bblackwood wrote on Fri, 14 May 2004 19:12 |
Well, I'm not GM, but resolution certainly isn't the answer as to why some feel analog sound better. IME, good analog machines have a S/N ratio of about 13 bit and a useable hi-freq limit of about 20kHz or so... |
jazzius wrote on Fri, 14 May 2004 14:38 |
why does it sound better then? |
jazzius wrote on Fri, 14 May 2004 19:38 |
why does it sound better then? Surely analog works at a higher resolution then digital?....electrons, atoms, eeeerrrrrr.....quarks?! (yeah, i don't actually know what the hell i'm talking about!) |
raw-tracks wrote on Fri, 14 May 2004 22:16 |
I could be way off base here, but shouldn't the correct answer be that analog recording has INFINITE resolution. Here is the way I see it: |
raw-tracks wrote on Fri, 14 May 2004 21:16 |
I could be way off base here, but shouldn't the correct answer be that analog recording has INFINITE resolution. |
Quote: |
posted by My World: analog has "infinite" resolution |
Chuck wrote on Sat, 15 May 2004 13:26 |
You can imagine that the larger the squares, the more difficult it is to filter them into a round wave-form. |
Nika Aldrich wrote on Sat, 15 May 2004 14:34 | ||
Chuck, This is a completely erroneous and misleading statement. Nika. |
Quote: |
posted by Charles: Nika Aldrich wrote on Sat, 15 May 2004 14:34 Chuck wrote on Sat, 15 May 2004 13:26 You can imagine that the larger the squares, the more difficult it is to filter them into a round wave-form. Chuck, This is a completely erroneous and misleading statement. Nika. Hi Nika,, Try to think about it again, but an octave higher. Charles |
Former Oceanway drone wrote on Sat, 15 May 2004 17:17 |
I certainly have opinions about analog vs. digital and I have no intention of talking about them right now. That said, I do have a problem with Chuck's connecting the declining recording quality of Eagles albums with a "decline" in analog. |
Ethan Winer wrote on Fri, 14 May 2004 13:16 |
In any way you care to measure, even a $25 SoundBlaster sound card beats pretty much any analog tape recorder. |
Chuck wrote on Sat, 15 May 2004 08:26 |
All that we try to capture are sinewaves, and all our converters put out are square waves. As you all know, squares consist of odd harmonics alltogether, so the main job of digital reproduction is filtering out those high-order odd stuff. |
Quote: |
I just wonder: "How did they manage to have that wonderful sound, almost half a century ago, and what are the reasons, that make it so difficult to achieve today ?" Charles |
My World wrote on Sat, 15 May 2004 03:13 |
I have been "begging" sound processing companies for several years to develop 28-bit processors! C'mon guys! |
jazzius wrote on Fri, 14 May 2004 12:56 |
George, we hear about the resolution of digital all the time.....24 bits, 44.1, 192.....2.8 million whatevers.... ...do you know if anyone has ever worked out the resolution of analog?.....how many bits would it be equivelent to?.....i know this is bit of a strange question, but i'd love to be able to give my customers a smart-arse answer for why analog sounds better then digital... ...cheers.....Darius |
chrisj wrote on Sat, 15 May 2004 20:40 |
There are no squares, really. You're looking at the visual representation of a bunch of numbers designed to tell the converter what kind of SINES to put out. The only time you'd be hearing squares is if you were using one of those funky DACs with no reconstruction filter. And that's actually not technically correct, though there are some things about it that count as advantages... The squares you're thinking of, you might be better served by thinking of them as lists of numbers, not as a waveform. Nothing is ever about trying to present that information as squares. The DAC wants to present it as sines again, that's what it's for. That's what a reconstruction filter is for. What you should be looking into is not 'squares', not increasing frequency range (necessarily), but the types of harmonic and inharmonic distortion generated by each kind of recording. |
sfdennis wrote on Sun, 16 May 2004 15:31 |
Hey Charles, that is a very funky converter indeed. What is it exactly? At first blush, it looks like there is absolutely no reconstruction filter working. Also please let us know the time & voltage division settings are on that picture. -Dennis |
gtphill wrote on Sat, 15 May 2004 21:20 |
[...] Let's start by trying to define "resolution." A common definition for this in physics is the Ralyeigh limit for optical "resolution." It is approximately equal to .66*lambda. It says that two objects closer together than .66*wavelength cannot be focused on well enough to tell one from the other. In some sense the Nyquist frequency can be thought of as the limit of the (frequency,wavelength,length-however you want to look at it) "resolution" of a discrete time system. Below the nyquist frequency all waveform content is completely captured, and above the nyquist frequency it aliases back into the lower frequencies (an erronius result). [...] I am tempted to talk about other subjects, such as electron thermalization, johnson noise, etc. but I don't want to devote further time to this before I see if it is heading in the direction you were looking to comprehend. |
dwoz wrote on Sun, 16 May 2004 18:35 |
when an analog signal is sampled, it is essentially 'sliced' into segments, each of which has a certain width in time. |
Quote: |
These 'slices' have to be made small enough that the signal being sampled cannot change direction TWICE inside the slice. (that's the nyquist thingy). Another way of saying this is that the frequency of the signal must be limited so it CAN'T go south then north inside a single slice. |
Quote: |
OK, so when we check the voltage of this sample, we really have to make TWO measurements...the absolute level, and also the slope of the wave. without these two, we don't really know the complete behavior of that signal. But the measure we take is only the level. BUT, because of the definition we've given, that the signal will only go up, or go down, or go up and down ONCE within the sample, we can check the samples on either side and infer a slope for the width of the sample. It isn't perfect, but its pretty much good enough for our purposes. |
Quote: |
What we're talking about here is QUANTIZATION ERROR. by knowing both the absolute value of the sample (either the leading edge, the exact middle, or the tail edge) and the interpolated slope, we can make some pretty good guesses about what "number" we should put in the output stream, so our DAC can reconstruct an analog signal that is very much like our input. |
Quote: |
So, the sample rate controls HOW WIDE (or long) that sample is in time. |
Quote: |
The bit depth controls how fine the discrete measurements we can make can be accurately captured. If we're trying to represent the whole dynamic range of the signal with just 8 bits, then we have to make some pretty gross approximations of the value of the particular we just measured. If we're using 28 bits, then the chance that the EXACT value of the sample that we just measured is only a tiny bit off a discrete integer that we can record, is much better. |
Quote: |
It should be obvious that BOTH the sample rate, and the bit rate, contribute to the QUANTIZATION ERROR. |
Quote: |
If I have a very high sample rate, then the more likely that my measurement of the slope is close to the actual value. I can rely on the absolute value of the sample, and less on the slope...and the higher the bit depth, the closer the measured value will be to one I can record. Remember, I have to round any 'inbetween' floating point value to an integer. |
Quote: |
So, resolution of digital signals is dependent on both the sample rate, and the bit depth, to some extent. |
Quote: |
In the context of what I've just said, the "resolution" of analog signals would be to all intents and purposes, infinite. Not REALLY infinite of course, and when I can count individual electrons and quanta, then we can enter that discussion of just HOW INFINITE it is. |
dwoz wrote on Sun, 16 May 2004 13:26 |
I stand by my assertion that sample rate has to do with quantization error... |
Bill Mueller wrote on Mon, 17 May 2004 09:56 |
dwoz, I'm not sure by your post if you were answering me regarding noise being a limiting factor to resolution, however if we define resolution as fidelity (the truest possible conversion of an acoustic event) then noise is a limiting factor. Any signal coming out of a device that is not present at the input, limits fidelity. |
Quote: |
To me analog noise is sort of like digital bit depth in its affect on the signal. While you can hear low level signal in both analog and digital systems, they are each distorted by the medium. |
Quote: |
Many don't seem to realize that recordings made 6 db below clip on digital systems, result in one less bit than maximum, ie -6 db on a 16 bit system results in a 15 bit (1/2 resolution) recording, -12 db results in a 14 bit recording (1/2 resolution of the 15 bit recording) and so on. Therefor as you fade out a signal, the resolution drops precipitously. Crank up your fades sometime for a truly horible experience. |
Quote: |
A similar thing happens in analog except that to my ear, it sounds significantly better. |
Chuck wrote on Sat, 15 May 2004 17:43 | ||
Hi Alan,, thanks for filling in the info on recording gear and engineers. Of course there are great recordings being made after the 70's. I did not want to say that available analog equipment is getting worse over time. But in general terms, while cruising through my record collection, I am expecting good sound from late 60's to late 70's some to '82, then it appears to become increasingly difficult to obtain that rich hi-fidelity sound quality. Btw. I also have the early Eagles records on CD with a sticker on them that says:"digitally remastered", and if you know and have the records, I would say the CDs are just unlistenable. I don't blame it on digital or the CD format. But actually, I don't know who or what I would want to blame for it. For me, subjectively, it is just a decline in audio quality. I have Joni Mitchell Hejira on vinyl, and I have the Travelogue album on CD. Both excellent, really very very good. Lately I bought Joni Mitchell 'Both sides now' as 24/96 DVD-A. As I have Chesky 24/96 DVDs that are really excellent, I expected something. But the Joni Mitchel DVD-A sounds like crap. I read the names Geoff Foster, Allen Sides and Bernie Grundman on the inlet. I don't know these guys personally, maybe you know them... I have just no comprehension and no words for the decline in audio quality, that I have to face today. I just wonder: "How did they manage to have that wonderful sound, almost half a century ago, and what are the reasons, that make it so difficult to achieve today ?" Charles |
hollywood_steve wrote on Mon, 17 May 2004 12:20 |
the only way that really matters is how it sounds. And I'll take the SOUND of classic analog gear any day, and so will a lot of people who might think otherwise. for example, I'll take mics: a pair of (you choose: Neumann u47/48/M49/M50, RCA 44/77d, AKG c12 / Telefunken 251, among others) console: API 1604 or EMI REDD37 recorder: Studer A800 1" 8trk and Ampex ATR102 2trk in any room at any time against the best of 2004 (and there is some AMAZING gear made by some small companies these days). At some point the only "improvements" are discerned by test equipment. I'm not the only one who thinks the best analog gear of the 70s was already as good as gear needed to be; it was just too expensive for all but the best studios. All we've done since then is made decent gear available to the masses; unfortunately, at the same time we have lowered the bar for what passes as acceptable commercial "product." |
Ethan Winer wrote on Sun, 16 May 2004 09:41 |
Charles, > How did they manage to have that wonderful sound, almost half a century ago, and what are the reasons, that make it so difficult to achieve today? < Years ago music was recorded and mixed in real recording studios that had large rooms. The better studios also had sufficient bass trapping and other acoustic treatment so the recordings were not permeated with the "small room" sound we hear so much today. And the mix engineers were able to hear what they were doing much more accurately than folks today who work in small untreated bedrooms. |
Quote: |
All of the gear used today beats all of the gear used years ago in every way you care to assess gear. But the rooms have gotten much smaller, and a lack of proper acoustic treatment compounds the problem. |
Brent wrote on Sun, 16 May 2004 19:44 |
I have listened to records from my youth, and remember them being much better than what they actually sound. Granted CD's do sound different than records, and you don't get the words that you can read without a magnifying glass, the cool artwork, and the smell of vinyl. I don't know about you, but I gave up the Grado turntable long ago. I enjoy pop free, rumble free, scratch free music. What I like best is that I can get 10x in the peach crates. LOL |
oudplayer wrote on Mon, 17 May 2004 07:20 |
I have a good-sized collection of orchestral, Chinese, and Armenian music on 78s - the pops and rumble do not at all get in the way of my enjoyment of the music. And nothing made today sounds like it (with the exception of the 78s that the Cheap Suit Serenaders released a few years back). Not that they're the best sonics - obviously the dynamic range is, like, 10dB, and the frequency response... you can count it on one hand. But the effect of the recording focuses your attention on certain aspects of the music - the effect of modern CDs is entirely different. |
dwoz wrote on Sun, 16 May 2004 19:26 |
I don't agree that noise is a limiter of resolution. I find that to be a red herring. Of course noise is a factor, but as many others have noted, and as many including myself have heard time and time again, content can be discerned under the noise floor. |
Quote: |
Noise is never good, but noise floor isn't necessarily a boundary. |
Quote: |
also, I completely understand that a sample is an instant, without width...thus my "leading edge, middle, trailing edge" qualification of the "width" of a sample. the sample represents a choice of a single value among perhaps many inside the span of time being represented by that sample. |
Quote: |
I stand by my assertion that sample rate has to do with quantization error... |
Quote: |
as the more time between samples, the more opportunity to pick a value that doesn't represent the activity within the sample. |
Quote: |
And again, I stated that we only record the absolute value, not the slope, because we can interpolate slope using adjacent values, not having to resort to "complex" calculus and defining slope and value....as you state, after all, that because of nyquist, we don't see real values of slopes that diverge significantly from our ability to interpolate their values. |
Jim Dugger wrote on Sun, 16 May 2004 21:35 |
May I ask Nika or one of the others here to help explain something to me in clearer (layman) terms? |
Quote: |
It is true that the time and frequency domain are quite linked, correct? Afterall, there are well known ways to covert information of one kind into the other. So, does it stand that increasing the sampling rate is somehow (within some reasonable limits) equal to adding one bit to the frequency domain? |
Nika Aldrich wrote on Mon, 17 May 2004 09:17 |
So going from 48KS/s to 192KS/s does, in theory, increase your dynamic range by 1 bit, but with today's converters this is mitigated in other ways. Further, with 24 bit protocols it is irrelevant as the bit depth already exceeds the dynamic range we can hear. |
Jim Dugger wrote on Mon, 17 May 2004 16:40 |
1. We can decrease the amount of quantization noise (error?) by doubling the sampling rate or adding bits to the sample quantity. |
Quote: |
2. The value of doing this tails off quickly as we push the level of the noise a few db below that of the analog electronics on either side and the sampling rate at just a bit over double where the energy is we want to capture. |
Quote: |
What this should tell you, is that there are some infinities that are bigger than others. How is that possible? Simple: Take simple numbers. There are an infinte number of simple numbers, right? You start counting from one...and you can count onto infinity. Now, take just the even numbers, and just count them exclusively. If you just count the even numbers...you can count them into infinity, right? But when you compare the infinite numbers against the infinite even-only numbers, which is the greater infinity? It's a brain-twister, right? Counter-intuitive as it is, it makes sense. Some infinities are indeed bigger than others. |
ted nightshade wrote on Mon, 17 May 2004 10:31 |
I'm very intrigued to find that it is in many ways superior to any more recent digital recordings I can find! My guess is, that they created a first generation master directly from recording, and that's bit-identical to what's on the CD. That and some fantastic analog engineering on the front end. |
ted nightshade wrote on Mon, 17 May 2004 17:36 |
I'm thinking, the digital recording system needs to be a goodly bit better than a Radar S-Nyquist |
Jim Dugger wrote on Mon, 17 May 2004 17:21 | ||
I've always approached recording with the mindset the tools were like paintbrushes and canvas. Ever notice how different a painting can be from the source that inspired it, and yet it is still an amazing, beautiful painting, as good as the source itself? To me, live performance and recordings are a very different media. And, as different media, there's a very high probability a best-of-class example from each might not be all that similiar. Maybe that helps explain how you feel? That said, I've got an early 80s Teac A3340 with Simul-sync around here I'll be happy to trade for that RADAR system. It's in great condition, all original. |
Quote: |
posted by Ernie: Quote: What this should tell you, is that there are some infinities that are bigger than others. How is that possible? Simple: Take simple numbers. There are an infinte number of simple numbers, right? You start counting from one...and you can count onto infinity. Now, take just the even numbers, and just count them exclusively. If you just count the even numbers...you can count them into infinity, right? But when you compare the infinite numbers against the infinite even-only numbers, which is the greater infinity? It's a brain-twister, right? Counter-intuitive as it is, it makes sense. Some infinities are indeed bigger than others. Yes and no. Some infinities are bigger than others, yes; this is the math of transfinite numbers or infinite set theory as developed by Georg Cantor. But no, the infinity of "simple" (natural) numbers is not greater than that of even-only numbers. Transfinite numbers can only be compared by placing the systems in one-to-one correspondence; in the case of natural (whole) numbers v. even numbers, you can draw the following correspondence: 1 | 2, 2 | 4, 3 | 6, 4 | 8, 5 | 10, 6 | 12, 7 | 14, etc. endlessly. In terms of infinity, a part is equal to the whole; divide infinity and you still have infinity. This extends to all integers and fractional numbers; this infinite set is known as aleph-0 (represented with the Hebrew letter aleph). However, there is no one-to-one correspondence between all rational numbers (all integers and fractional numbers as above) and all the points on a line, plane, or in a cube; this second set is therefore known as aleph-1. Curiously, the infinite set of points on a line one inch long and a line one lightyear long is mathematically equal (i.e., infinite and uncountable); and, more so, the number of points on a plane or in a cube is no greater than that on a line. Yet larger than the infinite set of points in a line, plane or cube is the set of all possible geometrical curves, or aleph-2. (c.f. Georg Cantor on infinite sets; one source: http://mathworld.wolfram.com/Infinity.html) How or if this applies to analog v. digital recording, I leave to you lot; this is fascinating reading to your humble webmaster. ------------------------------------------------------------ ------------ Ernie, PSW Webmaster |
Ethan Winer wrote on Tue, 18 May 2004 08:17 |
Ted, > The analog machine sounds a whole hell of a lot more like the vocalist there in the room. < That sure has not been my experience! When I used to own a pro studio with a 2-inch recorder I could always hear the degradation between monitoring the console while recording and tape playback later. Today, with even a modest digital sound card (Delta 66), I hear no difference. |
Quote: |
> I'm thinking, the digital recording system needs to be a goodly bit better than a Radar S-Nyquist < I admit I'm not familiar with that box. But I'm curious: In what way does it sound different / worse than the direct mike feed from the console? Maybe it's simply broken or faulty? --Ethan |
Quote: |
I've never heard a recording in any medium on any playback system that sounded like the real thing. At best, a recording can only create an illusion which requires a certain "willing suspension of disbelief". |
Quote: |
To engineers who have spent a lifetime developing the mental filters which enable them to create the best possible illusion using analog means, a digital recording may never sound "right"; they just bring a different set expectations to the listening experience. A generation raised on digital sound is likely hear things quite differently. Neither is right or wrong, they're just different. |
charles maynes wrote on Tue, 18 May 2004 22:54 |
I think it has been touched on here in this thread, but the thing I am really intrigued by is how certain analog recordings- mainly orchestral and non-rock recordings just totally leap out of your speakers and shout "I'M BAD"! |
Quote: |
In what way does it sound different / worse than the direct mike feed from the console? Maybe it's simply broken or faulty? |
dbock wrote on Wed, 19 May 2004 21:51 |
But not all audible changes can be measured! |
Ethan Winer wrote on Wed, 19 May 2004 15:25 |
Ted, > I'm talking about the sound of the vocalist live in the flesh through nothing but the air. That's what I'm trying to capture. < Then like me, you're looking for a recording medium that most closely reproduces what you recorded. Which takes us back to my original statement that in all ways you care to assess it, digital recorders beats analog every time. If you have a digital system that doesn't sound exactly what you sent it, then it's broken. Or it should be replaced with a newer and better recorder that lives up to digital's potential, which is total transparency. --Ethan |
Nika Aldrich wrote on Wed, 19 May 2004 16:56 | ||
Such as? |
Quote: |
posted by Harvey Gerst: Makes me wonder if changing them fast rising squares to a whole bunch of itty bitty ramping sine waves might be messing with the slew rate a bit. Ya reckon? |
Eric Vincent wrote on Wed, 19 May 2004 21:58 | ||
Harvey, I'm stumped. Please enlighten. How does A-D conversion affect slew? |
hargerst wrote on Thu, 20 May 2004 11:49 |
I remember back when transistor amplifiers started appearing on the market. Many of us heard differences between tube amps and transistor amps that measured "exactly the same". Turned out we had to learn how to measure things differently, and we came up with TIM, slew rates, problems with massive amounts of negative feedback, and the actual individual harmonics that were distorting (instead of THD). Seems our ears were right. Makes me wonder if changing them fast rising squares to a whole bunch of itty bitty ramping sine waves might be messing with the slew rate a bit. Ya reckon? |
Quote: |
posted by Harvey Gerst: I think we're all agreed that a 7.4kHz square wave (kinda looks like the side of a building) comes out as a sine wave (kinda looks like a ski slope). One of the main things that's changed is the rise time, isn't it? |
Quote: |
Call it slew rate, call it skew rate, call it rise time, but it ain't goin up as fast as it was. I don't know what to call it, but when something comes out different from when it went in (and it's down low enough to hear), what would you call it? |
Quote: |
Granted, I'm an old fart and I don't understand a lot of this digital stuff, but it seems logical to me that there is a difference between something that goes up like this: | and something that goes up like this: / - my only question is: at 7.4kHz, is it audible to some people? |
Han S. wrote on Thu, 20 May 2004 05:40 |
In a dutch pro audio magazine there was a test with an Ampex ATR102 and a DAW. They recorded a 15khz square wave on both and in playback the ATR showed an almost perfect square wave while the DAW showed a kind of between square and a sinus on the scope. I guess this phenomenon is what Harvey is pointing at. Most people can hear the difference between a square wave and a sinus. |
Quote: |
posted by Han S: In a dutch pro audio magazine there was a test with an Ampex ATR102 and a DAW. They recorded a 15khz square wave on both and in playback the ATR showed an almost perfect square wave while the DAW showed a kind of between square and a sinus on the scope. I guess this phenomenon is what Harvey is pointing at. Most people can hear the difference between a square wave and a sinus. |
hargerst wrote on Thu, 20 May 2004 02:49 | ||||
Seems our ears were right. Makes me wonder if changing them fast rising squares to a whole bunch of itty bitty ramping sine waves might be messing with the slew rate a bit. Ya reckon? |
hargerst wrote on Thu, 20 May 2004 04:36 |
I think we're all agreed that a 7.4kHz square wave (kinda looks like the side of a building) comes out as a sine wave (kinda looks like a ski slope). One of the main things that's changed is the rise time, isn't it? |
Quote: |
Call it slew rate, call it skew rate, call it rise time, but it ain't goin up as fast as it was. I don't know what to call it, but when something comes out different from when it went in (and it's down low enough to hear), what would you call it? |
Quote: |
Granted, I'm an old fart and I don't understand a lot of this digital stuff, but it seems logical to me that ... |
George Massenburg wrote on Thu, 20 May 2004 06:06 | ||
Well, yes. But not at 15kHz. Not unless there's something else wrong (like non-linearities). GM |
Level wrote on Thu, 20 May 2004 15:12 |
The audiophile cable rants that litter the Internet are a fine example of ... |
Quote: |
a majority of people hearing a difference in one cable over another but when took to task by the "lab heads" they have "not figured out what to measure to substantiate to perceived difference" |
Quote: |
In audio, many examples of audible differences that are not "proven" by current measurement techniques. Especially true when a majority can hear the differences and agree upon what they hear. |
Level wrote on Thu, 20 May 2004 18:15 |
Also, remember, a cable is part of a CIRCUIT. The associate equipment will provide different results on the same cable. |
Level wrote on Thu, 20 May 2004 18:15 |
Here you go Nick, I will start with this one...but the formal studies, I have to dig back in my archives. |
Johnny B wrote on Thu, 20 May 2004 12:03 |
Harvey, You described how minds were opened and new measurement techniques were the result. Can you give more details on how that process occurred? The arguments that were going on back then and how they came up with the new test procedures? Ya know, more details. Thanks. |
George Massenburg wrote on Sun, 16 May 2004 10:58 |
Thanks very much for your notes. Now, this is alot closer to the discussion that I think we should be having about digital audio. I'm currently having a enlightening discussion off-line with Jim Johnston about quantizing multi-channel audio and differential resolution in the time-domain. George |
the tests, and the conclusions, are invariably tilted toward the analog case.
George Massenburg wrote on Tue, 25 May 2004 02:46 |
NO calibration. NO A/B. NO objectivity. I'd have to call that "biased" towards analog. |
Quote: |
DVD players, and other electronics that are not affected by vibration. |
Ryan Moore wrote on Tue, 25 May 2004 15:49 |
FWIW - I bought some of the Lynn Fuston (sp?) CDS & thought on 1st listening that the PT in the box mix was the one I preferred / the non box mix.. And upon knowing which was which - somehow suddenly the analogue mix seemed to be the one with 'vibe' and the PT box mix was plasticky.. |
George Massenburg wrote on Tue, 25 May 2004 20:13 But if you go [i |
only[/i] by ear and not by fear of being shamed, most of the time you're really not going to like what all the extra fucked-up analog electronics are doing to your music. George |
George Massenburg wrote on Wed, 26 May 2004 03:13 | ||
THANK YOU. And this is why in-the-box is losing to analog mixing...because it supposedly "vibier", even though it doesn't sound as good. Definitely a case of the "King's New Clothes" But if you go only by ear and not by fear of being shamed, most of the time you're really not going to like what all the extra fucked-up analog electronics are doing to your music. George |
St. Domingo's wrote on Tue, 25 May 2004 21:06 | ||
Ethan, please help this bear of little brain out. Are you saying that DVD players are NOT affected by vibration? One would assume anything that could cause eccentricity in disc rotation would affect the performance of a DVD player in the same way it can in a CD player. Where am I going wrong? Cheers, Matt |
davidstewart wrote on Wed, 26 May 2004 21:48 |
The large buffers do correct for intermittent data due to vibrations. However I have heard of a related issue that may have some validity. When an optical player cannot correctly read a block of data it will either correct with error correction, or may try to re-read the data. The momentary power supply draw from the extra servo movements can potentially cause inconsistencies in the clock (voltage drop presumably??), which could conceivably manifest themselves sonically at the D/A. Obviously this (or the severity of it) would depend on the player (power supply, etc.). This is not something I've noticed personally, only heard about it, but it sounds plausible. David Stewart |
davidstewart wrote on Wed, 26 May 2004 13:48 |
The momentary power supply draw from the extra servo movements can potentially cause inconsistencies in the clock (voltage drop presumably??), which could conceivably manifest themselves sonically at the D/A. |
Quote: |
The stability of a CD or DVD drive mechanism is not important at all. These drives are not like a turntable, where instability of the drive motor directly impacts wow and flutter. Rather, the data is read from the disk into a memory buffer, and then a moment later it's sent out at a very uniform rate which is controlled by the clock circuitry. This is why walkman style CD players often boast a large memory buffer, for when you're jogging and the player is frequently bumped and jostled. Since the buffer always holds the next few seconds of music, it can keep outputting until the mechanism has a chance to find the track again and resume reading. --Ethan |
I'll have to strongly disagree with you on the suggestion about a CD mechanism's stability. Firstly, most domestic hifi CD players play at 1* speed. I'll come back to this in a second...The FIFO buffering you talk of eliminates playback jitter in the data stream coming off the disc, I'll accept that. However, jitter inherent on a badly pressed / burnt CD (and also caused by physical vibration of a playback mechanism) has two effects:1- timing errors in the data coming off the disc (which you will have eliminated in the buffering).2- The ability for the servo systems to ensure the correct data is retrieved.If the correct data cannot be read, the error correction will kick in. The more errors that need correcting, the more audible it becomes.
Paul Frindle wrote on Thu, 27 May 2004 00:17 |
I'll have to strongly disagree with you on the suggestion about a CD mechanism's stability. Firstly, most domestic hifi CD players play at 1* speed. I'll come back to this in a second...The FIFO buffering you talk of eliminates playback jitter in the data stream coming off the disc, I'll accept that. However, jitter inherent on a badly pressed / burnt CD (and also caused by physical vibration of a playback mechanism) has two effects:1- timing errors in the data coming off the disc (which you will have eliminated in the buffering).2- The ability for the servo systems to ensure the correct data is retrieved.If the correct data cannot be read, the error correction will kick in. The more errors that need correcting, the more audible it becomes. Quite apart from actual data errors and correction, another class of problem can indeed arise out of the efforts and machinations of the servos as they track. In this case the modulation of the power supply and internally generated PCB earth currents cause the clock transmission and converter stages to degrade - even though there are no data errors. Obviously the gravity of this effect is closely related to the quality of the design of the circuits and should not affect a player with an external DAC that additionally has good clock recovery stages (i.e. PLL). However, many even high costing players do not have this facility. |
St. Domingo's wrote on Wed, 26 May 2004 17:35 |
[...] I'll have to strongly disagree with you on the suggestion about a CD mechanism's stability. [...] The FIFO buffering you talk of eliminates playback jitter in the data stream coming off the disc, I'll accept that. |
Quote: |
However, jitter inherent on a badly pressed / burnt CD (and also caused by physical vibration of a playback mechanism) has two effects: 1- timing errors in the data coming off the disc (which you will have eliminated in the buffering). |
Quote: |
2- The ability for the servo systems to ensure the correct data is retrieved. [...] |
Former Oceanway drone wrote on Thu, 27 May 2004 17:09 |
Ethan Winer wrote: "Just because you can measure an error doesn't mean it's important or even audible." I would say two things: First, some of the people in this discussion are saying that just because it can not be measured does not mean that it is not audible. Some of the more "just give me the facts" people have even declared that if it is not measurable, it does not exist. But now to say that even if it is measurable it may not be important seems to me to be changing standards. |
Quote: |
"Just because you can measure an error doesn't mean it's important or even audible." |
interstellar sasquatch wrote on Fri, 28 May 2004 13:30 |
Well when I hear the term resolution mentioned it is usually referring to bits. Mike |
djui5 wrote on Sat, 29 May 2004 01:32 |
Will there ever be a set in stone definition for resolution in it's reference to audio whether that be digital or analog? |
Quote: |
djui5 wrote on Sat, 29 May 2004 01:32: Will there ever be a set in stone definition for resolution in it's reference to audio whether that be digital or analog? |
Giovanni Speranza wrote on Mon, 31 May 2004 14:03 |
Pro Analog machines have those specs: 25-25000 Hz, with DBX 100 dB dynamics |
Eric Vincent wrote on Sun, 30 May 2004 10:40 | ||
That's a question which intrinsically answers itself, no? In other words, if you have to ask that at this late stage of the game, it tells you that the term is too vague for engineering purposes, because no one has definitively answered it so far. So one can interpolate from that experience that "resolution" is an abstract term. No? So where is there room in the engineering sciences for abstraction? |
Giovanni Speranza wrote on Mon, 31 May 2004 17:04 |
Put a DBX and you will almost double the SN ratio, or use Dolby SR. We have to remember that most of the best recordings were made on those noisy tape machines, but never without Dolby or DBX nr. |
Han S. wrote on Tue, 01 June 2004 18:09 |
DBX sounds muddy in the lows and with the +9 tape you don't need any NR when recording pop music Giovanni. Dolby A, S or SR will sound better than DBX and my two inch machine records a 35 khz sine with ease. |
ranalog wrote on Tue, 01 June 2004 21:41 |
it sounds better because its "MAGIC!!" heres one, Digital is a flourescent light bulb, flashing so fast you cant tell, but it kind of annoys...Yes? anolog is like a smooth burning incandesent light bulb, just a thought, If all you people only knew..... Ranalog |
Eric Vincent wrote on Sun, 30 May 2004 06:40 | ||
That's a question which intrinsically answers itself, no? In other words, if you have to ask that at this late stage of the game, it tells you that the term is too vague for engineering purposes, because no one has definitively answered it so far. So one can interpolate from that experience that "resolution" is an abstract term. No? So where is there room in the engineering sciences for abstraction? Bit depth is definable. Sampling rate is definable. Dynamic range is definable. SNR is definable. One can sum these factors, and concievably choose to blanket that sum into a "resolution" factor, but that is still a very subjective exercise, because: 1) Who's authority is it to scale those factors into a "resolution" factor??, and 2) Isn't it insulting to the average engineer's intelligence that they cannot sort out the individual factors, and make up their own mind what is optimal "resolution" without third-party critics imposing vague and un-verifiable "resolution" standards? This term "resolution" is a cop-out for people who do not understand the basics of recording, yet seek to justify some abstract notion of what they cannot achieve. |
Paul Frindle wrote on Wed, 02 June 2004 09:11 |
Darned right on The term resolution cannot be applied in any definable sense to audio either analogue or digital. It means absolutely nothing what so ever in any defensible language definition. It has been hijacked and coined by marketeers (for dubious reasons) and people are searching manically to assign it a real life meaning - completely in vain. It's just so sad |
Zoesch wrote on Wed, 02 June 2004 04:41 |
...if you don't like the term resolution because it (correctly) paints a picture of a discrete time system in your mind then the problem is not the term, is your refusal of applying the terms properly. |
Zoesch wrote on Wed, 02 June 2004 14:33 |
Oh and to drive the point further... Q: What's the resolution of a digitized signal? A: One sample |
Quote: |
Paul Frindle wrote on Wed, 02 June 2004 09:11 Darned right on The term resolution cannot be applied in any definable sense to audio either analogue or digital. It means absolutely nothing what so ever in any defensible language definition. It has been hijacked and coined by marketeers (for dubious reasons) and people are searching manically to assign it a real life meaning - completely in vain. It's just so sad Zoesch responds: You know, I kinda find it insulting when non EE's try to rephrase terms to fit their understanding of the world, it's called defending the indefensible. |
Eric Rudd wrote on Thu, 03 June 2004 21:47 |
What I really don?t quite get about this thread is why people are getting there noses in a scrunchy about definitions. If a designer uses a particular word that happens to mean one thing in his world, but it means something different to me?oh well. |
Quote: |
One definition of resolution reads: ?The fineness of detail that can be distinguished in an image? Nika?s explanation of the 126db of dynamic range of a system as its resolution seems incorrect to me. You should call it bandwidth, or brackets, or ?gasp? range? |
Quote: |
Fineness of detail sounds pretty darn close to how I use the word resolution. |
Eric Rudd wrote on Thu, 03 June 2004 21:47 |
You should call it bandwidth, or brackets, or ?gasp? range? |
Nika Aldrich wrote on Thu, 03 June 2004 22:42 | ||||||
Because words are used for communication - to communicate messages. If two people use one word differently then when they speak the message that is encoded with the words will not get through.
But that is a very real use of the term in the floating point mathematics world and has been in use for some time. Neither of these definitions are "wrong" at all. When we talk about this stuff we need to keep our words straight to communicate effectively. The real problem comes in when people start talking about bit-depth in terms of "resolution." Bit depth neither directly relates to the depth-detail definition of resolution nor does it relate to the instantaneous dynamic range definition. The public communication of this, however, has people believing that a change in bit-depth inherently creates an increase in detail, which is a complete fallacy. It simply becomes a poor word to use in those circumstances. It does not effectively communicate and instead indicates that the user has a level of misunderstanding.
Just be aware that changing bit-depth does not inherently increase your "resolution" as you and many others use the term. Nika. |
Nika Aldrich wrote on Thu, 03 June 2004 16:42 |
Just be aware that changing bit-depth does not inherently increase your "resolution" as you and many others use the term. Nika. |
George Massenburg wrote on Fri, 04 June 2004 03:17 | ||
Incorrect. Well, in the well-intentioned converter (one that intends to deploy real "resolution" with meaning to it's claims), it should, and at least some of the time it actually can and does. |
Quote: |
Stop being a contrarian, Nika, it's not going to look good from the perspective of history. |
Quote: |
Even though there are still people commenting on the subject, I've come to the conclusion that the definition of resolution is in the eye of the beholder and a "set in stone" definition will never happen. |
xaMdaM wrote on Fri, 04 June 2004 12:00 |
First the "formula" for analog; What is the MINIMUM measurable density of flux that can be transferred to tape at each of the frequencies and then the maximum. (and maybe a "nominal" mid-range value) [snip] Resolution in imaging is accepted to indicate the smallest measuable unit. However, it has recently come to be scrutinized as having a secondary component of accuracy. e.g. how faithful is detail captured. |
xaMdaM wrote on Fri, 04 June 2004 06:00 |
[...] My point is that indeed there are ways to determine the resolution of analog just as there are ways to measure the resolution of digital. Does anyone have these number or the ability to perform any of these tests? I for one would really like to see the numbers and results. Max |
Quote: |
Geez, I wish audio/music/hearing were as clear-cut as video/pictures/seeing. |
George Massenburg wrote on Fri, 04 June 2004 11:59 |
What, in your opinion, would a "perception-sensitive" resolution test look like? |
Nika Aldrich wrote on Fri, 04 June 2004 23:25 |
Max, A point worth bringing forth is that video/photo and audio are inherently different in that digital images are sampled and then stay digital, all the way through the delivery stage. The photo does not, for example, undergo a "reconstruction" stage that turns it back into an analog photograph. What gets printed on silver paper and delivered to the eyes of the consumers is still a pixilated digital image. In that respect we can discuss the resolution of the photograph simply by discussing the photograph in its pixilated (and only) form. In audio the pixils, or samples, are only representations of the complete product. The complete product has to get reconciled from the samples, undergoing a reconstruction process to determine the actual analog waveform that is represented by the samples. Because of this, discussing the resolution of the samples can be unhelpful because we really need to discuss the effect on the resultant waveform. In digital photos this is again not an issue -the samples ARE the end result. In digital audio the samples only REPRESENT the end result. It is important that we discuss things in digital audio by discussing the effect on the net result - the product that is REPRESENTED by the samples, and not so much on the samples themselves. In digital photos you can change all of the samples in some way and we can still discuss that in terms of resolution. In digital audio you change all of the samples in some way and we now have to use other terms to talk about the net result on the resultant waveform - we may talk about it in terms of distortion, or dynamic range, etc. Does that make any sense? Nika. |
Quote: |
Resolution in imaging is accepted to indicate the smallest measurable unit. However, it has recently come to be scrutinized as having a secondary component of accuracy. e.g. how faithful is detail captured. |
Quote: |
If you define "resolution" this way, and stack on top of it "accuracy", then statistically you can only count resolution as meaningful if it is accurate. Looking at it from a mathematical point of view, simplistically we know that if you have two measurements that are accurate to two decimal places and you multiply them, you have a resultant number that may contain four decimal places of precision, but the last few decimal places are junk... they are not accurate because they result from an infinitely precise math function produced from rounded or truncated numbers. |
Quote: |
posted by Paul Frindle: Now if we look closer at the screen (and camera) quantisation we can see that if we wanted to remove the quantisation of the pixels statistically from the whole system, one method (rather than simply trying for infinite numbers of pixels) would be to shake them around at random physically whilst holding the actual image position constant wrt to the viewer. In this way all positional possibilities for the optical image would be serviced at one time or another. All quantisation would be removed from the imaging system and be replaced by amplitude noise. |
JDSStudios wrote on Sat, 05 June 2004 11:41 |
We could perhaps define audio analogue "resolution" as the difference, or lack of, between the recorded signal and the live natural audio. |
josh wrote on Sat, 05 June 2004 06:10 |
audio recording is not about "resolution", "accuracy", "precision", "frequency response", or any other measurable thing... our current state-of-the-art in both analog and digital is way beyond what is required in these categories. |
Quote: |
When we record, we intentionally distort the recordings because audio is all about perception, and human beings are extremely non-linear perceivers and processors of auditory stimuli. Making an accurate, linear recording will not result in one that is lifelike to listen to! To make a lifelike recording requires distortion. We would do better to focus on exactly what type of distortion is beneficial to making lifelike recordings, rather than the meaningless argument over the minutia of what does not make a real difference. |
josh wrote on Sat, 05 June 2004 14:10 |
When we record, we intentionally distort the recordings because audio is all about perception, and human beings are extremely non-linear perceivers and processors of auditory stimuli. Making an accurate, linear recording will not result in one that is lifelike to listen to! To make a lifelike recording requires distortion. We would do better to focus on exactly what type of distortion is beneficial to making lifelike recordings, rather than the meaningless argument over the minutia of what does not make a real difference. |
Eric Vincent wrote on Sat, 05 June 2004 06:57 | ||
Paul, In PhotoShop, there is a filter which does this, called Unsharp Mask. It is a digital emulation of a darkroom technique, as it happens. It's quite simple: On one layer, the program isolates contrasting pixels, and amplfies the contrast, to user specification. On the second layer, it blurs those contrasting regions in a Gaussian fashion (hence "Unsharp Mask"), to user specification. The program then overlays the two effects. The combination of the blurred image with the sharpened image, creates the desired effect. The trick is in the proper mixing of the sharpening and the blurring, and fine tuning the two so they work together optimally. As in audio?? ...kinda, sorta, but not really. My experiences with photography have not lent well to parallels with audio in regards to this "resolution" issue. Photography has inherent variables which exist outside the audio paradigm, and vice-versa. |
JDSStudios wrote on Sat, 05 June 2004 06:41 |
We could perhaps define audio analogue "resolution" as the difference, or lack of, between the recorded signal and the live natural audio. |
Quote: |
We could have a scale, with its maximum value when there would be no measurable or perceivable difference between the original live audio, and the recorded audio. |
josh wrote on Sat, 05 June 2004 09:10 |
One problem with this whole debate is that whatever we are comparing has to be measurable in some way. So if we isolate the problem deeper into the system... the tape head amp input to tape playback head amp output vs. the ADC input to the DAC output for example, then provided we have a controllable, repeatable and measurable way... <snip happens> ... In order to produce a repeatable audible event for a test like this, we would have to do it with a loudspeaker, an amplifier, and then the signal generator on the front end of the amplifier. In this scenario far and away the dominant feature of the test is the loudspeaker. |
Quote: |
So, using actual auditory information (a real sound you can hear), we cannot conduct a meaningful test regarding the resolution of a recording device. |
Quote: |
A recording device such as a digital recorder or an analog tape machine does not actually receive audible sound as input. It receives a time-varying voltage as input. The only meaningful measurement of a device like this is to measure it with respect only to the accuracy and precision of the recording and playback of the time-varying voltage. Input a known signal (some kind of broadband, transient, etc., agreed-upon signal source), record it, then play back the ouput, and compare it to the original signal. |
Quote: |
<snippage>...is a crude test and completely unreliable because our test equipment is the same equipment <snippage... again>...much of the most accurate audio measuring equipment we have access to is the very digital recording equipment we would like to measure. I don't know because I have not investigated digital capture equipment for audio other than that used for recording. If there were measuring equipment capable of determining this difference, then it means that somewhere there exists a recording and playback device that is orders of magnitude more precise and more accurate than even the high-end recording gear that we would like to test. I don't know if this exists. |
Quote: |
Again, though, this all is a moot point because no sane audio engineer is going to try and make a rational argument |
Quote: |
Making an accurate, linear recording will not result in one that is lifelike to listen to! To make a lifelike recording requires distortion. |
JDSStudios wrote on Sun, 06 June 2004 07:52 |
The flaw of the argument is some people wanting to nail down a singular definition for a particular word: "resolution" Don't like "resolution" because of the picture pixel comparison? Ok, then use the term "accuracy". Don't like accuracy?.. We could go on and on until you realize there is no perfect word either to communicate anything. |
JDSStudios wrote on Sun, 06 June 2004 08:24 |
Theory is when you know something, but it doesn't work. Practice is when something works, but you don't know why. Engineers combine theory and practice: Their questions and answers do not work and they don't know why. |
Quote: |
The difference between theory and practice is that in theory they are the same. Nika. |
xaMdaM wrote on Sun, 06 June 2004 02:42 |
Josh, take a look at may last post please. This quickly could become a pissin' contest that it isn't. Let me see if I can clear up why I think we are about to get into an apple-n-oranges thing... and how to avoid slipping into that mode of thinking. |
Quote: |
You had it! You had it!! You had IT!!! Now, go back and modify the thought this way... OK, you realize that the machine (analog OR digital) is receiving a voltage. The voltage-signal must contain the equivalent of a known quantity of a black and white image... e.g. DC/20kHz/DC/20kHz/DC/etc... Then measure the latent information/signal at the storage device root. (tape/harddisk/ram) To complete the circle of reproducibility, reverse the process. But the process can only be used within the realm of an electronic signal. You cannot for the sake of scientific evaluation leave that realm, unless all things are exactly the same. Otherwise yu are comparing apples to oranges. Do you see that point? |
Quote: | ||
I own a simple little 100MHz scope that'll measure just about anything that is in the analog domain. The last scope I had when I was working in the digital imaging manufacturing domain was a 500MHz 16 bit buss analyzer... 20 years ago... Check out this scope... http://www.picotech.com/pc_oscilloscope_specification.html Measurement is not the issue. It's what to measure that's at the heart of the debate. |
Quote: | ||
Can you expand on this a bit? I'm serious... I must not be getting something through my thik haid or something. My ears don't artificially distort a low volume audio signal... that I know of. I do know that my brain will attempt to process and clean up a compressed sound wave that the ear naturally compresses in cases of excessive sound pressure levels... but I don't necessarily think I buy in to the statement that to capture a lifelike recording that I purposefully want or need to add ANY kind of distortion. Max |
josh wrote on Mon, 07 June 2004 21:00 |
I wrote about this oh on about page 9 of this thread... When aided by your other senses, your brain is able to provide selective filtering and sensory effects to sound input... kind of like how in a noisy environment you can carry on a conversation with someone but if you made a recording of it you would not be able to hear it. You can "focus" on sounds, filter others... When you walk into a room, for example, your brain can quickly begin to filter out the reverberation of the room so you don't notice it. There are all kinds of psychoacoustic effects that should be taken into account when considering what's important about audio recording and reproduction. |
josh wrote on Mon, 07 June 2004 16:00 |
I have skipped the articles dealing with the analogy of audio to digital imaging... just not enough hours in the day and it didn't seem relevant... can you point me to the exact post? |
Quote: |
Maybe your point is that if in fact you could play it back through the mic pre and microphone it would be a usable test? I don't know if that was what you were trying to say or not. But if it were, well that wouldn't work either, since the microphone is the bottleneck to begin with, and having to ram information through it twice would just compound the distortion. |
Quote: |
bunny trail... |
Quote: |
During the early development of ADSL... |
Quote: |
Can you expand on this a bit? |
Quote: |
I wrote about this oh on about page 9 of this thread... When aided by your other senses, your brain is able to provide selective filtering and sensory effects to sound input... kind of like how in a noisy environment you can carry on a conversation with someone but if you made a recording of it you would not be able to hear it. You can "focus" on sounds, filter others... When you walk into a room, for example, your brain can quickly begin to filter out the reverberation of the room so you don't notice it. There are all kinds of psychoacoustic effects that should be taken into account when considering what's important about audio recording and reproduction. |
Nika Aldrich wrote on Mon, 07 June 2004 17:38 |
Again, I am trying to answer the question about why it would be necessary to intentionally distort something in order for it to sound un-distorted. Nika. |
lucey wrote on Thu, 24 June 2004 16:31 |
It's as basic as asking yourself this ... if you were a kid which waterslide would you rather ride? a) the smoothe and wet one with some friction and a few slight turns, or b) the one with tiny stair steps and perfectly even friction, straight down? Scientific minds are of great help to many aspects of music, but the scientific ear does not hear the subtleties of the stair steps as the catastophic loss of integrity that musicians hear. |
lucey |
Digital maintains linearity to varying smaller degrees with better technology, yet changes the fundamental quality of the sound from a vibe to a number. This simple fact is the whole biscuit. It's as basic as asking yourself this ... if you were a kid which waterslide would you rather ride? a) the smoothe and wet one with some friction and a few slight turns, or b) the one with tiny stair steps and perfectly even friction, straight down? |
Chuck wrote on Thu, 24 June 2004 11:31 |
Hi Brian,, I think you are - by intuition - mentioning one of the most important distortion mechanisms in digital audio. During playback we have to generate a analog signal out of squares. Even if the DAC chip has some inbuildt analog filter, it was squares in the beginning. And squares consist of 3rd order harmonics only and the faster the settling time of your DAC the sharper the edge, and the higher the harmonics go. |
RobertRandolph wrote on Thu, 24 June 2004 14:57 |
What the heck is all this talk about squares!?! I really dont see where a lot of you are getting information from... IF you plot a sine wave on a peice of paper, given 2 points... are you plotting squares?! NO. Because you have a mathmatic formula telling you it is a sine or cosine. Likewise, with digital audio conversions, we know it will be a sine. So the converter does not just "draw squares". |
Quote: |
Likewise, if it is a square wav, saw, triangle etc.. You must remember this is combonation of sine waves with various harmonics at specific amplitudes and intervals. Still sines. |
lucey wrote on Fri, 25 June 2004 04:22 |
Doesn't a converter draw sines of a slope that is a guess? |
Quote: |
How can they be when an A to D measures in samples and a D to A tries to remember the curve? |
Quote: |
And just for fun, what if a "perfect" and distortion free reproduction is not the most musical? what would you do then? |
Quote: |
Then what are those 3rd harmonics all about? When will they go away? |
Quote: |
And where is the full impulse energy of the wave as it was played in the air? It does go away with digital ... any digital. |
Nika Aldrich wrote on Fri, 25 June 2004 09:26 |
The task of digital is to present to the ear what the ear hears, so while ALL of the energy of the wave is indeed not captured by digital, since digital filters like the ear does, the entirety of that wave that the ear hears is indeed captured and not distorted. Nika. |
Nika Aldrich wrote on Fri, 25 June 2004 08:26 |
The task of digital is to present to the ear what the ear hears, so while ALL of the energy of the wave is indeed not captured by digital, since digital filters like the ear does, the entirety of that wave that the ear hears is indeed captured and not distorted. Nika. |
djui5 wrote on Fri, 25 June 2004 22:36 |
Where did you get this from anyways?....digital filters like the ear does? |
lucey wrote on Fri, 25 June 2004 21:50 | ||
So from this perspective which 2 channel converter sounds the most accurate and which one the most musical, if any different to you? |
fuze wrote on Fri, 25 June 2004 07:50 |
can you explain exactly what you mean? this sounds like you are suggesting that digital doesn't capture inital peaks as well as any analog? all the best! steve parker. |
lucey wrote on Wed, 30 June 2004 06:04 |
Steve, I admit to a limited understanding of the subtle technicalities of digital recording. I do know tone however, and specifically I know vibrations. |
Quote: |
To my ear any AD DA (even my amazing Pacific Microsonics) is fundamentally altering the waveform's quality. |
Johnny B wrote on Thu, 01 July 2004 08:02 |
Dan, Are you saying that real improvements could be made by going to 32-bit converter chips? |
Johnny B wrote on Thu, 01 July 2004 08:02 |
Dan, Are you saying that real improvements could be made by going to 32-bit converter chips? Sure some of those 32 bits could be wasted, but those extra bits could come in handy, at some point down the road, ya never know... LOL. |
danlavry wrote on Thu, 01 July 2004 19:28 | ||
Correction: I said (for 32 bits) So yes, make me a mic that can supply say .4 milliwatt of noise free signal into a 4 Ohms resistor, than have a mic pre amplifier with no noise at all and we are there I Should have said "make me a mic that can supply say 1/4 watt (1 volt into 4 Ohms) of noise free signal into a 4 Ohms resistor... That 32 bit was carzy enough at any power level, but .4 watt was the wrong number so I am correcting it. Sorry, i did it too fast thus the error. BR Dan Lavry Lavry Engineering |
Nika Aldrich wrote on Wed, 30 June 2004 09:37 | ||
I am not surprised that you hear differences between analog and digital systems. I've heard your music and work you've mixed, and my personal observation is that you tend toward very rich sounding analog material and equipment. |
Quote: | ||
Perhaps the PMi DOES alter the waveform's quality. How have you tested it and what other converters have you tested - and how? Nika |
lucey wrote on Thu, 01 July 2004 20:17 |
what are you implying? (as for my work you've heard one or two older mixes that were poor by current standards. you've heard no recordings made in the last 3 years, nor any mastering work) |
danlavry wrote on Fri, 02 July 2004 20:40 |
I say: Hi, I am Dan Lavry and I wish you listened to a Lavry |
Loco wrote on Tue, 25 May 2004 18:24 |
It's not that those boxes are bad. They sound good. Maybe it's just they don't sound all that money. You can improve your sound a lot more putting your money somewhere else. |
Brent wrote on Sat, 15 May 2004 15:08 | ||
I believe that there is a 28-bit processor. But it is not a true usable 28-bit any more than a 24 is really 24. Lets say that they get to 32, so that it performs at 28. Will the microphones used to record hinder the process? |
Sjoko wrote on Sun, 04 September 2005 17:06 | ||||
The only available 28 bit converter I know of is the Stagetec Nexus, a stage gaining converter which converts to 28 bit. The key advantage of this converter is that it keeps the integrity of the signal intact throughout the dynamic range, where in "normal" converter technology the lower levels of a recording get messed up. This "difference" becomes a big, easy to hear factor when recording acoustic instruments in particular. To my ears this converter is the only one that adequately represents the reality of the audio you try to capture (providing, of course, the million plus other factors involved). I have had them for a number of years, and the only problem I've had is that other factors pop their evil heads up - as in I had to replace all wiring to keep the signal intact etc. Next move will be to move the mic converter boards into the tracking rooms to minimise distance. |
Sjoko wrote on Sun, 04 September 2005 12:06 | ||||
The only available 28 bit converter I know of is the Stagetec Nexus, a stage gaining converter which converts to 28 bit. The key advantage of this converter is that it keeps the integrity of the signal intact throughout the dynamic range, where in "normal" converter technology the lower levels of a recording get messed up. This "difference" becomes a big, easy to hear factor when recording acoustic instruments in particular. To my ears this converter is the only one that adequately represents the reality of the audio you try to capture (providing, of course, the million plus other factors involved). I have had them for a number of years, and the only problem I've had is that other factors pop their evil heads up - as in I had to replace all wiring to keep the signal intact etc. Next move will be to move the mic converter boards into the tracking rooms to minimise distance. |
Sjoko wrote on Mon, 05 September 2005 11:22 |
Yup Ronny, you're right, but it ain't no recording gear, just something lacking a decent clock |
Thomas Lester wrote on Tue, 13 July 2004 15:27 |
Agree that it's not "better or worse". It's different. We hear analog every second of our lives... that's why digital sounds different to us. As far as noise floor goes... I worked in a large office environment once (a large cube farm). The noise from all the phone chatter was unbearable to most and they got head aches and couldn't concentrate on their jobs. So... they brought in an acoustician to kill the noise. I expected him to put up acoustic tiles, traps, etc... Nope... he actually ADDED to the noise. He put in ceiling speakers all around the place and piped white noise at a very low volume through the speakers. It was low enough that you barely heard it (an that was only while working late after everyone went home). The complaining stopped and everyone was happy. I asked him why, and basically it was enough "noise" to convince your brain to ignore the bad sounds in the room (i.e. other people talking). I'm guessing the noise floor in analog is what makes it sound so good or the lack there of in digital is why it doesn't sound "right" to us. -Tom |
Ronny wrote on Sun, 04 September 2005 20:59 |
Yamaha PM1D, 28 bit ADC's, 27 bit DAC's. |
maxdimario wrote on Tue, 06 September 2005 05:47 |
now I'm gonna ask a very simple question, and I'd like some comments from you guys technical and non technical. Why is it that when I listen to an analog recording with a chain consisting of: tube mic in perfect shape, tube pre (or exceptional low feedback solid state), discrete or tube analog recorder... I can hear the feel of the music more evidently than the equivalent path with digital recording?. It sounds like a real person in the room with you, especially when reproduced on a good amp. There must be an electronic distortion in the digital process that plays with what makes a sound realistic to our sense of hearing. let's leave easily audible distortion out of it and focus only on the phsycological effect, please. what could it be? |
maxdimario wrote on Tue, 06 September 2005 05:47 |
now I'm gonna ask a very simple question, and I'd like some comments from you guys technical and non technical. Why is it that when I listen to an analog recording with a chain consisting of: tube mic in perfect shape, tube pre (or exceptional low feedback solid state), discrete or tube analog recorder... I can hear the feel of the music more evidently than the equivalent path with digital recording?. It sounds like a real person in the room with you, especially when reproduced on a good amp. There must be an electronic distortion in the digital process that plays with what makes a sound realistic to our sense of hearing. let's leave easily audible distortion out of it and focus only on the phsycological effect, please. what could it be? |
Ronny wrote on Mon, 05 September 2005 22:24 |
This is an old post but thought I'd respond anyway as I find the method used at the office interesting. The sound tech dithered the room. |
Quote: |
It's the analog distortion that is euphonic to your ears |
Quote: |
I don't experience this same problem with high-end digital gear. Most digital gear is cheap and sounds cheap. |
Bob Olhsson wrote on Tue, 06 September 2005 22:54 |
I think the most damaging distortion comes from not leaving enough headroom. A lot of digital gear simply hasn't got enough analog headroom to work with full-scale signals. The first solid state mike preamp that I thought really stood up to the tube preamps we used at Motown was a prototype that my friend Pat Duran built for Deane Jensen. It clipped at something like +40. Upon realizing headroom was a common problem, I began experimenting with lower line levels and haven't looked back. I'm finding the same issues with digital multitracking. A lot of gear doesn't sound very good when you get within 6 dB of the top. |
Ronny wrote on Wed, 07 September 2005 01:50 |
What is the physical reason for not getting a good sound above -6dBFs, Bob? Not that I'm a stickler for squeezing every bit out, especially at 24 bit, but you'll find a trainload |
bobkatz wrote on Wed, 07 September 2005 07:36 |
I've built unbalanced IC-based mixers with + and - 18 volt power supplies, using Burr Brown and other high quality opamps and despite the pedigree, I found they just sound better running "0 VU" at lower than 0 dBu, even as low as -6 dBu (0.0775 volts). Clipping point of these, running unbalanced, is nominally +20 to +22 dBu. |
David Glasser wrote on Wed, 07 September 2005 12:08 | ||
I have a Calrec mixer, mid 80s vintage, that uses run-of-the-mill chips, but operates at -4dBu (ie, 4 db below 0 VU or 0 dBu). I've wondered if that's one of the reasons it sounds so good, despite the multitude of TL072s and NE5534s (which I will replace... someday). |
Eric Bridenbaker wrote on Wed, 07 September 2005 15:09 |
Paul Frindle has given a very comprehensive explanation as to as why it's a good idea to keep digital levels low, and why you can't always trust digital peak meters. It can be found here: http://recforums.prosoundweb.com/index.php/mv/msg/4918/0/64/ 4645/?SQ=14738cbb8d492e71f640277262efe15f |
Bob Olhsson wrote on Wed, 07 September 2005 05:54 |
I think the most damaging distortion comes from not leaving enough headroom. A lot of digital gear simply hasn't got enough analog headroom to work with full-scale signals. |
Ronny wrote on Mon, 05 September 2005 19:21 | ||
It's mainly an FOH board, but many, many live recordings have been made off of them. The tower can run 8 MY8-AT's which gives 64 channels of ADAT. Clock is excellent. |
Nika Aldrich wrote on Thu, 24 June 2004 16:40 | ||
Why are you ignoring the role of the reconstruction filter at the end of the D/A converter? Digital does NOT have all of the stair-steps that you describe. Digital is only a representation of the waveform - NOT the waveform itself. The original waveform still has to be reconstructed from the sample points. Simply doing a "dot to dot" or sample-hold reconstruction is clearly inadequate, as neither of those re-create the original waveform. When proper reconstruction filtering is done there is no "stair-stepping," and continuing to refer to such obfuscates the way in which digital actually works and confuses the questions. Nika. |
Ronny wrote on Fri, 09 September 2005 18:52 |
Infinity. |
dcollins wrote on Sat, 10 September 2005 12:30 | ||
Bzzzzzt! DC |
jazzius wrote on Fri, 14 May 2004 09:56 |
George, we hear about the resolution of digital all the time.....24 bits, 44.1, 192.....2.8 million whatevers.... ...do you know if anyone has ever worked out the resolution of analog?.....how many bits would it be equivelent to?.....i know this is bit of a strange question, but i'd love to be able to give my customers a smart-arse answer for why analog sounds better then digital... ...cheers.....Darius |
Nick Sevilla wrote on Fri, 09 January 2009 17:23 | ||
I can give you a resolution for a 1974 8048 Neve desk I helped restore : 10 to 100,000 cycles from the line / mic inputs through to the mix buss. That's 10 Hz to 100KHz. this is better than all semi-pro and even some pro desks out there. The EQ modules in the desk specd' out to : 1066 EQ modules - 10Hz to 97500Hz (average from 18 modules). Older modules, and well within spec. 1081 EQ modules - 10Hz to 100,000 Hz (average for 36 modules). Noise floor of -97 dB. A little better than CD quality, but not much. THD of 0.5% using 1kHz tone at +23 dB, a little less than the original spec, but still usable headroom. We only had a sine generator that went from 10Hz to 100,000Hz, so it could go higher and lower, but we kept referring to the original spec, and matched it. Cheers, Nick |
Andy Peters wrote on Sun, 11 January 2009 00:26 |
Frequency response is NOT resolution. -a |
Fig wrote on Fri, 06 February 2009 17:51 | ||
Hi Andy, To be honest, I have been thinking about this topic since it was posted. I have searched and searched for a definition of "resolution" as it pertains to audio - there is tons of BS regarding video, printers, faxes, photos, etc. While I agree that the freq response only tells us part of the story - how WOULD someone go about determining the other aspects and coming up with an "answer" for "resolution"? I'm picturing some kind of equation utilizing cycles-per-second (Hz) and inches per second (IPS) where maybe the "time" cancels out and perhaps cycles per inch somehow connected to the flux level, dynamic range, track width, blah-blah. With your experience and knowledge, Andy, how would you go about generating such a spec? Fig |
Tomas Danko wrote on Fri, 06 February 2009 12:09 | ||||
<snip> Please bare with me for a while here... |
Jon Hodgson wrote on Wed, 01 April 2009 17:30 |
Actually analogue systems can have quantization errors too. When I met Tim De Paravicini (a man who certainly seems to know his cutting lathes) I was surprised to discover that the error caused by the fact that you can either have a vinyl molecule or not (so the resolution of a cut is +/- half a molecule) was only 90dB below full signal, I'd always assumed that the molecules were too small for that to be a consideration... it seems not. The fact that you random variations in the matrix results in those errors being random and thus white noise. It amused me somewhat to discover that vinyl is actually a dithered system with a quantization step size approximately equivalent to a 15 bit system. |
Jay Kadis wrote on Thu, 02 April 2009 15:55 | ||
We have a similar issue in analog tape due to the finite dimensions of the individual magnetic domains. |
Jon Hodgson wrote on Wed, 01 April 2009 20:30 |
Actually analogue systems can have quantization errors too. |
Jon Hodgson wrote |
When I met Tim De Paravicini (a man who certainly seems to know his cutting lathes) I was surprised to discover that the error caused by the fact that you can either have a vinyl molecule or not... |
Jon Hodgson wrote |
The fact that you [have] random variations in the matrix results in those errors being random and thus white noise. It amused me somewhat to discover that vinyl is actually a dithered system with a quantization step size approximately equivalent to a 15 bit system. |
Andrew Hamilton wrote on Tue, 30 June 2009 03:11 | ||||||
Errors of truncation, when the lacquer molecules refuse to be cut in two. Not a quantum "decision," though, since the molecule _can_ be cut in two. Just a dance between inertia and momentum. No sampling means no quantization, though, doesn't it?
Aren't we talking about cellulose nitrate at the mastering lathe? Vinyl (or pvc with black pigmentation) would be a pressing concern, no?
So, the dynamic range is equivalent to that of a 14+ bit LPCM carrier? No actual steps in analog, though, right? Andrew |
Jon Hodgson wrote on Tue, 30 June 2009 13:29 |
Ever tried to cut a molecule? I don't think you're going to do it with a chisel (which is effectively what a cutting lathe is). |
Andrew Hamilton wrote on Thu, 02 July 2009 08:08 | ||
The molecules in the lacquer may refuse to be cut by a mere heated wiggling blade... ...more nitro-cellulose? (; If I add uncorrelated dither to an analog signal, that would make the signal have slightly more hiss, but I hesitate to call the result quantized audio, since the available amplitude modulation values are still continuous in scope. Any error, as you have said, is heard as part of the noise floor, although at 90 dB into vinyl, I suspect that other surface noises are predominant. Dither is supposed to be the remedy for the audibility of quantization error. However, in your analogy, the molecular auto-dither is what's causing the "error." |
Jay Kadis wrote on Thu, 02 July 2009 15:38 |
The dimensions of a vinyl chloride molecule are on the nanometer scale. Polymerized vinyl chloride chains may consist of hundreds to thousands of monomer units and complex side-chains may form producing a very complex molecule. How the surface of a groove may interact with a stylus depends on the particular composition of the polyvinyl chloride and would be quite complex to analyze, requiring scanning electron microscopy to image the surface. Given the dimensions of a stylus (microns), the molecular size of the particles themselves (nanometers) would not matter as much as the way the molecules pack together as there's a thousand-fold difference in size between the stylus and the surface molecular arrangement. I doubt the stylus could follow individual molecules and would rather respond to the macro-level surface that would depend on the aggregation of vinyl molecules. |
Jay Kadis wrote on Thu, 02 July 2009 15:38 |
Given the dimensions of a stylus (microns), the molecular size of the particles themselves (nanometers) would not matter as much as the way the molecules pack together as there's a thousand-fold difference in size between the stylus and the surface molecular arrangement. I doubt the stylus could follow individual molecules and would rather respond to the macro-level surface that would depend on the aggregation of vinyl molecules. |
johnR wrote on Fri, 03 July 2009 04:04 | ||
Excellent points, but bear in mind that a thousand-fold difference in linear movement is only 60dB. |
Bob Olhsson wrote on Wed, 07 September 2005 20:19 |
I think Bob Katz treats this issue very well. Since I can't rebuild most of the gear I have to use, I simply take the time to find the sweetest sounding gain structure for the system I'm using. |
sfdennis wrote on Fri, 14 May 2004 13:35 |
I'm surprised the thread got this far without anyone mentioning this: Digital Word Length <=> Analog Noise Floor Digital Sample Rate <=> Analog Bandwidth So for example, in an ideal converter system at 96kHz would correspond with an analog bandwidth of 0-48kHz. A digital word length of 24 bits would correspond with an analog noise floor of about -132dBFS, notwithstanding dither. -Dennis |
Dusk Bennett wrote on Wed, 18 November 2009 22:01 | ||
Dennis, I want to hit this point harder (and I may repost this separately since this is what I came here to ask). I got involved in an academic discussion today with a student regarding proper digital recording technique. He claims his professor (a Pohlmann student who's quite knowledgeable in her own right) stated that to obtain the highest resolution possible it is a MUST to record your sample as high up against 0dBfs as possible. (Which is a really bad idea IMHO). It's as if he's suggesting amplitude and "Word Length" are mutually exclusive (and you imply they are). I have argued with the students here that there is no reason to "over record" in digital because the medium is much more forgiving than analog, and analog hardware cannot handle such hot levels coming out of digital converters. So I recommend that they print near 0VU (-20dBfs). There are many benefits to this in my mind (which I'll get into later), aside from the obvious. So what is it? Is it really necessary in digital to print all the way up to 0dBfs to maximize your word length? Are amplitude and word length in digital truly related in that way? What do the facts say in digital theory on this? The problem I have with recording so hot going into the box (especially if you end up interfacing with the outside analog world) is how much gain reduction one must apply to actually get your plug-ins and inserts to actually avoid distortion. Recording a guitar at -2dbfs leaves you no head room to compress and add eq after the fact without altering the "word length" of your inserts by reducing it's input to compensate for a very hot output. At that point aren't you just trading one set of problems for another? I still believe (based on alot of practical experience) that there is no reason to print so fricking hot to digital. As long as you fall within the range of a typical VU meter your headroom in digital will allow for more than enough gain processing after the fact and not destroy the quality of the signal. How would you respond to this? |
Dusk Bennett wrote on Wed, 18 November 2009 16:01 |
So what is it? Is it really necessary in digital to print all the way up to 0dBfs to maximize your word length? Are amplitude and word length in digital truly related in that way? What do the facts say in digital theory on this? |
Quote: |
The problem I have with recording so hot going into the box (especially if you end up interfacing with the outside analog world) is how much gain reduction one must apply to actually get your plug-ins and inserts to actually avoid distortion. Recording a guitar at -2dbfs leaves you no head room to compress and add eq after the fact without altering the "word length" of your inserts by reducing it's input to compensate for a very hot output. At that point aren't you just trading one set of problems for another? |
Quote: |
I still believe (based on alot of practical experience) that there is no reason to print so fricking hot to digital. |
Quote: |
As long as you fall within the range of a typical VU meter your headroom in digital will allow for more than enough gain processing after the fact and not destroy the quality of the signal. |
Geoff Emerick de Fake wrote on Sat, 12 December 2009 20:01 | ||
|
Tomas Danko wrote on Sun, 13 December 2009 09:31 |
So all in all, advocating staying as close to full scale based on theory will only hurt actual recordings due to the way it's been practically implemented. |
Geoff Emerick de Fake wrote on Mon, 14 December 2009 10:12 |
Please remember, the question was: "Is it really necessary in digital to print all the way up to 0dBfs to maximize your word length? Are amplitude and word length in digital truly related in that way?". The answer to that simple question is simple. Now, the poor implementation of the analogue counterparts may impose some limitations and operating workarounds, but the simple relation: "More level = more resolution" is a mathematical fact. Fortunately, 24bit bit-depth offers so much headroom that it is possible to displace the operating range in an area where both analog and digital gear are happy. |
minister wrote on Wed, 23 December 2009 19:58 | ||
This is the myth of Digital Audio that permeates the internet. More level ≠ More Resolution. It is simply not a Mathematic Fact. |
Quote: |
It give you more "foot room", not "head room". |
Quote: |
Recording close to 0dBFS invites not only inter-sample peak distortions, it is forcing your pre-DAW analog gear to operate beyond it's optimal range. |
Geoff Emerick de Fake wrote on Sat, 26 December 2009 18:28 | ||||||||
|
compasspnt wrote on Mon, 18 January 2010 14:08 |
I will inform George Massenburg to STOP his recording on analogue tape at significantly lowered levels. |
Edvaard wrote on Mon, 18 January 2010 14:25 |
If someone were about to go out driving for the first time and asked the simple question; "is it true that I need to have the accelerator all the way to the floor to reach top speed?" should I just give as answer a simple (and true) "yes" and leave it at that? |
Quote: |
There is a series of sharp curves a half mile down the road, not to mention a speed limit whatever the road is like, and it being the first drive, no one knows how well the brakes work yet. But he didn't ask about any of that, so I'll only answer what he asked. |
Quote: |
People here assumed by the very simplicity of the question that more information beyond the simple question might be warranted. All people are trying to do is to say an obvious newcomer "well, there's more to it than that." |
Geoff Emerick de Fake wrote on Mon, 18 January 2010 13:41 |
And again: Please remember, the question was: "Is it really necessary in digital to print all the way up to 0dBfs to maximize your word length? Are amplitude and word length in digital truly related in that way?". There's only one answer to that: Yes Because the question is exactly the same as: "Is it really necessary in analog to print all the way up to maximize S/N ratio? Are amplitude and S/N truly related in that way?". Again, yes. I'm not saying that there won't be any problems, because there will be distortion, but the QUESTION doesn't mention distortion. To a unidimensional question, I bring a unidimensional answer. If distortion had been mentioned in the question I would have said that S/N and THD are mutually exclusive. If the question was about quality, I would mention the compromise between wordlength and loss of headroom. |
Geoff Emerick de Fake wrote on Mon, 18 January 2010 12:41 |
And I don't really need to read any "digital for dummies" book after 37 years of product development... |
minister wrote on Mon, 25 January 2010 15:15 |
Calling this book "digital for dummies" just reveals your arrogance which does not impress me in the least. |
Quote: |
It is hardly a book for dummies. I suggest perusing it. |