Head-Fi.org › Forums › Equipment Forums › Sound Science › Infrasonics / Sub-Bass and reproduction (and I may rant a bit)
New Posts  All Forums:Forum Nav:

Infrasonics / Sub-Bass and reproduction (and I may rant a bit)

post #1 of 8
Thread Starter 
First off, apologies all around for ranting, and I really would like a discussion on this. I have a question, which I shall ask in a round about way. The recordings I lissen to on a regular basis "should" include infrasonics/sub-bass (those frequencies below 30Hz). I know for a fact that I have a recording of Diane Bish playing Anton Bruckner's Two Improvisations on an unnamed Germanic organ. This specific recording produces a C0 which (depending on the tuning of the 64-foot) pipe, should be around 16Hz. My headphones will reproduce this tone, which sounds fantastic. More examples: those fantastic Beethoven piano recordings were played on a Bösendorfer Imperial Grand, which has a low C of 16.35Hz, this is also the theoretical limit of a Tuba. My question regards the frequency limitations of playback devices in whatever medium you choose. I want to talk specifically about CD's and SACD's. Some (probably) old dead dude decided just to limit everything from 20Hz to 20,000Hz... The threshold of human hearing. That's all fine, but I want more. There are microphones that are rated to record sub-bass frequencies down to 4Hz. Vinyl is reported to reproduce well up past 60khz. But that's analogue you say! Yes, I do want this discussion to focus on digital reproduction. I have been trying to find a stand alone CD player that will give me the fidelity I want. I also strive to find recordings that are faithful to the instrument. Since I'm a classically trained musician, I am intimately familiar with the tone qualities and sonics of specific instruments. While I am most interested in two channel audio, I don't really need the multi-channel capabilities of Super-Audio CD's. The SACD is reported to sweep from 20-50khz. While the expanded headroom is nice, why has the bass frequencies not been extended as well? Is there some sort of electronic or solid state law that stops chips from reproducing low frequencies? Or is it a function of digital clock divisibility? Of course, there could be a very simple answer to my problems. If someone knows of digital equipment with rediclous specifications, such as the Schiit Lyr amp, or most of the Carver line of CD players, which sadly are hard to find for me. Also, since I'm new here, I'm sorry if this has been discussed thousands of times. I have searched for any references regarding this topic, but I haven't found any specific answers.
post #2 of 8
Generate a 16hz tone in audacity or whatever sound editor you use, burn it to an ACD, play it on your receiver and/or rip it back, to see whether 20hz is a real limitation. Above 20k is usually filtered out for ACD, I'm not so sure about below 20hz...

It seems 20hz was chosen as it's commonly cited as the audible limit for most people and frequencies below are reportedly felt, not heard. (also, chances are not a lot (very few) speakers or HP will be able to reproduce it)
post #3 of 8

It's more likely your speakers/headphones can't reproduce a 16hz tone at a audible level, you need at least a 10" sub or large headphones to play 16hz at a good level of volume, my old philips cd630 cd player i brought second hand is rated at 2hz-20khz , my cheapo samson servo speaker amp is 5hz-60khz it can go lower and higher but thats it's -3db rating at it's max power output of 300watts, at normal listening levels it can do 1hz without a volume drop, my dac/headphone amp the fiio E9/E7 is 10hz - 100khz and half the price of the Schiit Lyr amp. 

post #4 of 8
JohnRichard, 20Hz to 20kHz is a range that is described as audio frequencies, because this is approx. the range of frequencies people can hear.

When some piece of equipment is specified with '20Hz - 20kHz' it doesn't meant that outside of this range the signal is completely filtered out. Microphones can easily pick up frequencies above 20 kHz and also below 20 Hz. A DC coupled amp can output signals down to a 1 Hz with negligible drop in volume level. Most speakers and headphones have a hard time reproducing such low frequencies though. If you play a 16 Hz sine wave through your headphones and hear something earthquake-like doesn't mean that you're hearing the 16 Hz sine wave. Usually harmonic distortion increases a lot with lower frequencies and you might also hear mechanical noise coming from the driver. Another problem can be power requirements (see sub-woofers).

The standard format of CD audio uses a sampling rate of 44100 Hz and due to the sampling theorem only frequencies up to half of the sampling rate (=> 22050 Hz) can be contained within the sampled signal. Depending on your equipment you might see a high frequency roll-off starting with around 20kHz due to a low pass filter. On the low end an AC coupled amp with small output capacitors will cause severe roll-off (like a high pass filter). If you want music that contains 'something' above that 22.05 kHz limit you can buy audiophile records that use sample rates up to 96 or even 192 kHz.

Another thing to mention is the mixing/mastering process. If the record is targeted at the average consumer you have to optimize it for playback on ear-buds, integrated TV-speakers etc.. And if you don't filter out the lowest frequencies you cannot compress the dynamics (and therefore increase loudness) as much as some people in the industry are forced or want to.
Edited by xnor - 4/27/11 at 9:39am
post #5 of 8
Thread Starter 
Quote:
Originally Posted by xnor View Post

JohnRichard, 20Hz to 20kHz is a range that is described as audio frequencies, because this is approx. the range of frequencies people can hear.

When some piece of equipment is specified with '20Hz - 20kHz' it doesn't meant that outside of this range the signal is completely filtered out. Microphones can easily pick up frequencies above 20 kHz and also below 20 Hz. A DC coupled amp can output signals down to a 1 Hz with negligible drop in volume level. Most speakers and headphones have a hard time reproducing such low frequencies though. If you play a 16 Hz sine wave through your headphones and hear something earthquake-like doesn't mean that you're hearing the 16 Hz sine wave. Usually harmonic distortion increases a lot with lower frequencies and you might also hear mechanical noise coming from the driver. Another problem can be power requirements (see sub-woofers).

The standard format of CD audio uses a sampling rate of 44100 Hz and due to the sampling theorem only frequencies up to half of the sampling rate (=> 22050 Hz) can be contained within the sampled signal. Depending on your equipment you might see a high frequency roll-off starting with around 20kHz due to a low pass filter. On the low end an AC coupled amp with small output capacitors will cause severe roll-off (like a high pass filter). If you want music that contains 'something' above that 22.05 kHz limit you can buy audiophile records that use sample rates up to 96 or even 192 kHz.

Another thing to mention is the mixing/mastering process. If the record is targeted at the average consumer you have to optimize it for playback on ear-buds, integrated TV-speakers etc.. And if you don't filter out the lowest frequencies you cannot compress the dynamics (and therefore increase loudness) as much as some people in the industry are forced or want to.

I did take a look at the audiophile recordings with the higher sample rate, and find them very nice!  I mostly record and playback analogue on old fantastic equipment. Not dealing with sample rates (usually), I never fully understood the deal with the complexities.  Thanks for the brief explanation. 

 

I guess I can just go about doing what I have always done, finding the very best recordings possible, and if they are at the higher sample rates (if digital), I'll know that, even if it is pointless, the stuff I want in my recordings is there.
 

 

post #6 of 8
>On the low end an AC coupled amp with small output capacitors will cause severe roll-off (like a high pass filter).

That's a good point as well. Any chance you remember the i/o impedance,capacitance and lf roll-off formula?
post #7 of 8
Quote:
Originally Posted by JohnRichard View Post

I did take a look at the audiophile recordings with the higher sample rate, and find them very nice!  I mostly record and playback analogue on old fantastic equipment. Not dealing with sample rates (usually), I never fully understood the deal with the complexities.  Thanks for the brief explanation. 

 

I guess I can just go about doing what I have always done, finding the very best recordings possible, and if they are at the higher sample rates (if digital), I'll know that, even if it is pointless, the stuff I want in my recordings is there.

It's not really complex. The analogue waveform is measured every T seconds, this is called sampling. The sampling rate is defined as 1/T (with the unit Hz) and tells how many times per second a sample is obtained.
 

300px-Signal_Sampling.png

 

Imagine a signal that changes very quickly from high to low to high to low and so on (= high frequency). We need at least two samples to represent that waveform: one for the high and one for the low value.

If we used a sampling rate of 44100 Hz this highest possible frequency would be (44100 Hz / 2 samples) = 22050 Hz.

If we used a sampling rate of 96000 Hz we could store frequencies up to 48000 Hz which is way above our hearing range.

 

If your music is not digital then this doesn't matter of course. Anyway, how well it sounds is up to the guys who produced the record, pretty much regardless of the format I'd say.

 


@svyr:

Take a look at http://sim.okawa-denshi.jp/en/Fkeisan.htm.

 

 


Edited by xnor - 4/28/11 at 6:02am
post #8 of 8

Most of the replies so far have been great - but I'd like to add a bit.

 

So far, everyone has established that the maximum reproduceable freqency from a digital medium is limited by the Nyquist-Shannon sampling theorem.  It's worth noting, however, that DACs usually have low-pass filters to prevent high frequency transmission beyond the maximum allowed by the intended sampling rate - and this can be/is in amplifiers as well.

 

The reason there's a lower bound for frequencies - other than the meaningless slap-on "20 Hz - 20 kHz" label - is because preventing DC pass-through is important.  A high power constant DC voltage can damage drivers (Someone correct me if I'm wrong, but this is what I understand: constant DC current = no effect of the driver impedance on the DC resistance = high current (thus power) transmitted = possible driver burn-out.) - often loudspeakers have a high-pass DC decoupling capacitor in the crossover to remove that risk, but headphones don't have that luxury.

 

So, a high-pass filter that gets rid of everything below 5-10 Hz or so isn't uncommon, because those really low frequency waves start to act like DC.  Some rolloff before that certainly isn't uncommon, especially in extremely inexpensive equipment.

 

However, you could encode a DC signal (i.e. 0 Hz) into PCM audio (i.e. CDs, etc.) if you really wanted...  There's nothing stopping that.  It's just the following equipment.  Most of the time, all of your components (if competent) will be fine down to below the lowest 16 Hz musical content - except the transducers.  Oh, and your ears' greatly reduced sensitivity to such low frequencies.


Edited by BlackbeardBen - 4/28/11 at 7:10am
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Infrasonics / Sub-Bass and reproduction (and I may rant a bit)