frequency information than does the average classical piece.

If you ask me, some of the brightest sounding recordings are small ensemble jazz from the 50s and 60s. Think Dave Brubeck Quartet's Take Five which has piercing cymbals in one channel.

Applause: agreed.

 

[Thad E Ginathom]

The question was asked and answered in a different context, but maybe it throws some light on this complexity business. Paraphrasing...

Why do two instruments playing the same note always sound louder than one, and never cancel each other out? Because, although they may be playing the same fundamental note, and even although they are the same sort of instrument, the harmonics of each one will vary slightly.

There was quite a lot more possible reasons, but this one, in this context, makes me think: a violin, has all the overtones that make it sound violin. Add another twenty, all with minutely different overtones. Is this not massively complex? It is a bit extreme, but add several hundred more performers in a Mahler symphony, including whole choirs.

Because of my current experiments with eq, I was watching some wave forms yesterday. Mahler: all over the place. Incredible String Band: mostly surprisingly flat.

 

[alpha1]

JLS:
Since I also don't know all the mathematics involved, I did that simple observation / test.

Why doesn't mp3 or ogg or aac not use more bits for classical type of music compared to Electronic music when given freedom?

 

[Thad E Ginathom]

You're saying that a variable-bit-rate compressions system should produce bigger files for "more complex" music, and that you find that it doesn't?

Sounds like a good question! But I don't think it helps to guess the answer: we'd need to hear it from someone who knows how the encoding internals work. Failing that, a cryptologist make make a better guess, because this sort of stuff is really closely allied to cryptology.

I might well not understand the answer, though

BTW... I used to use minidisc a lot when I was trying to be a music student, and I was always impressed by the recording quality of my portable unit with a GBP-100 stereo mic. I think Sony's compression system (forget the name now) for minidisc was really excellent. Curiuosly, there was one thing I found it bad at handling: the drone from a sruthi box.

 

[jls001]

JLS:
Since I also don't know all the mathematics involved, I did that simple observation / test.

Why doesn't mp3 or ogg or aac not use more bits for classical type of music compared to Electronic music when given freedom?

Please correct me if I'm wrong as I'm no expert here:

Assuming you're referring to Variable Bit Rate encoding, the variability has an upper ceiling in terms of bit rate - which is what one chooses as the maximum bit rate, say 192, 256, or 320 kbps. All parts of the song will be encoded at or below the max bit rate.

What is complexity from encoding point of view?

First and foremost is whether there is change in the music from one moment to the next. If there is lot of change, more bits are required to describe that change from one moment to the next. If there is lesser change, lesser number of bits are needed to describe the changed music. This is adaptive differential PCM. It describes the difference of one moment of musical passage from the preceding moment. No need to describe the whole moment. Thus lesser number of bits required.

The second kind of complexity that the encoder sees is wide dynamic swings. The trick used is to push down loud passages closer to the average level of music. This saves some more bits needed to encode higher amplitude sound.

The third complexity is sudden temporal changes that most music reproduction equipment any way cannot reproduce well. Eliminate such passages or tone them down.

A fourth complexity the encoder deals with is soft passages too low compared to the average level, or too close to the noise floor. Eliminate those as well.

Now, one must understand that all these are double-edged swords, they can cut both ways. Use too much of a good thing and you get a reproduction that is more far away from the original. This is why different algorithms, for instance a Fraunhoffer Labs MP3 and a LAME MP3, do not produce the same sound to the discerning listener. So finding a balance is a tightrope walk for the algorithm designer. I wouldn't know how a designer decides the limits or sweet spots of a particular parameter but I would hazard that it is by ear. I could be totally wrong here.

To come to your query, I don't have a definitive answer. But I would presume if one were to minutely do side-by-side comparison of simple passage of music and a more complex one (comparing instantaneous bit rates), the complex passage ought to be encoded at higher bit rate. And this would be genre-agnostic. You can try this at home when you have time. We'd be very interested in your findings.

 

[Thad E Ginathom]

Very interesting.

Can you enlarge on "sudden temporal changes" for me, please? That's only thing where I'm not sure what you mean

 

[jls001]

Very interesting.

Can you enlarge on "sudden temporal changes" for me, please? That's only thing where I'm not sure what you mean

Oh, that's just a fancy way of saying sudden changes from quiet or average to very loud, or vice versa.

 

[Thad E Ginathom]

dynamic changes?

I quibble ... fast dynamic changes, I guess.

Thanks

 

[ajay124]

"Rock music is essentially high pitched noise + juvenile lyrics."

I would like to take back at least the first part of this statement. As alpha 1 said: it sounds like a tirade. I have had a love/hate relationship with rock music for many years now. The day I wrote that post was probably one of the hate days Rock music may be noisy and (arguably) inferior to classical music or jazz, but it does have its own space in the history of the 20th century. By rock music I mean the music which was recorded between 1960-1980. I have no idea about rock which came after 1980 and I am not interested in finding out. For me rock music (and all the vocabulary and life style which went with it) is enclosed within these two decades. The high point being from '62-'72. Once the spirit of the times has been lost the music of those times cannot be replicated. The music which Bob Dylan, Joni Mitchell, Jimi Hendrix, Joan Baez, Leonard Cohen, Neil Young, Bob Marley, David Bowie, Beatles, Stones, Who, Cream, Grateful Dead, Led Zep, Pink Floyd, Traffic, Steely Dan, Santana, Doors, S&G and many others made in those years cannot be replicated.

But the lyrics were nothing to write home about even in those years. Although they were definitely superior to the nursery rhymes which generally pass for songwriting these days. The man who is widely considered as the spokesman of his generation was Bob Dylan. He is credited with breaking down the barriers of what a song could do. In terms of content and in terms of running time. "Like A Rolling Stone" is considered to be one of the seminal and path breaking song of the 60's. It demolished the trend of popular music being nothing more than three minute love ditties. Popular music grew up after this song. Songs with a moderately more grown up content started appearing on the scene.

Personally I used to believe that Bob Dylan's music from the 60's and 70's was the high watermark of these times. Albums like: Bob Dylan (1962)/The Freewheelin' Bob Dylan (1963)/The Times They Are a-Changin' (1964)/Another Side of Bob Dylan (1964)/Bringing It All Back Home (1965)/Highway 61 Revisited (1965)/Blonde on Blonde (1966)/John Wesley Harding (1967)/Blood on the Tracks (1975)/Desire (1976)/Street Legal (1978)/Slow Train Coming (1979). But in the last decade I have gradually stopped listening to Bob Dylan. His voice, strumming and harmonica no longer make me giddy with happiness. His words seem to be piled up ramshackle into a towering monument. A monument I can no longer relate to because one of his lesser known contemporaries has demolished all my notions of what popular music can become.

http://www.youtube.com/watch?v=NrmNDZNmxIk

 

[jls001]

dynamic changes?

I quibble ... fast dynamic changes, I guess.

Thanks

You are right. I should have mentioned "fast".

 

[Thad E Ginathom]

"Rock music is essentially high pitched noise + juvenile lyrics."

I would like to take back at least the first part of this statement. As alpha 1 said: it sounds like a tirade. I have had a love/hate relationship with rock music for many years now. The day I wrote that post was probably one of the hate days Rock music may be noisy and (arguably) inferior to classical music or jazz, but it does have its own space in the history of the 20th century. By rock music I mean the music which was recorded between 1960-1980. I have no idea about rock which came after 1980 and I am not interested in finding out.

I'm in the same camp. If anything, a few years narrower. Friends tell me that I should be interested in finding out, as there is good music being made. Oh well!

For me rock music (and all the vocabulary and life style which went with it) is enclosed within these two decades. The high point being from '62-'72. Once the spirit of the times has been lost the music of those times cannot be replicated. The music which Bob Dylan, Joni Mitchell, Jimi Hendrix, Joan Baez, Leonard Cohen, Neil Young, Bob Marley, David Bowie, Beatles, Stones, Who, Cream, Grateful Dead, Led Zep, Pink Floyd, Traffic, Steely Dan, Santana, Doors, S&G and many others made in those years cannot be replicated.

Apart from its musical merits, or lack thereof, it is (as a friend of mine put it) the folk music of our times.

But the lyrics were nothing to write home about even in those years. Although they were definitely superior to the nursery rhymes which generally pass for songwriting these days.

Even some of my favourite Grateful Dead songs really are absolute crap ("Transative nightfall of diamonds?" WTF!) and, at sixty now, I often feel that many of the young writers of forty years ago sound like pretentious fools who had, in fact, barely discovered life! People such as Crosby, Stills etc are exceptions.

The man who is widely considered as the spokesman of his generation was Bobby Dylan. He is credited with breaking down the barriers of what a song could do. In terms of content and in terms of running time. "Like A Rolling Stone" is considered to be one of the seminal and path breaking song of the 60's. It demolished the trend of popular music being nothing more than three minute love ditties. Popular music grew up after this song. Songs with a moderately more grown up content started appearing on the scene.

Have to say that I wish he had stuck to writing. Maybe his whining drawl was, itself, a thing of the times, but I used to go for a walk when someone put on a Dylan album.

What was song writing ever about? Look at folk. Mostly glorifyng the ordinary issues of human life. The record industry meant compressing it into three minute doses. The period of which you speak brought in social issues (although Dylan followed others), the mystical, the psychadelic --- and, sometimes, tried to take a slightly more informed view on relationships.

The best lyrics stand in their own right and can be read as poetry. That is not to say that, even then, they necessarily have anything of value to say or even make sense!

In the secret space of dreams, where,
Dreaming, I lay amazed...

(My favourite song line)

 

[jls001]

Even some of my favourite Grateful Dead songs really are absolute crap ("Transative nightfall of diamonds?" WTF!)

The other day I picked up a Grateful Dead album (Terrapin Station) for the nice cover. I haven't listened to it yet.

If anyone's heard it, please give a quick and dirty review - outstanding, excellent, good, average, bad, crappy, avoidable?

 

[gobble]

Why do two instruments playing the same note always sound louder than one, and never cancel each other out? Because, although they may be playing the same fundamental note, and even although they are the same sort of instrument, the harmonics of each one will vary slightly.

I believe only if they are exactly 180* out of phase... Else the will always add up. Harmonics being the same or not.

Sent from mobile on a crappy keyboard. Pls excuse typos.

 

[ajay124]

Joshua

Terrapin Station and Blues For Allah are fairly nice albums. But personally the Dead albums I still listen to are An American Beauty, Workingman's Dead and Live Dead.

 

[Thad E Ginathom]

The other day I picked up a Grateful Dead album (Terrapin Station) for the nice cover. I haven't listened to it yet.

If anyone's heard it, please give a quick and dirty review - outstanding, excellent, good, average, bad, crappy, avoidable?

I don't even remember the songs on it. Really not sure if that is a reflection of their worth or my memory!

Wake of the Flood is my favourite of later albums.

My Dead Essentials:

Trippy psychadelic: Aoxa Moxoa; Anthem of The Sun

Sweet Country/Folk: American Beauty; Workingman's Dead

Live: Grateful Dead; Live Dead

 

[alpha1]

Oh, that's just a fancy way of saying sudden changes from quiet or average to very loud, or vice versa.

Acoustic music doesn't have any difficult to encode fast transients.
Except for castanets.
Electro music - since is artificial - it can have very difficult to encode transients limited only by the sampling rate of the encoded signal.

Anyway, since none of us here knows about the real encoding method,
we'll have to wait unless someone comes along and explains why classical music uses less bits compared to electro music.

I am trying to read Perceptual Coding: How Mp3 Compression Works
Hope it clears up certain things.

 

[doors666]

The other day I picked up a Grateful Dead album (Terrapin Station) for the nice cover. I haven't listened to it yet.

If anyone's heard it, please give a quick and dirty review - outstanding, excellent, good, average, bad, crappy, avoidable?

Almost all of the dead studio albums are pretty much crap. you want to listen to dead, listen to their live shows. You can buy some as albums (dicks picks, europe 72 etc). most of the dead live shows are available as free music online on various sites (its legal). archive.org has thousands of shows, you can download them or stream them. iirc soundboards are only streamable from archive.org, but can be downloaded from elsewhere (again, legally).

oh, and its all flac.