apl
Stand Up Comity
I routinely make digital measurements down to DC.
w said:
Apparently you missed the proof that any signal can be described as a sum of pure sine waves, so that argument is also invalid.
Why analogue and not digital?
- Thread starter cjacek
- Start date
I routinely make digital measurements down to DC.
Apparently you missed the proof that any signal can be described as a sum of pure sine waves, so that argument is also invalid.
No, not all sounds are a sum of sine waves. Have you tried creating a square wave or even a triangle wave using sine waves? You can create a similar sound using sine waves but the outcome is not the same. It's just another outdated scientific theory that needs to be updated along with many others including tacheons and hydogen economy.
At any rate, I haven't even had time to follow this most excellent thread amidst several mastering jobs and doing location dialogue for a movie, not to mention playing with my band. I certainly haven't had time to listen to or recreate your examples much less create my own. In fact, this will have to be my last post for a while because it's another 14 hour day tomorrow.
But I must state that no recording format is perfect. Analogue is not perfect and no matter how much digitalphiles whine about Nyquist or dither, digital isn't perfect either. I prefer analogue, I think it sounds better as do almost everybody who hears it. But I use both technologies on a daily basis. For instance, the dialogue for the movie I'm booming is recorded on a digital medium. The reason for that is I can wear the equipment around my neck all day, then take it to my studio and drag the audio directly into the timeline of the project while simultaneously beeming it across the city to the editor's studio where he can do his work. I also can't afford a Nagra when I do so little film/video field work. When it comes time to record a rock band like next weekend, it'll be on 1" because I don't need portability but rather sound quality. So I don't care how much anybody whines about one format being superior, it's not. I prefer sound quality over convinience for most stuff. That's great for me. Some people can't handle a medium that forces people to try and perform rather than somebody else double clicking and typing their performance for them. That's fine for them I guess.
I'm just so sick of people making false claims that digital is perfect. It's not and never will be just like analogue is not and never will be. Anybody who makes such claims is an idiot, I don't care where they went to school. It's not just problems with converters, it's an enherant shortcomming of the format itself. Nothing will ever change that. DSD is a step in the right direction. It reduces most problems with digital down to jitter and long slew rates, oh and the fact that square waves are still loaded with ripple. But at least the nasty ringing and brick wall LPF goes away with DSD. It's a much better format than PCM. Though I must state I admire that many of the people here in this thread are willing to admit the shortcommings of this medium in its infancy. I've been witness to many such threads and this is the only one where people were not only willing to provide evidence but also take to heart other people's evidence. I remember one such thread a while ago where somebody just insisted he was right and offered no proof or reason why. He just quoted a book about recording stormed away never to be seen again.
Anyways, it's been fun discussing this stuff. When I get some more time, I'll generate some samples and post them here but it'll be a while. The problem is, everybody will be using different converters and monitors so they'll all hear something different. But in the mean time, I gotta get some sleep.
Night all.
At any rate, I haven't even had time to follow this most excellent thread amidst several mastering jobs and doing location dialogue for a movie, not to mention playing with my band. I certainly haven't had time to listen to or recreate your examples much less create my own. In fact, this will have to be my last post for a while because it's another 14 hour day tomorrow.
But I must state that no recording format is perfect. Analogue is not perfect and no matter how much digitalphiles whine about Nyquist or dither, digital isn't perfect either. I prefer analogue, I think it sounds better as do almost everybody who hears it. But I use both technologies on a daily basis. For instance, the dialogue for the movie I'm booming is recorded on a digital medium. The reason for that is I can wear the equipment around my neck all day, then take it to my studio and drag the audio directly into the timeline of the project while simultaneously beeming it across the city to the editor's studio where he can do his work. I also can't afford a Nagra when I do so little film/video field work. When it comes time to record a rock band like next weekend, it'll be on 1" because I don't need portability but rather sound quality. So I don't care how much anybody whines about one format being superior, it's not. I prefer sound quality over convinience for most stuff. That's great for me. Some people can't handle a medium that forces people to try and perform rather than somebody else double clicking and typing their performance for them. That's fine for them I guess.
I'm just so sick of people making false claims that digital is perfect. It's not and never will be just like analogue is not and never will be. Anybody who makes such claims is an idiot, I don't care where they went to school. It's not just problems with converters, it's an enherant shortcomming of the format itself. Nothing will ever change that. DSD is a step in the right direction. It reduces most problems with digital down to jitter and long slew rates, oh and the fact that square waves are still loaded with ripple. But at least the nasty ringing and brick wall LPF goes away with DSD. It's a much better format than PCM. Though I must state I admire that many of the people here in this thread are willing to admit the shortcommings of this medium in its infancy. I've been witness to many such threads and this is the only one where people were not only willing to provide evidence but also take to heart other people's evidence. I remember one such thread a while ago where somebody just insisted he was right and offered no proof or reason why. He just quoted a book about recording stormed away never to be seen again.
Anyways, it's been fun discussing this stuff. When I get some more time, I'll generate some samples and post them here but it'll be a while. The problem is, everybody will be using different converters and monitors so they'll all hear something different. But in the mean time, I gotta get some sleep.
Night all.
cjacek
Analogue Enthusiast
BINGO! The 'coherency' we oft speak about!
It's like analogue premixes everything which goes to tape, in a most pleasing and natural way. It's like everything gels nicely together, be it only one track or even within a narrow frequency range.Another point, which you've reminded me of, is that when recording straight to digital and then dumping to analogue, it pretty much sounds as a digital recording with some 'tape effects', I've found but when dumping a direct to tape recording to digital, it sounds as a reasonably nice representation of 'the' analogue recording, with perhaps the 'staging' or 'dimensions' being affected somewhat, depending on sampling rate etc.... BUT the analogue recording, 'coherence' and all, seems to be captured fairly nicely, even on the CD format. Again, I can tell that the overall sound picture has been reduced somewhat [like going from 3 dimensions to 2 to 1, for instance] but one gets a nice idea of how the original [analogue] signal sounds like.
And that's the crux of it, at least for me, that digital gives a good 'idea' of the source but will never represent it as well as analogue and it's not how low or high, on the frequency scale it shows up but rather what happens in between and what may get 'omitted' [during sampling]. Anyway, just my 2 bits.
---
Dr ZEE
Anti-Pro Circles Insider
Or better yet, apl (or who ever may apply) , please, describe the difference between description and what ever you are describing.
Well, first, before even trying to describe it (the difference), you'd actually have to be able to recognize such difference, and, well, the only trusted "tool" you can use in the process of the recognition (and the only tool you'd need) is the one inside of your skull. That's life, man.
In case you are wondering: "Why do I need to recognize and/or describe such difference ?"
Here's why:
If you understand such difference (or at least recognize it), then you would not view (or use, or present) a description of a phenomenon as a proof of the notion that the actual phenomenon under the question is exactly the same as the imaginary fenomenon that is being proposed by the description. In short: You would not confuse a description with what is being described.
Meanwhile, keep your analyzer hot and keep on measuring stuff
/respects
Something my Dad told me about this sort of stuff a long time ago.
"The buying public knows little about any of these differences and they care even less. If it sounds good, it sounds good. The differences that engineers hear mean nothing whatsoever when it comes to the CD at the cashier's register. All of the chest beating simply leaves the buyer with a blank stare".
Latigo
"The buying public knows little about any of these differences and they care even less. If it sounds good, it sounds good. The differences that engineers hear mean nothing whatsoever when it comes to the CD at the cashier's register. All of the chest beating simply leaves the buyer with a blank stare".
Latigo
P
pianodano
Guest
Yikes ! I'm pretty sure people would have a better chance of understanding Einstein's theory of relativity than most of this stuff.
http://en.wikipedia.org/wiki/Theory_of_relativity
http://en.wikipedia.org/wiki/Theory_of_relativity
evm1024
New member
limits, understanding and engineering
I asked APL to describe a square wave train in terms of a fourier series simply because it cannot be done. It can only be approximated. There is no summation of a finite number of pure sine waves that can completely describe a discontinuous function (read square wave). (look up Gibbs for example)
Once again the psudo-engineening types trot out quite complex math which works quite well in a purely math domain , that they do not understand in ALL it aspects and REQUIREMENTS for validity, to prove someone's generalization is wrong all the while missing the point.
I hope that they did not defend their thesis in these terms. It would have been a blood bath.
But what really frosts me in this is that weakness or minor errors in others arguments are pounced upon all the while the weaknesses in their arguments are dismissed as "in-audible" or not significant.
To compound these errors it is assumed that 32 or 64 bit floating point math is exact, that our measuring devices have no errors and introduce no artifacts and that we can read a plot with such precision so as to SEE any audible features that may exist. My what great eyes you have.
Once again the error of "I know what I know and there is nothing else to consider" rears it's ugly head and prevents people from moving closer to a greater understanding.
In theory you could describe In-A-Gadda-Da-Vida by a fourier series. Sure, thats theory. The reality being that you would need an infinite number of terms and an infinite frequency bandwidth and you would need to start the sine wave series an infinite time before the beginning of the performance and continue the sine wave series for a infinite time past the performance.
Of course you could approximate the full solution....and thus lose your exactness.
One last comment to those who wield nyquist or fourier like a club. Let us disregard the stringent requirements for their proper use. I release you from reality.
When you take nyquist and transpose it from signal theory to information theory you run into problems. Classic problem, Phone lines are samples at 8k samples per second, 8 bits (oh gosh I'm sure this must have changed over the years, just hang in there). So the highest signal that can be reproduced on a phone line is 1/2 f (8k/2) = 4 kHz! And yet we have modems that signal at a much higher rate than 4 kbps. 28.8 kbps or more.
The point is that recording hardware be it analog tape or digital sampling as limits based on signal theory. But the information contained whatever was recorded on you machine is subject to another incompletely understood set of rules. (rules I might add that the psudo-engineering types like to discount or ignore)
I asked APL to describe a square wave train in terms of a fourier series simply because it cannot be done. It can only be approximated. There is no summation of a finite number of pure sine waves that can completely describe a discontinuous function (read square wave). (look up Gibbs for example)
Once again the psudo-engineening types trot out quite complex math which works quite well in a purely math domain , that they do not understand in ALL it aspects and REQUIREMENTS for validity, to prove someone's generalization is wrong all the while missing the point.
I hope that they did not defend their thesis in these terms. It would have been a blood bath.
But what really frosts me in this is that weakness or minor errors in others arguments are pounced upon all the while the weaknesses in their arguments are dismissed as "in-audible" or not significant.
To compound these errors it is assumed that 32 or 64 bit floating point math is exact, that our measuring devices have no errors and introduce no artifacts and that we can read a plot with such precision so as to SEE any audible features that may exist. My what great eyes you have.
Once again the error of "I know what I know and there is nothing else to consider" rears it's ugly head and prevents people from moving closer to a greater understanding.
In theory you could describe In-A-Gadda-Da-Vida by a fourier series. Sure, thats theory. The reality being that you would need an infinite number of terms and an infinite frequency bandwidth and you would need to start the sine wave series an infinite time before the beginning of the performance and continue the sine wave series for a infinite time past the performance.
Of course you could approximate the full solution....and thus lose your exactness.
One last comment to those who wield nyquist or fourier like a club. Let us disregard the stringent requirements for their proper use. I release you from reality.
When you take nyquist and transpose it from signal theory to information theory you run into problems. Classic problem, Phone lines are samples at 8k samples per second, 8 bits (oh gosh I'm sure this must have changed over the years, just hang in there). So the highest signal that can be reproduced on a phone line is 1/2 f (8k/2) = 4 kHz! And yet we have modems that signal at a much higher rate than 4 kbps. 28.8 kbps or more.
The point is that recording hardware be it analog tape or digital sampling as limits based on signal theory. But the information contained whatever was recorded on you machine is subject to another incompletely understood set of rules. (rules I might add that the psudo-engineering types like to discount or ignore)
_brian_
Member
I think people are kind of losing sight of the real reasons people would choose analogue over digital or digital over analogue. Man, you can recite any amount of formulae or bullshit you want, but if it sounds good that means it's good if it sounds bad then it's bad. It's dead simple. I don't really care if something is more or less perfect but if it adds a certain quality in which I can work better with or if it adds something better to the musical field as a whole i'm all for it. With analogue i see things being more audio related and less complex with time spent on the things that matter instead of ploughing through a screen. Because of digital, automation & product cheapness & reliability jobs, skills and the like are going down the drain.
Assistant engineer? bugger off automation.
Reel to Reel broken down - send for a skilled technician? bugger off digital - if it breaks we pull out the faulty card & send it back to the company.
Really try and figure out what is BETTER for OUR TRADE.
People are losing jobs, bands don't have to be as talented, engineers don't have to be as talented, producers don't have to be as talented.
Now digiheads go & pick faults in what i've said, I know I'm hardly the best at saying it but I'm getting annoyed that noone else even gets close to mentioning it. And I'm new to analogue!!
ANALOGUE FOR THE SAKE OF MUSIC - OUR TRADE
Assistant engineer? bugger off automation.
Reel to Reel broken down - send for a skilled technician? bugger off digital - if it breaks we pull out the faulty card & send it back to the company.
Really try and figure out what is BETTER for OUR TRADE.
People are losing jobs, bands don't have to be as talented, engineers don't have to be as talented, producers don't have to be as talented.
Now digiheads go & pick faults in what i've said, I know I'm hardly the best at saying it but I'm getting annoyed that noone else even gets close to mentioning it. And I'm new to analogue!!
ANALOGUE FOR THE SAKE OF MUSIC - OUR TRADE
P
pianodano
Guest
Interesting the way you put it.
And dang, anybody can be taught to arm tracks and hit record on a tape recorder in 2 mins and go off to record some pretty amazing stuff.
Otoh, 13 pages of gobblegook and their still discussing the principles behind digital and whether it's perfect or not.
I'm responsibe for our analog recoder every day. I clean heads, run the frequency test tape once a week and clean all 456 card and insert points every six months.
We do have a 24 track Hard Drive recorder too, but I really love the Sony/MCI JH24.
Latigo
I'd post a picture but I don't think this board allows it or at least I can't find how to do it here.
We do have a 24 track Hard Drive recorder too, but I really love the Sony/MCI JH24.
Latigo
I'd post a picture but I don't think this board allows it or at least I can't find how to do it here.
mshilarious
Banned
Has anyone ever denied this? apl and I have said repeatedly the implementation of digital audio is limited both to its theoretical constraints of bandwidth and dynamic range as well as the real-world performance of its circuits. We have admitted that REPEATEDLY.
Truth is, no electronic circuit can faithfully reproduce a square wave. None. I am not talking about tape or digital, I'm talking about an amplifier. You'd need an amp with unlimited bandwidth and no noise, and those don't exist in the real world.
But you keep making the mistake of saying we are trying to prove digital is perfect. I am doing no such thing. Rather, I am arguing that it does not possess the specific flaws that have been mentioned in this thread, which are:
- A digital sample is capable of less than 6 * bit - 42dB dynamic range due to QD. Wado claimed that, and for any properly dithered (read real world) signal, that's false.
- A 1/4 sample rate sine wave is reproduced as an (imperfect) square wave. Heck, make a little less than 1/2 sample rate. Theory or real world, that's false. It will lack any harmonics typical of a square wave within 1/2 sample rate limit, and the harmonics above that will be filtered out.
The goalposts keep shifting. Once I falsify the above claims, somebody will say OK, it can accurately record high-frequency signals (less the filter attenuation I must have mentioned 100 times, another instance of theoretical digital being imperfect that I BROUGHT UP), but it's missing the "meat" of lower frequencies. Well, give me some specifics, and I'll look at that.
In the meantime, don't keep insisting we are arguing digital is perfect. It's not, but if it were so badly broken as many here seem to believe, then this board would be called "Digital Only", and this BBS would be called "Home Tape Recording".
PS to _brian_, there was a ton of absolute analog garbage produced in the 1980s, and I have to endure radio stations lately thinking for some reason I wanted to hear it for a second time in my life
And there are a lot of people who think grunge was the downfall, but most of that was analog. It's the message not the medium.
Dr ZEE
Anti-Pro Circles Insider
heh heh,
Nice "hair pulling" monologue!
On that note shall we "roll" a digital clip of the way things used to be, sort of...
Watching a clip like that and at the same time thinking about what has been done to once great "form of entertainment" known as Musical Records makes my blood boil.
Let's "roll" the clip, shall we (thanks to digital technology, so we can
): https://www.youtube.com/watch?v=BYc4uRo9maQ
*********
and here are some memories to scroll through (don't miss memory#15!!!!, read description
) : http://www.nickblagona.com/index.php?page=memories/later
_brian_
Member
i expected as much, i thought i wouldn't have to explain further but obviousley some people of a certain opinion will only try to understand something and not take the piss only if that something is water tight & in black & white.
by the way i'm 25 and wasn't even alive when the 80's started, give me a little credit here i'm not talking in anyway about the message i'm talking only about the medium.
And hey pianodano, well smart ass i was talking about getting people into the studio to learn. Generally most people started by becoming an ass. engineer and the main jobs they had to put up while watching & learning in the control room was resetting the mixer, lending an extra couple of hands at the desk etc. Now it's almost impossible getting your foot in the door (In scotland anyway). These jobs arent needed anymore. so there are no ass engineers to
move up to engineer status. And quite frankly no i do not think digital has provided any jobs of note. Things are manufactured AND repaired in poor countries.
I was originally an inst. tech. for british energy and the same thing's happening there. If a transmitter was outwith tolerance it was calibrated and if out of spec could be repaired on site. Now SMART transmitters are all the rage, you plug in a laptop and within 5 seconds the transmitter is either calibrated or you are informed that it needs a repair. Less work for people to do, less employees, less jobs. It is then boxed and sent away. Used to be a skilled job - not anymore. Musical engineering is heading the same way.
Try and make your mind up what really matters
In my opinion you need more skill to make a good record in analogue, and generally as most people have posted you end up with something that sounds better anyway. Don't forget that there is more pressure involved as well (keeping generations down, if you make a mistake your f*cked) But in digital all you need to is revert back to a saved session or mabye hit "undo" a few times if you don't like the mess you've just made.
Apart from that in my opinion it sounds better & is more fun
by the way i'm 25 and wasn't even alive when the 80's started, give me a little credit here i'm not talking in anyway about the message i'm talking only about the medium.
And hey pianodano, well smart ass i was talking about getting people into the studio to learn. Generally most people started by becoming an ass. engineer
and the main jobs they had to put up while watching & learning in the control room was resetting the mixer, lending an extra couple of hands at the desk etc. Now it's almost impossible getting your foot in the door (In scotland anyway). These jobs arent needed anymore. so there are no ass engineers to move up to engineer status. And quite frankly no i do not think digital has provided any jobs of note. Things are manufactured AND repaired in poor countries.
I was originally an inst. tech. for british energy and the same thing's happening there. If a transmitter was outwith tolerance it was calibrated and if out of spec could be repaired on site. Now SMART transmitters are all the rage, you plug in a laptop and within 5 seconds the transmitter is either calibrated or you are informed that it needs a repair. Less work for people to do, less employees, less jobs. It is then boxed and sent away. Used to be a skilled job - not anymore. Musical engineering is heading the same way.
Try and make your mind up what really matters
In my opinion you need more skill to make a good record in analogue, and generally as most people have posted you end up with something that sounds better anyway. Don't forget that there is more pressure involved as well (keeping generations down, if you make a mistake your f*cked) But in digital all you need to is revert back to a saved session or mabye hit "undo" a few times if you don't like the mess you've just made.
Apart from that in my opinion it sounds better & is more fun
_brian_
Member
and quite frankly a doubt there would be as much contemporary crap gumming up our ears if analogue was the main medium. sure the 80's were a major low but there was still some very talented bands in the 80's. Put it this way, record companies would have to pay considerably more money to produce an analogue recording of pop crap than a digital one.
Is this the way you want to go dr zee?
http://en.wikipedia.org/wiki/Gotta_Get_Thru_This_(single)
You see in my opinion analogue acted as, not a perfect, but substantial filtration system for seperating the crap from the good or the talentless from the talented. Record companies instead of paying wads of cash to talented people (people able to produce a good analogue record) to write crap were as well paying wads of cash to talented people (people able to produce a good analogue record) to write something good (because they are more likely to be musically talented aswell). Now you just need egits like daniel dedingfield writing catchy crap on their DAW ready for the record companies to swallow whole. If you try & try & try and you have the option to erradicate all your mistakes with a touch of a button, at least once you'll come up with something catchy, or annoying like "gotta get thru this is". And guess what you don't have to play that guitar part over let's just duplicate that bar over and over. And due to the fact that a good quality demo tape was harder to create (not nowadays lads). Companies spend substantially more effort trying to uncover a new artist that could make them money - grass roots level. Lets rewind the tape of time and plop our favourite artist back afew decades "Wait, have you heard of bedingfield", nope "he sit's in his room editing crap into something catchy... but crappy", "probably why i haven't heard of him". "Wait, have you heard of a band in manchester called joy division?" "yes, i've heard they are performing some good stuff. They have a substantial fanbase and they have only been gigging for less than a year", "i've also read some promising reviews in the local paper, mabye we should keep an eye on them".
Just something else to think about, remember i'm trying to answer in the main of "why analog & not digital"
Is this the way you want to go dr zee?
http://en.wikipedia.org/wiki/Gotta_Get_Thru_This_(single)
You see in my opinion analogue acted as, not a perfect, but substantial filtration system for seperating the crap from the good or the talentless from the talented. Record companies instead of paying wads of cash to talented people (people able to produce a good analogue record) to write crap were as well paying wads of cash to talented people (people able to produce a good analogue record) to write something good (because they are more likely to be musically talented aswell). Now you just need egits like daniel dedingfield writing catchy crap on their DAW ready for the record companies to swallow whole. If you try & try & try and you have the option to erradicate all your mistakes with a touch of a button, at least once you'll come up with something catchy, or annoying like "gotta get thru this is". And guess what you don't have to play that guitar part over let's just duplicate that bar over and over. And due to the fact that a good quality demo tape was harder to create (not nowadays lads
). Companies spend substantially more effort trying to uncover a new artist that could make them money - grass roots level. Lets rewind the tape of time and plop our favourite artist back afew decades "Wait, have you heard of bedingfield", nope "he sit's in his room editing crap into something catchy... but crappy", "probably why i haven't heard of him". "Wait, have you heard of a band in manchester called joy division?" "yes, i've heard they are performing some good stuff. They have a substantial fanbase and they have only been gigging for less than a year", "i've also read some promising reviews in the local paper, mabye we should keep an eye on them".Just something else to think about, remember i'm trying to answer in the main of "why analog & not digital"
mshilarious
Banned
Consider yourself fortunate that whatever reunions you attend will not be punctuated by A-Ha, and I mean that sincerely!
I understand your point and I agree, but it's just the acceleration of a trend that started long before digital recording came into vogue. The PSW thread I linked blamed the multitrack; it's not an inherently evil technology but in that sense it was the beginning of the end. Sure, there was lots of bubblegum crap pop music in the early 1960s, but it was bubblegum crap done in a live take. And these days I listen to disco and thing, hey actual musicians performed that, and the singing is naturally in tune! But in the day, it was the root of all evil. People burned vinyl, yes, precious vinyl, in anger!
Others have blamed synthesizers, but the early Moog stuff (Walter/Wendy, etc.) was good; it didn't start sucking bad until about ten years after that. And I think at that point it didn't matter if it was an analog or digital synth, the usage of both was pretty bad for a number of years there.
The main difference I see in digital vs. analog so far as quality of talent is that if Milli Vanilli came along these days, they'd be singing on the CD, (no) thanks to Autotune. So with that in mind, I agree 100%.
But what has gutted the industry is digital distribution, not production. Given that mp3s suck bad, people still want them, and there is nothing to do about it, even for the most confirmed Luddite. And I think we all need a bit of Ned Ludd in our lives.
mshilarious
Banned
Oh sure, but those bands never left the airwaves, at least on college/alt radio. I'm talking about the stuff that no one has heard in 20 years, and no one should ever have to hear again! It's back! Auugggh!
.
.
.
A great irony of my favorite recording, Brian Wilson's Smile, is that he drove himself insane splicing the original effort, but his band can perform it live now. Yeah, they used ProTools for the CD, but they didn't need it. I went to the concert, and they played it straight through, start to finish, flawlessly. It took me a couple of days to reset my jaw in its proper non-dropped position
The only concession I think they made to a live setting was they didn't have real farm animals on stage. They did have live vegetables though . . .