the importance of analog sound in prog

 

If we convert the 1:10,000,000,000,000 range into dB (see previous link for details) that would equate to 130dB - to put that into context 24-bit audio is 144dB, 16-bit is 97dB, pro-tape is 60dB and vinyl is less than 50dB. We can just about detect a change in volume of 3dB (this figure is used a lot in audio engineering for this reason). This means if we slowly increase the volume from the threshold of hearing to the threshold of pain and asked people to make a tick on a piece of paper when they noticed a difference in volume at the end of the experiment we would average around 43 tick marks (130÷3) - obviously this is a lot less than the 65,535 different signal levels that 16-bit audio can produce. If we replaced the volume control with a rotary switch with 65,535 positions we would go through 1,500 switch positions between tick marks. In reality if you wanted to replace your standard stereo volume control with a rotary switch then fewer positions are needed - since music is not listened to over the full range of hearing 24 positions is sufficient (see here for engineering theory put into audiophilist practice).

 

______________________________________________________________________________________

 

I don't know whether the ear can detect differences as small as 6ns, I can find nothing on the interweb to support this and it's outside the wealth if useless knowledge that my brain is filled with - let's say I'm highly sceptical.

 

6ns (6 nano-seconds or 0.000000006 seconds) is an incredibly small time duration. I think that we tend to get a little blasé about incredibly big and incredibly small numbers because they are difficult to imagine - 6ns or 6µs or 6ms are all "small" and that's as far as we comprehend them even if 6ns is a million times smaller than 6ms because the concept of 1 million is too big to comprehend most of the time too. Nothing I know of in physical nature can react in 6ns: travelling at the speed of light (300 million m/s) light travels 1.8m (~6 feet) which means that in 6ns the light from your ceiling light hasn't reached the floor yet; a sound wave travelling through air at 300 m/s will have travelled 1.8 micron (1.8µm or 0.0018mm) ...  1.8 micron is at least ten times smaller than the thickness of a human hair. To be able (or capable) of detecting a 6nS difference in sound waves the ear would need to be capable of measuring the distance between two pressure waves of 1.8 micron and I admit I do not know whether that is possible, I am highly sceptical, but receptive to plausible explanations.

 

 

Ah, I started to listen to the video and gave up before they got to that part (people make glaring errors in interviews that they would not make in the written word, for example Sean Bevan said 44.1 sampling samples 44.1 times a second, when he meant 44,100 times a second - a simple slip of the tongue that he would not have made in print). Sampling at 44.1K equates to a sampling interval of 22.7µs which would yield 264 samples within 6ms, therefore 16/44K1 is perfectly capable of capturing differences as fast as 6ms.

 

However in the interview (which I have now listened to) he said 6µs, not 6ms (another slip of the tongue perhaps, but perhaps not because he is correct in saying 44.1KHz cannot capture 6µs), but that is also incorrect because our ears cannot resolve differences of 6µs - the best we can manage is around 600µs and only at a narrow range of high frequencies (certainly not drum kicks) - sampling at 44.1KHz can capture differences 600µs without problem.

 

Our ears are separated by roughly 0.15m and that corresponds to a time difference of 2ms for sound travelling at 300 m/s - how well we can locate sound (which is what we mean by detecting those differences) is determined by how much phase-change can occur in that time. So how well we can locate sound using that 2ms "resolution" is dependant on the frequency of the sound - we cannot locate low to lower-mid frequencies at all, (really, we're rubbish at it - that's why you only need 1 sub-woofer in a system) - because the phase-change in 2ms of lower frequencies is too small for us to detect - this is related to the wavelengths of sound but I won't bore everyone with that.

Aquiring the Taste wrote: Much bigger considerations like, the quality of the digital transfer, the mid to late 70s Analogue  " loudness Wars" , the Digital loudness Wars, error correction, block-wall at 22khz, & the threat of Digital Remastering.

I think the effect of 22KHz brick-wall is over-stated - we're talking about psychoacoustic phenomena that has already happened before the recording was made, removing the supersonic frequencies that cause those "subtle perceptions" after recording will not remove the effect of them.

 

The quality of the digital transfer, the mid to late 70s Analogue  " loudness Wars" , the Digital loudness Wars, error correction, & the threat of Digital Remastering are all to do with how the technology is used, they are not limitations of the technology itself. But in principle I agree with you - using a technology badly (regardless of what that technology is) is much more of a concern.

 

Remastering and remixing are two different things that unfortunately have become colloquially termed "remastering" in marketing speak. The mastering stage is the final step in the process that occurs after the mixing that tailors the sound to suit the media that it will be recorded on to, ideally this should be done by a mastering engineer who wasn't involved in the mixing process - the mastering for vinyl is different to the mastering for CD - when transferring from one media to another the final mix should be remastered for the end media. [listen to the Sean Beavan interview that Mr "Aquiring the Taste" posted - he explains it rather well, and so he should - he's a brilliant producer even if he did make a few slip-of-the-tongue numerical errors in the interview].

 

Remixing is different process, and while I think that granny's stew can be altered to suit personal tastes, I do agree that antique recordings should not be altered to suit modern tastes of drum sound or guitar tone of whatever - unless the original artist or producer is present at the mix - they know what sound they wanted and they should be permitted to achieving or maintaining that on any remixing. The Picasso analogy is inaccurate because the remix/remaster does not destroy the original - you can always go back to an earlier release, you don't have to buy the remaster.

 

Some analogue mixing desks have automation (that's where the idea came from) and it is just a tool for the mixing engineer to use - live mixing is such fun, but having that "performance" recorded as automation is no different to recording and using the best take of a guitar solo.

 

But in general I agree with you: Feel the musicians, feel the music, feel the time.. and listen on whatever format you damn well like.

Nothing you can do can make that system pump out a continuous 1280 W of power. It isn't a matter of efficiency or cooling because even running at 100% efficiency the amplifiers would only deliver 220 W into the speakers. Also the speakers themselves are not designed to handle that much continuous power ... the speaker in your sub is 5.25 inches in diameter ... do you really believe that can handle 550 W of continuous power?

 

The amplifiers can deliver 1280W into the speakers for a very short duration of time - for a sharp peak in the music. This instantaneous power comes from reservoir capacitors in the power supply not from the incoming supply - these capacitors cannot provide continuous power ... you use all their energy in the peak then recharge them during the quieter bits ready for the next peak.

 

I believe the manufacturers are being dishonest when they quote peak music power - this is a marketing ploy that is pretty much exclusively confined to the home theatre market and 5.1 speaker system - the hi-fi market generally quotes RMS or continuous power.

 

 

further reading: http://en.wikipedia.org/wiki/Audio_power#PMPO