Knowing when or not to buy vinyl

[Joe_caithness]

If you like the band/music?

[ds11]

Originally Posted by Plush

it is this noise that swamps any proposed benefit of using hi-res files for LP mastering.

High-res as in high sample rate, or high bitrate? Are you saying 16 bit and 24 bit digital pressed to vinyl would sound the same?

[Rudiger]

I don't understand the vinyl craziness going on. Wouldn't mastering the work without smashing the dynamics and releasing it in digital be a better solution?

I can't see the point in releasing a vinyl cut from a CD master.

Maybe I'm off topic here but dynamic range of 7dB sounds nasty regardless of analog or digital.

[ds11]

Originally Posted by Rudiger

I don't understand the vinyl craziness going on. Wouldn't mastering the work without smashing the dynamics and releasing it in digital be a better solution?

I can't see the point in releasing a vinyl cut from a CD master.

Maybe I'm off topic here but dynamic range of 7dB sounds nasty regardless of analog or digital.

Well that's the question now.. you'd have to compare 96khz from master tapes to a vinyl copy that was pressed before converters and digital delay were used.

[duomo]

Originally Posted by ds11

Well that's the question now.. you'd have to compare 96khz from master tapes to a vinyl copy that was pressed before converters and digital delay were used.

its much more important that the sound of the source is mixed and mastered well.
Is it compressed too much ?
too much High frequency content ?
is the playing time per side not too long ?
All this, and of course the play back situation, is much more important than if its a 44,1 or 86 khz file. If everything is perfect you'll hear a difference, otherwise the faults of the medium are bigger than the faults of 44,1 compared to 86 Khz.

[Plush]

Originally Posted by Paul Gold

I haven't read the paper but if it takes 12 bits to describe the LP signal at the noise floor you need more of those to describe the full dynamic range.

The paper offers that 12 bits describes the entire capability of the LP.
This makes sense since we know that the LP maxes out around 74 dB dynamic range. As you know, most are far less.

I know that you're a vinyl expert. Has it been your experience that hi-res material makes the LP sound better?

p.s. for the OP--hi-res refers to anything at 24/44.1 and above.

[Paul Gold]

Originally Posted by Plush

The paper offers that 12 bits describes the entire capability of the LP.
This makes sense since we know that the LP maxes out around 74 dB dynamic range. As you know, most are far less.

You have said that when the paper says "bit" they are not only refering to dynamic range. There is some sort of "equivalency" going on. This smacks of some sort of agenda. Using terminology for digital signals to describe analog signals is pretty useless to start with. Ain't no bits in analog. I have no problem comparing the two. I have no problem with digital being better on paper. I have no problem with someone liking or not liking either one. I do have a problem misapplying terms with a dash of fairy dust to prove a political point.

I know that you're a vinyl expert. Has it been your experience that hi-res material makes the LP sound better?

Sometimes yes, sometimes no.

[Plush]

For goodness sake--my posts aren't about an agenda at all and neither does the paper I cited have an agenda. It is a paper written to inform. So you can step off of that line of thought. I am a pro who loves analog playback from tape and at home from LP. But I make my living with hi-res digital in classical recording. I am a recording scientist.

What I was attempting to do was to offer a description of LP playback limitations. Then the posters can better understand what mastering sources can make a difference.

Audio luminaries and pioneers like Stuart, Craven and Gerzon have done the same so I hardly think that it is a sacrilege to point people to their work here in this thread.

[jeff_free69]

Originally Posted by audioaddict

Not sure if it is crushed.....I gave a listen on youtube and it didn't appear crushed.

I bought the Foo Fighters CD (sorry if I'm getting OT about vinyl) and it sounded totally fatiguing to me. Starts off nice, but by the time you get to the second chorus of the first song, it gets fatiguing and there's no place left for it go. Its maxed out with a DDR of 6, if I remember correctly

Originally Posted by audioaddict

They want all out to record and edit to tape.....would be a colossal waste of time to ruin it in mastering.

Thats why I refer to "Wasting Light" as "Wasting Tape". it actually came with a sliver of 2" tape in the package. Good material and great playing; but sonically, what a disappointment!


OTOH I'm listening to Jack White "Blunderbuss" now, and that is all analog recroding and NO loudness mastering. Individual tracks still have plenty of JW 's trademark saturated, screech guitar and hyped vocals, but there's plenty of dynamic range on the stereo mix.

[Paul Gold]

Originally Posted by Plush

The paper offers that 12 bits describes the entire capability of the LP.
This makes sense since we know that the LP maxes out around 74 dB dynamic range. As you know, most are far less.

So if you are contracted to do a transfer of an LP you will bust out the 12bit converter with the best analog front end and clocking ever built. Then send it out knowing as a scientist you did the best job possible? Tape has a similar dynamic range. 12 bit capture must be good enough for that too.

[Trakworx]

It seems to me that this whole notion (of high bit depth sources not improving the sound of vinyl pressings) rests on the assumption that the noise floor is the only audible difference between, say, a 12 bit recording and a 24 bit recording. I've been operating under the impression that there is more to it than that...

[huejahfink]

I'm also of the opinion that the depth resolution (whether measured in bits or dynamic range) of digital audio, and the signal to noise ratio of an analogue signal / recording are two ENTIRELY different things. Further signal exists below the noise floor on a tape or vinyl, but nothing exists below the least significant bit.

[Rudiger]

Originally Posted by huejahfink

I'm also of the opinion that the depth resolution (whether measured in bits or dynamic range) of digital audio, and the signal to noise ratio of an analogue signal / recording are two ENTIRELY different things. Further signal exists below the noise floor on a tape or vinyl, but nothing exists below the least significant bit.

Well, nothing exists in the upper 8 bits of most modern CDs except a bunch of unchanging 1's. :-)

[Plush]

Pretty funny responses here when one tries to enliven the thread and hopefully point out relevant info. This is especially the case when the OP is seeking to base his buying decisions on information that is often not available. (What type of source for the LP or cd??)

Probably best to take a cursory look at the paper I cited before dismissing it all. I am summarizing what Robert Stuart said. I am NOT making up facts or reporting in a delusional manner as is so common now on GS.

As far as Paul's recommendation to me for use of the 12 bit converter, I done already retired my dat machines, baby. I do still have a Sony F-1 with a 14 bit switch here though.

[Paul Gold]

Originally Posted by Plush

Probably best to take a cursory look at the paper I cited before dismissing it all. I am summarizing what Robert Stuart said.

I will take a look at the paper. I am not dismissing it. I am dismissing the summary because it makes no sense.

[Hic&Nunc]

As far as I'm concerned I think people who care about music prefers vinyl over cd because;

a) the vinyl won't be as squashed.

b) in most case the cutting engeneer D/A will beat those fond on consumer cd players and computers.

c) the specifications inherent to vinyl reproduction.

And of course all the more subjective reasons why vinyl may be a more immersive ways of consumming music.

[Paul Gold]

Originally Posted by Plush

Probably best to take a cursory look at the paper I cited before dismissing it all.

I read the paper over dinner. Have you? It makes a single passing reference to the dynamic range of vinyl being about equivalent to a 12 bit system. No argument there. I don't have any arguments with the paper at all. It's a 'best practices' paper on digitally capturing and storing audio. Unsurprisingly it recommends capturing at no less than 20bit and to pay close attention to low level non linear distortion and the other weak spots in PCM audio. Nothing remotely controversial. It never equates total performance of a disk recording system with 12 bit performance.

[dcollins]

Originally Posted by huejahfink

I'm also of the opinion that the depth resolution (whether measured in bits or dynamic range) of digital audio, and the signal to noise ratio of an analogue signal / recording are two ENTIRELY different things. Further signal exists below the noise floor on a tape or vinyl, but nothing exists below the least significant bit.

The two are completely equivalent............

There is audio below the least significant bit..............

Not opinions.

One way to look at the digital/vinyl comparison:

Divide the groove into 65536 (16 bit) slices. Does your stylus have that kind of amplitude resolution?

What is a typical "minimum" size feature that can be played?

You can work back from there - also using the real-world dynamic range of vinyl - to get a type of bit-equivalency.



But, there is certainly no disadvantage to using >44/16 for cutting masters if the source is available.


DC

[Plush]

Glad you liked it, Paul. It is a very interesting paper from 1997--just when higher res digital was being adopted.

Yes I have read the paper many times
and I was also there when it was presented at the AES convention.

Please go to the ending appendix where Stuart graphs CD, LP and FM radio noise vs. hearing threshold. It is from those graphs and from talking to Stuart about them that I learned that a bit comparison can be made. They are illustrations about some of the differences between the formats.

I wonder why you would ask me if I have read the paper? I am not an unknown person here.

[Paul Gold]

Originally Posted by Plush

Please go to the ending appendix where Stuart graphs CD, LP and FM radio noise vs. hearing threshold. It is from those graphs and from talking to Stuart about them that I learned that a bit comparison can be made. They are illustrations about some of the differences between the formats.

I agree that a comparison can be made. In the same way you can compare apples and oranges. As long as you restrict the comparison to single parameters like vitamin C level. My gripe is a broad brush comparison which the paper never makes and isn't the point of the graphs. The graph refers strictly to dynamic range.

I wonder why you would ask me if I have read the paper?

Because I draw none of the same conclusions as you.

[Plush]

No big deal. I'm not here to debate only with one cynic.

I stepped in as a vinyl lover and high quality digital lover to point out some relevant factors for the OP.

That the LP has specific limitations has never bothered me. LP playback is enjoyed often here.

As Paul has mentioned, the dynamic range of tape and of the LP match well.

That's why I am in the process of issuing a 30 year retrospective of my recordings on LP. The retrospective will be mastered from an IEC 1/4" master recorded on a Stellavox SM-8.

For those who don't come from the LP era, and who have misconceptions about what the LP can offer, it is essential to understand that it is not a high dynamic range carrier. High dynamic range is one of the first, and most sought after, benefits of hi-res digital and a major reason that it is used at all.

[ds11]

Originally Posted by Plush

That's why I am in the process of issuing a 30 year retrospective of my recordings on LP. The retrospective will be mastered from an IEC 1/4" master recorded on a Stellavox SM-8.

Very cool! Have you digitally archived the master tape? at 96k? It would be great to hear both that and the vinyl A/B'd on the same sound system.

[ds11]

One other related question for you vinyl mastering engineers -

Is it worth searching out original pressings if vinyl reissues were done in the 80's? For example, Bob Marley's Catch a Fire was released in 1973, 74, and 76 in various countries, then reissued in '86.

From what I've read, this likely went through some sort of digital delay for syncing purposes? I can't imagine ADC as great in the 80's (am I wrong?), so should this be avoided in favour of a 70's pressing?

[huejahfink]

Originally Posted by dcollins

There is audio below the least significant bit..............

OK, I'm intrigued. How is there audio beneath the least significant bit in a digital audio file?

(I'd like to point out that this question is born of genuine interest to hear from you as an esteemed expert because I am genuinely here to learn, and not some kind of simple minded backlash because you disagreed with my post.)

[dcollins]

Originally Posted by huejahfink

OK, I'm intrigued. How is there audio beneath the least significant bit in a digital audio file?

http://www.drewdaniels.com/dither.pdf

Because d*ther modulates the LSB. If you are talking about a 3k tone, it's audible way below the supposed -96dB level of 16 bits. Obviously, the type of signal will define just how much actual audible response you will get (transients less, steady tones more).

(I'd like to point out that this question is born of genuine interest to hear from you as an esteemed expert because I am genuinely here to learn, and not some kind of simple minded backlash because you disagreed with my post.)

Either reason is fine, I was merely making a technical correction to your post.


DC

[huejahfink]

Well if you are talking about dither (I had a feeling you might be referring to that), I would have thought that the purpose of that was to pull some of the audio information from the larger wordlength file into the least significant bit of the shorter wordlength file, so that 'perceptibly' it's still audible.
I certainly understand the point you are making, but to my mind (at least right now) that information that has been saved is still contained within the bits, not magically existing beyond the bits themselves.
Maybe I've got the wrong end of the stick here... but that still sounds like it's not actually existing beyond the LSB.

Well whatever - thanks for taking the time to explain your reasoning. What you say fits perfectly well into my current realm of understanding, my world has not been shattered - back to work

[dcollins]

Originally Posted by huejahfink

I certainly understand the point you are making, but to my mind (at least right now) that information that has been saved is still contained within the bits, not magically existing beyond the bits themselves.
Maybe I've got the wrong end of the stick here... but that still sounds like it's not existing beyond the LSB.

It's still in the LSB, but now pulse-width-modulated. Because your ear works in a similar way to an averaging FFT spectrum analyzer, (Hi EV!) over time the signal can be extracted. That's why continuous tones come through better.

There are graphs in that paper showing this effect.

DC

[wado1942]

It's still in the LSB, but now pulse-width-modulated.

Not pulse width modulated, the pulse is always the same width in digital. A better term is merely dither modulation.

As for a previous post (I quickly skimmed the thread since I don't have much time), I wouldn't call myself a vinyl expert by any means or even an audio expert, though I know more than most. On the subject of dynamic range, vinyl uses emphasis that is sort of a simplified inverse Fletcher-Munson curve (called the RIAA curve) while the lacquer is being made, which is reversed on replay. When the RIAA repro curve is applied in your turntable preamp, the music is essentially of flat frequency response, but the noise is the opposite of our hearing sensitivity. The noise floor has not been significantly reduced, but the noise is much less audible as it conforms to our threshold of hearing at a given listening level. Remember, you can hear audio up to 20dB below the noise floor as well, so the total dynamic range from noise masking to max level is more like 93dB.

If use used 12-bit, 44.1K, non-dithered storage, you have a max dynamic range of 72dB, plain and simple. I'll note that audio approaching the threshold of the LSBs crackles and is unlistenable, so I'd argue that the dynamic range is even less than 72dB. If you introduce noise-shaped dither before truncation, you're more on the lines of vinyl as far as usable dynamic range goes. I'll argue, though, that even dithered 12-bit digital only has 4096 sample values while a stylus has infinite variations. Analogue audio cannot be summarized in "bits" because the theory doesn't apply.


On another note, I got a hold of Wasting Light on vinyl, digitized it at 24/96 and analyzed it. It is full of audible limiting and digital clipping, showing that it was taken from the same master that was used to make the CD. I'm just glad I didn't actually buy it, because I'd be writing Dave Grohl personally to demand my $30 back.

[dcollins]

Originally Posted by wado1942

Not pulse width modulated, the pulse is always the same width in digital. A better term is merely dither modulation.

No, it's pulse width modulation. You should read the paper I referenced. The authors actually call it "duty-cycle modulation" which is the same thing, but it's forgivable as they are Canadian.

Here is a picture of the actual waveforms, showing both the PWM waveform and the recovery of the signal after averaging.

DC

[dcollins]

Originally Posted by wado1942

I'll note that audio approaching the threshold of the LSBs crackles and is unlistenable,

Sigh...................


DC