Vinyl Conversion

[Harpy]

I have some music converted from Vinyl to 24 bit digital. I was just wondering if anything was lost in the conversion process. I can hear the limitations of the vinyl recordings like mono bass with stereo subs. I did like the vinyl version of one album over it's Redbook CD counter part (AIFF).

I wasn't sure if there many computer audiophiles that owned a table and spun some vinyl from time to time and had done comparisons between the Lp and a digital copy of the Lp. I was wondering if the two sounded similar or if the Dac added it's own signature.

[Guidof]

I have some music converted from Vinyl to 24 bit digital. I was just wondering if anything was lost in the conversion process. I can hear the limitations of the vinyl recordings like mono bass with stereo subs. I did like the vinyl version of one album over it's Redbook CD counter part (AIFF).

I wasn't sure if there many computer audiophiles that owned a table and spun some vinyl from time to time and had done comparisons between the Lp and a digital copy of the Lp. I was wondering if the two sounded similar or if the Dac added it's own signature.

 

I have done that, and in my system the digitized LPs (at 16/48) sound slightly better than, or indistinguishable from, the original LP.

 

Regards,

 

Guido F.

[crisnee]

I have done that, and in my system the digitized LPs (at 16/48) sound slightly better than, or indistinguishable from, the original LP.

Regards,

Guido F.

 

Slightly better? That seems odd; what's your theory on that?

 

Chris

[Mike Mcsweeney]

Slightly better? That seems odd; what's your theory on that?

 

Chris

 

a lot of people like digitized music better...to me it always sounds "digitized" or unnatural or bright or as most purists say "not as warm"....but whatever floats your boat...i am about digital because of conveniency....when technology is able to digitize and convert back to same analog output, then i will be happy.

[esldude]

As a purely subjective evaluation, I think 44/24 conversion is pretty good, but not quite there. But 48/24 simply leaves nothing. You don't need more than that. That could be down to the particulars of the equipment I have used. I have a friend who insists on 88/24. Certainly 88/24 or 96/24 should leave nothing wanting.

 

I haven't done 48/16 in some time. It too may be enough. But you can do 48/24 then do RIAA digitally and you are into some of the finest sound possible from vinyl. I would caution that you need to record with speaker playing at full volume. My friend was unhappy with his vinyl conversions. He did them with speaker off assuming cleaner was better. But the acoustic feedback is part of the experience. I had him play music with his TT sitting on a LP not spinning while we recorded the output. He could hear the acoustic feedback. Once he recorded while playing including the acoustic feedback he found the results indistinguishable from the LP itself. This was with some high quality gear throughout the LP chain as well.

 

Another oddity is FM conversion. I found 44/24 just a bit lacking. But 48/24 super good. As in an FM tuner being converted to digital and played back. I once used tube McIntosh tuners. Another obviously colored but enjoyable tuner is a Fisher 50B which I have and digitize to this day. It isn't transparent, but it fun to listen to I assure you. As FM has nothing beyond 15 khz it shouldn't matter, but seems like it does. My other tuner (also digitized for playback) is a late 80's Denon which seems very transparent. When listening to blues or orchestra from public radio stations though it is the Fisher that gets the nod. Got that devil for $6 at a yard sale, and put all new tubes garnered from a hamfest into it. Lots of musical enjoyment for around $50 total.

[esldude]

a lot of people like digitized music better...to me it always sounds "digitized" or unnatural or bright or as most purists say "not as warm"....but whatever floats your boat...i am about digital because of conveniency....when technology is able to digitize and convert back to same analog output, then i will be happy.

 

That day has been here for awhile my friend. If it isn't you are using lousy digital equipment. AD/DA is darned transparent. If it sounds bright look to the source. Digital by itself is not at all bright or unnatural.

[Mike Mcsweeney]

That day has been here for awhile my friend. If it isn't you are using lousy digital equipment. AD/DA is darned transparent. If it sounds bright look to the source. Digital by itself is not at all bright or unnatural.

 

The truth is that an analog signal is one continuous sign wave, whereas digital has breaks,but is "stitched" back together with different algorithms by programmers each thinking their algorithm is more correct or accurate than the other.... They will never sound 100% the same...the debate will go on forever. I am not going to say one sounds better than the other, just that they will sound different.

 

HowStuffWorks "Does digital sound better than analog?"

[kennyb123]

I have done that, and in my system the digitized LPs (at 16/48) sound slightly better than, or indistinguishable from, the original LP.

 

I recorded a dozen or so of my better sounding LPs in DSD using a friend's Korg MR2000s. This probably captured about 90%. I say probably because the fault was most likely with my playback device. My Mytek DAC is not on par with my phono stage. It would take a better DAC for me to gain more insight into how good of a job the Korg really did as an A/D recorder.

[esldude]

The truth is that an analog signal is one continuous sign wave, whereas digital has breaks,but is "stitched" back together with different algorithms by programmers each thinking their algorithm is more correct or accurate than the other.... They will never sound 100% the same...the debate will go on forever. I am not going to say one sounds better than the other, just that they will sound different.

 

HowStuffWorks "Does digital sound better than analog?"

 

Mike, I am sorry, but no this is not nor has it ever been true of digital. It has no breaks, it puts out a fully continuous sine wave or whatever waveform is supposed to be output. Your 'truth' that digital has breaks is simply not true. The article you link to is garbage. The whole bit about whistling and in early digital it would get broken up is someone's imagination.

 

I am not trying to be unfriendly, but repeating mistaken ideas about how something works is not helping anything. One can prefer either, but at least be truthful in saying how it works. If you really want to learn the basics of how digital works watch this highly informative video. It is 24 minutes, but very much worth your time.

 

Xiph.Org Video Presentations: Digital Show & Tell

 

Among other things he will use analog signal generators which get sampled digitally and reproduced then sent to an analog scope where you can see, contrary to what you have said that in fact the waves are fully continuous after an AD/DA conversion stage in the middle.

 

BTW, the reason I say digital doesn't sound bright, is recording LP's. They aren't bright and neither are the digital recordings of them. That alone should show digital is not inherently bright. Some CD's of older material are bright. There are many possible reasons, but just being digital isn't one of them.

[PopPop]

+1 To what esldude said! The difference can be nearly transparent but it depends on the quality of the components in playback system. In the case of ripping vinyl, the TT, tone arm, cartridge, and interconnects probably are more important to the process than the sample rate. After that, the digital conversion should be done with gear that is designed to give you what you are looking for. If you want to listen to files on your old iPod while you are out jogging it doesn't matter much what you use for ripping..

[Mike Mcsweeney]

Mike, I am sorry, but no this is not nor has it ever been true of digital. It has no breaks, it puts out a fully continuous sine wave or whatever waveform is supposed to be output.

 

That is the funniest thing I have ever heard. A digital signal puts out a fully continuous sine wave (grin)...it is only AFTER the conversion back to analog that you have a sine wave...and the method it does it by is a mathematical formula to "guesstimate" what the original sine wave would look like.

 

I am no expert at either audio or electronics, but i do know some "FUNDAMENTALS" that you clearly don't.

[PopPop]

OK then, popcorn anyone?

[esldude]

That is the funniest thing I have ever heard. A digital signal puts out a fully continuous sine wave (grin)...it is only AFTER the conversion back to analog that you have a sine wave...and the method it does it by is a mathematical formula to "guesstimate" what the original sine wave would look like.

 

I am no expert at either audio or electronics, but i do know some "FUNDAMENTALS" that you clearly don't.

 

First you proclaim you are no expert, and then use that status to make a claim. Cute. Stay ignorant or learn. You are wrong about this. You could either view the video or do the experiment yourself. Those digital waveforms are not guessed at, they are reconstructed. Yes, they used mathematical principles to develop and build digital systems that work. So what if the result is the appropriately reconstructed analog wave which it is?

[Guidof]

Slightly better? That seems odd; what's your theory on that?

 

Chris

 

Yes, I thought it was odd too.

 

I have no theory. But note two things: the equipment chain when playing an LP is different from the chain playing the digitized version, and (more likely a cause) there are no cartridge microphonics due to acoustic feedback when a digitized LP is played back. Also note that, on my system at least, the effect is proportional to the dynamics of the original LP: records with the wider dynamics tend to sound just a tad better (cleaner) when digitized.

 

Regards,

 

Guido F.

[firedog]

That is the funniest thing I have ever heard. A digital signal puts out a fully continuous sine wave (grin)...it is only AFTER the conversion back to analog that you have a sine wave...and the method it does it by is a mathematical formula to "guesstimate" what the original sine wave would look like.

 

I am no expert at either audio or electronics, but i do know some "FUNDAMENTALS" that you clearly don't.

 

Mike, again, apparently you don't understand the fundamentals. There is no break in a digital signal, and the "stepped" representation you often see is an incorrect oversimplification. The mathematics involved aren't a "guesstimate"; it can be proven that no information is lost.

 

If digital doesn't sound right, it has to do with filtering and the quality of the equipment doing the reproduction, not digital itself.

 

When you record an LP, even at 16/44.1, you don't hear a 'digital' sound on playback, you hear something that sounds like an analogue LP, with all those aspects analog lovers say belong only to analog.

 

Analog recording is actually not truly continuous - it misses all sorts of information because of its inherent limitations. That's what analog advocates don't get.

[Mike Mcsweeney]

First you proclaim you are no expert, and then use that status to make a claim. Cute. Stay ignorant or learn. You are wrong about this. You could either view the video or do the experiment yourself. Those digital waveforms are not guessed at, they are reconstructed. Yes, they used mathematical principles to develop and build digital systems that work. So what if the result is the appropriately reconstructed analog wave which it is?

 

Who is wrong, he or me that claims a digital signal is a sine wave?

 

The fact is that there are different algorithms that produce different outputs.

The fact is a digital signal is not a sine wave.

The fact is that the output will never be 100% the same as the original analog signal, it can only get more accurate with a higher sampling rate. I am not going to argue what is audible and what is not...but facts are facts.

[Mike Mcsweeney]

Mike, again, apparently you don't understand the fundamentals. There is no break in a digital signal, and the "stepped" representation you often see is an incorrect oversimplification. The mathematics involved aren't a "guesstimate"; it can be proven that no information is lost.

 

If digital doesn't sound right, it has to do with filtering and the quality of the equipment doing the reproduction, not digital itself.

 

When you record an LP, even at 16/44.1, you don't hear a 'digital' sound on playback, you hear something that sounds like an analogue LP, with all those aspects analog lovers say belong only to analog.

 

you are wrong.

You cannot play digital..it must be converted to analog...you hear something that sounds like analog, because IT IS analog. That is the entire purpose of the DAC.

draw me as many square waves as necessary to create a perfect sine wave.

digits (or ones and zeros) are either on or off. analog varies by amplitude..this is as basic and fundamental as it can get.

Ask anyone that knows logic and electronics and they will set you straight.

 

Show me what a 20hz sine wave looks like, then show me what a 20hz digital signal looks like. The digital signal has to be converted to analog before it can even be played.

 

If algorithms aren't an estimate, but a perfect science, there wouldn't be so many different algorithms. The fact is, they can only estimate based on the samples. If there was dead silence for one pico second and there wasn't a sampling during that pico second, the silence would be lost in the output, because the algorithm would fail. Again, i am not going to say what is audible and what is not, but it is not a perfect reproduction, and never can be.

 

Also, why do you think technology is still changing the sampling rate....why are millions of dollars being spent on 5.6mhz sampling rate engineering today...why not just be satisfied with your 44.1khz if there isnt a difference. Whatever....

[PopPop]

Who is wrong, he or me that claims a digital signal is a sine wave?

 

The fact is that there are different algorithms that produce different outputs.

The fact is a digital signal is not a sine wave.

The fact is that the output will never be 100% the same as the original analog signal, it can only get more accurate with a higher sampling rate. I am not going to argue what is audible and what is not...but facts are facts.

 

 

An audio signal is quite complex, oversimplified demonstrations of pure tones can mislead your understanding. Yes, it's true that an insufficient sample rate can result in lost information and sound degradation but would be for rates far below 44.1khz. But at some point increasing the sample only creates more data and doesn't improve the conversion process. As I understand it, the sample rate need only be sufficient to capture the highest frequencies in the source material along with the transitions between alternate sounds.

 

EDIT: I buy high resolution files because my experience has told me that in many cases they provide superior sound. But not always.

[Teresa]

When you record an LP, even at 16/44.1, you don't hear a 'digital' sound on playback, you hear something that sounds like an analogue LP, with all those aspects analog lovers say belong only to analog.

 

LPs recorded at 16/44.1 sound "digital" to me, it's sort of the worst of both worlds. LPs at 16/44.1 WAV do sound better to me than CDs, but that is not saying much. At 24/96 I get about 80% of the resolution and color of the LP using WAV, slightly less if saved as FLAC or ALAC. Even at 24/96 I can't capture all the ambiance, warmth and high frequency sparkle of a well made audiophile LP with good insolation from its environment played through a pure analog signal path. Perhaps you should have said "I" instead of "You"

 

Also when I record an LP I use headphones and have the output to the speakers turned off, to my ears reel to reel mastering tape at 7 1/2 ips or even better at 15 ips does a more accurate job of capturing the resolution of an LP, but new good quality reel to reel tape is just too damn expensive. Computers don't give me perfect LP playback but I prefer it for the convenience.

 

Analog recording is actually not truly continuous - it misses all sorts of information because of its inherent limitations. That's what analog advocates don't get.

 

Analog actually is a continuous musical waveform whereas digital are 1's and 0's which need to be reconstructed to analog to be a musical waveform again and be heard as music.

 

However, we understand this analog continuous musical waveform is not perfect. In the case of tape the better the tape (more particles per inch), faster speed, wider track width, etc. the more accurately this continuous waveform is captured.

[Mike Mcsweeney]

An audio signal is quite complex, oversimplified demonstrations of pure tones can mislead your understanding. Yes, it's true that an insufficient sample rate can result in lost information and sound degradation but would be for rates far below 44.1khz. But at some point increasing the sample only creates more data and doesn't improve the conversion process. As I understand it, the sample rate need only be sufficient to capture the highest frequencies in the source material along with the transitions between alternate sounds.

 

EDIT: I buy high resolution files because my experience has told me that in many cases they provide superior sound. But not always.

 

I am not going to argue which "sounds better" or what is "audible". That debate will never die, so there is no reason to debate it. I am into digital for convenience and because that is the method in which files are stored on computers, but the facts i outlined above are facts, that no one "in the know" can argue.

1. You cannot play a digital file

2. It must be converted to an analog signal

3. Different algorithms and different sampling rates will produce different outputs (hence why it is so subjective as to what sounds better)

4. It is impossible to reproduce the original sine wave perfectly (whether it can be reproduced as accurately as audible, will remain a debate).

5. The method to reproduce the original sine wave is a calculation (algorithm) based on the samples. It is an estimate and is probably pretty darn accurate most of the time....but you will find ad's that state their algorithm is the best for accuracy.

 

Why the heck do you think it is so subjective (grin).

 

DSD is here now, and there will be something better in the future.....as long as you try to stay on top of technology, you will pay for that technology...wait and reap the rewards of yesterday's best technology....

[Teresa]

...I buy high resolution files because my experience has told me that in many cases they provide superior sound. But not always.

 

I also have been disappointed with poor "digital" sound of some SACDs and downloads sold as high resolution. Before I was into computer music I guessed these harsh sounding SACDs were from 16/44.1 masters or converted to 16/44.1 at some point in mastering.

 

Since using a computer for music I have discovered most of the poor sounding so-called high resolution downloads have a brickwall filter with the high frequencies dropping hard like a ton of bricks from 20-22kHz using Audacity software.

 

My solution may not work for you, I now buy only high resolution downloads and physical formats (SACD, DVD-Audio, LPs) from audiophile labels, either original high resolution recordings or remasters of commercial recordings. It is the commercial companies trying to pass off low resolution digital as high resolution so I avoid them.

[PopPop]

I am not going to argue which "sounds better" or what is "audible". That debate will never die, so there is no reason to debate it. I am into digital for convenience and because that is the method in which files are stored on computers, but the facts i outlined above are facts, that no one "in the know" can argue.

1. You cannot play a digital file

2. It must be converted to an analog signal

3. Different algorithms and different sampling rates will produce different outputs (hence why it is so subjective as to what sounds better)

4. It is impossible to reproduce the original sine wave perfectly (whether it can be reproduced as accurately as audible, will remain a debate).

5. The method to reproduce the original sine wave is a calculation (algorithm) based on the samples. It is an estimate and is probably pretty darn accurate most of the time....but you will find ad's that state their algorithm is the best for accuracy.

 

Why the heck do you think it is so subjective (grin).

 

 

DSD is here now, and there will be something better in the future.....as long as you try to stay on top of technology, you will pay for that technology...wait and reap the rewards of yesterday's best technology....

 

So what's your point (grin)

 

As to your item 4. I's not all so subjective. The sinuosity of a wave form for e natural, played on an open violin string is quite different than that of the same note played on the 4th string in a higher position. It is possible to create an algorithm that will be up to the task in theory but perhaps not practical. Designing electronics that represent the correct balance between perfect and practical is entirely dependent on ones ability to sense the difference.

 

Have you done any reading about the sampling of sinusoidal functions?

[Don Hills]

...

If algorithms aren't an estimate, but a perfect science, there wouldn't be so many different algorithms. The fact is, they can only estimate based on the samples. If there was dead silence for one pico second and there wasn't a sampling during that pico second, the silence would be lost in the output, because the algorithm would fail. Again, i am not going to say what is audible and what is not, but it is not a perfect reproduction, and never can be. ...

 

If there was dead silence for one pico second, it would require a sampling rate in the terahertz range to correctly capture. This is not an "algorithm failure". Not even the most obsessive audiophile would argue that audio extends into the terahertz range.

 

Algorithms for digital audio are well understood, and are indeed a "perfect science". You do get exactly the result the algorithms predict. The alogrithm for reconstructing the output of a D/A converter is the sinc function. The problem is that for perfect accuracy you need an infinite series. You'd never get a result. So in practice you have to stop calculating and produce a somewhat less than perfect result. It's like using a calculator - you have to decide how many digits of accuracy are enough for your purposes. DAC filter designers have to strike a balance between the accuracy of their calculations and the amount of resources (processing power, chip complexity etc) available. In most cases, they push the resulting error well below the noise level. Remember this:

Analogue signal --> filter to Nyquist --> Add dither --> A/D --> D/A -- Reconstruct --> Listen

... is theoretically identical to:

Analogue signal --> filter to Nyquist --> Add dither --> Listen

So long as the errors from less than perfect A/D - DA - Reconstruct are well below the dither, and this in turn is well below audibility, this is generally considered to be "good enough". If the resulting error signal is lower than the noise caused by Brownian motion of the air molecules hitting your eardrums, you're unlikely to hear it.

[Mike Mcsweeney]

YOU>>>The problem is that for perfect accuracy you need an infinite series.

 

My exact point. I said i wasn't going to argue what is audible and what wasnt.

You agreed with my fact, that you cannot have 100% accuracy.

 

Do you deny my other facts, that

1. A digital signal must be converted to analog before you can play it.

 

2. It is impossible to reproduce the original sine wave perfectly (whether it can be reproduced as accurately as audible, will remain a debate).

 

3. The method to reproduce the original sine wave is a calculation (algorithm) based on the samples. It is an estimate and is probably pretty darn accurate most of the time....but you will find ad's that state their algorithm is the best for accuracy.

 

I think the only thing you and i disagree on is the algorithms used are a "perfect science" to me that is a very funny statement.

 

And my MAIN point was the OP that suggestion that a digital output is the same as an analog output...or that there aren't breaks in a digital signal....snicker...

[Mike Mcsweeney]

So what's your point (grin)

Designing electronics that represent the correct balance between perfect and practical is entirely dependent on ones ability to sense the difference.

sounds subjective to me....

as far as what i have read, no. i know what i know and that is enough for me..i have better things to waste my time on. the facts that i do know is as deep as i will ever care to go into it. I have no desire to become someone who can debate or present facts on the topic beyond the fundamental facts that I am already aware of...you will need to find someone else to debate with. I just know that once an analog signal is digitized it will never be reproduced exactly to it's original waveform....just as OP above stated, that would require infinite sampling...to which i agree. Again, i am not going to debate what is audible and what is not, as that debate can never be won....so read as much as you want, so you can win others over...you won't win me (wink).