Jump to content
IGNORED

Lies about vinyl vs digital


Recommended Posts

4 hours ago, Paul R said:

All the tracks on iTunes are high resolution already. They are simply delivering them as AAC 256 right now.

 

There's no way of knowing this with certainty unless you have inside Apple info...  it wouldn't even make sense to guess the number.

 

The Mastered for iTunes process recommends tracks are submitted in 24/96kHz ideally, but it doesn't mean they always are or have to be...

 

1918885215_ScreenShot2019-05-21at9_49_16pm.thumb.png.5a0aa0cff87143b267e11f04bce8ec66.png

 

Many aren't aware that even Spotify recommends tracks are submitted to them in FLAC or WAV... could be RBCD or could be 24bit...

 

1669563781_ScreenShot2019-05-21at9_49_30pm.thumb.png.3b693baa79b475aa84dbf9ed939af047.png

 

Link to comment
28 minutes ago, PeterSt said:

 

I'll grant you that.

Still I'd rather have Joe Jackson sound right. Or Elvis Costello. Or Ian Dury and a couple of Blockheads. And they sure do. Also made for the masses but different masses.

Dire Straits always sounded right everywhere. And therefore it was wrong. I loved the band back in the days. But today I never play them. There's nothing much in it. Too easy. No challenge. No ABBA but no challenge either.

Try Grand Funk. War. And if you really don't know where to go to, the very first of Supertramp. Or King Crimson's first, if you can find the not-remaster sh*t. And a couple of 1000 others.

 

Btw, Bleeper has no works. Madonna just died too, on that stairs. I hope you missed it.

 

I will try The Stranger tonight. See what's in it these days. See whether you are right, never mind we agree already.

Maybe after that I'll try The Moody Blues.

 

Unremastered "Who Are You" at high volume.  The bridge on that will knock you against the back wall!  For the masses, kiddo! 😆

Link to comment
On 5/19/2019 at 8:26 AM, davide256 said:

 

 

That's not true compression... its clipping. If compression is done correctly  the wave form should resemble the original but with a lower peak level and reduced differences

between soft and loud passages. Fire the engineer for incompetent use of technology.

 

Compression is nothing new... most vinyl is compressed because otherwise too much distance between grooves would be required to avoid perturbing a neighboring groove

during master cutting

Modifying the comment about the 'lower peak level'.  Not all compression is to lower the peak/avg/RMS level.  Some compression (e.g. my own experimental one) as an option, can keep the peaks the same, but brings up lower level material.  The advantage of doing that (when it is applicable) is that the 'squeze' sound is less.  Compressing the higher signal levels also tends to be a little more obvious.  Some people might do a sidechain kind of thing with a compressor also to get similar effects.

 

Limiters modify the peak level when exceeded (or sometimes do a bit of high ratio compression above a certain level.)  It would seem to be silly to have a limiter bring up the lower levels and not 'compress' the high levels.  'Compressors' can do either/both depending on the design.

 

Otherwise, I agree with your statement.

 

Link to comment
3 minutes ago, John Dyson said:

Modifying the comment about the 'lower peak level'.  Not all compression is to lower the peak/avg/RMS level.  Some compression (e.g. my own experimental one) as an option, can keep the peaks the same, but brings up lower level material.  The advantage of doing that (when it is applicable) is that the 'squeze' sound is less.  Compressing the higher signal levels also tends to be a little more obvious.  Some people might do a sidechain kind of thing with a compressor also to get similar effects.

 

Limiters modify the peak level when exceeded (or sometimes do a bit of high ratio compression above a certain level.)  It would seem to be silly to have a limiter bring up the lower levels and not 'compress' the high levels.  'Compressors' can do either/both depending on the design.

 

Otherwise, I agree with your statement.

 

Most of what I observe in badly recorded digital music is music  recorded too close to peak limit so that downstream DAC's misbehave. So perhaps not technically required but from a

"plays well with others" perspective, something that ought to be done.

Regards,

Dave

 

Audio system

Link to comment
Just now, davide256 said:

Most of what I observe in badly recorded digital music is music  recorded too close to peak limit so that downstream DAC's misbehave. So perhaps not technically required but from a

"plays well with others" perspective, something that ought to be done.

In the context that you are talking about -- yep, the kind of compression/limiting that is used is UGLY.  There is one set of recordings in my repository (relegated to my backup media) seem to be processed by a FM broadcast processor -- REALLY.  Not just a nice, mild compressor, but really extreme processing (e.g. crest factors under 4 and peak-RMS in the 12.5-14dB, where 14dB would be optimistic.)

(The complete studio recordings -- ABBA.)  Normally, ABBA would do a crest factor of 7-12 and 17-21dB peak-RMS on clean/direct DolbyA decodes...

 

So, I do agree with you about your main sentiment -- I was trying to clarify about compressors in general.  Sometimes subjects are so complex that it is almost impossible to be technically correct and all inclusive.  I did NOT intend to be a know-it-all.

 

John

Link to comment
8 minutes ago, davide256 said:

Most of what I observe in badly recorded digital music is music  recorded too close to peak limit so that downstream DAC's misbehave

 

If you were describing analog tape being pushed, in your sentence, you'd be correct.

 

Digital does not distort - until you attempt to exceed MSB(full scale) - when it clips.

 

In that token I do believe that the input stage of a typical stereo receiver, or integrated amp, can be strained by a recording with too high and too constant a RMS value.  

 

That's why I have  a pair of these:

 

https://www.parts-express.com/ViewProductImage.aspx?productid=8559&productzoomimage=True&img=266-242_HR_0.jpg

 

between the RCAs from my CD player and the corresponding inputs on my receiver.

Link to comment
4 minutes ago, The_K-Man said:

 

If you were describing analog tape being pushed, in your sentence, you'd be correct.

 

Digital does not distort - until you attempt to exceed MSB(full scale) - when it clips.

I had interpreted his comment to be about the problem of a low pass filter chopping the spectrum of square waveforms when running tpo close to the limit.  (Gibbs effect, paradoxically the peak level can be increased when doing an LPF in certain cases.)  One does need to be mindful when running too close to clipping.  I always normalize at least 1dB below clipping, and that might be too aggressive in some cases.  I start worrying when closer than 1dB.

 

John

Link to comment
5 minutes ago, John Dyson said:

I had interpreted his comment to be about the problem of a low pass filter chopping the spectrum of square waveforms when running tpo close to the limit.  (Gibbs effect, paradoxically the peak level can be increased when doing an LPF in certain cases.)  One does need to be mindful when running too close to clipping.  I always normalize at least 1dB below clipping, and that might be too aggressive in some cases.  I start worrying when closer than 1dB.

 

John

 

I was once told by a supposed mastering engineer on Gearslutz that a waveform that is not picket-fenced(constantly against full scale) is "not properly or professionally mastered".  He uttered something about "not making effective use of all the bits" or some such other ... .

Link to comment
21 minutes ago, The_K-Man said:

 

I was once told by a supposed mastering engineer on Gearslutz that a waveform that is not picket-fenced(constantly against full scale) is "not properly or professionally mastered".  He uttered something about "not making effective use of all the bits" or some such other ... .

AIEEE!!!   I have heard that kind of thing also.

I do see a lot of consumer recordings that are pinned up to 0dB, but not normally for every selection on an album.  It seems like that sometimes they keep the original relative levels from the master tape.

It can be safe to run up to almost 0dB -- but better be no processing until the signal is converted or brought down a little bit. The signal is a bit more fragile when so close to clipping, and any minor change can push over the limit.  It often doesn't hurt THAT much to clip a little bit, but shouldn't happen and shouldnt be allowed to happen (just my opinon.)


On my favorite subject, but relevant to the clipping/maximum scale thing -- different DolbyA decoders can produce different extreme peaks.  That means, when normalizing the output (looking for the peaks, then basing the signal levels on that peak), that for the same *apparent/audible* sound level, that the peak level can be signifcantly different.  Part of the issue is the varying gains and the IIR/analog style non-linear phase filters.


Probably just as important is a well tuned DolbyA unit (on OLD material.)  New stuff -- just ramrod the signal 🙂

 

Link to comment
1 hour ago, John Dyson said:

AIEEE!!!   I have heard that kind of thing also.

I do see a lot of consumer recordings that are pinned up to 0dB, but not normally for every selection on an album.  It seems like that sometimes they keep the original relative levels from the master tape.

It can be safe to run up to almost 0dB -- but better be no processing until the signal is converted or brought down a little bit. The signal is a bit more fragile when so close to clipping, and any minor change can push over the limit.  It often doesn't hurt THAT much to clip a little bit, but shouldn't happen and shouldnt be allowed to happen (just my opinon.)


On my favorite subject, but relevant to the clipping/maximum scale thing -- different DolbyA decoders can produce different extreme peaks.  That means, when normalizing the output (looking for the peaks, then basing the signal levels on that peak), that for the same *apparent/audible* sound level, that the peak level can be signifcantly different.  Part of the issue is the varying gains and the IIR/analog style non-linear phase filters.


Probably just as important is a well tuned DolbyA unit (on OLD material.)  New stuff -- just ramrod the signal 🙂

 

 

I personally don't do it.  

 

If I have 8 songs, of different dynamic ranges and loudness levels, I'd align them to, IE -12LUFS(loudness units, so they all sound the same volume without adjusting between tracks), I figure I've done my job.

 

The more dynamic ones might peak, during certain parts of the song, at -0.5dB full scale, and other less dynamic ones might never peak above -6 to -4.  As long as the listener doesn't have to touch their volume during the entire album(except to answer the phone or when significant other wants to sleep) that's what mastering is for.

 

I'm not obsessed with DRC'ing and limiting everything to within a dB of it's life just to compete with another artist or label. Guess I wouldn't make many friends in the recording business! lol

Link to comment
7 hours ago, mansr said:

That has nothing to do with 16 vs 24 bits. Perhaps it's not so simple after all.

 

But our friend is not merely talking about 16 vs 24 bit depth, he is also talking about 44.1k sample rates. Otherwise, I would probably agree with you. 

Anyone who considers protocol unimportant has never dealt with a cat DAC.

Robert A. Heinlein

Link to comment
5 hours ago, Em2016 said:

 

There's no way of knowing this with certainty unless you have inside Apple info...  it wouldn't even make sense to guess the number.

 

The Mastered for iTunes process recommends tracks are submitted in 24/96kHz ideally, but it doesn't mean they always are or have to be...

 

1918885215_ScreenShot2019-05-21at9_49_16pm.thumb.png.5a0aa0cff87143b267e11f04bce8ec66.png

 

Many aren't aware that even Spotify recommends tracks are submitted to them in FLAC or WAV... could be RBCD or could be 24bit...

 

1669563781_ScreenShot2019-05-21at9_49_30pm.thumb.png.3b693baa79b475aa84dbf9ed939af047.png

 

 

Okay, I grant you are correct in your conclusions. Let’s say then, I strongly believe all, or at least the vast majority of all music on iTunes has a high res original, be it 24/48 or 24/96, or above. 

 

I also strongly believe, but cannot provide absolute proof of, all new music is recorded at higher than CD resolution, usually much higher, and so providing Apple with high resolutions originals is neither difficult or any extra expense. 

 

How is that?  😎

Anyone who considers protocol unimportant has never dealt with a cat DAC.

Robert A. Heinlein

Link to comment
13 minutes ago, Paul R said:

But our friend is not merely talking about 16 vs 24 bit depth, he is also talking about 44.1k sample rates. Otherwise, I would probably agree with you. 

44.1 kHz is challenging with all-analogue filters. With digital filters there's no real problem.

Link to comment

My $million$ question, and one for which an answer seems to not exist on this wonderful thing we call the internet...

 

I understand Shannon-Nyquist theorum, I know what ADCs and DACs are for, but...  

 

Assuming a 44.1kHz sampling rate at 16bits: 

 

A 20kHz sine wave contains about 1/1,000th as many "dots"(sampling points) as does as 20Hz sine wave.   

 

Given that figure, is there a document somewhere explaining how a DAC accurately recreates that higher frequency wave from that many fewer points?  Succinctly, how exactly does a DAC do its job?

Link to comment
4 minutes ago, The_K-Man said:

My $million$ question, and one for which an answer seems to not exist on this wonderful thing we call the internet...

 

I understand Shannon-Nyquist theorum, I know what ADCs and DACs are for, but...  

 

Assuming a 44.1kHz sampling rate at 16bits: 

 

A 20kHz sine wave contains about 1/1,000th as many "dots"(sampling points) as does as a 20Hz sine wave.   

 

Given that figure, is there a document somewhere explaining how a DAC accurately recreates that higher frequency wave from that many fewer points?  Succinctly, how exactly does a DAC do its job?

 

Link to comment
15 minutes ago, mansr said:

That's what the sampling theorem is all about. Maybe you don't understand it quite as well as you think.

 

It's simple connect-the-dots, right?

 

IE: A child's connect the dots drawing of the most recognizeable ship: Titanic.

 

One drawing has 50 dots(samples) that properly connected will render those four smoke stacks, the vertical bow, and S-shaped counter stern.

 

The other drawing: 500 dots.

 

Which completed connect-the-dots will more strongly resemble the Titanic's silhouette?

 

Translated to a DAC 'connecting the dots' to 'redraw' a  complex violin quartet, how does the DAC do it? In electronic terms?

 

Aren't there complex algorithms that prevent dots from being 'skipped'?  Or the wrong dots from being connected:

 

IE our Titanic has two realllllly wiiiide funnels instead of four skinnier ones, because the child(the DAC!), went across from the top of one funnel to the next one instead of back down to the ship's deck, essentially missing a few dots(samples).

Link to comment
10 minutes ago, The_K-Man said:

 

It's simple connect-the-dots, right?

 

IE: A child's connect the dots drawing of the most recognizeable ship: Titanic.

 

One drawing has 50 dots(samples) that properly connected will render those four smoke stacks, the vertical bow, and S-shaped counter stern.

 

The other drawing: 500 dots.

 

Which completed connect-the-dots will more strongly resemble the Titanic's silhouette?

 

Translated to a DAC 'connecting the dots' to 'redraw' a  complex violin quartet, how does the DAC do it? In electronic terms?

 

Aren't there complex algorithms that prevent dots from being 'skipped'?  Or the wrong dots from being connected:

 

IE our Titanic has two realllllly wide funnels instead of four skinny ones, because the child(the DAC!), went across from the top of one funnel to the next one instead of back down to the ship's deck, essentially missing a few dots(samples).

 

Your "connect-the-dots" view of the world is incorrect but you are not the only who believes it.

 

 

Sometimes it's like someone took a knife, baby
Edgy and dull and cut a six inch valley
Through the middle of my skull

Link to comment
1 minute ago, The_K-Man said:

It's simple connect-the-dots, right?

Wrong. It's band-limited interpolation. If you simply connect the dots with straight lines (linear interpolation), you introduce frequencies above the Nyquist limit. To stay below the limit, a smooth curve passing through the points must be used. According to the sampling theorem, only one such curve is possible for a given set of sample points.

 

4 minutes ago, The_K-Man said:

Translated to a DAC 'connecting the dots' to 'redraw' a  complex violin quartet, how does the DAC do it? In electronic terms?

The DAC works by first connecting the dots in whatever crude manner, then removing the illegal frequencies this created. What remains is the correct waveform.

Link to comment
4 minutes ago, mansr said:

To stay below the limit, a smooth curve passing through the points must be used.

How does the DAC determine the radius of said smooth curve?

 

4 minutes ago, mansr said:

The DAC works by first connecting the dots in whatever crude manner, then removing the illegal frequencies this created. What remains is the correct waveform.

Juicy!  Mmmmhh!  🤤  Technical detail, yum!

 

So I suppose a good DAC has algorithms programmed in to it to "remove the illegal"(non-Nyquist) frequencies?

Link to comment
1 minute ago, The_K-Man said:

How does the DAC determine the radius of said smooth curve?

Again, the sampling theorem covers that.

 

1 minute ago, The_K-Man said:

Juicy!  Mmmmhh!  🤤  Technical detail, yum!

 

So I suppose a good DAC has algorithms programmed in to it to "remove the illegal"(non-Nyquist) frequencies?

Yes, it's called a low-pass filter. These days most DACs do part of this filtering digitally and the remainder using analogue circuits.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...