View previous topic | View next topic

Music

Page 6 of 8
Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next

suze
1108507.  Thu Jan 01, 2015 8:57 am Reply with quote

That statement isn't very precise, is it? But what I think is meant is this.

Let us suppose that we have our trumpet (let us say) connected to an oscilloscope, and so we can see the waveform on screen. It's easy enough for us to measure the fundamental period - that is to say, the time period between successive peaks of the fundamental frequency.

The fundamental frequency of a note played on a trumpet is of the order of 500 Hz, so that fundamental period will be around 0.002 sec.

We can then use software to delete the first .002 / 25 sec of each cycle. I'm not quite sure which software - Audacity is nothing like so precise - but I suppose we could slow the recording down by a factor of one thousand and then speed it up again when we're done deleting.

The next question: what does first mean here? To me it ought to mean that part of the cycle immediately after the line crosses the x-axis with positive gradient, but I'd want clarification of that.

Anyways, the contention is that if we do this, we will no longer be able to recognize the instrument as being a trumpet. I'll defer to the musicians among us as whether this is true or not, but I have to say that I'd be surprised if it is true.

 
WordLover
1108529.  Thu Jan 01, 2015 1:52 pm Reply with quote

.....


Last edited by WordLover on Fri Sep 16, 2016 12:21 pm; edited 1 time in total

 
Leith
1108539.  Thu Jan 01, 2015 3:21 pm Reply with quote

The transients referred to are the onset, or 'attack' of a musical note, suze, rather than the beginning of a wavelength period.

For a plucked or hammered instrument like a guitar or piano, the character of the sound is heavily dependent on that initial percussive part of the note's waveform (as you might imagine).

I can run my guitar through a filter that suppresses the volume during the attack of a note, with a result that sounds more like a violin than a guitar. I don't know exactly what timescale this filter runs over, but I could believe that 40 ms would be sufficient for a pretty noticeable effect.

http://en.wikipedia.org/wiki/Synthesizer#ADSR_envelope

Suppressing or excising a segment from the repeating waveform at a note's fundamental frequency would have a quite different effect on the timbre of a note. I don't know what sort of effect exactly, but probably one that would be more noticeable on non-percussive instruments like a church organ.

 
suze
1108615.  Fri Jan 02, 2015 7:56 am Reply with quote

Thanks Leith, that makes sense.

I'm still dubious that excising 1/25 of the sound - whatever that does exactly mean - would render the instrument unrecognizable.

If the 4% excisions were at random rather than at prescribed point, I'm confident that we'd still be able to recognize the instrument.

Speech can survive the random excision of one quarter to one third of the content before it becomes incomprehensible. This is the reason why air traffic controllers use standardized terms which in some circumstances the pilot repeats back, and it's also why BBC World Service radio on shortwave was a thing.

The way that mathematical linguists - an eccentric lot who have a tendency towards wearing bow ties and/or being Russian - interpret Zipf's Law would suggest that this ought to be true of other forms of sound as well.

 
ConorOberstIsGo
1108665.  Fri Jan 02, 2015 2:34 pm Reply with quote

Anybody ever heard of Binaural Beat? It's a kind of wowowowowow effect that you get when playing two constant tones of very slightly differing frequencies. It's a lovely a bizarre effect; if you take out one earphone there's no wobble to the sound at all; only listening to them simultaneously in each ear causes this. It's used in hypnosis and other odd West-coast therapies.

 
14-11-2014
1108725.  Sat Jan 03, 2015 1:55 am Reply with quote

Wikipedia wrote:
Mia

After "Anja", several other singles were released from the debut album 'Gorky'. Among them was 'Soms vraagt een mens zich af' (Sometimes a man wonders), the b-side of which was 'Mia'. By making number one in the 'Tijdloze' (the Timeless), the 'best of all times' charts in Flanders by the music station Studio Brussel, for three consecutive years, 'Mia' became known as their largest hit. Likewise, it made number 1 in a similar chart of Radio 1 in Flanders, and named best Belgian song on the chart by Radio Donna in 2005.

In 2006, it was elected the best song about girls on a TV-show on the public network één. In 2008, the early music consort Capilla Flamenca issued 'Rosa (Mia)', a polyphonic adaptation of 'Mia' sung in Latin. In 2008 a new award show was founded called "de MIA's" (the MIAs). It was named after the best Flemish song of all times and officially stands for the Music Industry Awards.


If Flemish is a language on its own, then the lyrics of the B-side song Mia are 100% Dutch. If you know that, and you know that the theme of the song isn't a Belgian theme, and you know that Mia apparently is the best Flemish (pop)song ever, then it's quite remarkable that this song is not known in the Netherlands. It's almost like having to explain in Scotland what The Beatles or Oasis are.

Milow's Mia, in English with lyrics. The language or the border doesn't explain why Mia is almost unknown there, because e.g. Dutch tv did broadcast a long interview with Stromae (@ 5:29) in English. Perhaps Stromae's famous songs are better and more international, but that still doesn't explain why the song Mia isn't known in both countries.

 
WordLover
1108726.  Sat Jan 03, 2015 2:32 am Reply with quote

.....


Last edited by WordLover on Fri Sep 16, 2016 12:20 pm; edited 1 time in total

 
14-11-2014
1108728.  Sat Jan 03, 2015 2:40 am Reply with quote

suze wrote:
Anyways, the contention is that if we do this, we will no longer be able to recognize the instrument as being a trumpet. I'll defer to the musicians among us as whether this is true or not, but I have to say that I'd be surprised if it is true.


Yes, and perhaps be surprised? It was a reviewed science quiz question of a scientist, compiled by an association of scientists, with broadcasted evidence (link broken). I'm sorry I don't have a better source.

Q, sloppy translation: you play a well-known melody with an instrument and record it. Next you cut off 1/25'th of the transients of the tones. Someone listening to the recording ...

A, sloppy translation: ... cannot recognize the instrument anymore.

The remaining part of each tone is too stable to recognize it, and the first 1/25'th of the transient of a tone contains the most information. You will be able to recognize the melody, and so on, but not the instrument.

 
Leith
1108744.  Sat Jan 03, 2015 6:21 am Reply with quote

suze wrote:
Thanks Leith, that makes sense.

I'm still dubious that excising 1/25 of the sound - whatever that does exactly mean - would render the instrument unrecognizable.

If the 4% excisions were at random rather than at prescribed point, I'm confident that we'd still be able to recognize the instrument.

Speech can survive the random excision of one quarter to one third of the content before it becomes incomprehensible. This is the reason why air traffic controllers use standardized terms which in some circumstances the pilot repeats back, and it's also why BBC World Service radio on shortwave was a thing.

The way that mathematical linguists - an eccentric lot who have a tendency towards wearing bow ties and/or being Russian - interpret Zipf's Law would suggest that this ought to be true of other forms of sound as well.


Yes, the effect I'm familiar with only works by removing specifically the percussive bit of a percussive note. This is perhaps not surprising when expressed in those terms, but the effect is quite striking nonetheless. The amount of sound you'd have to suppress would probably vary by instrument.

Interesting statistic about speech recognition. I get the impression that this works precisely because the listener has a vocabulary of expected word frequencies and sentence forms that they can use to 'fill in the gaps' of missing tonal information. I imagine that recognising the specific sound qualities of a particular instrument really requires the raw tones themselves - they're not being used as a 'carrier' for a higher form of information in quite the same way here.

 
Leith
1108746.  Sat Jan 03, 2015 6:33 am Reply with quote

WordLover wrote:
ConorOberstIsGo wrote:
Anybody ever heard of Binaural Beat? It's a kind of wowowowowow effect that you get when playing two constant tones of very slightly differing frequencies. It's a lovely a bizarre effect; if you take out one earphone there's no wobble to the sound at all; only listening to them simultaneously in each ear causes this. It's used in hypnosis and other odd West-coast therapies.
Yes, the phenomenon is well-known among instrumentalists and instrument-tuners: it's "beating" (you don't have to feed one pitch into just the left ear and the other into just the right).

If binaural beating is used to hypnotise people, that's QI -- I wonder if the idea is that the person to be hypnotised is to fixate on the beating.


I was wondering what was special about the 'binaural' bit, too. Having just looked it up, I see that binaural beats aren't caused by audio interference patterns; they're a mental effect in which the brain appears to recreate the beats that would occur if the sounds from each headphone speaker were played together monophonically.

The wiki article mentions a study finding that stroke victims with aphasia couldn't hear binaural beats.

 
suze
1108753.  Sat Jan 03, 2015 7:28 am Reply with quote

14-11-2014 wrote:
If you know that, and you know that the theme of the song isn't a Belgian theme, and you know that Mia apparently is the best Flemish (pop)song ever, then it's quite remarkable that this song is not known in the Netherlands. It's almost like having to explain in Scotland what The Beatles or Oasis are.


Isn't it more like a song by a Scottish or Irish band being well known in Scotland or Ireland but not in England?

And there absolutely are Irish musicians who are well known in Ireland but not in England. For sure, many of those play traditional Irish music or country music - genres which are popular in Ireland but not in England - but there are plenty of unremarkable pop singers whose popularity is confined to Ireland too.

As for why Stromae is well known internationally but Gorki are not, isn't that simply because Stromae performs in French rather than Dutch? You're doing well if you have a hit in (say) Switzerland with a song in Dutch, but French is rather easier. (Mind you, Alors on danse was actually a minor hit in the UK as well. Songs in French have rarely done well in the UK, so maybe it is just that he's good!)

 
Spud McLaren
1108785.  Sat Jan 03, 2015 10:13 am Reply with quote

WordLover wrote:
ConorOberstIsGo wrote:
Anybody ever heard of Binaural Beat? It's a kind of wowowowowow effect that you get when playing two constant tones of very slightly differing frequencies. It's a lovely a bizarre effect; if you take out one earphone there's no wobble to the sound at all; only listening to them simultaneously in each ear causes this. It's used in hypnosis and other odd West-coast therapies.
Yes, the phenomenon is well-known among instrumentalists and instrument-tuners: it's "beating" (you don't have to feed one pitch into just the left ear and the other into just the right).

If binaural beating is used to hypnotise people, that's QI -- I wonder if the idea is that the person to be hypnotised is to fixate on the beating.
I haven't heard of it being used for hypnosis, but recordings are available that are used to induce particular brainwave patterns, notably for relaxation. I used to have one such recording, but seem to have lost it*. I don't know whether it actually works, or is quackery by placebo.

* see post 677617.

Regarding removing the attack of a note, it seems to be true for many instruments - but I think only those that employ a fair bit of attack in the first place. And you can (often) infer at least which type of instrument is playing by the decay of the note. However, I heard tenThing's BBC Proms concert and the leader, who's a trumpeter, is able to modify the attack on her instrument to make it easily mistaken for a saxophone. Youtube videos are available, but I can't remember the specific piece I heard.

 
'yorz
1108787.  Sat Jan 03, 2015 10:29 am Reply with quote

Found this binaural beat meditation music.

It comes with a warning.

 
Leith
1108788.  Sat Jan 03, 2015 10:39 am Reply with quote

'yorz wrote:
this

No headphones to hand, but I'm quite enjoying the comments :)

 
'yorz
1108793.  Sat Jan 03, 2015 11:05 am Reply with quote

Yeah. They're quite irreverent. :-)

 

Page 6 of 8
Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next

All times are GMT - 5 Hours


Display posts from previous:   

Search Search Forums

Powered by phpBB © 2001, 2002 phpBB Group