Rickresource Rickenbacker Forum

A few disjoint thoughts:

A huge percentage of the music I love was recorded to tape. It sounds fabulous.

The only full-time professional producer I know insists on using digital technology exclusively. He has a huge commercial workload and loves the reliability and convenience that digital offers. He could not meet his deadlines and feed his family if he was using tape. (He is also a fabulous musician and has some great album credits to his name.)

My current favourite album was recorded on an ipad.

Choice of medium is not going to make anyone a better artist. If I'm a dud on Protools I'll be pretty much the same dud on tape (experience speaking here). If I'm a truly great artist, I'll be able to harness the quirks of a given medium to my advantage, and my understanding of these quirks may well influence my choice of medium.

What with people listening to music downloaded to their computer then to their iPods through ear buds, does fidelity even matter to most?

Just a quick thought here on reliability and convenience of computer recording. Can't computers break? Files get lost or accidentally deleted? Most equipment is likely as reliable or convenient as the person operating it.

marc61 wrote:Just a quick thought here on reliability and convenience of computer recording. Can't computers break? Files get lost or accidentally deleted? Most equipment is likely as reliable or convenient as the person operating it.

Yes, but one beauty of digital storage is the ease with which you can make perfect backup copies to take care of such eventualities...

cjj wrote:

marc61 wrote:Just a quick thought here on reliability and convenience of computer recording. Can't computers break? Files get lost or accidentally deleted? Most equipment is likely as reliable or convenient as the person operating it.

Yes, but one beauty of digital storage is the ease with which you can make perfect backup copies to take care of such eventualities...

Absolutely, plus computer hardware is readily available, relatively cheap, compact, and even I can replace a hard drive. Digital studios have layers of redundancy so if one piece of hardware fails, (typically a disk drive) another will be brought on line and down time is minimal. This is harder to do with analogue equipment. It's these features that have made digital almost ubiquitous, despite the disadvantage in fidelity that Mitch and Jeff have identified.

Ah, but there's absolutely no reason for a loss of fidelity with digital recording. Sure, MP3 uses lossy compression, but there are numerous lossless formats out there. And with a high enough sample rate, you can reproduce anything within the hearing of even the best of dogs, let alone people.

Of course, this does take up more storage space. But, I just bought a 1TB (that's TeraByte, 1000 GigaByte) USB disk drive for about $100, so storage is CHEAP these days. In other words, the only reason to use space saving formats is when you don't have a lot of space to begin with - devices such as MP3 players with limited NAND Flash memory. But NFlash is getting to the point where it's capacity vs. price is pretty darned good too, and that will only get better in the coming years...

Yes, isn't it weird that nearly 30 years after the introduction of the CD the standard of digital audio delivery is a lossy format? mp3s do sound drastically better than they used to, of course. A shame DSD didn't have a chance. Even in the studio world it's hard to find anybody using sampling rates above 48KHz, which is completely mysterious when vast storage is essentially free! With good converters, all the usual sampling rates sound fine, but I use 96KHz because, why not? And most Pro Tools engineers I know seem to think that using 96KHz is some kind of wild, spendthrift behavior. I halfway expect them to start telling me how tough it was in the Depression!

Well, sampling rate is a strange thing. The Nyquist theorem states that in order to accurately reproduce a signal you need to sample at twice te maximum frequency content of the material that you wish to reproduce. Now, that is for a sine wave, higher harmonics will require higher sample rates.

So, the accepted standard maximum frequency that humans can hear is 20kHz, so a 40 kHz sample rate should be fine to reproduce a 20 kHz sine wave. But wait! You can't produce any harmonic content above 20kHz with a 40kHz sample rate!

Well, if you can't hear anything above 20kHz, it doesn't really matter does it? Hence the 44.1kHz sample rate used for CD audio. That can accurately reproduce frequencies up to 22.05kHz, which is near the top end of what most humans can hear (I used to be able to hear up to about 30kHz, but the closer I get to 60, the less I am able to hear that high frequency stuff, I mean those high frequency ring things for smart phones that only young folks can hear? Well I can tell something is there, but can't really hear it anymore).

So, does it really make sense to sample much above 44.1kHz? Well, for the vast majority of people on this planet, probably not. If you lose some higher harmonic content, it's probably safe to assume that most people couldn't hear it anyway, so there's no real loss...

Ah yes! There's frequency response and there's there's something else... I certainly don't claim to have fantastic frequency response in my hearing, but things that spec out the same don't necessarily sound the same, and to some extent that's what keeps all this fiddly stuff interesting. The most intriguing article I read about sampling rates and perceived sound quality was a test done in the UK with a wide range of listeners, and sampling rates up to 192 KHz. There was a pretty strong correlation between higher sampling rates and perceived good sound quality even among the oldest listeners, but what people said was surprising. A lot of them liked the low-end quality of the recordings made with higher sampling rates. A finding like this might cause some people to reject the entire test, of course, but from what I read, it was pretty carefully arranged to give meaningful results.

In the end, an expensive converter at 44.1 is likely to sound better than a cheap one at 192.

But yeah, I'm the 1st to admit that I just want to listen to whatever is pleasing. I have some Altec amps that have a lovely sound and they are rated at something like 5% THD. A graph of analog tape playback of a short-duration pulse looks like a haystack, it has so many harmonic artifacts, but it can sound very clear! I think some of this randomness is what we like to hear. It's kind of like strumming a guitar vs. repeating a sample of the perfect strum...

There's a lot of important detail that goes along with digital recording technology and most of it seems to be ignored or misunderstood. The Nyquist limit will bite you if you have any frequency content above 2x prior to sampling. Input filter design is critical to avoid aliasing in the recording which will be impossible to remove later. Playback (reconstruction) also has these requirements. Nyquist also only guarantees frequency, not phase. To accurately record the signal including the relative phase of the higher frequency components requires more than 2x the highest input frequency. Of course none of this applies with an analog signal path and tape for recording.

A couple of friends and I have a studio set up with some nice gear we've cobbled together. One guy has an MC JH24 tape machine, the other has a ProTools setup with some nice converters. We have all pitched in on mic preamps, outboard gear, mics, etc. It's nowhere near the league of the Fidelitorium (would love to see it some day), but it beats a bedroom/living room arrangement by quite a bit. Good equipment and live rooms with high ceilings and some acoustic treatment makes a huge difference. We often track to tape and then transfer to digital for overdubs and mixing. Best of both worlds?

MP3s are a bummer. ****** cell phone voice quality also gets me down. Seems like everything has gone cheap and convenient with little to no attention to quality and fidelity. I'd pay a little extra to be able to buy SACD (or other better than CD quality) format music. For 99% of the music out there it just isn't available. It's not like we're asking for flying cars or rocket packs, either. This stuff is not rocket science in the age of Gigahertz processors in your pocket.

bumpity bump

I just bought an Akai GX-646 reel to reel from a friend who is selling off all ther old studio gear. It's near mint and works great. Think I might try to sell it though. Not sure yet. I listed it on CL but no bites yet.

A friend of mine in the Tampa Bay area runs a recording studio and explained the whole 96K thing to me once when I asked similar questions. I can't remember it well enough to repeat it, but it totally made sense when he explained it. Something to do with what happens to low frequencies, I think.

The sampling rate affect the high frequencies as more sample per cycle/Hz produces greater accuracy. With the CD Red Book (RB) standard of 16 bit/44.1 kHz sampling rate you only get down to around two samples at the extreme high frequencies. With a 96 kHz sampling rate, that more than doubles (still not fantastic), although down converting the the CD RB standard is not as accurate as, for example, converting from 88.2 kHz.