[Chris J]

Well this is ONLY my opinion so here goes.........several months ago a remastered version of Bruce Springsteen's Darkness on the Edge of Town was released on CD. Truthfully I was hoping this would shed new light on an old recording. unfortunately I was disappointed, as I found that I still preferred my old CBS half speed mastered LP from 1978 and my Japanese pressing from 1978. Sure the vinyl has ticks, pops and surface noise, but the vinyl versions of this particular album still give me a greater emotional response. Who knows, maybe it's just nostalgia, but I just don't FEEL the music on the remastered CD.
I also have a remastered version of another album ( which shall go nameless...I'm too embarrassed to name it) where I find that I do prefer the modern CD over the 1987 vinyl.

At the end of the day, it's all about the music, which is why we are all here debating this.

 

[gregorio]

Unfortunately, when it comes to remastering there appears to be two main philosophies: 1. If it's worth doing, it's worth doing right or 2. If it's worth doing, it's worth doing as cheaply as possible to ensure a decent profit"!

G

 

[Chris J]

 
Quote:

Unfortunately, when it comes to remastering there appears to be two main philosophies: 1. If it's worth doing, it's worth doing right or 2. If it's worth doing, it's worth doing as cheaply as possible to ensure a decent profit"!

G

I thought philosophy #2 also stated that if it's worth doing, then it's worth compressing the life out of it!  LOL
 
 
BTW, I understand that more than a few dollars were spent remastering Darkness On The Edge Of Town,  I can only guess (and it is only a guess on my part) that somehow the LP mastering fattened up a somewhat dry and thin recording.
 

 

[Chris J]

Quote:

Absolutely not, that's exactly the same as saying some paintings are better viewed through rose tinted sunglasses. Regardless of whether or not I think a painting has been well painted, even if I think the painting could do with a bit of pink tint, I still want to view it as close to the intention of the artist as I can. Maybe the painting has been poorly painted or maybe the artist deliberately wanted a lack of pink to create a particular effect. It's up to me to interpret and judge the painting as it is, not to repaint it!




OK, you're right, the painting analogy is wearing a bit thin. Music isn't like a painting, it's been designed to be listened to on speakers. Unless your system is actually adding something in addition to what is on the recording, you cannot have too much detail in a recording. Unless it's been badly made, the only detail on a recording which is not designed to be heard is the noise floor, everything else is designed to be heard. I might not like the detail and I might think the recording sounds better without it but I don't try to re-master it, I just avoid that combination of artists in the future.




What low level linearity of digital? What phase shift added by AD and DA? There can be no doubt that CD as a format is orders of magnitude more accurate than vinyl in every conceivable way. What I would say, is that regardless of the fact that CD is clearly the better format, knowing this gives us no indication whatsoever of the quality of what is on the format. I would far rather listen to a great recording on vinyl than I would to a rubbish recording on CD.

G

 
C
 

 

[gregorio]

G,
It is an artifact of the digital recording process that as you record signal at a lower level that 0 dfs, distortion increases.  Distortion is higher if you were to record a signal at (for example) -80 dB below dfs. This is one of the reasons why professional standard is usually 24 bits.

Mmm, not really: All digital audio recording is below 0dBFS, as 0dBFS is the theoretical maximum of any fixed point digital audio system. There is no distortion in digital audio until you hit the noise floor (-96dB approx for 16bit). It's not like analogue which gradually gets worse as you near the limits. With digital audio it's perfect until you hit the limit and then it's suddenly phuqed! The reason we use 24bit for professional recording is to provide 18dB (or more) of headroom during recording. As we do not want to hit 0dBFS and as we can't predict how loud the musician is going to play, 18dB of headroom is essential. Of course we don't need any headroom on mix down, so 24bit is pointless for the consumer.

As for the phase shift, for a D/A converter to recreate the original signal  of a CD (i.e. convert it back into analog), a lot of low pass filtering is normally used. If this is done as an analog process then the more filtering you use, the more phase shift you will get at 20 khz relative to the midband (say 1 kHz).  There are numerous ways to get around this, like oversampling to 88.2 kHz or 176.4 kHz and/or filtering in the digital domain.  This is also one of the reasons why professional digital recording sampling rates are usually 96 kHz or even as high as 192 kHz.
This phase shift can also be added in the A/D process used in mastering and/or recording.
C

While you are correct in stating that analogue filters can cause phase issues, you have misunderstood what happens in the AD process. All professional ADCs oversample 256 or 512 times, so the analogue filter has a very gentle slope with the transition band not starting until well above the mega hertz range, which therefore causes zero phase problems. At a certain point in the circuitry, this 22mHz (yes, mega hertz!) or so sample rate is put through a decimation filter to reduce it down to the required rate, say 44.1kHz. The decimation filter (and anti-alias filter) is a FIR filter (or minimum phase filter), so phase really is not an issue in any half decent pro ADC.

When it comes to DACs, I would expect a decent upsampling DAC to have pretty much zero phase issues, for the reasons you mentioned. Although they should never upsample (or be set) beyond 96kHz. 176.4kHz and higher sample rates causes problems, not so much with phase but with alias-image rejection. If you see 192kHz recordings, they are compromised and made only to keep the marketing departments happy. No educated self respecting engineer would record or mix at 192kHz unless forced. For this reason, some of the highest end pro-gear does not allow for 192kHz.

Compared to the phase problems (in the HF) and non-linearities of analogue and vinyl, digital audio is orders of magnitude more accurate (linear).

G

 

[jiiteepee]

Regarding compression/loudness war ... I've checked DR (effective dynamc range) of few vinyl rips (my own rips) and compared those against their CD versions with the result DR value in vinyl rip can be few decibels higher (not because of subsonics/crackle/pops).
 
There are tools for to measure DR at
http://www.pleasurizemusic.com/
http://www.jokhan.demon.nl/DynamicRange/index.htm
 
jiiteepee
 

 

[Edoardo]

 
Quote:

What you are saying is absolutely true in theory. In practice though not necessarily. There is unfortunately a great deal of extremely poor mastering going on, mainly because the basic technology is so cheap and easily available compared to say 15 or more years ago. Given that they often re-master when selling 24/96 versions and that it's often done on the cheap (some kid in a bedroom) rather than by real mastering engineers; it's entirely possible that a modern 24/96 commercial rip is of poorer quality than a straight 16bit rip of an old (very well recorded) vinyl disk.

However, given a good mastering engineer with the right equipment and a high quality master to work with, the commercial rip should be of higher quality.

G

 
 
Do you really think that in the same studio recording that can afford a tape recorder or an analog recording they have cheap USB 1.0 soundcard and kind of Windows Vista SP 1 PCs?
 

 
 

 

[gregorio]

Do you really think that in the same studio recording that can afford a tape recorder or an analog recording they have cheap USB 1.0 soundcard and kind of Windows Vista SP 1 PCs?

The vast majority of modern vinyl releases are not actually recorded in analogue but are cut from digital masters. The master has to be EQ'ed to take into account the non-linear nature of vinyl. If this same master was released (without change) as a 24/96 recording it would sound harsh and tinny. I'm not saying this is definitely what is happening but there is no doubt that some 24/96 releases and re-releases are poorly mastered relative to the vinyl versions. The inherent superiority of 24/96 over vinyl should result in the 24/96 version always sounding better, the fact that this is not always the case indicates some incompetency or some deliberate act to reduce the quality of the 24/96 version.

G

 

[Elysian]

A properly mastered digital recording, on a good digital setup, will allow a consumer to hear something pretty close to what the producer was hearing.
 
Vinyl won't come as close.  I'm slightly biased from dealing with a lot of white label and acetone 12" recordings, but unless there's been a complete screw-up in the studio, CDMs will be much more faithful.
 
Your personal preference on the kind of sound you like is an entirely different matter altogether, but analog, by nature of the medium, cannot be as accurate as digital.

 

[Chris J]

 
Quote:

Mmm, not really: All digital audio recording is below 0dBFS, as 0dBFS is the theoretical maximum of any fixed point digital audio system. There is no distortion in digital audio until you hit the noise floor (-96dB approx for 16bit). It's not like analogue which gradually gets worse as you near the limits. With digital audio it's perfect until you hit the limit and then it's suddenly phuqed! The reason we use 24bit for professional recording is to provide 18dB (or more) of headroom during recording. As we do not want to hit 0dBFS and as we can't predict how loud the musician is going to play, 18dB of headroom is essential. Of course we don't need any headroom on mix down, so 24bit is pointless for the consumer.



While you are correct in stating that analogue filters can cause phase issues, you have misunderstood what happens in the AD process. All professional ADCs oversample 256 or 512 times, so the analogue filter has a very gentle slope with the transition band not starting until well above the mega hertz range, which therefore causes zero phase problems. At a certain point in the circuitry, this 22mHz (yes, mega hertz!) or so sample rate is put through a decimation filter to reduce it down to the required rate, say 44.1kHz. The decimation filter (and anti-alias filter) is a FIR filter (or minimum phase filter), so phase really is not an issue in any half decent pro ADC.

When it comes to DACs, I would expect a decent upsampling DAC to have pretty much zero phase issues, for the reasons you mentioned. Although they should never upsample (or be set) beyond 96kHz. 176.4kHz and higher sample rates causes problems, not so much with phase but with alias-image rejection. If you see 192kHz recordings, they are compromised and made only to keep the marketing departments happy. No educated self respecting engineer would record or mix at 192kHz unless forced. For this reason, some of the highest end pro-gear does not allow for 192kHz.

Compared to the phase problems (in the HF) and non-linearities of analogue and vinyl, digital audio is orders of magnitude more accurate (linear).

G

G,
 
I suspect what most audiophiles really want to know is why, oh why does a mono Jazz recording from 1955 sound virtually as good as a stereo digital Jazz recording from the present day? To be specific, I have a new pressing of Miles Davis Walkin on 45 RPM vinyl which I have compared to a present day Roy Hargrove CD. I'm not trying to compare the quality of the music, just the quality of the recording, and I am certainly not trying to scientific, just expressing the (very subjective) opinion of this music lover.
I know that is a completely subjective question, and I'm not trying to confrontational or argumentative, I'm just curious, where is the progress? Sounds like modern recording technology has really hit the fine point of diminishing returns. I understand that the ability to process sound has taken some massive leaps forward, but as a dumb ***** consumer, I don't hear a massive leap forward in fidelity. I suppose you could argue that my turntable is of "higher quality" than my CD player, but again, I do understand that my turntable still drags a rock thru a piece of plastic to reproduce sound.
 
Again, I'm really not trying be argumentative............just curious. Are we being phuqued over by the record companies and the mastering engineers?
 
C.
 
PS  if you look at my personal profile, I am an electrical engineer by trade, maybe I should add that I am an electrical power distribution engineer.
 
 
 

 
 

 

[gregorio]

G,
I did some digging, re-read this entire thread and found your post on "16 vs 24 bit, the myth exploded". Interesting, informative stuff. Thanks.
I think my data is probably 20 years out of date, but then I don't record music, I listen to it. Some of the technical stuff I read in the Hi-Fi mags is probably a bit suspect.
So I guess the whole phase shift problem and low level distortion problem was solved many years ago. Maybe I should give a few state of the art classical CDs another chance (I got so disillusioned that I stopped buying classical CDs years ago unless they got very high recommendations from reasonably reputable rags).
 
I suspect what most audiophiles really want to know is why, oh why does a mono Jazz recording from 1955 sound virtually as good as a stereo digital Jazz recording from the present day? To be specific, I have a new pressing of Miles Davis Walkin on 45 RPM vinyl which I have compared to a present day Roy Hargrove CD. I'm not trying to compare the quality of the music, just the quality of the recording, and I am certainly not trying to scientific, just expressing the (very subjective) opinion of this music lover.
I know that is a completely subjective question, and I'm not trying to confrontational or argumentative, I'm just curious, where is the progress? Sounds like modern recording technology has really hit the fine point of diminishing returns. I understand that the ability to process sound has taken some massive leaps forward, but as a dumb ***** consumer, I don't hear a massive leap forward in fidelity. I suppose you could argue that my turntable is of "higher quality" than my CD player, but again, I do understand that my turntable still drags a rock thru a piece of plastic to reproduce sound.
 
Again, I'm really not trying be argumentative............just curious. Are we being phuqued over by the record companies and the mastering engineers?
 
C.
 
PS  if you look at my personal profile, I am an electrical engineer by trade, maybe I should add that I am an electrical power distribution engineer.

There are a number of reasons which explain your observations, covered to a great extent already in this thread, but I'll recap. Taking your example of your 1955 mono jazz recording. There are several differences between now and 1955 when it comes to commercial recording.

1. In 1955 the only recording studios were commercial studios. Very expensive with highly trained and experienced engineers and producers working in teams.
2. Record sales were very high in the 1950's, big profit margins, plenty of money to invest in new product and not many different quality options.
3. Piracy, single track purchase, diversified music market and many other forms of entertainment has drastically reduced profit for the record companies in the C21st. Less money to invest in new product.
4. Technology has resulted in tens of thousands of smaller cheaper studios, staffed by less well trained and less experienced engineers, teams are virtually a thing of the past. Often the various engineers and producer is now the same person.
5. In an attempt to get more sales record companies have caught themselves in a vicious circle of louder and louder recordings. This started to affect sound quality 20 years ago and has only gotten worse since then.
6. An album would often take 6 months to create in the 60's and 70's, today 6 weeks is quite generous.

Even though digital is far better as a format, what is on the format is even more important. A high quality digital recording is better than a high quality vinyl recording. However, a high quality vinyl recording is better than a low quality digital recording. The difficult part is finding the high quality digital recordings amongst so much rubbish!

G

 

[Edoardo]

Quote:

A properly mastered digital recording, on a good digital setup, will allow a consumer to hear something pretty close to what the producer was hearing.
 
Vinyl won't come as close.  I'm slightly biased from dealing with a lot of white label and acetone 12" recordings, but unless there's been a complete screw-up in the studio, CDMs will be much more faithful.
 
Your personal preference on the kind of sound you like is an entirely different matter altogether, but analog, by nature of the medium, cannot be as accurate as digital.

Nice to see you here too! Fully agree with you
 

 

[Edoardo]

Quote:

The vast majority of modern vinyl releases are not actually recorded in analogue but are cut from digital masters. The master has to be EQ'ed to take into account the non-linear nature of vinyl. If this same master was released (without change) as a 24/96 recording it would sound harsh and tinny. I'm not saying this is definitely what is happening but there is no doubt that some 24/96 releases and re-releases are poorly mastered relative to the vinyl versions. The inherent superiority of 24/96 over vinyl should result in the 24/96 version always sounding better, the fact that this is not always the case indicates some incompetency or some deliberate act to reduce the quality of the 24/96 version.

G

 
You know, there are also producers who make 24/96 masters or take a 24/96 digital track of the master tape and don't equalize anything before selling... Because not everybody means to put it on vinyl... By the way I'd like to know who does equalize the master tape or the master file for vinyl and then sells it in a digital format???
 
HD audio gives us the chance to listen to recordings just as the master engineer did.
 
I don't know why we should insist on claiming the superiority of a format which is very very charming and cool looking, but full of physical limitations, especially now that digital conversion technology allows us to listen to something as near as possible to the original master tapes, if not the very original master file. And files don't deteriorates. I don't know what are DG and Decca waiting to release those files out!
 
 
 

 

[mrspeakers]

G,
It is an artifact of the digital recording process that as you record signal at a lower level that 0 dfs, distortion increases.  Distortion is higher if you were to record a signal at (for example) -80 dB below dfs. This is one of the reasons why professional standard is usually 24 bits.
As for the phase shift, for a D/A converter to recreate the original signal  of a CD (i.e. convert it back into analog), a lot of low pass filtering is normally used. If this is done as an analog process then the more filtering you use, the more phase shift you will get at 20 khz relative to the midband (say 1 kHz).  There are numerous ways to get around this, like oversampling to 88.2 kHz or 176.4 kHz and/or filtering in the digital domain.  This is also one of the reasons why professional digital recording sampling rates are usually 96 kHz or even as high as 192 kHz.
This phase shift can also be added in the A/D process used in mastering and/or recording.
C
 

The reason to use higher sample rates is to reduce the need for a brick wall filter near 20KHz, thus reducing phase shift, as noted above, but also to reduce aliasing artifacts. The neat thing is higher sample rates let you use gentler filters which also reduce ringing.

By the way, most people here seem to assume digital is perfect or all digital systems are created equal. This is not the case at all. For example, there are huge variations in how accurate the least significant bit(s) linearity is, especially as bit depth and sample rate increase. Also, the designers' choices in filter type, slope, etc all affect sound.

My $.02 is that I have heard awesome analog and digital recordings. It has as much to do with the studio setup and the engineering as it does analog vs. Digital. I used to be an analog bigot, but that has changed as digital has matured. Now, all I care about is "does this draw me in and engage me?"

 

[gregorio]

You know, there are also producers who make 24/96 masters or take a 24/96 digital track of the master tape and don't equalize anything before selling... Because not everybody means to put it on vinyl...

I don't know why we should insist on claiming the superiority of a format which is very very charming and cool looking, but full of physical limitations, especially now that digital conversion technology allows us to listen to something as near as possible to the original master tapes...

Whatever format the recording is being releasing on, the pre-master created by the producer still needs mastering, which includes compression, EQ and various other processes. The point of mastering is to alter the original mix so that it sounds good on a variety of consumer equipment, rather than sounding good just in the studio in which it was produced.

I hope, after reading my posts that you do not think I am arguing that vinyl is better?

By the way, most people here seem to assume digital is perfect or all digital systems are created equal. This is not the case at all. For example, there are huge variations in how accurate the least significant bit(s) linearity is, especially as bit depth and sample rate increase.

The LSB should have nothing much in it except for random noise. It's all a bit irrelevant though as the noise floor of even the most dynamic recording is going to be at least 30 times higher than the digital noise floor.

G