effects of "unnecessary" Bandwidth in a piece of playback electronics....

[louped garouv]

so i was thinking.... (a rare occurance, I can assure you)

i was pondering what effects would be of having "unnecessary" bandwidth available in a piece of audio electronics...

i notice people claiming frequency response of 5Hz to over 110KHz with gear a lot now-a-days.....

it got me to wondering what, if any, negative effects an extremely wide, and mostly unused, amount of bandwith could potentially have on a playback system...

it's not like any "normal" set up can do 5Hz... nor can most folks set-ups reproduce 45KHz, much less anything over 100KHz....




thoughts anyone?

[edgewound]

I guess that would get into the never ending argument of digital vs analog and which one is better, requires more filtering, computing power, 10 inch wide tape and a player to play it on, the effects...or affects....of quarks on phase delay, the relative position of black holes to the tangential orbital rotation of outlying galaxies, and how your wife's menstrual cycle alters your own auditory sensitivities.

In other words...let your ears ...and bank balance or credit limit decide.

[louped garouv]

:lmao:

let us say for just an instance, that the format (digital or analog) can be ignored...
purely an available, but largely unused, bandwidth issue....


and for the record, the piece i was thinking of is actually touted as being analog....
really expensive analog, but analog....

[Harvey Gerst]

Increased bandwidth help the slew rate, which in turn, helps the transient response,and improves clarity. A good thing.

Poorly designed equipment with increased bandwidth can oscillate at high frequencies, blowing out tweeters. A bad thing.

[louped garouv]

Thanks Harvey, that helps alot...

David

[JBL 4645]

Ether someone has ears like an elephant for lows and ears like a bat for highs LOL.

I heard there really isn’t anything below in the infrasonic range expect for garbage of noise? Has for highs I can barely still hear 16 KHz.

I had on lone once back in the early 1990's a dbx frequency expander that was quite nifty.

[Ian Mackenzie]

In terms of the worst case scenario the consensus for mic preamps is they should be designed to adequately handle the full gain (and voltage output) of the preamp @10 times the highest intended frequency to ensure the signal is clean at the upper end of the audio band (refer Benchmark Audio).

The problem is that without adequate GBP the pre will clip where the feeback and open loop gain curves cross and this can scew up the integrity of the pre amp at much lower frequencies.

Deane Jensen apparently measured up to 30 Khertz from a mic on some recordings so go figure.

[whizzer]

In Tchaikovsky's 1812 Overture, the fundamental frequency of the brass cannon shots is 6 Hz, which, in a live performance, is experienced as a physical shock wave. The human-perceived "sound" is all in the upper harmonics, and few systems will be able to reproduce even those harmonics (a least the lower couple) at sufficient amplitude for a "realistic" reproduction, but even ordinary musical speakers can give the "idea" of the cannon shots. At the other end of the spectrum, while 20 kHz is usually given as the upper "cutoff" of the human hearing range, there are some few who can sense sounds above that frequency, but, more often, there are very few who can hear tones that high. The high frequency range of most adults tops out in the 15-17 kHz region--my own trails off rather steeply above 13 kHz--they didn't use ear plugs on the rifle range when I was in the Army back in 1970. So, should the equipment itself need a bandwidth exceeding these limits? As usual, there's more to it than what we can hear, although what we can hear is the main thing that concerns us lovers of music. My experience has always been that the wider the bandwidth of the electronic equipment between the source and the speakers, the cleaner and more detailed the sound, but one would think that there must be some practical limit above which diminishing returns set in, or, worse yet, some sort of degradation comes along as a side effect. In practical terms, I've yet to hear an amplifier that sounds any better playing a given set of loudspeakers than the old Crown DC series; evidently, for at least my range of hearing, the bandwidth is wide enough.

[Jody]

Just to point out that although you may not be able to hear sine waves above (say) 16k, you can absolutely hear the DIFFERENCE between WITH the faerie frequencies, and without by using a low pass at, say, 30 khrtz and 16/17 khertz. vocals sound much more natural and less like a dodgey MP3.

[whizzer]

Exactly; the bandwidth of the equipment used for reproducing the music, for the very reasons stated above ought to exceed by some margin the bandwidth of both the music itself and the maximum range of hearing that can be expected of the listener; the question is "by how much ought it exceed that range?" The old (analogue) rule of thumb was that the equipment's bandwidth ought to be at least double that of the range encompassed by the range of the signal through it. In the digital age, the bandwidth for music reproduction is in terms of sampling rate. At first, 2x oversampling (approximately 44 kHz) was thought to be sufficient--it wasn't--strings sounded steely and compressed. The standard soon became 8x. Is that sufficient? Is 16x or 32X overkill? Is there a benefit to going beyond, and if so, by how much beyond? The idea is usually that more is better, but is that always the case? Or is there a point at which additional increases become pointless?

[Beowulf]

The problem with the 'how much bandwidth is enough' question is that transition-band behavior and in-band transient response are fundamentally joined at the hip . . . in particular, I remember the aforementioned Deane Jensen's work in defining that for optimum in-band transient response, the transition-band should correspond to a Bessel function.

So when bandwidth is increased and sonic benefits are claimed, I think it's probably mostly because transient and phase response is improved in the frequency range that is inarguably audible. This is definately, measurably the case with digital - virtually every filtering scheme introduces Gibbs-phenomonon ringing and/or aliasing into the audible spectrum . . . it's just a matter of how much, and where.

I once roughly calculated what would be required to have a linear PCM system with extremely flat response to 20K, aliasing artifacts buried in a 16-bit noise floor, and a Bessel transition-region response, and came up with a minimum of something around 800KHz. Don't know how accurate that is, but it's definately way higher than anything we currently have - which means that it's plausible for us to be able to hear sampling-rate improvements for quite some time.