Vinyl > Digital on HowStuffWorks, Is their explanation incorrect? Options

[snoozer]

My friend won't give in that he just likes the motions of putting his records on his turntable but he found this website:

http://electronics.howstuffworks.com/question487.htm

This means that the waveforms from a vinyl recording can be much more accurate, and that can be heard in the richness of the sound.

From the graph above you can see that CD quality audio does not do a very good job of replicating the original signal. (at 16/44.1)

I referred him to this:

http://wiki.hydrogenaudio.org/index.php?title=Vinyl_Myths
(http://www.stevehoffman.tv/forums/showthread.php?t=85436)

I'm trying to understand the specifics but I'm a little lost. Are these two contradictory? Could someone clear up where HowStuffWorks might be disseminating wrong info?

This post has been edited by Peter: Feb 19 2009, 02:32

Reason for edit: Link now less search engine friendly.

[saratoga]

I'm trying to understand the specifics but I'm a little lost. Are these two contradictory? Could someone clear up where HowStuffWorks might be disseminating wrong info?

That site is completely wrong. I clicked the comment box ages ago and said something like "none of this is true" followed by some links but they don't really seem to care.

[Canar]

It is completely, horribly wrong, down to the graphic, which does not even attempt to linearly interpolate. That article was written by someone who does not have the slightest understanding of the technical nature of audio playback and reproduction.

Your friend may never come around, sorry to say. I've given up on many people. Some people don't want to listen to science.

[Axon]

Here's what I put down on the contact us page....

Your vinyl vs cd article, question487.htm, is completely wrong, and is misleading a lot of people.The waveform is a lie (CDs do not output square waves), vinyl recordings lose lots of information, etc. I will let as many people as possible know how incompetent the authorship is on this site, to ensure you don't mislead anybody else.

This is definitely worth a blog post and facebook remark/digg thumbs down/etc for anybody who can swing it. After some googling I can see this page is very influential.

[DVDdoug]

The waveform is a lie (CDs do not output square waves)...

Hopefully, most CD players don't output "square waves". But, two of my 3 computers at work do put-out completely unfiltered stair-stepped waves! And, they still sound better than vinyl to me. (I can't stand the "snap", "crackle", and "pop" from vinyl.)

On the oscilloscope, the waveforms from these 2 computers look exactly like the waveforms on "how stuff works", and exactly like the display in a wave editor. With a 1kHz sine wave, you can clearly see 48 steps per cycle, and with a 24kHz cosine wave, you get a nice square wave! (At a 48kHz sample rate.)

[AndyH-ha]

Perhaps these have exceptionally low cost soundcards. There isn't actually a DAC, they depend on the intrinsic low-pass filter quality of speaker coils to sort-of produce continuous waveforms from the pulses they output.

[greynol]

Well, there's gotta be a DAC, but it appears that it's missing a low-pass filter.

Even if your speaker does reproduce the higher-frequency images, your ears will do the necessary filtering since they also have a low-pass response.

[Axon]

If it really does look like that then it will have a significantly compromised high frequency response, so the issues involved are not merely theoretical.

Doug, can you please dig into your work computers and identify the chipsets?

[greynol]

If it really does look like that then it will have a significantly compromised high frequency response

I don't understand why you would say this. All that happens is the frequency response repeats itself every fs Hz to infinity. It's not like there's any aliasing of the signal or anything.

Or are you trying to say that speakers will have a hard time reproducing high frequencies when there is energy present at even higher frequencies?

[Axon]

Lavry's sampling paper goes over inherent attenuation in all sample-and-hold converters.

http://lavryengineering.com/documents/Sampling_Theory.pdf

[seangan]

There are two things you can tell your friend. Firstly, the graph is grossly misrepresentative particularly with regard to scale. The analogue waveform depicted would actually be wholly inaudible. A 16-bit ADC does indeed provide 65,536 possible values but this results in a signal-to-noise ratio of 98.09 dB which is pretty is negligible in comparison to the noise introduced by the process of mastering and cutting vinyl.

Ditto, for the sampling frequency, which can capture frequencies of up to 22kHz while the human ear can only hear up to 20kHz (when very young – most adults are unable to hear above 16 kHz). There’s probably no need to mention the need for low-pass filtering before the sampling stage as you don’t want to baffle them with too much information.

There are very good reasons why audio engineers picked out 16 bits and 44.1kHz. It’s simply because more audio information is beyond the capabilities of the human ear to detect. Even still, much of CD quality audio information is redundant as far as human ears are concerned – as the use of lossy compression techniques demonstrates.

Maybe, refer your friend to http://en.wikipedia.org/wiki/Analog-to-digital_converter , http://en.wikipedia.org/wiki/Sampling_frequency , and http://en.wikipedia.org/wiki/Quantization_noise where he/she can check the maths for themselves. They could also read http://en.wikipedia.org/wiki/Psychoacoustics and other articles and web pages linked to from those articles.

I’m an audiophile and I love vinyl but it’s not because the sound is better. I like the whole experience of sitting down and putting on the record, just sitting there listening to it and then getting up again to change to the other side. While digital media is far superior, it seems a little more disposable and less precious.

Other vinyl loving friends think the sound is better because they compared a record on their turntable with the same recording on a CD deck played through the same amplifier / speaker set-up. The vinyl sounded better to their ears but I explained that they weren’t actually comparing like with like as the turntable output passed through a pre-amp stage. I also explained that differences in the mastering processes didn’t mean that the vinyl carried a more faithful reproduction of the original.

This post has been edited by seangan: Feb 18 2009, 03:26

[greynol]

That's an informative paper, Axon.

Thanks!

This post has been edited by greynol: Feb 18 2009, 03:29

[2Bdecided]

Can't we break the link in the first post? I don't see why HA should help maintain the Google rank of this nonsense.

I notice that the article itself links on to another article that says 44.1/16 is audible indistinguishable from the analogue original. Different author.

If the first article is so bad that it causes people to visit the (advertising funded?) site to have a laugh, it's probably done its job for the site owner.

Cheers,
David.

[krabapple]

This is definitely worth a blog post and facebook remark/digg thumbs down/etc for anybody who can swing it. After some googling I can see this page is very influential.

It is. I've had it thrown at me more than a few times by vinylphiles on other forums. Another misleading favorite is Christine Tham's articles on LP vs CD at Audioholics. The editors there have been TOLD why that article is wrong, but they refuse to do anything -- they said critics should write and submit a better one.

This post has been edited by krabapple: Feb 18 2009, 23:13

[krabapple]

Perhaps these have exceptionally low cost soundcards. There isn't actually a DAC, they depend on the intrinsic low-pass filter quality of speaker coils to sort-of produce continuous waveforms from the pulses they output.

oh my god.

[krabapple]

I’m an audiophile and I love vinyl but it’s not because the sound is better. I like the whole experience of sitting down and putting on the record, just sitting there listening to it and then getting up again to change to the other side. While digital media is far superior, it seems a little more disposable and less precious.

Other vinyl loving friends think the sound is better because they compared a record on their turntable with the same recording on a CD deck played through the same amplifier / speaker set-up. The vinyl sounded better to their ears but I explained that they weren’t actually comparing like with like as the turntable output passed through a pre-amp stage. I also explained that differences in the mastering processes didn’t mean that the vinyl carried a more faithful reproduction of the original.

You, sir, are a mensch . (In case you aren't familiar with Yiddish, that's a good thing to be)

[Axon]

This is definitely worth a blog post and facebook remark/digg thumbs down/etc for anybody who can swing it. After some googling I can see this page is very influential.

It is. I've had it thrown at my more than a few times by vinylphiles on other forums. Another misleading favorite is Christine Tham's articles on LP vs CD at Audioholics. The editors there have been TOLD why that article is wrong, but they refuse to do anything -- they said critics should write and submit a better one.

This is indeed a compelling argument for Audioholics and HowStuffWorks being ad whores who have no qualms lying to make a buck/pay the bills/etc. I used to be more open-minded about this, but seriously - editors of a site have a responsibility for the content of the site, and if they can't get somebody to fix it, they should fix it themselves.

But David points out an important issue... linking to them is not a good thing. HA has a significant amount of PageRank cachet (google "vinyl myths" sometime - I didn't even try!) and referencing them directly is really, well, helping teh terrorists win, so to speak.

Mods: is there any way to implement this?

java script:document.location="http://hydrogenaudio.org/"

[DVDdoug]

Doug, can you please dig into your work computers and identify the chipsets?

As reported by Windows Device Manager:

"RealTek AC'97 Audio" (Motherboard soundchip)

"CMI 8738/C3DX PCI Audio Device" (Separate PCI soundcard)

And, if you'd like to see the waveforms, send me a PM with your email. (I have several captured waveform images.)


P.S.
Of course, these are cheap soundcards/soundchips. Audio has nothing to do with my job, and none of these computers are intended for audio use. Another one of my computers (also with a motherboard soundchip) does put-out sine waves. (I haven't tested my home computers.)

This post has been edited by DVDdoug: Feb 18 2009, 22:23

[Axon]

Those are some pretty common cards. I did not expect them to be that bad. The whole stairstep thing was believed to be a complete urban legend, so even though we can confidently say that these cards are inferior, it's a significant counterexample to the notion that virtually every sound card in existence does oversampling (or high order output filtering).

I was not able to find direct confirmation of the CMedia card having this issue, but judging from the frequency response plot on this page, the CMedia card tested certainly has the rolloff as if it was a sample-and-hold NRZ ADC without image filtering...

http://vychodil.sweb.cz/zvukovky.htm

[Arnold B. Kruege...]

Doug, can you please dig into your work computers and identify the chipsets?

As reported by Windows Device Manager:

"RealTek AC'97 Audio" (Motherboard soundchip)

Covers a wide range of chips. Most have adequate but not sparkling performance. Hard to tell exactly what's going on.

"CMI 8738/C3DX PCI Audio Device" (Separate PCI soundcard)

An oldie-moldie with poor performance. In some tests it almost seems like the guts have only 8 bit resolution.

[Dynamic]

Sorry to resurrect an oldish topic, but OMG!

I've just googled for...
Vinyl versus CD

Top link is to http://answers.google.com/answers/threadview?id=375592 - a Google Answers response which I didn't provide a proper link to for fear of improving its Google rating - and is completely misguided or patently false. All responses are pretty old (2004).

If you wanted to know the objectivist right answer, you're unlikely to search for Vinyl myths and would struggle to come across Hydrogen Audio's rational, sound reasoning (no pun intended) that CD is superior to Vinyl as a delivery format to human listening as long as one sticks to the rules.

The rules include pre-filtering to avoid aliasing below the Nyquist frequency, quantization with appropriate dither and reconstruction with appropriate reconstruction filtering (whether done by analogue filters or oversampling).

The google search...
LP vinyl compare CD
...isn't much better and the same Google Answers result is now second highest.

Is CD better than LP?
...is a poor search
Is CD better than vinyl?
...provides a better second link, Audio myth: Vinyl better than CD? | Audio DesignLine

A myth so popular is hard to supercede in terms of answer-finding, and SearchWiki appears to be pretty-much useless to warn people of false information.

From a blind testing perspective, it's possible make a demonstration by means of properly digitising a good Vinyl source, that the analogue sound and its digitial reproduction are indisinguishable to the human ear. I'm pretty sure Pio2001 spent a lot of time doing this and reporting his results on Hydrogen Audio some years ago. Therefore, even if pops and crackles, hiss etc are desired, they can be transparently reproduced on CD.

An example is to feed an analogue source of any quality we choose in case A, directly to the amplifier, or in case B, via a good properly filtered ADC and DAC at CD specifications (i.e. at some stage, the digital signal must be reduced to 44.1kHz, 16-bit per channel with proper dither) and then to the amplifier. Switching in (case B) and out (case A) the digital stage one can attempt to determine whether or not the digital stage is transparent (subject to negligible delay). If the switch position is kept secret to the experimental participants, we have an ABX test. We could even use a repeating loop of studio magnetic tape with a suitably high tape speed that should satisfy the most stringent analogue enthusiasts.

[Axon]

There was a blind test conducted by a bunch of Vinyl Asylum members a while back of vinyl vs CD-R, on a thoroughly redonkulous rig - the preamp alone cost $23,000. It's hard to figure out what's more ironic - that the one person who "passed" (6/6) was the lone skeptic who thought CD-Rs and vinyl are indistinguishable, or that the likely culprit in the audible difference was in the $23,000 preamp's profoundly poor performance.

http://db.audioasylum.com/cgi/m.mpl?forum=vinyl&n=667283

Moral: never let anybody tell you that you need merely a more expensive rig to hear the difference.

This post has been edited by Axon: Mar 3 2009, 22:01

[Knowzy]

Mods: is there any way to implement this?

java script:document.location="http://hydrogenaudio.org/"

A better way to ensure the search engines don't see a link on HA as an endorsement is adding a "nofollow" in the hyperlink. This way, people who don't have JavaScript or have disabled it can still visit the link.

Google and all major search engine don't count links with the nofollow attribute in their ranking algorithms. Nearly all links on Wikipedia are nofollow.

The HTML would look like this: <a rel="nofollow" href="http://www.HowWeWantYouToThinkStuffWorks.com">Don't rank this site</a>

Edit: Peter edited the OP to add nofollow attribute to the HowStuffWorks link. I tried doing this myself the "Toggle HTML Source" feature- second button from the left on the top row. It didn't work. It looks like only mods can add no follow.

This post has been edited by Knowzy: Mar 4 2009, 20:06