I don't know if you can make the statement that 16 bit and 24 bit are equal.
I didn't imply that. You've misinterpreted something. I said that when you're only dealing with levels in the upper part of the range there is no advantage to 24 bit, especially as a delivery to consumer medium, because the extra bits only come into play in the lower realm of the range. A/D or D/A it doesn't matter. If the signal never drops to the range where it makes a difference then there's no advantage to it. More bits, as a format, is all about increased headroom and less grit at the lower end of the meter and nothing else. Once a song has had the transients knocked off (decreasing the dynamic range) and raised to the complete top of the meter, as is done in literally nearly all mastering (even when not overly compressed) the song stays in the realm where 16 bits gets all the info that's needed for playback.
I'm not saying it's a worthless folly for all music, but for literally 95% of what is released it is.
Quote:
Originally Posted by drblank
It all depends on the converters. Some aren't that much different between different bit levels, some are. So, i don't like those kinds of generalities. But if the converter is designed properly, clocked properly, etc. etc. 24 bit will generally sound better, but again, the converters have a LOT to do with it. Converters have been getting MUCH better and it's not just the conversion, it's better clocking, better noise isolation, better power supplies, input and output stages. So, it's hard to really make a generality when it comes to conversion. These engineers know a LOT more know about conversion than they did 20+ years ago. In fact, one of the first designers of one of the first mastering converters said publicly that he changed his attitude towards 16 bit vs 24 bit. He first said 16 bit was all we would need and then years later admitted he was wrong. That person was Bob Stuart of Meridian. They had one of the first digital converters used in Mastering studios. He also said initially that 24/96 was all that was needed, now he changed his attitude towards 24/192 and a lot of this has to do with new ways to improve their product designs that there weren't aware of before. Meridian is also the company behind MLP, which is used in DVD-A, and True HD from Dolby, so he kind of knows a little bit about converters and file compression. He's still up in the air towards DSD vs PCM though. But there is another camp that swears by DSD over PCM recordings. Personally, I haven't compared the two, so I can't comment on that.
This is off the rails now, on to a completely different topic here. This isn't about converter qualities, clocking, input or output stages or anything else. That's all irrelevant to the issue, which is solely, technically "Is 24 bits a valid improvement as a delivery medium for commercial music". You're talking about the value of 24/96-192 bit as a recording format and advances in conversion, which don't apply here.
Don't want to come off as harsh or argumentative because that's not it at all. It's a muddy enough topic when it stays on course : )
To be clear, increasing the sample rate has one and only one effect: It extends the highest frequency the system can record. It does NOT, contrary to some opinions, make any difference whatsoever to lower frequencies. It's not a "resolution enhancer." Increasing the sample rate does not magically make stuff at 5 KHz sound better, or even any different at all.
Quality of sound is measured any number of ways and Nyquist's Theorem doesn't explain everything. Why do different pieces of equipment produce different quality of sound? top recording studios have done listening tests and they've changed converters when they find something that sounds better than another product they've already used. So, I don't know how you can use a THEORY to explain how something sounds in REALITY.
Have Nyquist give me a call and I'll invite him over to do some listening tests of 16 bit and 24 bit masters that we can purchase today to see if he hears a difference in quality of sound. And what A/D and D/A converter products that are available on the market today that Nyquist designed and built? Are his the best sounding out of all converters on the market?
Please forgive my skepticism, but I consider such statements at least suspect and perhaps completely disingenuous. All that story says to me is that a vendor of Mastering products, a very small market, discovered after reaching market saturation that "changing his mind" created a reason for all of his clients to buy new equipment from him.
Bob Stuart is highly regarded by a LOT of engineers in the industry. Meridian stopped making Mastering converters a LONG time ago and they focus on 2 channel stereo and home theater products on the market.
Well, I guess you only work with clients that only work with 16 bit recordings? Sounds like you are very bitter about something. I know studios that don't have lots of money always complain about having to spend more money on upgrading studio equipment and I completely understand that mentality. It's a pain in the butt to have to spend more money to get a small incremental improvement. But sometimes things are better. I was involved with one home studio recording and we went from 16 bit to 24 bit just to see if there was a difference and in the relatively inexpensive converters we were using, it did sound better. It was noticeable to me and the other person involved. But I didn't have a lot of time to spend on that system, but I've heard improvements in studio recordings when comparing tracks recorded at 24 bit vs 16 bit. I think we were comparing 16/44 to 24/96 at the time during the tracking process, but this was a number of years ago. I do recall 24/96 was a little better than 16/44 on that particular system and it was just a Digi OO2R setup with Presonus mic pre. It wasn't all that expensive of a system, just a little project studio in a spare bedroom around 11 years ago.
Improvements in recording technologies primarily benefit styles that feature a wide sonic range. By definition, that is classical music and to a lesser degree jazz.
If you listen to contemporary music (rock, pop, rap, country, whatever), the benefits are nearly non-existent.
Well, it all depends on the recording process. You can get someone playing a pop song with acoustic instruments that's recorded in the same manner as a classical or jazz recording. But commercial projects usually tend to use a lot more signal processing. It's not the style of music necessarily as it is the recording, mixing, and mastering techniques used.
Blue Coast Records records all of their recordings in the same basic manner regarding of style of music. They try to capture a natural sound recording of a live performance without processing everything to death. Their goal is to be able to listen to a live performance where nothing is altered and overdubbed and they record lots of different styles of music so the listener can listen to it as close to being in the room where the musicians were performing. It's just a different recording technique, only most recordings that are on the market aren't done that way in the pop/rock/rap world. Rap music isn't recording acoustic instruments for the most part as they are typically using drum machines or some other form of synthesized instrument or sound effect, so I don't classify it as even music in many ways. Modern commercial music recordings certainly aren't always done the way they USED to do it. It's a shame. They are more pumping out something to sell as quickly as possible to kids that barely know what a REAL piano, etc. sound like. Not much honesty in a lot of recordings these days.
I didn't imply that. You've misinterpreted something. I said that when you're only dealing with levels in the upper part of the range there is no advantage to 24 bit, especially as a delivery to consumer medium, because the extra bits only come into play in the lower realm of the range. A/D or D/A it doesn't matter. If the signal never drops to the range where it makes a difference then there's no advantage to it. More bits, as a format, is all about increased headroom and less grit at the lower end of the meter and nothing else. Once a song has had the transients knocked off (decreasing the dynamic range) and raised to the complete top of the meter, as is done in literally nearly all mastering (even when not overly compressed) the song stays in the realm where 16 bits gets all the info that's needed for playback.
I'm not saying it's a worthless folly for all music, but for literally 95% of what is released it is.
This is off the rails now, on to a completely different topic here. This isn't about converter qualities, clocking, input or output stages or anything else. That's all irrelevant to the issue, which is solely, technically "Is 24 bits a valid improvement as a delivery medium for commercial music". You're talking about the value of 24/96-192 bit as a recording format and advances in conversion, which don't apply here.
Don't want to come off as harsh or argumentative because that's not it at all. It's a muddy enough topic when it stays on course : )
If you are only working with simple wave forms like a sine wave, then it doesn't matter, but with music, we are dealing with complex wave forms.
So by your way of thinking ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms. oh, ok. Next time I have a discussion with an engineer that does mastering for HD Tracks, I'll let them know. I do know of at least one I can email and he responds fairly quickly and he's well respected in the audio recording industry. So i guess I'll have to break the news to him.
I'm glad to see [@]winterspan[/@] posting. I learned a great deal about cellular connectivity from him back around 2007 and 2008.
Ha, thanks for the mention, [@]SolipsismX[/@] .
You are actually one of the few rationally-minded regulars who I remember around here. Due to a change in employment, I have been less involved with Apple technology and so haven't been real active on the forums. But Apple products still hold a special place in my heart, so I'll be around...
Btw, are you the solipsism on sites like Arstechnica/TheVerge/Anandtech?
Btw, are you the solipsism on sites like Arstechnica/TheVerge/Anandtech?
I don't recall the last time I posted on Ars but the other two are definitely me. If I'm not mistaken I think I see you post on AnandTech from time to time.
Why do different pieces of equipment produce different quality of sound?
What does this have to do with a discussion of sample rate? A better converter is a better converter, but for reasons that have nothing to do with increasing the sample rate. Obviously better designs have quieter, cleaner analog stages in and out and smoother anti-aliasing filters. Some converters sound better than others, but not because the sample rate fairy raised the upper limit.
It's really hard to have a meaningful discussion with you when you keep changing the focus. No one has said all gear sounds the same or that there are no conditions under which one device is better than another. This all began with your comment that 24 bit sounds better than 16 bit, with no basis for that conclusion other than comparisons of dissimilar recordings. Every time someone gives you a glimpse into how digital recording really works you go off on some tangent about which famous engineer ate a bagel in which famous studio.
Digital recording is just math, pure and simple. Saying that sampling theory doesn't explain it all is ridiculous because of course it does. There's no magic, no pixie dust, but lots and lots of marketing bullshit. If you wanna believe the crap manufacturers feed you to keep you buying new gear, be my guest. The rest of us will benefit from more critical analysis. By dispelling the misinformation and concentrating on the areas that REALLY result in better sound we can make sure the dollars we spend on upgrades yield actual results, not just meaningless numbers.
What does this have to do with a discussion of sample rate? A better converter is a better converter, but for reasons that have nothing to do with increasing the sample rate. Obviously better designs have quieter, cleaner analog stages in and out and smoother anti-aliasing filters. Some converters sound better than others, but not because the sample rate fairy raised the upper limit.
It's really hard to have a meaningful discussion with you when you keep changing the focus. No one has said all gear sounds the same or that there are no conditions under which one device is better than another. This all began with your comment that 24 bit sounds better than 16 bit, with no basis for that conclusion other than comparisons of dissimilar recordings. Every time someone gives you a glimpse into how digital recording really works you go off on some tangent about which famous engineer ate a bagel in which famous studio.
Digital recording is just math, pure and simple. Saying that sampling theory doesn't explain it all is ridiculous because of course it does. There's no magic, no pixie dust, but lots and lots of marketing bullshit. If you wanna believe the crap manufacturers feed you to keep you buying new gear, be my guest. The rest of us will benefit from more critical analysis. By dispelling the misinformation and concentrating on the areas that REALLY result in better sound we can make sure the dollars we spend on upgrades yield actual results, not just meaningless numbers.
So why are there so many different converter chip designs? There should only be the need for one and it should be perfect. Right? But why don't you just use the absolute cheapest A/D and D/A converters in all recording and mastering studios? I would think that any recording studio owner would LOVE to prove that there is absolutely no difference in audio quality between 16 bit and 24 bit on the cheapest converters to the most expensive converters. How come when Abby Road or Ocean Way, or any other reputable recording studio replaces converters every so often after performing countless measurement and listening tests? Why is that? I thought that it's just math that explains it and therefor there is no difference, yet there are reputable recording studio engineers that will claim otherwise.
Have you any reputable recording engineer that will claim there is absolutely no sound difference between 16 bit and 24 bit? What does Bob Ludwig, George Massenburg, Ken Scott, other any other legendary engineer say about this? I'm sure they've all have LOTS of experience with a variety of converters over the years in both tracking and mastering production in what we would call pristine listening conditions with people that do have a lot more experience in listening to subtleties from a recording. And why do studios that have unlimited means don't all use the same brand/model for mastering? I do see some trends in what top recording studio sometimes do when a new product is released that get high praises, but they aren't always used by every top studio all of the time. Some engineers have their favorites.
One thing I was told a LONG time ago about theories. They usually only exist in a PERFECT world with PERFECT conditions. However we all know that A/D and D/A converters aren't perfect. Some are not even CLOSE to being perfect. That's why I don't just rely on theories to explain something, especially when I can hear a difference and you say the theory says there isn't. Well, so much for theories.
So in the mean time, I will buy whatever version of content I will buy despite what YOU or Nyquist think is true, when I can plainly hear differently. Thankfully we can listen to samples of tracks before we buy them to see for ourselves if we hear a difference, and we have other people to rely on to obtain a loaner of something to see for ourselves. And 24 bit WILL become more prevalent in the consumer marketplace despite what you say or think or what Nyquist's theory says, means, or left out that explains why people can and do hear differences between 16 bit and 24 bit recordings. Whatever reason WHY they sound different is of NO meaning to me, because I'll probably never really know what all of those differences are since I wasn't in the studio at the time the masters were created. They don't normally give the consumer full knowledge of every little detail of their mastering process.
Have a nice day and enjoy your theories. I'll just sit back and enjoy the recordings, like I've been doing, knowing there is a VERY noticeable difference in the recordings I have.
The iTunes Store would then be the place to buy classical music.
Well, you can certainly buy 24 bit recordings on a variety of sites currently. You can get a great deal at Bower's & Wilkins site. It costs about $100 a year and they post new recordings every month. They already have a bunch of London Symphony Orchestra recordings posted in 24 bit, 16 bit. HD Tracks has them as well as a few other sites.
The biggest problem EVERYONE faces is how quickly these record labels can spit out remasters of older analog recordings and the release of new recordings at 24 bit. So far, there isn't much content, and so far they are far more expensive.
I'm hoping the recording industry gets off their high horse and realizes that if they want 24 bit to become the defacto standard for new recordings and remasters of analog recordings, that they need to simply charge the same amount as a 16 bit version and then people will buy whatever they want to store since some people have different storage abilities. 24 bit files are much larger than 16 bit and some people simply don't want to use up their storage if they really aren't that concerned about sound quality.
That's why SACD and DVD-A didn't really take off. Lack of content and high retail costs prevented mainstream acceptance and I'm hoping Apple can get the current premium pricing model changed to something more affordable for ALL, not just the financially privileged.
What does this have to do with a discussion of sample rate? A better converter is a better converter, but for reasons that have nothing to do with increasing the sample rate. Obviously better designs have quieter, cleaner analog stages in and out and smoother anti-aliasing filters. Some converters sound better than others, but not because the sample rate fairy raised the upper limit.
It's really hard to have a meaningful discussion with you when you keep changing the focus. No one has said all gear sounds the same or that there are no conditions under which one device is better than another. This all began with your comment that 24 bit sounds better than 16 bit, with no basis for that conclusion other than comparisons of dissimilar recordings. Every time someone gives you a glimpse into how digital recording really works you go off on some tangent about which famous engineer ate a bagel in which famous studio.
Digital recording is just math, pure and simple. Saying that sampling theory doesn't explain it all is ridiculous because of course it does. There's no magic, no pixie dust, but lots and lots of marketing bullshit. If you wanna believe the crap manufacturers feed you to keep you buying new gear, be my guest. The rest of us will benefit from more critical analysis. By dispelling the misinformation and concentrating on the areas that REALLY result in better sound we can make sure the dollars we spend on upgrades yield actual results, not just meaningless numbers.
Here's a little proof that theories aren't always proven to be correct. What's the speed of sound? Well, most people that know about sound will tell you it's 1130 Feet Per Second or they'll tell you 340.29 meters per second. Either would probably be correct on an exam. Or at least they are commonly used numbers. Some will even shorten it to 340 meters per second because it's close enough. But the real answer is that it will change due to temperature and even altitude conditions. So, what is a theoretical speed isn't really REALITY ALL OF THE TIME. But they didn't know about these other conditions when they derived the original THEORY. Did they?
Just because someone wrote down a bunch of equations, they kind of have to prove it and it has to be repeatable consistently time and time again. So why can't all A/D and D/A converters measure and PROVE the theory repeatable no matter which chips are used, no matter what brand, model and design on the market? Why is that? Because Nyquist's theory is simply just a THEORY. It's not reality. Nyquist didn't know everything about A/D and D/A conversion because he didn't have a variety of different equipment to test using modern day testing methods and equipment which have improved dramatically. Heck, there is no such thing as a PERFECT resistor, capacitor, cable, etc. where it doesn't fluctuate in it's measurements. With the most precise test measurement equipment, they can only get the tolerances so good. So a PERFECT A/D and D/A converter PROBABLY will NEVER exist. But they are getting pretty close if you have an extra $30K to $100K for a 2 channel DAC. Yeah, like the masses can afford them.
If you are only working with simple wave forms like a sine wave, then it doesn't matter, but with music, we are dealing with complex wave forms.
So by your way of thinking ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms. oh, ok. Next time I have a discussion with an engineer that does mastering for HD Tracks, I'll let them know. I do know of at least one I can email and he responds fairly quickly and he's well respected in the audio recording industry. So i guess I'll have to break the news to him.
Respectfully, drblank, again you are missing the jist of the issue, and in the process not getting the information right. I hate adding this in, because it doesn't add much, but I've been making my living as a recording engineer for 30 years and have gone through enough converters and the rest to fill a closet, and I'm well versed in the subject. It is NOT my contention that "ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms". It IS my contention that one gains nothing with file playback at 24 vs 16 if the level never goes below the point at which the 8 bits are used. You are hung up on the differences between different converters, which is a whole different discussion. We are not talking about the rest of the filters and signal path involved in getting audio into the digital realm, just what happens once you have the audio and will put in onto a delivery medium.
Again, once the file is done, finished, mastered, 24/192, whatever, doesn't matter. If the dynamic range is lessened and raised to the limit (in mastering with, say a 64 bit engine) does the output take advantage of the full 24 bits, which are located lower than its floor. I'd add "in any way that matters to the consumer" but it's not necessary. That's it. Please don't bring up how good a $10,000 converter sounds at 24 bit next to a $200 one at 16. ; )
Best,
j
The value of inputting at 24 bit is also not relevant to this topic. No one records at 16 bit anymore (except in certain circumstances). I certainly haven't since it was available. Different discussion.
Respectfully, drblank, again you are missing the jist of the issue, and in the process not getting the information right. I hate adding this in, because it doesn't add much, but I've been making my living as a recording engineer for 30 years and have gone through enough converters and the rest to fill a closet, and I'm well versed in the subject. It is NOT my contention that "ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms". It IS my contention that one gains nothing with file playback at 24 vs 16 if the level never goes below the point at which the 8 bits are used. You are hung up on the differences between different converters, which is a whole different discussion. We are not talking about the rest of the filters and signal path involved in getting audio into the digital realm, just what happens once you have the audio and will put in onto a delivery medium.
Again, once the file is done, finished, mastered, 24/192, whatever, doesn't matter. If the dynamic range is lessened and raised to the limit (in mastering with, say a 64 bit engine) does the output take advantage of the full 24 bits, which are located lower than its floor. I'd add "in any way that matters to the consumer" but it's not necessary. That's it. Please don't bring up how good a $10,000 converter sounds at 24 bit next to a $200 one at 16. ; )
Best,
j
The value of inputting at 24 bit is also not relevant to this topic. No one records at 16 bit anymore (except in certain circumstances). I certainly haven't since it was available. Different discussion.
The topic is that Apple going to add 24 bit files to iTunes. I don't think they are going to add 24 bit files that sound identical to the current 16 bit files they currently have. I think the 24 bit files will sound better than their 16 bit counterparts more often than not. Obviously time will tell, but that's what WE the consumers care about. And what equipment we use is OUR decision. Personally, I don't have the money to spend (or waste) $10K, $30K or $100K on a D/A converter. I leave that decision to people that do. :-)
I just have a growing number of 24 bit recordings that I've compared against 16 bit versions and in every case the 24 bit sounds better. noticeably better. What and how they did this, I could give rip. What you claim has nothing to do with anything unless you are involved with creating both the 16 bit and 24 bit versions and I can listen to both to compare the two on my system. That's all I can go by.
So, what it sounds like is that you will never make any 24 bit recordings, since you claim there isn't a difference. I wonder how that will impact your ability to get future business if the world is moving towards 24 bit and you won't do 24 bit. Sounds like a career limiting move. :-)
I tell you what. Here's a test for you to do. Go conduct your listening tests in your studio using a pair of TAD Reference monitors and some decent amps and cables and then tell me afterwards what you hear when comparing the tests you speak of.
Oh, and back when they were doing hearing tests, I don't even think they had headphones that could accurately reproduce more than 20KHz. It wasn't until the 60's until they broke the barrier of speakers going past 10KHz. They NOW have headphones that go past 30KHz, and speakers that can reproduce 100KHz. Maybe you need those to hear the difference. :-) I'm kind of being a little sarcastic, but there are top engineers that would give their right Testicle to own a pair of TAD Reference monitors. If you can't afford them, then get some Pioneer S-1EX or S-2EX. I know someone that has the smaller units available for about $3800 a pair with stands if he still wants to sell them and you want to buy them. They are unbelievably accurate lower priced versions of the TAD. Same technology designed by the same person. They just don't go down as low in the bottom frequencies (34Hz instead of 32Hz for the smaller monitors), but they do go to 100KHz. :-)
I don't recall the last time I posted on Ars but the other two are definitely me. If I'm not mistaken I think I see you post on AnandTech from time to time.
Yeah, I tend to comment around... I also use "Loosely Coupled" on a lot of sites, so if you ever see that, its me
I'm looking forward to experiencing music from the Pono music player. It will be better quality than this alleged Mastered for iTunes product. Pono will use FLAC files and be capable of using other industry standard files of lesser quality.
I doubt any process will be as good as original vinyl recordings on a good system but Pono will certainly be the top of the line standard for a while to come. They will debut in the summer of 2014.
You're right. Digital is better in every conceivable way. Except that some people who don't understand it convince themselves that it must be worse than vinyl and go on to perform poorly or not-at-all controlled comparisons which - surprise, surprise - reinforce their original viewpoint.
This.
This can quite often be the cause of vinyl sounding better - whilst it is worse in every way as a medium compared to CD (dynamic range, frequency response, wow & flutter, distortion) vinyl masters do not usually have their dynamic range compressed into oblivion.
No. Analogue is better than digital. But iTunes is fine for me.
Comments
I don't know if you can make the statement that 16 bit and 24 bit are equal.
I didn't imply that. You've misinterpreted something. I said that when you're only dealing with levels in the upper part of the range there is no advantage to 24 bit, especially as a delivery to consumer medium, because the extra bits only come into play in the lower realm of the range. A/D or D/A it doesn't matter. If the signal never drops to the range where it makes a difference then there's no advantage to it. More bits, as a format, is all about increased headroom and less grit at the lower end of the meter and nothing else. Once a song has had the transients knocked off (decreasing the dynamic range) and raised to the complete top of the meter, as is done in literally nearly all mastering (even when not overly compressed) the song stays in the realm where 16 bits gets all the info that's needed for playback.
I'm not saying it's a worthless folly for all music, but for literally 95% of what is released it is.
Originally Posted by drblank
It all depends on the converters. Some aren't that much different between different bit levels, some are. So, i don't like those kinds of generalities. But if the converter is designed properly, clocked properly, etc. etc. 24 bit will generally sound better, but again, the converters have a LOT to do with it. Converters have been getting MUCH better and it's not just the conversion, it's better clocking, better noise isolation, better power supplies, input and output stages. So, it's hard to really make a generality when it comes to conversion. These engineers know a LOT more know about conversion than they did 20+ years ago. In fact, one of the first designers of one of the first mastering converters said publicly that he changed his attitude towards 16 bit vs 24 bit. He first said 16 bit was all we would need and then years later admitted he was wrong. That person was Bob Stuart of Meridian. They had one of the first digital converters used in Mastering studios. He also said initially that 24/96 was all that was needed, now he changed his attitude towards 24/192 and a lot of this has to do with new ways to improve their product designs that there weren't aware of before. Meridian is also the company behind MLP, which is used in DVD-A, and True HD from Dolby, so he kind of knows a little bit about converters and file compression. He's still up in the air towards DSD vs PCM though. But there is another camp that swears by DSD over PCM recordings. Personally, I haven't compared the two, so I can't comment on that.
This is off the rails now, on to a completely different topic here. This isn't about converter qualities, clocking, input or output stages or anything else. That's all irrelevant to the issue, which is solely, technically "Is 24 bits a valid improvement as a delivery medium for commercial music". You're talking about the value of 24/96-192 bit as a recording format and advances in conversion, which don't apply here.
Don't want to come off as harsh or argumentative because that's not it at all. It's a muddy enough topic when it stays on course : )
To be clear, increasing the sample rate has one and only one effect: It extends the highest frequency the system can record. It does NOT, contrary to some opinions, make any difference whatsoever to lower frequencies. It's not a "resolution enhancer." Increasing the sample rate does not magically make stuff at 5 KHz sound better, or even any different at all.
Quality of sound is measured any number of ways and Nyquist's Theorem doesn't explain everything. Why do different pieces of equipment produce different quality of sound? top recording studios have done listening tests and they've changed converters when they find something that sounds better than another product they've already used. So, I don't know how you can use a THEORY to explain how something sounds in REALITY.
Have Nyquist give me a call and I'll invite him over to do some listening tests of 16 bit and 24 bit masters that we can purchase today to see if he hears a difference in quality of sound. And what A/D and D/A converter products that are available on the market today that Nyquist designed and built? Are his the best sounding out of all converters on the market?
Please forgive my skepticism, but I consider such statements at least suspect and perhaps completely disingenuous. All that story says to me is that a vendor of Mastering products, a very small market, discovered after reaching market saturation that "changing his mind" created a reason for all of his clients to buy new equipment from him.
Bob Stuart is highly regarded by a LOT of engineers in the industry. Meridian stopped making Mastering converters a LONG time ago and they focus on 2 channel stereo and home theater products on the market.
Well, I guess you only work with clients that only work with 16 bit recordings? Sounds like you are very bitter about something. I know studios that don't have lots of money always complain about having to spend more money on upgrading studio equipment and I completely understand that mentality. It's a pain in the butt to have to spend more money to get a small incremental improvement. But sometimes things are better. I was involved with one home studio recording and we went from 16 bit to 24 bit just to see if there was a difference and in the relatively inexpensive converters we were using, it did sound better. It was noticeable to me and the other person involved. But I didn't have a lot of time to spend on that system, but I've heard improvements in studio recordings when comparing tracks recorded at 24 bit vs 16 bit. I think we were comparing 16/44 to 24/96 at the time during the tracking process, but this was a number of years ago. I do recall 24/96 was a little better than 16/44 on that particular system and it was just a Digi OO2R setup with Presonus mic pre. It wasn't all that expensive of a system, just a little project studio in a spare bedroom around 11 years ago.
Improvements in recording technologies primarily benefit styles that feature a wide sonic range. By definition, that is classical music and to a lesser degree jazz.
If you listen to contemporary music (rock, pop, rap, country, whatever), the benefits are nearly non-existent.
Well, it all depends on the recording process. You can get someone playing a pop song with acoustic instruments that's recorded in the same manner as a classical or jazz recording. But commercial projects usually tend to use a lot more signal processing. It's not the style of music necessarily as it is the recording, mixing, and mastering techniques used.
Blue Coast Records records all of their recordings in the same basic manner regarding of style of music. They try to capture a natural sound recording of a live performance without processing everything to death. Their goal is to be able to listen to a live performance where nothing is altered and overdubbed and they record lots of different styles of music so the listener can listen to it as close to being in the room where the musicians were performing. It's just a different recording technique, only most recordings that are on the market aren't done that way in the pop/rock/rap world. Rap music isn't recording acoustic instruments for the most part as they are typically using drum machines or some other form of synthesized instrument or sound effect, so I don't classify it as even music in many ways. Modern commercial music recordings certainly aren't always done the way they USED to do it. It's a shame. They are more pumping out something to sell as quickly as possible to kids that barely know what a REAL piano, etc. sound like. Not much honesty in a lot of recordings these days.
I didn't imply that. You've misinterpreted something. I said that when you're only dealing with levels in the upper part of the range there is no advantage to 24 bit, especially as a delivery to consumer medium, because the extra bits only come into play in the lower realm of the range. A/D or D/A it doesn't matter. If the signal never drops to the range where it makes a difference then there's no advantage to it. More bits, as a format, is all about increased headroom and less grit at the lower end of the meter and nothing else. Once a song has had the transients knocked off (decreasing the dynamic range) and raised to the complete top of the meter, as is done in literally nearly all mastering (even when not overly compressed) the song stays in the realm where 16 bits gets all the info that's needed for playback.
I'm not saying it's a worthless folly for all music, but for literally 95% of what is released it is.
This is off the rails now, on to a completely different topic here. This isn't about converter qualities, clocking, input or output stages or anything else. That's all irrelevant to the issue, which is solely, technically "Is 24 bits a valid improvement as a delivery medium for commercial music". You're talking about the value of 24/96-192 bit as a recording format and advances in conversion, which don't apply here.
Don't want to come off as harsh or argumentative because that's not it at all. It's a muddy enough topic when it stays on course : )
If you are only working with simple wave forms like a sine wave, then it doesn't matter, but with music, we are dealing with complex wave forms.
So by your way of thinking ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms. oh, ok. Next time I have a discussion with an engineer that does mastering for HD Tracks, I'll let them know. I do know of at least one I can email and he responds fairly quickly and he's well respected in the audio recording industry. So i guess I'll have to break the news to him.
Ha, thanks for the mention, [@]SolipsismX[/@] .
You are actually one of the few rationally-minded regulars who I remember around here. Due to a change in employment, I have been less involved with Apple technology and so haven't been real active on the forums. But Apple products still hold a special place in my heart, so I'll be around...
Btw, are you the solipsism on sites like Arstechnica/TheVerge/Anandtech?
I don't recall the last time I posted on Ars but the other two are definitely me. If I'm not mistaken I think I see you post on AnandTech from time to time.
Why do different pieces of equipment produce different quality of sound?
What does this have to do with a discussion of sample rate? A better converter is a better converter, but for reasons that have nothing to do with increasing the sample rate. Obviously better designs have quieter, cleaner analog stages in and out and smoother anti-aliasing filters. Some converters sound better than others, but not because the sample rate fairy raised the upper limit.
It's really hard to have a meaningful discussion with you when you keep changing the focus. No one has said all gear sounds the same or that there are no conditions under which one device is better than another. This all began with your comment that 24 bit sounds better than 16 bit, with no basis for that conclusion other than comparisons of dissimilar recordings. Every time someone gives you a glimpse into how digital recording really works you go off on some tangent about which famous engineer ate a bagel in which famous studio.
Digital recording is just math, pure and simple. Saying that sampling theory doesn't explain it all is ridiculous because of course it does. There's no magic, no pixie dust, but lots and lots of marketing bullshit. If you wanna believe the crap manufacturers feed you to keep you buying new gear, be my guest. The rest of us will benefit from more critical analysis. By dispelling the misinformation and concentrating on the areas that REALLY result in better sound we can make sure the dollars we spend on upgrades yield actual results, not just meaningless numbers.
What does this have to do with a discussion of sample rate? A better converter is a better converter, but for reasons that have nothing to do with increasing the sample rate. Obviously better designs have quieter, cleaner analog stages in and out and smoother anti-aliasing filters. Some converters sound better than others, but not because the sample rate fairy raised the upper limit.
It's really hard to have a meaningful discussion with you when you keep changing the focus. No one has said all gear sounds the same or that there are no conditions under which one device is better than another. This all began with your comment that 24 bit sounds better than 16 bit, with no basis for that conclusion other than comparisons of dissimilar recordings. Every time someone gives you a glimpse into how digital recording really works you go off on some tangent about which famous engineer ate a bagel in which famous studio.
Digital recording is just math, pure and simple. Saying that sampling theory doesn't explain it all is ridiculous because of course it does. There's no magic, no pixie dust, but lots and lots of marketing bullshit. If you wanna believe the crap manufacturers feed you to keep you buying new gear, be my guest. The rest of us will benefit from more critical analysis. By dispelling the misinformation and concentrating on the areas that REALLY result in better sound we can make sure the dollars we spend on upgrades yield actual results, not just meaningless numbers.
So why are there so many different converter chip designs? There should only be the need for one and it should be perfect. Right? But why don't you just use the absolute cheapest A/D and D/A converters in all recording and mastering studios? I would think that any recording studio owner would LOVE to prove that there is absolutely no difference in audio quality between 16 bit and 24 bit on the cheapest converters to the most expensive converters. How come when Abby Road or Ocean Way, or any other reputable recording studio replaces converters every so often after performing countless measurement and listening tests? Why is that? I thought that it's just math that explains it and therefor there is no difference, yet there are reputable recording studio engineers that will claim otherwise.
Have you any reputable recording engineer that will claim there is absolutely no sound difference between 16 bit and 24 bit? What does Bob Ludwig, George Massenburg, Ken Scott, other any other legendary engineer say about this? I'm sure they've all have LOTS of experience with a variety of converters over the years in both tracking and mastering production in what we would call pristine listening conditions with people that do have a lot more experience in listening to subtleties from a recording. And why do studios that have unlimited means don't all use the same brand/model for mastering? I do see some trends in what top recording studio sometimes do when a new product is released that get high praises, but they aren't always used by every top studio all of the time. Some engineers have their favorites.
One thing I was told a LONG time ago about theories. They usually only exist in a PERFECT world with PERFECT conditions. However we all know that A/D and D/A converters aren't perfect. Some are not even CLOSE to being perfect. That's why I don't just rely on theories to explain something, especially when I can hear a difference and you say the theory says there isn't. Well, so much for theories.
So in the mean time, I will buy whatever version of content I will buy despite what YOU or Nyquist think is true, when I can plainly hear differently. Thankfully we can listen to samples of tracks before we buy them to see for ourselves if we hear a difference, and we have other people to rely on to obtain a loaner of something to see for ourselves. And 24 bit WILL become more prevalent in the consumer marketplace despite what you say or think or what Nyquist's theory says, means, or left out that explains why people can and do hear differences between 16 bit and 24 bit recordings. Whatever reason WHY they sound different is of NO meaning to me, because I'll probably never really know what all of those differences are since I wasn't in the studio at the time the masters were created. They don't normally give the consumer full knowledge of every little detail of their mastering process.
Have a nice day and enjoy your theories. I'll just sit back and enjoy the recordings, like I've been doing, knowing there is a VERY noticeable difference in the recordings I have.
I sure hope this rumor is true.
The iTunes Store would then be the place to buy classical music.
Well, you can certainly buy 24 bit recordings on a variety of sites currently. You can get a great deal at Bower's & Wilkins site. It costs about $100 a year and they post new recordings every month. They already have a bunch of London Symphony Orchestra recordings posted in 24 bit, 16 bit. HD Tracks has them as well as a few other sites.
The biggest problem EVERYONE faces is how quickly these record labels can spit out remasters of older analog recordings and the release of new recordings at 24 bit. So far, there isn't much content, and so far they are far more expensive.
I'm hoping the recording industry gets off their high horse and realizes that if they want 24 bit to become the defacto standard for new recordings and remasters of analog recordings, that they need to simply charge the same amount as a 16 bit version and then people will buy whatever they want to store since some people have different storage abilities. 24 bit files are much larger than 16 bit and some people simply don't want to use up their storage if they really aren't that concerned about sound quality.
That's why SACD and DVD-A didn't really take off. Lack of content and high retail costs prevented mainstream acceptance and I'm hoping Apple can get the current premium pricing model changed to something more affordable for ALL, not just the financially privileged.
What does this have to do with a discussion of sample rate? A better converter is a better converter, but for reasons that have nothing to do with increasing the sample rate. Obviously better designs have quieter, cleaner analog stages in and out and smoother anti-aliasing filters. Some converters sound better than others, but not because the sample rate fairy raised the upper limit.
It's really hard to have a meaningful discussion with you when you keep changing the focus. No one has said all gear sounds the same or that there are no conditions under which one device is better than another. This all began with your comment that 24 bit sounds better than 16 bit, with no basis for that conclusion other than comparisons of dissimilar recordings. Every time someone gives you a glimpse into how digital recording really works you go off on some tangent about which famous engineer ate a bagel in which famous studio.
Digital recording is just math, pure and simple. Saying that sampling theory doesn't explain it all is ridiculous because of course it does. There's no magic, no pixie dust, but lots and lots of marketing bullshit. If you wanna believe the crap manufacturers feed you to keep you buying new gear, be my guest. The rest of us will benefit from more critical analysis. By dispelling the misinformation and concentrating on the areas that REALLY result in better sound we can make sure the dollars we spend on upgrades yield actual results, not just meaningless numbers.
Here's a little proof that theories aren't always proven to be correct. What's the speed of sound? Well, most people that know about sound will tell you it's 1130 Feet Per Second or they'll tell you 340.29 meters per second. Either would probably be correct on an exam. Or at least they are commonly used numbers. Some will even shorten it to 340 meters per second because it's close enough. But the real answer is that it will change due to temperature and even altitude conditions. So, what is a theoretical speed isn't really REALITY ALL OF THE TIME. But they didn't know about these other conditions when they derived the original THEORY. Did they?
Just because someone wrote down a bunch of equations, they kind of have to prove it and it has to be repeatable consistently time and time again. So why can't all A/D and D/A converters measure and PROVE the theory repeatable no matter which chips are used, no matter what brand, model and design on the market? Why is that? Because Nyquist's theory is simply just a THEORY. It's not reality. Nyquist didn't know everything about A/D and D/A conversion because he didn't have a variety of different equipment to test using modern day testing methods and equipment which have improved dramatically. Heck, there is no such thing as a PERFECT resistor, capacitor, cable, etc. where it doesn't fluctuate in it's measurements. With the most precise test measurement equipment, they can only get the tolerances so good. So a PERFECT A/D and D/A converter PROBABLY will NEVER exist. But they are getting pretty close if you have an extra $30K to $100K for a 2 channel DAC. Yeah, like the masses can afford them.
If you are only working with simple wave forms like a sine wave, then it doesn't matter, but with music, we are dealing with complex wave forms.
So by your way of thinking ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms. oh, ok. Next time I have a discussion with an engineer that does mastering for HD Tracks, I'll let them know. I do know of at least one I can email and he responds fairly quickly and he's well respected in the audio recording industry. So i guess I'll have to break the news to him.
Respectfully, drblank, again you are missing the jist of the issue, and in the process not getting the information right. I hate adding this in, because it doesn't add much, but I've been making my living as a recording engineer for 30 years and have gone through enough converters and the rest to fill a closet, and I'm well versed in the subject. It is NOT my contention that "ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms". It IS my contention that one gains nothing with file playback at 24 vs 16 if the level never goes below the point at which the 8 bits are used. You are hung up on the differences between different converters, which is a whole different discussion. We are not talking about the rest of the filters and signal path involved in getting audio into the digital realm, just what happens once you have the audio and will put in onto a delivery medium.
Again, once the file is done, finished, mastered, 24/192, whatever, doesn't matter. If the dynamic range is lessened and raised to the limit (in mastering with, say a 64 bit engine) does the output take advantage of the full 24 bits, which are located lower than its floor. I'd add "in any way that matters to the consumer" but it's not necessary. That's it. Please don't bring up how good a $10,000 converter sounds at 24 bit next to a $200 one at 16. ; )
Best,
j
The value of inputting at 24 bit is also not relevant to this topic. No one records at 16 bit anymore (except in certain circumstances). I certainly haven't since it was available. Different discussion.
Respectfully, drblank, again you are missing the jist of the issue, and in the process not getting the information right. I hate adding this in, because it doesn't add much, but I've been making my living as a recording engineer for 30 years and have gone through enough converters and the rest to fill a closet, and I'm well versed in the subject. It is NOT my contention that "ALL 16 bit converters A/D and D/A and all 24 bit converters regardless of sample rate all sound equal with complex wave forms". It IS my contention that one gains nothing with file playback at 24 vs 16 if the level never goes below the point at which the 8 bits are used. You are hung up on the differences between different converters, which is a whole different discussion. We are not talking about the rest of the filters and signal path involved in getting audio into the digital realm, just what happens once you have the audio and will put in onto a delivery medium.
Again, once the file is done, finished, mastered, 24/192, whatever, doesn't matter. If the dynamic range is lessened and raised to the limit (in mastering with, say a 64 bit engine) does the output take advantage of the full 24 bits, which are located lower than its floor. I'd add "in any way that matters to the consumer" but it's not necessary. That's it. Please don't bring up how good a $10,000 converter sounds at 24 bit next to a $200 one at 16. ; )
Best,
j
The value of inputting at 24 bit is also not relevant to this topic. No one records at 16 bit anymore (except in certain circumstances). I certainly haven't since it was available. Different discussion.
The topic is that Apple going to add 24 bit files to iTunes. I don't think they are going to add 24 bit files that sound identical to the current 16 bit files they currently have. I think the 24 bit files will sound better than their 16 bit counterparts more often than not. Obviously time will tell, but that's what WE the consumers care about. And what equipment we use is OUR decision. Personally, I don't have the money to spend (or waste) $10K, $30K or $100K on a D/A converter. I leave that decision to people that do. :-)
I just have a growing number of 24 bit recordings that I've compared against 16 bit versions and in every case the 24 bit sounds better. noticeably better. What and how they did this, I could give rip. What you claim has nothing to do with anything unless you are involved with creating both the 16 bit and 24 bit versions and I can listen to both to compare the two on my system. That's all I can go by.
So, what it sounds like is that you will never make any 24 bit recordings, since you claim there isn't a difference. I wonder how that will impact your ability to get future business if the world is moving towards 24 bit and you won't do 24 bit. Sounds like a career limiting move. :-)
I tell you what. Here's a test for you to do. Go conduct your listening tests in your studio using a pair of TAD Reference monitors and some decent amps and cables and then tell me afterwards what you hear when comparing the tests you speak of.
Oh, and back when they were doing hearing tests, I don't even think they had headphones that could accurately reproduce more than 20KHz. It wasn't until the 60's until they broke the barrier of speakers going past 10KHz. They NOW have headphones that go past 30KHz, and speakers that can reproduce 100KHz. Maybe you need those to hear the difference. :-) I'm kind of being a little sarcastic, but there are top engineers that would give their right Testicle to own a pair of TAD Reference monitors. If you can't afford them, then get some Pioneer S-1EX or S-2EX. I know someone that has the smaller units available for about $3800 a pair with stands if he still wants to sell them and you want to buy them. They are unbelievably accurate lower priced versions of the TAD. Same technology designed by the same person. They just don't go down as low in the bottom frequencies (34Hz instead of 32Hz for the smaller monitors), but they do go to 100KHz. :-)
Oy vey.
I don't recall the last time I posted on Ars but the other two are definitely me. If I'm not mistaken I think I see you post on AnandTech from time to time.
Yeah, I tend to comment around... I also use "Loosely Coupled" on a lot of sites, so if you ever see that, its me
It already is.
Ponoschmono.
I agree. It would be an incentive to purchase iTunes Match. No-one wants to pay to upgrade all their albums separately.
No. Analogue is better than digital. But iTunes is fine for me.
It's the place for that, too.