This is probably exactly your problem, but even so, I feel your research is far to subjective without adding quite a bit of statistical research (which has already been done, so don't completely reinvent the wheel). The biggest problem is that any mathematical computations about modelling light are done in radiometric space, where light is considered pure energy of varying frequency (Joules, Watts, and other simple(?) units). In reality, we see on a different, completely NONLINEAR scale, known as the photometric scale (Talbots, lumens, etc. (very complex)). Converting between the two involves the use of what are called Spectral Luminous Efficiency Curves. These are the conversion factors that take into account that the eye has widely varying response to different frequencies of light. Not only are they highly non-linear, the only reason they appear to be smooth is that they are a statistical average of many, many different individuals' efficiency curves.
In other words, if your data isn't lining up, the fundamental problem is that the response of individuals' eyes are rarely comparable. Then again, there's all kinds of psycho-visual effects that can screw up this kind of research without even taking this into account.
Sorry you chose to pick such an incredibly complex subject to research, but good luck in your results!
Black and White is ONE-color mode... We're not talking about palettes here, we're talking about dimensions (or more/less specifically, degrees of freedom) in color-spaces. Two color dimensions are often referred to in printing terms as 'duotone' images, because they only need to mix two colors of ink to print. What the colors are specifically depends on what the primary colors in the image are.
The '4-color' computer displays you are referring to is the classic 'CGA' graphics palette. White, Black, Ugly Cyan, and Uglier Pink were the four colors you had to choose from. Anything dealing with higher colors, and you're probably talking about an 'EGA' palette, which indexes 16 colors to use from a palette of 8red*8blue*8green=512 total colors.
I promise you the physics of such a collapse has been worked out in enormous detail. A rope-like mass is usually thought of as an infinite string of coupled rigid bodies. You're never gonna get any exact analytical expressions for how a rope is going to behave under those circumstances, but you can definitely predict its behaviour.
I would imagine that any part of the rope above a certain altitude would burn up before impact. However, I would not worry much about huge destruction, as the majority of the rope would simply pile itself (relatively) peacefully at it's base. You can calculate how long it would take an object (say, the top of the rope) to fall to the earth (it's actually a tricky problem, as the bottom of the rope may have a slightly larger pull gravity, and will pull the top of the rope faster than it might fall on its own). In that time, you can easily tell how much distance the earth has moved beneath the falling rope, and therefore determine the field of damage.
Unless you expect the rope to take an entire day to fall (which I can't see happening), then there's no way it could wrap itself around the earth. The mass of the entire rope may be significant, but even if it all did hit the ground at the same time, because it's spread out along a huge length it is very unlikely that it will cause any significant damage (i.e. no tsunamis...).
It seems to me that this is what I remember getting a physics degree was all about;-)
What's more important in an experiment is understanding where your errors come from. All of my undergraduate labs (nuclear and optics) were based around using whatever rediculously ancient and decrepit pieces of equipment we had lying around, and learning all the clever little tricks we could ween from our professors about how we could get accurate results from them.
It seems to me like there is no answer to your question unless we know exactly what it is that you're trying to measure. I accurately measured diffraction patterns in optics labs with a cheap CCD video camera and a framegrabber card. Sure, we had to program a filter to convert the frame into raw data for analysis, but I remember that just using our eyes, we were able to determine correct contrast settings.
It also seems to me that if you're working on a senior project, what your professors are more concerned about is not your results, but how well you statistically analyze the nonlinearities that are actually there. Trying to find a more accurate measurement tool usually just means that you're going to have to use more sensetive calibration tools to determine nonlinearities.
Now if you could post what your experiment actually is (although it sounds like you're trying to revamp many experiments), someone here may be able to propose a solution to you that allows you to ignore the nonlinearities in a device.
"It's the industry standard! Standard does not mean better. Like the Imperial mesuring system vs the Metric..."
"Better" does not mean more productive.
I prefer smaller PC applications, such as SoundForge or CoolEdit Pro, but like the good sound engineer I am, I'm perfectly comfortable working with ProTools on a mac as well.
Nothing will ever be 100 percent stable, and nothing will ever work exactly the way you want it to. However, if I can get the job done with a mac and ProTools, I will.
The Industry standard argument was not conjured up by mac users. It's the argument you're going to get directly from the industry itself.
P.S. SoundForge doesn't sample anything. Your audio hardware does that. All the software does is push bits around.
The quality of the A/D and other electronics in your sound card is what manufacturers usually use to label something 'Pro Audio'. Often, manufacturers will release multiple versions of the same product, but with higher quality parts in the 'professional' versions.
Sounds a bit like VCR's doesn't it? Or cars, for that matter...
True, ProTools may be overpriced and not worth it to do many even medium sized audio tasks.
But "pro audio" is not defined by bit depth and sample rates. Pro audio isn't really even defined by low s/n ratios and quality connectors. Pro audio, when it comes down to it, is defined simply by what the 'professionals' choose to use. And when I say professionals, I mean people who are qualified to make decisions as to what equipment better suits them for a recording task. This requires much more knowledge than how to use a piece of software. It requires knowledge of all the other equipment in the studio and it's history, knowledge of the physics of sound and the mathematics of sampling theory, and usually, knowledge of the genre of music you're trying to record. Chances are if you need a big professional solution to a recording problem, you're recording a large orchestra, or a complicated and dense rock track. Otherwise, all it takes is a little imagination to learn to use some really basic audio software to do accomplish smaller tasks. Very few recording applications actually neccessitate the use of a full 36 channel digital mixing console and ProTools software. But then again, you get what you pay for (which is generally an engineer that is very familiar with the ProTools and thus saves you a ton of money by being more productive).
I think too many people confuse 'screwing around' and 'learning'.
It takes time to learn a new system, and it takes even more time to learn new software. It takes 1000 times longer than both of those combined to learn how to be a decent recording engineer, who may use digital software based editing as one of the many, many tools at his disposal.
What we have here is an individual that is looking for a new tool to learn to help him with his digital audio work. I'm not going to question his ability to be an audio engineer, but seeing as how the post was on slashdot, I'd be willing to assume that he has the time to learn new systems and software.
Please don't disrespect all of the prefectly good audio software and hardware that's been developed for PCs. A lot of research time and energy and a lot of great programming went into the PC based software packages.
I know of plenty of well respected studios that are PC based. In fact, a decent studio will have BOTH types of hardware around, to make sure that they can be productive to more of their clients.
It's the same reason that major studios don't buy into a single brand of signal processors, or speakers, or better yet, microphones. If they can afford it (and they can in most cases, simply from the additional client response) they will buy multiple devices that perform the same function from many manufacturers.
~Loren
(why does no one respond to my posts??? I feel so unloved...:-)
Something most/.ers don't realize about the audio engineering industry is that money isn't the primary concern. The primary concern is getting the job done without pissing off your customers (the artists/producers/labels). I'm sure programmers can find that analogy a little easier to chew. Studios (not home project studios) will use a mac because that's where the industry is, simply because macs had a headstart in audio software. Recording schools will instruct you using macs, therefore engineers and producers understand macs.
A big idea in the recording industry is compatability between studios. A producer will choose to use a studio if it has the equipment he knows how to use, or knows will do the job for him. You think there's a pair of Yamaha NS10's in every studio in the world because they're a decent speaker? Far from it. They probably add more color to the sound than anything else in a major recording studio. But the fact is, there's a pair in every major studio, and everyone knows what they sound like, so they've become an industry standard.
The same idea holds true with mac hardware in studios. It's there, they know how to use it. The software is there for it. Most importantly, it works, and if it ain't broke, studios will see very little reason to fix it.
I think what you can expect to see regarding Linux in pro studios is that the recording engineers won't start using Linux solutions until it becomes a 'big' topic in the recording industry. And that won't happen until a 'big' product like ProTools gets ported to Linux.
Cooledit Pro is not a destructive editor when used in multitrack mode. I do quite a bit of non-destructive editing with CoolEdit and other software packages like it.
I'll give you that CoolEdit has the ease of use in their noise reduction algorithm, but professional audio solutions are in an entire class of their own in terms of the quality that they produce. You can bet that there was a plethora of DSP and psychoacoustical research that went into the creation of, say, a Sonic Solutions package. They're in a class all their own, and therefore worth whatever the manufacturer wants to charge for them.
A well trained, experienced engineer familiar with a Sonic Solutions workstation can do things to audio that would blow your mind.
ProTools definitely has a high quality noise reduction algorith, as well. The difference between theirs and CoolEdit is that they don't display a nifty little noise signature and they don't give you complete control over all the FFT parameters (which I still see people fiddle with, even when they've never even heard of an FFT).
Pro Tools systems are sold and supported by Digidesign, complete.
ProTools is just a software package, and is usually a solo system only in professional studios. If you buy DigiDesign's hardware, then you've got but one interface with which to use ProTools, but it supports quite a bit of other audio hardware as well.
For Apple's part, they spent the last five years writing a completely new operating system that doesn't crash,
Does this furrow anyone else's brow? No? Just me? hmmm...
What more do you want from them, to rewrite Pro Tools as well?
Well, it seems so, I believe the question posed was for professional audio software to be available for Linux. I know of many many audio professionals that would absolutely love to have ProTools for Linux. I'd even be willing to pay for it;-)
They did in a way, because Mac OS X includes a complete modern, multichannel audio and MIDI subsystem that supports 32-bit floating point files (infinite headroom, no clipping,
Infinite headroom? No clipping? Again, my eyebrows are having trouble staying in place.
easier to process, they are the new standard).
Oh great... Another standard 'Pro Audio' format...
As if Windows never crashes... sheesh.
This may come as a surprise to some of us, but Linux crashes too... The question isn't only about stability, although that is definitely the key issue. I'll still pick a Linux desktop over Windows 2000 or MacOS X for stability purposes. PC's have a slight advantage due to available hardware drivers. Mac's have a slight advantage due to available, tested, industry approved pro audio software. Linux, I hope, will gain popularity enough to attract the hardware and software industries and perhaps set a new 'standard' of pro audio editing workstations. I for one would love to not have to worry that all my expensive software is installed on a questionably stable foundation.
I'm just glad there's finally a MacOS with memory management...
Like it or not, most music consumers listen to these days IS sampled/stretched/mutilated to start with, and therefore already contains the worst aspects of the digital artifacts you are trying to get rid of by increasing bitrate.
Being "true to the original" has lost its meaning, I'm afraid.
You should be more realistic in your posts. A Screemerstucken VM75^3 168 channel Virtual Matrix Control System with Warp(r) Fader technology will color the sound far less than your AudioStar...
If you want a really rough test to see what you're missing when you encode with a lossy compression. Try this:
Decode your MP3/ogg file back to a wav, input the file into any decent audio editing software along with the original wav file. Reverse the polarity (not the L/R channels, the actual polarity of the audio. Look it up if you're confused), and mix the two tracks together. The result will be a digital representation of what 'audible' material is lost in the compression/decompression. If you normalize that data, you'll be able to hear how little audible data is actually discernable from the resulting noise.
I feel that as more people get used to listening to compressed audio, they will begin to feel like the compressed audio is the norm for the way things should sound. It's going to be an interesting regression if so. Kindof the same way some people feel about vinyl.
Keep in mind, this test in no way tells you anything about the psychoacoustical effects of listening to the differencing of two digital audio files, which could be substantial, for all I know. What you listen to will pretty much be noise and the harmonic distortion created by the codec. I'd imagine many of you will be surprised at the sound, seeing as how people rarely listen to pure harmonic distortion. This will be much more apparent in music that has low noise content to start with (classical, as opposed to grunge).
Still, have fun with it. If you're making a career out of digital compression, please tell me you've got something to fall back on.
And to think, all those engineers and developers went through all that time to make sure that jpeg had a lossless mode from the very beginning. Kinda makes you sad to know that software rarely implements it.
I hardly notice MPEG artifacting when I'm engrossed in a "high-octane thriller" [ouch], but if I'm watching a long dramatic dialogue I will see every digital flaw.
That's great;-) Nothing's worse than watching an MPEG2 movie where a whole bunch of color is thrown into a low contrast scene... Sure puts the encoder to the test, doesn't it?
~Loren
So people who've learned a little bit about sound equipment suddenly (for example) like to think that they're now qualified to make broad statements about sound equipment on/. Theres a name for it..:/ can't think of it at the moment.
I think the term is called "Talking out your ass.";-)
If you want to know what the actual issues are in the recording industry concerning sound quality between speakers/hardware/software/other equipment, then ask a sound engineer or an experienced recording technician. Even a producer usually has enough engineering background to really understand the subtle concepts behind audio reproduction technology.
It's immediately evident to me that there's never going to be a shortage of opinions about sound quality from people that don't have any experience or basis to judge by. Just because you listen to CDs all the time does not give you an ear for whether something sounds 'good' or not. Everyone has a different opinion of what sounds 'good' to them. Recording engineers will have a much more standard set of opinions on what sounds 'good'.
I think that the reason you'll find very few bona fide audio engineers responding to threads like these is that audio engineers know that recordings are rarely mixed for extremely high fidelity playback. The sound quality is tailored from the beginning of the mastering process to the medium that it's most expected to be listened through (car, home, theater, headphones, etc.). So many of you expect higher fidelity from your recordings than is really there in the first place.
I'm a sound engineer first, and a programmer/linux dude second. I decided to respond to this thread because someone finally mentioned the real issues behind all of this, which are purely psychological (not to belittle MP3/ogg themselves, which are mainly psycho-acoustical). I for one am sick of reading posts about how 'crappy' one format is over the other sonically. More useful comment to the end user AS WELL as the recording industry will be along the lines of what formats are more convenient to you, and why.
I'm willing to bet that I could set up many listening tests that would fool many of you into thinking that CD audio is less quality that the resulting MP3/ogg/wma files, simply from the very nature of the compression itself.
In any case, not to sound disenheartened. I've found this topic very enjoyable, and I'm looking forward to seeing what the outcome of the format war will be.
~Loren
(P.S. I enjoy ogg for the sole purpose that I can watch the code develop and see what psychoacoustical research they've implemented. I can only do this partly through MP3)
Slashdot Mirror
This is probably exactly your problem, but even so, I feel your research is far to subjective without adding quite a bit of statistical research (which has already been done, so don't completely reinvent the wheel). The biggest problem is that any mathematical computations about modelling light are done in radiometric space, where light is considered pure energy of varying frequency (Joules, Watts, and other simple(?) units). In reality, we see on a different, completely NONLINEAR scale, known as the photometric scale (Talbots, lumens, etc. (very complex)). Converting between the two involves the use of what are called Spectral Luminous Efficiency Curves. These are the conversion factors that take into account that the eye has widely varying response to different frequencies of light. Not only are they highly non-linear, the only reason they appear to be smooth is that they are a statistical average of many, many different individuals' efficiency curves.
In other words, if your data isn't lining up, the fundamental problem is that the response of individuals' eyes are rarely comparable. Then again, there's all kinds of psycho-visual effects that can screw up this kind of research without even taking this into account.
Sorry you chose to pick such an incredibly complex subject to research, but good luck in your results!
~Loren
Black and White is ONE-color mode... We're not talking about palettes here, we're talking about dimensions (or more/less specifically, degrees of freedom) in color-spaces. Two color dimensions are often referred to in printing terms as 'duotone' images, because they only need to mix two colors of ink to print. What the colors are specifically depends on what the primary colors in the image are.
The '4-color' computer displays you are referring to is the classic 'CGA' graphics palette. White, Black, Ugly Cyan, and Uglier Pink were the four colors you had to choose from. Anything dealing with higher colors, and you're probably talking about an 'EGA' palette, which indexes 16 colors to use from a palette of 8red*8blue*8green=512 total colors.
~Loren
I promise you the physics of such a collapse has been worked out in enormous detail. A rope-like mass is usually thought of as an infinite string of coupled rigid bodies. You're never gonna get any exact analytical expressions for how a rope is going to behave under those circumstances, but you can definitely predict its behaviour.
I would imagine that any part of the rope above a certain altitude would burn up before impact. However, I would not worry much about huge destruction, as the majority of the rope would simply pile itself (relatively) peacefully at it's base. You can calculate how long it would take an object (say, the top of the rope) to fall to the earth (it's actually a tricky problem, as the bottom of the rope may have a slightly larger pull gravity, and will pull the top of the rope faster than it might fall on its own). In that time, you can easily tell how much distance the earth has moved beneath the falling rope, and therefore determine the field of damage.
Unless you expect the rope to take an entire day to fall (which I can't see happening), then there's no way it could wrap itself around the earth. The mass of the entire rope may be significant, but even if it all did hit the ground at the same time, because it's spread out along a huge length it is very unlikely that it will cause any significant damage (i.e. no tsunamis...).
~Loren
It seems to me that this is what I remember getting a physics degree was all about ;-)
What's more important in an experiment is understanding where your errors come from. All of my undergraduate labs (nuclear and optics) were based around using whatever rediculously ancient and decrepit pieces of equipment we had lying around, and learning all the clever little tricks we could ween from our professors about how we could get accurate results from them.
It seems to me like there is no answer to your question unless we know exactly what it is that you're trying to measure. I accurately measured diffraction patterns in optics labs with a cheap CCD video camera and a framegrabber card. Sure, we had to program a filter to convert the frame into raw data for analysis, but I remember that just using our eyes, we were able to determine correct contrast settings.
It also seems to me that if you're working on a senior project, what your professors are more concerned about is not your results, but how well you statistically analyze the nonlinearities that are actually there. Trying to find a more accurate measurement tool usually just means that you're going to have to use more sensetive calibration tools to determine nonlinearities.
Now if you could post what your experiment actually is (although it sounds like you're trying to revamp many experiments), someone here may be able to propose a solution to you that allows you to ignore the nonlinearities in a device.
~Loren
"It's the industry standard! Standard does not mean better. Like the Imperial mesuring system vs the Metric..."
"Better" does not mean more productive.
I prefer smaller PC applications, such as SoundForge or CoolEdit Pro, but like the good sound engineer I am, I'm perfectly comfortable working with ProTools on a mac as well.
Nothing will ever be 100 percent stable, and nothing will ever work exactly the way you want it to. However, if I can get the job done with a mac and ProTools, I will.
The Industry standard argument was not conjured up by mac users. It's the argument you're going to get directly from the industry itself.
Loren
btw, I use a PC... with Linux... at home...
P.S. SoundForge doesn't sample anything. Your audio hardware does that. All the software does is push bits around.
The quality of the A/D and other electronics in your sound card is what manufacturers usually use to label something 'Pro Audio'. Often, manufacturers will release multiple versions of the same product, but with higher quality parts in the 'professional' versions.
Sounds a bit like VCR's doesn't it? Or cars, for that matter...
~Loren
True, ProTools may be overpriced and not worth it to do many even medium sized audio tasks.
But "pro audio" is not defined by bit depth and sample rates. Pro audio isn't really even defined by low s/n ratios and quality connectors. Pro audio, when it comes down to it, is defined simply by what the 'professionals' choose to use. And when I say professionals, I mean people who are qualified to make decisions as to what equipment better suits them for a recording task. This requires much more knowledge than how to use a piece of software. It requires knowledge of all the other equipment in the studio and it's history, knowledge of the physics of sound and the mathematics of sampling theory, and usually, knowledge of the genre of music you're trying to record. Chances are if you need a big professional solution to a recording problem, you're recording a large orchestra, or a complicated and dense rock track. Otherwise, all it takes is a little imagination to learn to use some really basic audio software to do accomplish smaller tasks. Very few recording applications actually neccessitate the use of a full 36 channel digital mixing console and ProTools software. But then again, you get what you pay for (which is generally an engineer that is very familiar with the ProTools and thus saves you a ton of money by being more productive).
~Loren
No, what you really need to write music is a decent instrument and a lot of time practicing and studying orchestration and theory.
Yes, this applies to techno and pop, too.
~Loren
I believe the threshold of human hearing to notice latency is about 15-20ms.
;-)
Offtopic, but the human ear can be trained to detect latencies of as little as 5ms, for many different types of sounds. White noise excluded
~Loren
I think too many people confuse 'screwing around' and 'learning'.
It takes time to learn a new system, and it takes even more time to learn new software. It takes 1000 times longer than both of those combined to learn how to be a decent recording engineer, who may use digital software based editing as one of the many, many tools at his disposal.
What we have here is an individual that is looking for a new tool to learn to help him with his digital audio work. I'm not going to question his ability to be an audio engineer, but seeing as how the post was on slashdot, I'd be willing to assume that he has the time to learn new systems and software.
~Loren
Please don't disrespect all of the prefectly good audio software and hardware that's been developed for PCs. A lot of research time and energy and a lot of great programming went into the PC based software packages.
:-)
I know of plenty of well respected studios that are PC based. In fact, a decent studio will have BOTH types of hardware around, to make sure that they can be productive to more of their clients.
It's the same reason that major studios don't buy into a single brand of signal processors, or speakers, or better yet, microphones. If they can afford it (and they can in most cases, simply from the additional client response) they will buy multiple devices that perform the same function from many manufacturers.
~Loren
(why does no one respond to my posts??? I feel so unloved...
No one but hobbiest use Linux for audio
:-)
And professionals curious about how Linux developments are going to change their field
~Loren
Something most /.ers don't realize about the audio engineering industry is that money isn't the primary concern. The primary concern is getting the job done without pissing off your customers (the artists/producers/labels). I'm sure programmers can find that analogy a little easier to chew. Studios (not home project studios) will use a mac because that's where the industry is, simply because macs had a headstart in audio software. Recording schools will instruct you using macs, therefore engineers and producers understand macs.
A big idea in the recording industry is compatability between studios. A producer will choose to use a studio if it has the equipment he knows how to use, or knows will do the job for him. You think there's a pair of Yamaha NS10's in every studio in the world because they're a decent speaker? Far from it. They probably add more color to the sound than anything else in a major recording studio. But the fact is, there's a pair in every major studio, and everyone knows what they sound like, so they've become an industry standard.
The same idea holds true with mac hardware in studios. It's there, they know how to use it. The software is there for it. Most importantly, it works, and if it ain't broke, studios will see very little reason to fix it.
I think what you can expect to see regarding Linux in pro studios is that the recording engineers won't start using Linux solutions until it becomes a 'big' topic in the recording industry. And that won't happen until a 'big' product like ProTools gets ported to Linux.
Loren
Cooledit Pro is not a destructive editor when used in multitrack mode. I do quite a bit of non-destructive editing with CoolEdit and other software packages like it.
I'll give you that CoolEdit has the ease of use in their noise reduction algorithm, but professional audio solutions are in an entire class of their own in terms of the quality that they produce. You can bet that there was a plethora of DSP and psychoacoustical research that went into the creation of, say, a Sonic Solutions package. They're in a class all their own, and therefore worth whatever the manufacturer wants to charge for them.
A well trained, experienced engineer familiar with a Sonic Solutions workstation can do things to audio that would blow your mind.
ProTools definitely has a high quality noise reduction algorith, as well. The difference between theirs and CoolEdit is that they don't display a nifty little noise signature and they don't give you complete control over all the FFT parameters (which I still see people fiddle with, even when they've never even heard of an FFT).
Pro Tools systems are sold and supported by Digidesign, complete.
;-)
... sheesh.
ProTools is just a software package, and is usually a solo system only in professional studios. If you buy DigiDesign's hardware, then you've got but one interface with which to use ProTools, but it supports quite a bit of other audio hardware as well.
For Apple's part, they spent the last five years writing a completely new operating system that doesn't crash,
Does this furrow anyone else's brow? No? Just me? hmmm...
What more do you want from them, to rewrite Pro Tools as well?
Well, it seems so, I believe the question posed was for professional audio software to be available for Linux. I know of many many audio professionals that would absolutely love to have ProTools for Linux. I'd even be willing to pay for it
They did in a way, because Mac OS X includes a complete modern, multichannel audio and MIDI subsystem that supports 32-bit floating point files (infinite headroom, no clipping,
Infinite headroom? No clipping? Again, my eyebrows are having trouble staying in place.
easier to process, they are the new standard).
Oh great... Another standard 'Pro Audio' format...
As if Windows never crashes
This may come as a surprise to some of us, but Linux crashes too... The question isn't only about stability, although that is definitely the key issue. I'll still pick a Linux desktop over Windows 2000 or MacOS X for stability purposes. PC's have a slight advantage due to available hardware drivers. Mac's have a slight advantage due to available, tested, industry approved pro audio software. Linux, I hope, will gain popularity enough to attract the hardware and software industries and perhaps set a new 'standard' of pro audio editing workstations. I for one would love to not have to worry that all my expensive software is installed on a questionably stable foundation.
I'm just glad there's finally a MacOS with memory management...
~Loren
Thank you :-)
Like it or not, most music consumers listen to these days IS sampled/stretched/mutilated to start with, and therefore already contains the worst aspects of the digital artifacts you are trying to get rid of by increasing bitrate.
Being "true to the original" has lost its meaning, I'm afraid.
~Loren
Oh cmon... Amps don't go to 11, silly...
You should be more realistic in your posts. A Screemerstucken VM75^3 168 channel Virtual Matrix Control System with Warp(r) Fader technology will color the sound far less than your AudioStar...
~Loren
If you want a really rough test to see what you're missing when you encode with a lossy compression. Try this:
Decode your MP3/ogg file back to a wav, input the file into any decent audio editing software along with the original wav file. Reverse the polarity (not the L/R channels, the actual polarity of the audio. Look it up if you're confused), and mix the two tracks together. The result will be a digital representation of what 'audible' material is lost in the compression/decompression. If you normalize that data, you'll be able to hear how little audible data is actually discernable from the resulting noise.
I feel that as more people get used to listening to compressed audio, they will begin to feel like the compressed audio is the norm for the way things should sound. It's going to be an interesting regression if so. Kindof the same way some people feel about vinyl.
Keep in mind, this test in no way tells you anything about the psychoacoustical effects of listening to the differencing of two digital audio files, which could be substantial, for all I know. What you listen to will pretty much be noise and the harmonic distortion created by the codec. I'd imagine many of you will be surprised at the sound, seeing as how people rarely listen to pure harmonic distortion. This will be much more apparent in music that has low noise content to start with (classical, as opposed to grunge).
Still, have fun with it. If you're making a career out of digital compression, please tell me you've got something to fall back on.
~Loren
Hmmmm...
And to think, all those engineers and developers went through all that time to make sure that jpeg had a lossless mode from the very beginning. Kinda makes you sad to know that software rarely implements it.
If only the same could be said about MP3.
~Loren
I hardly notice MPEG artifacting when I'm engrossed in a "high-octane thriller" [ouch], but if I'm watching a long dramatic dialogue I will see every digital flaw.
;-) Nothing's worse than watching an MPEG2 movie where a whole bunch of color is thrown into a low contrast scene... Sure puts the encoder to the test, doesn't it?
~Loren
That's great
So people who've learned a little bit about sound equipment suddenly (for example) like to think that they're now qualified to make broad statements about sound equipment on /. Theres a name for it .. :/ can't think of it at the moment.
I think the term is called "Talking out your ass." ;-)
If you want to know what the actual issues are in the recording industry concerning sound quality between speakers/hardware/software/other equipment, then ask a sound engineer or an experienced recording technician. Even a producer usually has enough engineering background to really understand the subtle concepts behind audio reproduction technology.
It's immediately evident to me that there's never going to be a shortage of opinions about sound quality from people that don't have any experience or basis to judge by. Just because you listen to CDs all the time does not give you an ear for whether something sounds 'good' or not. Everyone has a different opinion of what sounds 'good' to them. Recording engineers will have a much more standard set of opinions on what sounds 'good'.
I think that the reason you'll find very few bona fide audio engineers responding to threads like these is that audio engineers know that recordings are rarely mixed for extremely high fidelity playback. The sound quality is tailored from the beginning of the mastering process to the medium that it's most expected to be listened through (car, home, theater, headphones, etc.). So many of you expect higher fidelity from your recordings than is really there in the first place.
I'm a sound engineer first, and a programmer/linux dude second. I decided to respond to this thread because someone finally mentioned the real issues behind all of this, which are purely psychological (not to belittle MP3/ogg themselves, which are mainly psycho-acoustical). I for one am sick of reading posts about how 'crappy' one format is over the other sonically. More useful comment to the end user AS WELL as the recording industry will be along the lines of what formats are more convenient to you, and why.
I'm willing to bet that I could set up many listening tests that would fool many of you into thinking that CD audio is less quality that the resulting MP3/ogg/wma files, simply from the very nature of the compression itself.
In any case, not to sound disenheartened. I've found this topic very enjoyable, and I'm looking forward to seeing what the outcome of the format war will be.
~Loren
(P.S. I enjoy ogg for the sole purpose that I can watch the code develop and see what psychoacoustical research they've implemented. I can only do this partly through MP3)
mp32ogg... useful little program. Although I find it interesting that there is no ogg2mp3 equivalent developed simultaneously.
Although I'm sure with some clever fifo piping you can accomplish the same task with ogg123 and LAME.
~Loren