Skip to main content ↑ ↓
Topic: TT Dynamic Range meter "lies" about vinyl's dynamic rang (Read 49773 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #50 –
quickie question -- does foo_dynamic_range 1.1.1  use the same loudness metric as TT DR Meter?

IOW , is it more 'reliable' as in an indicator of sound quality?

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #51 –
Quote
IOW , is it more 'reliable' as in an indicator of sound quality?
I don't know if it's a better indicator or dynamic range, someone else can answer that...

I would NOT say dynamic range (or dynamic contrast) is necessarily a measure of "quality".  It's one characteristic of the sound.  You can't say higher dynamic range is always better. 

One of the most common effects used in audio production is dynamic compression (reduction of dynamic range) and that's not done to make the quality worse!    Sometimes it's overdone in my opinion, and the opinion of many others, but it's done because that's what sells.       

On the other hand, dynamic expansion is almost never used in audio production (except in the special case of a noisegate).    And most popular music is not performed dynamically...  Classical music tends to be very dynamic and it's generally recorded with little compression, somtimes no compression.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #52 –
With all due respect, the argument is illogical. What is missing is consideration of the DR of the original master file. The author of the YouTube post has missed the possibility that the vinyl cutting process may have faithfully preserved more of the DR of the master than the CD burning process did.

The superior DR demonstrated in the vinyl may not be distortion, it may be that the CD DR is lower than the master. He doesn't demonstrate this at all.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #53 –
Quote from: Thegreyghost on
With all due respect, the argument is illogical. What is missing is consideration of the DR of the original master file. The author of the YouTube post has missed the possibility that the vinyl cutting process may have faithfully preserved more of the DR of the master than the CD burning process did.

The superior DR demonstrated in the vinyl may not be distortion, it may be that the CD DR is lower than the master. He doesn't demonstrate this at all.
He was clearly comparing the dynamics of the vinyl against the original digital source, from which the vinyl was made of and not against a CD rip. So the TT meter sais, that the vinyl copy has more dynamics than the original digital source, which should not be possible. That is the statement. Nothing more.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #54 –
Quote from: Thegreyghost on
What is missing is consideration of the DR of the original master file.

This consideration was directly addressed at 4:50 into the video:

https://www.youtube.com/watch?feature=playe...AE9dL5FG8#t=291
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #55 –
Quote from: Thegreyghost on
With all due respect, the argument is illogical. What is missing is consideration of the DR of the original master file. The author of the YouTube post has missed the possibility that the vinyl cutting process may have faithfully preserved more of the DR of the master than the CD burning process did.

The superior DR demonstrated in the vinyl may not be distortion, it may be that the CD DR is lower than the master. He doesn't demonstrate this at all.

That is illogical in itself.

How would the CD lower such dynamic ranges? Only vinyl, the format itself, the mechanic playback system, the creation process of vinyls can introduce such levels of distortion.

This should be clear even if you compare a dirt cheap CD player to a high end vinyl system.
Vinyl cannot directly store any digital signal. The added distortion by this required pre-processing alone... Add to that the rumble, speed inaccuracies, wow, flutter, the distortions added by the tonearm, low output level requires a pre that further changes the sound (add additional noise, distortion, crosstalk, frequency response deviations) ...
"I hear it when I see it."

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #56 –
We've been using electronic level meters now for many decades however the old style, mechanical meters used predominantly before the 80's, as well as modern day LP record cutting lathe heads (and TT styli?) have a property in common referred to as "ballistics". This refers to the tendency of the physically moving structure (tracing or representing the analog input waveform as best it can) to both be sluggish to fast and sudden alterations as well as a tendency to overshoot the peak values it is instructed to, especially on quick, transient peaks in the music (easily seen from percussion instruments), mostly because of its physical inertia due to its mass, which of course an electronic meter lacks. Meter ballistics can be desirable at times and they can be emulated or "dialed in" to modern day electronic meters, however they can't be "dialed out" of physical meters.

LP records should actually be expected to have transient peaks in the recording which exceed the input signals' true values, hence "added dynamic range" (albeit artificial) due to this inertial mass induced overshooting of the cutter head's needle, when the recording is physically etched onto the blank surface. I guess this issue would go away if we were to convert to laser cutting lathes to go along with our laser record players [which have actually existed] but for the time being the added hashiness seen [less "flat topping" I believe they call it] and the artificially greater dynamic range values seen in these simplistic measuring algorithms is a true manifestation of the actual (distorted) wave form etched and played back from the vinyl, due to cutting head ballistics.

It is an artifact and a distortion to the true signal, although it is quite possible some (many?) people actually "prefer" it. [If it is even audible. I'm not claiming it is; I don't know.] Me? I'd rather have an indistinguishably different, exact replication of the studio master, levels intact, so I'm sticking to digital.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #57 –
Quote from: xnor on
Quote from: Thegreyghost on
With all due respect, the argument is illogical. What is missing is consideration of the DR of the original master file. The author of the YouTube post has missed the possibility that the vinyl cutting process may have faithfully preserved more of the DR of the master than the CD burning process did.

The superior DR demonstrated in the vinyl may not be distortion, it may be that the CD DR is lower than the master. He doesn't demonstrate this at all.

That is illogical in itself.

How would the CD lower such dynamic ranges? Only vinyl, the format itself, the mechanic playback system, the creation process of vinyls can introduce such levels of distortion.

This should be clear even if you compare a dirt cheap CD player to a high end vinyl system.
Vinyl cannot directly store any digital signal. The added distortion by this required pre-processing alone... Add to that the rumble, speed inaccuracies, wow, flutter, the distortions added by the tonearm, low output level requires a pre that further changes the sound (add additional noise, distortion, crosstalk, frequency response deviations) ...


The process of determining discrepancy must be scientific, not based on dogma and assumption. We need to see the DR of the master. The CD creation process is algorithmic and digital,mans assumptions about its purity may be misguided.

Talk of the "pure signal" and the assumption that the CD press is a viable proxy may be erroneous,

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #58 –
Quote from: greynol on
Quote from: Thegreyghost on
What is missing is consideration of the DR of the original master file.

This consideration was directly addressed at 4:50 into the video:

https://www.youtube.com/watch?feature=playe...AE9dL5FG8#t=291


No it's not. He says it was produced from the swm master file, and he says "the wave form looked liked this..." But there is no proof St is the case. Why not? Let's see it. I'm not saying it's not true, just that it isn't proven here,

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #59 –
The DR value of the master used to create the vinyl in the video was 8, or are you accusing Ian Shepherd of lying?

"This vinyl master was made from the same digital file that was used to create this CD master"
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #60 –
Quote from: Thegreyghost on
The process of determining discrepancy must be scientific, not based on dogma and assumption.  We need to see the DR of the master.

You can verify everything I said by doing measurements. Nothing I said is dogma and assumption, but what you say and imply definitely is:
- assumption that CD changed the dynamic range
- assumption that vinyl preserves dynamic range more faithfully
- assumption that superior DR (which is a wrong assumption in itself, the TT DR meter simply yields higher numbers) in the vinyl is not due to distortion
- assumption that Ian is lying?
..

These technical assumptions of yours are demonstrably wrong.

I guess you have a vinyl rig at home. Have you never even done a measurement as simple as platter speed? If you had, you'd see the gross imperfections of vinyl systems.
"I hear it when I see it."

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #61 –
Quote from: xnor on
Quote from: Thegreyghost on
The process of determining discrepancy must be scientific, not based on dogma and assumption.  We need to see the DR of the master.

You can verify everything I said by doing measurements. Nothing I said is dogma and assumption, but what you say and imply definitely is:
- assumption that CD changed the dynamic range
- assumption that vinyl preserves dynamic range more faithfully
- assumption that superior DR (which is a wrong assumption in itself, the TT DR meter simply yields higher numbers) in the vinyl is not due to distortion
- assumption that Ian is lying?
..

These technical assumptions of yours are demonstrably wrong.

I guess you have a vinyl rig at home. Have you never even done a measurement as simple as platter speed? If you had, you'd see the gross imperfections of vinyl systems.


No, I didn't say these things were definitely true. I said they couldn't be ruled out as possibilities without access to, or visibility on the DR of the original master.  They may be demonstrably wrong, I accept that, but I'm asking that they be actually demonstrated to be wrong in this case.  This is just scientific process.

In fact, I'm not even asking for it, just suggesting it would be more rigorous proof to do so.

Personally,minding care. I'm know what I like to hear, and don't obsess on derived numbers to prove it. Trust your ears is the best advice in the clip in my humble opinion. sorry to cause offence to anyone, was entirely unintentional!

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #62 –
Quote from: greynol on
The DR value of the master used to create the vinyl in the video was 8, or are you accusing Ian Shepherd of lying?

"This vinyl master was made from the same digital file that was used to create this CD master"


Not accusing anyone if anything. He didn't say that at all. He said that the same digital file was used to master both the cd burn and vinyl press. He didn't specify the observed DR in that master file, although he may have implied it when he said the waveform "looked like the cd" image I admit. 

Now we all know that the CD waveform is futsed with,  it's possible, just possible that the two have different DR. That's all, small point, nobody get frantic,

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #63 –
Sorry, but I'm not going to demonstrate how digital audio and vinyl works for you here.

We do know that vinyl has many imperfections. And we know that the TT DR meter has a bad algorithm. Heck, even inaudible phase shifts can significantly alter TT DR readings..
"I hear it when I see it."

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #64 –
Quote from: xnor on
Sorry, but I'm not going to demonstrate how digital audio and vinyl works for you here.

We do know that vinyl has many imperfections. And we know that the TT DR meter has a bad algorithm. Heck, even inaudible phase shifts can significantly alter TT DR readings..


Of those imperfections that you previously stated, what effect does the wow & flutter of any competently-engineered turntable have to do with artificially "boosting" the DR? (As a sidenote I really think the wow & flutter thing is grossly exaggerated. Of course it's not the precision of digital but I've never owned a TT that exhibited perceptible wow & flutter...unlike tape decks.)

I'm asking in earnest and I apologize if there's something I'm missing.
The Loudness War is over. Now it's a hopeless occupation.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #65 –
I didn't say anything about audibility.
Regarding DR, well, both wow and flutter will change the waveform. Wow will change the number of samples analyzed in a fixed window. Flutter will add intermodulation.

With a good turntable neither should make a significant difference, but the TT meter seems to use a very naive algorithm (please someone send me a link to an implementation) that is susceptible to all kinds of "distortions" that really shouldn't make any difference in dynamic range.
"I hear it when I see it."

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #66 –
Quote from: xnor on
I didn't say anything about audibility.
Regarding DR, well, both wow and flutter will change the waveform. Wow will change the number of samples analyzed in a fixed window. Flutter will add intermodulation.

With a good turntable neither should make a significant difference, but the TT meter seems to use a very naive algorithm (please someone send me a link to an implementation) that is susceptible to all kinds of "distortions" that really shouldn't make any difference in dynamic range.


Ah, ok. I understand now. Thanks
The Loudness War is over. Now it's a hopeless occupation.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #67 –
I am not very familar with yinyl gear and how bad it really works but again i think it helps to link to Archimago's blog:
Roksan TMS (1) Wow & Flutter - PlatterSpeed test
When this is what Wow&Flutter does it is no wonder all waveforms look completely different when vinylized.
If this is something different, sorry. I have no clue in this direction.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #68 –
Quote from: Thegreyghost on
Now we all know that the CD waveform is futsed with

We suspect the CD master was derived from a higher-resolution master which would have entailed a reduction in bit-depth and sample rate.  This would not have altered the DR value.

To suggest there was some other mechanism that reduced the value of the DR measurement of the CD is to say that Ian Shepherd is either incompetent (supposing the mechanism was unintentionally put into play) or dishonest (supposing the mechanism was intentionally put into play).
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #69 –
Quote from: Wombat on
If this is something different, sorry. I have no clue in this direction.

As far as the video is concerned, the exact mechanism that results in an increase in DR value doesn't matter. The takeaway here is that DR figures can't be used reliably to determine whether vinyl and CD were sourced from different masters.

FWIW, I don't believe wow and flutter have much, if anything, to do with the DR figures.  I'm pretty sure xnor was just throwing them out as examples of the many forms of distortion inherent in vinyl transcription and playback.

I think mzil's explanation cuts closest to what is going on.  In the past I've characterized it as a predominantly underdamped highpass response, which has similar characteristics when viewed from a control systems point of view.  At least this is what I recall from a previous life.  Consider whether a stylus can provide a DC response.  Then consider whether it can provide a square wave and what the output will look like instead.  Then have a gander at some of the zoomed waveforms in the other threads on this subject, some of which certainly predate this topic.

This one was particularly entertaining (for me at least):
http://www.hydrogenaud.io/forums/index.php...st&p=875605
http://www.hydrogenaud.io/forums/index.php...st&p=875637
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #71 –
Regarding algorithm, one should IMHO consider (I do not make a claim it will work in the end, only that it should help against certain unwanted issues and is worth a try) a "robust statistics" approach. Those methods using median rather than mean in order to not let extremes influence too much.

For "mean" one could for example use median volume rather than RMS. Or calculate RMS over each short window and then take the median of these RMS windows; that has a chance to eliminate bursts from single clicks/pops in the vinyl. But I guess that the mean loudness isn't the critical part (so heck, why not use ReplayGain or EBU-R128?).

Then for the dynamic range, the worst to use would be file peak - it is prone to a single error.  But for example median excess over mean, dropping those which are lower than mean, because those may be prone to surface noise?
Memento: this is Hydrogenaudio. Do not assume good faith.

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #72 –
Of course the large frequency response (including phase!) deviations from flat is what will by far have the largest influence on the TT DR numbers.

@greynol: thanks for the link, that's what I found too
So given that even just changing the volume of a track changes the TT DR readings ... why would anyone still use that anyway?

Just take an EBU-R128 loudness meter?
"I hear it when I see it."

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #73 –
Median over mean won't fix the vinyl DR inflation problem, as it is not caused by a few large peaks; rather, all the horizontal envelopes on the master resulting from heavy compression/"brickwall" limiting become angled after being subjected to vinyl.  In addition, gain must be reduced in order to prevent this new shape from clipping.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

TT Dynamic Range meter "lies" about vinyl's dynamic rang

Reply #74 –
Quote from: Thegreyghost on
Quote from: xnor on
Quote from: Thegreyghost on
With all due respect, the argument is illogical. What is missing is consideration of the DR of the original master file. The author of the YouTube post has missed the possibility that the vinyl cutting process may have faithfully preserved more of the DR of the master than the CD burning process did.

The superior DR demonstrated in the vinyl may not be distortion, it may be that the CD DR is lower than the master. He doesn't demonstrate this at all.

That is illogical in itself.

How would the CD lower such dynamic ranges? Only vinyl, the format itself, the mechanic playback system, the creation process of vinyls can introduce such levels of distortion.

This should be clear even if you compare a dirt cheap CD player to a high end vinyl system.
Vinyl cannot directly store any digital signal. The added distortion by this required pre-processing alone... Add to that the rumble, speed inaccuracies, wow, flutter, the distortions added by the tonearm, low output level requires a pre that further changes the sound (add additional noise, distortion, crosstalk, frequency response deviations) ...


The process of determining discrepancy must be scientific, not based on dogma and assumption. We need to see the DR of the master. The CD creation process is algorithmic and digital,mans assumptions about its purity may be misguided.

Talk of the "pure signal" and the assumption that the CD press is a viable proxy may be erroneous,


so when comparing a format full of heavy fidelity weaknesses, and a digital conversion, your incentive is to blame the digital file. funny to see you talking about assumption^_^.

it's really not hard for you to demonstrate your strange theory. take a high res file, and convert it to CD resolution with whatever resampler and dither you like. then see how much the operation changed the values of the DR meter. and good luck getting something like 4db less on the 16bit file. even with all the already mentioned reliability problems from that DR meter I doubt you'll get there.

to play the devil's advocate, I could imagine a messed up situation where some ultrasounds would be louder than the audible range(scary!!!!), so cutting them out when converting to CD could maybe affect the DR. in that silly hypothesis the "loss of dynamic" would be in fact saving our ears.
now imagine what ultrasounds of high amplitude(higher than the audible range in my example) would do when playing the vinyl  . the needle would go all "I believe I can fly, I believe I can touch the sky!"
 
SimplePortal 1.0.0 RC1 © 2008-2018