mallcore pop
Hard Nipples!!!
DJL said:Isn't the transformers a kids cartoon?
That's what I've been trying to tell you!!!
NOBODY brickwalls Optimus Prime's square waves!!!
Transformers
- Thread starter chessrock
- Start date
That's what I've been trying to tell you!!!
NOBODY brickwalls Optimus Prime's square waves!!!
DJL said:Come on cress, this is a good thread, don't put the fire out yet.
Well, I'm sure you like it now . . . since it's now your thread, basically.
What's the title of the thread again? I forgot.Hey, I'm just kidding. This is interesting. Carry on, guys. It's been a while, actually, since I've seen a good thread about the merits of transformers hijacked and turned in to an analogy versus digital or a POD versus tube amp debate.
Hey, DJL . . . I still still say a Stratocaster kicks a Les Paul's ass any day of the week . . . Soundcraft is better than Mackie . . . PC's over Macs, Salma Hayek over Angelina Jolie, Christina over Britany . . .
Oh, and Joe Satriani's licks eat Stevie Ray Vaughan for dinner. How 'bout them apples, huh?
Harvey Gerst
New member
Okay, continuing on with the thread, part of my concern is a little thing called rise time, transient response, slew rate, bandwidth, and a host of other names. It's all about the same thing; how fast the device will respond to a signal with a sharp upswing.
One of the characteristics of a square wave is a sharp, almost right angle turn at the start of the signal. This sharp angle is the result of all the odd harmonic content added to the fundamental frequency.
If I'm right about the Nyquist filter not passing the odd harmonics, then I'm guessing that the rise time suffers greatly even with a 3,000 Hz signal. Rather than a sharp pivot point, where the signal goes almost straight up, the signal would just gradually climb, severly changing the character of the note.
Wide band alnalog equipment handles this kind of signal a lot easier than CD does. Most good analog equipment can pass a 10 kHz square wave with surprisingly little signal degradation - it comes out still looking kinda like a square wave. If I'm right, a CD chops off all the square waves harmonics above 7,500 Hz, leaving basically a sine wave.
To me, a square wave ain't a sine wave. Most of our signal chain at our studio is analog and wide band analog at that. It goes out pretty flat to 100 kHz until it hits our recording devices. My Telex tape deck was flat out to almost 40kHz, and still responded to signals above that.
When I do a recording that will eventually end up on a CD, I expect to hear a difference between the analog sound in the studio and the CD - and I do. It's different. It's perhaps better in some ways and a lot worse in other ways.
I'm not ready to give up digital waveform editing and the other advantages digital offers, but I realise (and accept) that something is lost in the translation.
I mourn that loss.
One of the characteristics of a square wave is a sharp, almost right angle turn at the start of the signal. This sharp angle is the result of all the odd harmonic content added to the fundamental frequency.
If I'm right about the Nyquist filter not passing the odd harmonics, then I'm guessing that the rise time suffers greatly even with a 3,000 Hz signal. Rather than a sharp pivot point, where the signal goes almost straight up, the signal would just gradually climb, severly changing the character of the note.
Wide band alnalog equipment handles this kind of signal a lot easier than CD does. Most good analog equipment can pass a 10 kHz square wave with surprisingly little signal degradation - it comes out still looking kinda like a square wave. If I'm right, a CD chops off all the square waves harmonics above 7,500 Hz, leaving basically a sine wave.
To me, a square wave ain't a sine wave. Most of our signal chain at our studio is analog and wide band analog at that. It goes out pretty flat to 100 kHz until it hits our recording devices. My Telex tape deck was flat out to almost 40kHz, and still responded to signals above that.
When I do a recording that will eventually end up on a CD, I expect to hear a difference between the analog sound in the studio and the CD - and I do. It's different. It's perhaps better in some ways and a lot worse in other ways.
I'm not ready to give up digital waveform editing and the other advantages digital offers, but I realise (and accept) that something is lost in the translation.
I mourn that loss.
Rossi
New member
Harvey, I think you're right and not right at the same time.
Talking about a square wave and about harmonics is two ways of looking at the same phenomenon. Thus, a square wave is or is not a sine wave, depending on how you look at it. If you remain in the realm of harmonics and limit yourself to the audible frequencies, a 7,500 Hz square wave is the same as a 7,500 Hz sine wave. Rise time, too, is a way of looking at frequency response. Digital has exactly the rise time it takes to reproduce frequencies within its bandwith. No more no less.
The question remains: do the frequencies above the audible range (which for most people is far less than 20 kHz) make a difference? I think this question is very difficult to answer. I don't think you could hear a difference with a simple square wave or any test signal. Also keep in mind, that few instruments produce frequencies above the audible range.
Nonetheless, I think it's perfectly possible that you could hear a difference in a full blown production with many instruments and lots of harmonics - audible or not - interacting with each other.
And yes, from a pure guts standpoint, well engineered analog productions do sound better to me than digital productions, even when "reduced" to 16bit 44,1 CD format in the final stage.
Talking about a square wave and about harmonics is two ways of looking at the same phenomenon. Thus, a square wave is or is not a sine wave, depending on how you look at it. If you remain in the realm of harmonics and limit yourself to the audible frequencies, a 7,500 Hz square wave is the same as a 7,500 Hz sine wave. Rise time, too, is a way of looking at frequency response. Digital has exactly the rise time it takes to reproduce frequencies within its bandwith. No more no less.
The question remains: do the frequencies above the audible range (which for most people is far less than 20 kHz) make a difference? I think this question is very difficult to answer. I don't think you could hear a difference with a simple square wave or any test signal. Also keep in mind, that few instruments produce frequencies above the audible range.
Nonetheless, I think it's perfectly possible that you could hear a difference in a full blown production with many instruments and lots of harmonics - audible or not - interacting with each other.
And yes, from a pure guts standpoint, well engineered analog productions do sound better to me than digital productions, even when "reduced" to 16bit 44,1 CD format in the final stage.
nwsoundman
Vintage Gear Junkie
Interesting thread!!
This is something that I am pondering regarding sample rates. My issue with digital is the lack of depth of the sound recorded. Is this and some other issues people have with digital caused by the time between samples in a digital recording being greater than the time of a sound traveling from one ear to the other. What I am trying to say is that we as humans rely greatly on the time differences between what one ear hears and the other to determine where the source is coming from and also some properties of that source. What do you all think? I think tape has a sample rate of infinity.
NWSM
This is something that I am pondering regarding sample rates. My issue with digital is the lack of depth of the sound recorded. Is this and some other issues people have with digital caused by the time between samples in a digital recording being greater than the time of a sound traveling from one ear to the other. What I am trying to say is that we as humans rely greatly on the time differences between what one ear hears and the other to determine where the source is coming from and also some properties of that source. What do you all think? I think tape has a sample rate of infinity.
NWSM
Rossi
New member
I didn't do exact calculations, but the speed of sound is about 1000 km/h or roughly 300 m/sec. Let's say your ears are about 15 cm apart. The sound can travel that distance about 2000 times per second. CD sampling frequency is 44100 Hz (i.e. samples per second). So there's roughly 20 samples while the sound is travelling from one ear to the next. At least if my computations are correct; I'm not exactly a math genious.
I also think there can be a lot of depth to a digital recording provided it's done properly. I friend of mine once had a theory that makes a lot of sense to me. Analog recordings, especially those on vinyl, have noise artifacts, hiss or minor crackle across the stereo image. These artifacts kind of spread the perceived stereo image. It seems wider than it really is because there is always something going on, and if it's only some crackle.
If you want depth for your digital recordings get a convolution reverb and add some high quality ambience for instance. Here's an excellent free one (read the FAQ, it's important because of the latency issue):
http://www.knufinke.de/sir/index_en.html
[edited for bad math]
I also think there can be a lot of depth to a digital recording provided it's done properly. I friend of mine once had a theory that makes a lot of sense to me. Analog recordings, especially those on vinyl, have noise artifacts, hiss or minor crackle across the stereo image. These artifacts kind of spread the perceived stereo image. It seems wider than it really is because there is always something going on, and if it's only some crackle.
If you want depth for your digital recordings get a convolution reverb and add some high quality ambience for instance. Here's an excellent free one (read the FAQ, it's important because of the latency issue):
http://www.knufinke.de/sir/index_en.html
[edited for bad math]
Last edited:
nwsoundman
Vintage Gear Junkie
I should have read my post before I posted it. Too busy. In the end I mean to sugest that our ears are sensitve to sample rate.
There are a couple of good analogies I have heard from R. Kaplan from Indigo Ranch and G. Massenberge. Mr. Kaplan compares sample rate and digital to a Lythograph (sp?). When you look at it from a distance you see solid lines and somewhat of a sharp image. When you look closer you see all the dots. G. Massenberge says it is like listening through a screen door.
As for using a plug to get depth, I really don't feel I should have to apply ambiance electronicly to get depth. I can record to tape and hard drive at the same time from the same source, blah, blah etc and still get way more depth, dimension in analog.
So if there are 2000 samples I guess that is worse than not having one sample between the ears. Maybe this whole idea of mine is crazy but I don't think so.
There are a couple of good analogies I have heard from R. Kaplan from Indigo Ranch and G. Massenberge. Mr. Kaplan compares sample rate and digital to a Lythograph (sp?). When you look at it from a distance you see solid lines and somewhat of a sharp image. When you look closer you see all the dots. G. Massenberge says it is like listening through a screen door.
As for using a plug to get depth, I really don't feel I should have to apply ambiance electronicly to get depth. I can record to tape and hard drive at the same time from the same source, blah, blah etc and still get way more depth, dimension in analog.
So if there are 2000 samples I guess that is worse than not having one sample between the ears. Maybe this whole idea of mine is crazy but I don't think so.
regebro said:No it doesn't. It doesn't have sample rate at all. Neither does the sound that comes out of the speakers and are travelling beteween your ears. Your theories are entertaining, but have nothing with reality to do.
In a manner of speaking, analog does have a sample rate. The bias on the record/playback heads are the carrier frequency and the audio signal modulates the bias signal. It is like FM radio waves. In digital, the AD/DA converters have a very precise voltage reference that is relational to the signal voltage.
The two formats have simliar theory.
regebro said:
What people fail to realize is that the signal that is output from a digital system (D/A) is a finished analog signal. It is the very same kind of signal that comes from an analog system. "Sample Rate" is how many "samples" per second a system can store information.
What is the information?
What people call a "sample" is what we engineers call a voltage level. A signal gets to an A/D converter and the A/D converter takes single voltage levels at various points in time. Each "sample" is formatted into a 16 or 32 bit word (there is other information in this word like volume).
Analog tape stores information. The information is a voltage level also. Analog tape stores 100% of the "samples" of the signal.
Digital stores 44100 "samples" per second (for 44.1 Khz).
When a digital signal goes through the D/A converter, the "samples" are looked at, converted back to the original voltage levels. But, what about the "lost, in between" samples?
There is circuitry that adds voltages between the actual samples to create a complete analog waveform. It is these filters, btw, that usually make some digital systems sound "harsh". The filters are a "best-guesstimate" of the missing voltages.
Example: 1, 3, 5, 7, 9
what is your best guess as to complete this set of numbers?
2, 4, 6, 8 fits the best. The filters would add the missing "samples" to complete the signal.
So, remembering that a "sample" is a voltage level, then a digital system and an analog system are the same idea. It is just that one uses the digital format to store the information and one uses magnetic tape.
Magnetic tape has an *almost* infinite "sample" rate. You can show this to be true using Calculus. Basically, A/D conversion is differential math. D/A conversion is integral math. Since limits can only approach zero, but not ever be zero, infinite does not exist.
regebro
Insane Genious!
No, this is just fancy wording to confuse things up. Analog has nothing that is in any way equivalent to a sample rate, and the bias voltage has nothing to do with it.
Yes, you are right. People tend to forget that what comes out of the speakers is an analog signal. It has no sample rate. Neither has the tape deck. I think you too have forgotten that what we hear is analog.
Trying to force a description of analog into a digital vocabulary will only confuse things.
However, since signals are converted to and from an analog signal even if they are stored digitally inbetween, it is perfectly possible to describe a digital system in analog terms, such as frequency response curve, slope times, phase distortion and all that.
Yes, you are right. People tend to forget that what comes out of the speakers is an analog signal. It has no sample rate. Neither has the tape deck. I think you too have forgotten that what we hear is analog.
Trying to force a description of analog into a digital vocabulary will only confuse things.However, since signals are converted to and from an analog signal even if they are stored digitally inbetween, it is perfectly possible to describe a digital system in analog terms, such as frequency response curve, slope times, phase distortion and all that.
nwsoundman
Vintage Gear Junkie
Ok, I have time now to clarify my original thought. What I meant is that we use the time differences from what both ears hear to determine where a sound is coming from and other similar properties. When a signal is sampled in the conversion process both from analog to digital and vise versa the sample rate when too low fucks with our perception of sound. This is a fact as many engineers feel that a sample rate of 192kHz is closer to analog.
This I totally agree with. It's the lost parts that cause the lack of depth, etc.
This I totally agree with. It's the lost parts that cause the lack of depth, etc.
acorec said:
Unless you think you can really hear frequencies above 22.05Hz, who cares? From what I understand, a well designed reconstruction filter can *perfectly* reconstruct those waveforms as long as you have a sample rate of more than 2x, ala Nyquist.
nwsoundman said:
What the hell does sample rate have to do with this? Are we somehow slowing down the speed of sound or something? I think your logic is a bit flawed here. The time differences from one ear to the other will not change based on sample rate. Maybe I'm misunderstanding what you are talking about....
Man, this debate keeps being rehashed. There are some really good threads on Nika's forum on prosoundweb about digital audio and on rec.audio.pro that helped me clear up some of these myths.....maybe a little reading and little less conjecture might help.