LP with better dynamic range than digital

I read Jim LeSurf's article and I have no idea where his formulas come from and what they mean. But I love a good challenge... so let's settle this once and for all ; wikipedia knows it all http://en.wikipedia.org/wiki/Comparison_of_analog_and_digital_recording

The dynamic range of digital audio systems can exceed that of analog audio systems. Typically, a 16 bit analog-to-digital converter may have a dynamic range of between 90 to 95 dB (Metzler 2005:132), whereas the signal-to-noise ratio (roughly the equivalent of dynamic range, noting the absence of quantization noise but presence of tape hiss) of a professional reel-to-reel 1/4 inch tape recorder would be between 60 and 70 dB at the recorder's rated output (Metzler 2005:111).
The benefits of using digital recorders with greater than 16 bit accuracy can be applied to the 16 bits of audio CD. Stuart (n.d.:3) stresses that with the correct dither, the resolution of a digital system is theoretically infinite, and that it is possible, for example, to resolve sounds at -110 dB (below digital full-scale) in a well-designed 16 bit channel.

Click to expand...

Unfortunately, you will have to buy Metzler's book http://www.scribd.com/doc/50438292/Audio-Measurement-Handbook-2nd-ed-2005-Bob-Metzler - he works (worked?) for Audio Precision. But notice, Keith Johnson said a "direct-cut LP" has a high dynamic range.

EDIT: Found a free copy of the book http://www.uldis.info/jvc/Audio_Measurement_Handbook.pdf

... and see what you make of all the data here http://www.audioholics.com/audio-technologies/dynamic-comparison-of-lps-vs-cds-part-4

[h=3] Conclusions
[/h] It appears that the vinylphile claims are not as outrageous as they seem: LPs do have a usable dynamic range far greater than the measured dynamic range would suggest, and LPs consistently have higher relative dynamics over digital formats. But it is also true that LPs have higher distortion levels which translate to ultrasonic frequency harmonics.

ack said:

But notice, Keith Johnson said a "direct-cut LP" has a high dynamic range.[/url]

Click to expand...

Makes sense to me. Some of the D2D records I have are very dynamic. The most dynamic recordings I have in fact.

tomelex said:

Yes, but we are talking real world here where the needle is large, very large in respect to the digital chunks in the vinyl, and thus like I said, unless you have a laser that can read the peak chunk and the deepest valley chunk, the needle will average all this, as in lost information, and then, in the real world, the needle will add a whole lot of self distortion which is quite pleasing to many, including me, but not for the low end of LP, where distortions are quite audible to me. The upper range is where LP does its magic in its distortion mix for plain old stereo.

Click to expand...

My point was simply that LeSurf's article has not much to say about whether or not LP has better dynamic range than digital unless one can determine with some precision what is the molecular size of vinyl and what is the relevance of his other simplifying assumptions (an impossible task). Distortion is yet another topic.

For what it matters, I am in the camp of those who subjectively find vinyl to have better dynamic range than digital.

ack said:

... and see what you make of all the data here http://www.audioholics.com/audio-technologies/dynamic-comparison-of-lps-vs-cds-part-4

[h=3] Conclusions
[/h] It appears that the vinylphile claims are not as outrageous as they seem: LPs do have a usable dynamic range far greater than the measured dynamic range would suggest, and LPs consistently have higher relative dynamics over digital formats. But it is also true that LPs have higher distortion levels which translate to ultrasonic frequency harmonics.

Click to expand...

And what is really of interest is that the investigation used retail mass produced 2nd hand LPs; also is when he also shows the real world noise floor analysis of the recording as CD and LP.

As a couple of us hinted to earlier, many would be very surprised just how good the retained test LP batch and true-best early white label pressed batch are in terms of noise....
One reason I was eventually put off LPs was that unfortunately I had a mix of standard retail and these "best" pressed LPs (that include mainstream genre such as Prince), and for me there is no comparison and was eventually put off my standard LPs; this was going back to 80s.
The generation and batch involved in terms of mother-daughter plating and iteration/pressing/cooling time/etc does have a notable effect on LP noise-warping-etc.
Much arguments against LPs does not take into consideration this variability on its performance and finding the best LPs, which was one of the positives with digital mass production.

Another aspect from what I understand also comes down to cost and that being how often the mastering disk lathe cutting head was retooled/sharpened in the studio; it is surprising just how few master sides can be done optimally before requiring retooling-sharpening.

Cheers
Orb

Orb said:

And what is really of interest is that the investigation used retail mass produced 2nd hand LPs; also is when he also shows the real world noise floor analysis of the recording as CD and LP.

Click to expand...

Much more than that, the more interesting thing is the relatively average analog equipment (by today's high end standards) he used: Dynavector DV-20xL cartridge, Rega P3 turntable, Dynavector P-75 phono stage. In other words, a relatively noisy table, a cartridge not at the top of the heap in dynamics, and a phono whose noise characteristics are unknown to me.

Therefore, I would contend that, today, with a high end untra-quiet table, a very dynamic cartridge, a modern quiet and more phono, and quiet phono cables, a repeat of his tests and measurements would probably result in more convincing evidence and "data" of the LP's dynamic capabilities, and probably even more so if one used a direct-cut LP (e.g. Sheffield, of which I have a few). Or as said earlier, just visit someone with worthy analog and digital set ups and take a simple audition.

People

The point is not that a LP are not dynamics. heck! 50 dB is plenty dynamic in any sane person's book . Few CDs or LPs achieve that anyway. You would take any highly dynamic LPs or CDs and the measurements will fall in this neighborhood. The point I have insisted on and that is being sidestepped is the claim that LP has intrinsically, dynamic range superior to that of Redbook CD. That remains to be proven . Be it at the pressing stage or the final record. The physics do not seem to match that claim. I could be wrong. But so far no one has proven me to be. Aside from the RR LPs which by the way I will make sure of listening personally on my rig compared to the CD or their HRx (acquiring them as soon as possible if need be) . I will take Reference Recording words for it and assume they came from the same master but ended up on different media. To repeat: Aside from RR I have seen people posting examples of very dynamic LPs. I could counter with example of extremely dynamic CDs and we will fall on subjective impression of mine is more dynamics than yours.

Once a claim is made and in numbers. We need more than subjective impressions we need proofs. The claim was 120 dB I lowered it to 100 dB and would not even mind going down to 80 dB which is already an extraordinary number that few systems would attain anyway, especially with SPL-limited speakers, low (relative) power amplifier and noisy environments. Let's be simplistic and throw these numbers: if we assume a quiet environment of 40 db ( in a very quiet Living Room) 80 dB of dynamic means 120 dB at the listening position !! Assuming 10 ft listening position ( ~ 3meters) from the plane of the speakers that would require about 600 Watts per channel ...

FrantzM said:

Aside from the RR LPs which by the way I will make sure of listening personally on my rig compared to the CD or their HRx (acquiring them as soon as possible if need be) . I will take Reference Recording words for it and assume they came from the same master but ended up on different media.

Click to expand...

Keep in mind that not all the RR LPs are sourced from analog tape and are instead from 176/24 digital files. In order to test the claimed superior dynamic range of analog, you should select LPs cut from analog tape.

FrantzM said:

The point I have insisted on and that is being sidestepped is the claim that LP has intrinsically, dynamic range superior to that of Redbook CD. That remains to be proven

Click to expand...

Did you find the data in the audioholics article unconvincing? When he shows peak amplitude (relative to 0VU reference I assumed) and average RMS power (both in dB) to be higher in LP than CD (pages 2 and 3), plus LP noise lower than CD in the vast majority of the spectrum, what do you conclude? Or am I missing something?

esldude said:

For what it is worth, in one $10k TT setup a couple years ago, I played music and measured acoustical feedback coming from the phono pre. With normal listening levels it was somewhere around -60 db or so of upper levels on discs. This was with a tubed phono and replacing it with a SS phono stage lowered the amount by 3 or 4 db. Meaning the amount of acoustical feedback was roughly the same for the tubes and the physical turntable. With a good reference level difficult to determine it seemed acoustical feedback was near or just barely above surface noise of some of the cleaner vinyl on hand. So not a certain measure of noise, but something of a data point.

Click to expand...

Hence the reason there are so many anti-vibration products- reduction of resonance and vibration is one of the bigger things that has happened in the last 20 years in audio. I use a custom Sound Anchors equipment stand, which allows the use of a pair of anti-vibration platforms, massive enough to sink vibration of anything on top of them. One is for the 'table and the other for the preamp (which has vibration isolation technology of its own). The difference of the stand and platforms was both measurable and instantly audible.

tomelex said:

Guys, I just used the term ball, as there are plenty of ideas of how a stylus should be cut, as seen in Atmaspheres link, which should be studied some by the newcomers to the hobby. I also appreciate that we can record and playback more than 20Khz on vinyl, but FR is not dynamic range, which is what we are discussing here. All one has to do is look at the fine measurments shown in Hi-Fi News for cartridges to see the measured distortions produced which add up to a pleasing rendition of poor old stereo. Look at September 2014 article on the $10K Techdas TDC01 TI for some information, facts, for example.

http://www.absolutesounds.com/pdf/main/press/TDC01Ti_HFN_Sept2014.pdf

Click to expand...

FWIW you used the word 'resolution' and bandwidth is part of that. The article you linked does state a value of +12db so we can safely add 10 db to the figure I was saying earlier. That puts the LP with more range than a Redbook CD: conservatively 100db. 4db is not a lot I know, and in practice does not matter a hoot since compression is the name of the game anyway.

Orb said:

And what is really of interest is that the investigation used retail mass produced 2nd hand LPs; also is when he also shows the real world noise floor analysis of the recording as CD and LP.

As a couple of us hinted to earlier, many would be very surprised just how good the retained test LP batch and true-best early white label pressed batch are in terms of noise....
One reason I was eventually put off LPs was that unfortunately I had a mix of standard retail and these "best" pressed LPs (that include mainstream genre such as Prince), and for me there is no comparison and was eventually put off my standard LPs; this was going back to 80s.

Click to expand...

FWIW these days all LPs are special limited editions! Its not like it was the 80s...

FrantzM said:

Once a claim is made and in numbers. We need more than subjective impressions we need proofs. The claim was 120 dB I lowered it to 100 dB and would not even mind going down to 80 dB which is already an extraordinary number that few systems would attain anyway, especially with SPL-limited speakers, low (relative) power amplifier and noisy environments. Let's be simplistic and throw these numbers: if we assume a quiet environment of 40 db ( in a very quiet Living Room) 80 dB of dynamic means 120 dB at the listening position !! Assuming 10 ft listening position ( ~ 3meters) from the plane of the speakers that would require about 600 Watts per channel ...

Click to expand...

If you missed my last post is was right after one of yours. In that post (http://www.whatsbestforum.com/showt...e-than-digital&p=291096&viewfull=1#post291096) we saw that 90 db is possible, now add 10 db to that. How? please see the article linked by Tomelex quoted above- in it we see that the cartridge is just beginning to mistrack at +12 db, so I am calling that 10db to be safe.

ack said:

Did you find the data in the audioholics article unconvincing? When he shows peak amplitude (relative to 0VU reference I assumed) and average RMS power (both in dB) to be higher in LP than CD (pages 2 and 3), plus LP noise lower than CD in the vast majority of the spectrum, what do you conclude? Or am I missing something?

Click to expand...

What your missing are the fine points of interpreting FFT plots.

Firstly the Max-min RMS power was lower for the LP in that article. He then shows FFT plots of the noise in silent portions for two recordings. The LP shows lower noise if you ignore the 500 hz and below portion. But you shouldn't get the idea that is 105 -108 db noise levels from that plot are what we are looking at.

Lots of people know this, but for any who don't a simple explanation.

If we take total RMS power over 20-20khz we get one number. If the spectrum were even (like white noise), splitting that into two bands the number we get is -3 db lower. If we split that into 4 bands it is another 3 db lower and 6 db lower versus the entire band. FFT's split the spectrum up into many bands. This lets you look at narrow slices of spectrum and see signals of lower level as it reduces the noise level versus looking at the entire band.

Now in the article we don't know what size FFT was used as it doesn't tell us. But as a for instance a 4096 point FFT of -80 db white noise (which we know CD is able to do) you would see a graph bouncing around -113 db. If this were a 16k point FFT it would bounce around -119 db. So the upper frequency portions in the article showing -108 db are impossible to interpret without knowing the FFT size.

Needle drops I have on file with good clean quiet vinyl show around 96-100 db with 16k point FFT's in those upper frequencies. So he could be using 32k or 64 k FFT plots or he could be using 16k and have some quieter vinyl. In any case, the CD he has for comparison isn't exercising the noise floor of the CD medium. We of course don't know the way in which the CD was remastered vs the LP or why it would have in some areas a higher noise floor. But that CD noise floor isn't likely anywhere close to the limit of the medium. Could be room noise, tape noise or who knows. We also don't know that these two LP's represent the state of art vinyl noise figures. But for these reasons the info in the article isn't good data showing LP can be effectively as quiet as CD. Only that in these cases the resulting recording in some ways is in some frequency ranges.

Atmasphere said:

If you missed my last post is was right after one of yours. In that post (http://www.whatsbestforum.com/showt...e-than-digital&p=291096&viewfull=1#post291096) we saw that 90 db is possible, now add 10 db to that. How? please see the article linked by Tomelex quoted above- in it we see that the cartridge is just beginning to mistrack at +12 db, so I am calling that 10db to be safe.

So Frantz, there is your 100db.
​

Click to expand...

Which would all be great if you weren't talking about lacquers rather than the available vinyl.

Atmasphere said:

Hence the reason there are so many anti-vibration products- reduction of resonance and vibration is one of the bigger things that has happened in the last 20 years in audio. I use a custom Sound Anchors equipment stand, which allows the use of a pair of anti-vibration platforms, massive enough to sink vibration of anything on top of them. One is for the 'table and the other for the preamp (which has vibration isolation technology of its own). The difference of the stand and platforms was both measurable and instantly audible.

Click to expand...

Yes, well the TT I measured this on also had custom stands and equipment to reduce and sink vibration. The effects were also measurable, and would have been worse without them. Whether it was best available, just okay or barely helpful I don't know. But it wasn't a naked TT and preamp sitting on a bookshelf directly behind a speaker.

esldude said:

The LP shows lower noise if you ignore the 500 hz and below portion. But you shouldn't get the idea that is 105 -108 db noise levels from that plot are what we are looking at.

Click to expand...

I didn't make a reference to 105 or 108dB; ONLY that LP is lower than CD, over 500Hz (and possibly under 500Hz if he had a better turntable).

esldude said:

We also don't know that these two LP's represent the state of art vinyl noise figures.

Click to expand...

Actually, that's the beauty of it - he pulled seemingly average LPs, nothing special, to do the research (or as Orb put it, "retail mass produced 2nd hand LPs"). I can imagine what he would have found had he be using mint "audiophile" pressings.

esldude said:

But for these reasons the info in the article isn't good data showing LP can be effectively as quiet as CD. Only that in these cases the resulting recording in some ways is in some frequency ranges.

Click to expand...

I would disagree.

Atmasphere said:

....
FWIW these days all LPs are special limited editions! Its not like it was the 80s...

Click to expand...

I would say yes and no tbh, still plenty of modern releases that do not match the best test LP and best white label batch from the 80s.
I agree in principal it should be better these days due to not being mass production, but there are plenty of room for variables to still influence performance and I have read quite a few examples of listeners/reviewers/those involved in the manufacturing/pressing plant industry aspect from 80s and still now being rather critical of a fair number of modern LPs in this regard (I agree there are also stunning releases).
So it is fair to say not all modern/existing LP mastering and manufacturing plants are equal.

Cheers
Orb

esldude said:

Which would all be great if you weren't talking about lacquers rather than the available vinyl.

Click to expand...

Actually I was referring to vinyl. But to be clear: what is possible, not what one gets all the time.

Orb said:

I would say yes and no tbh, still plenty of modern releases that do not match the best test LP and best white label batch from the 80s.
I agree in principal it should be better these days due to not being mass production, but there are plenty of room for variables to still influence performance and I have read quite a few examples of listeners/reviewers/those involved in the manufacturing/pressing plant industry aspect from 80s and still now being rather critical of a fair number of modern LPs in this regard (I agree there are also stunning releases).
So it is fair to say not all modern/existing LP mastering and manufacturing plants are equal.

Cheers
Orb

Click to expand...

In bold: Ain't **that** the truth!

Watch Keith Johnson talk about dynamic range, including RBCD's, starting at 19:00


He claims we hear into the noise floor around 10dB at ~32:30

ack said:

Watch Keith Johnson talk about dynamic range, including RBCD's, starting at 19:00



He claims we hear into the noise floor around 10dB at ~32:30

Click to expand...

http://ethanwiner.com/audibility.html#part2

In part 2 of this page are some wav files to let you listen to tones and speech below the noise floor at -10, -20 and -30 db.

The reason the ear hears into noise is not unlike a filterbank. Like in the FFT, the ear works like maybe a 25 or 30 filter bank. So it splits noise into these smaller bands letting it hear into noise more than it could if it registered all noise around a signal. It fits at least roughly we should detect signals to somewhere around 15 db below the noise level with the ear splitting this into 30 bands. But it is a simplified view of hearing. The spectrum of the noise and the signal being listened for will effect exactly how far into noise you can hear as well.

Sounds like we finally all agree LP>CD. I can sleep better now...lol

rockitman said:

Sounds like we finally all agree LP>CD. I can sleep better now...lol

Click to expand...

You must already be asleep and dreaming.