Skip to main content ↑ ↓
Topic: TT Dynamic Range meter "lies" about vinyl's dynamic rang (Read 50361 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

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

Reply #100 –
Quote from: Isabelxxx on
And again, if this thread is about the Tool NO ONE can make a claim about it lying or reporting wrong results. That's false and misleading.
...
Even if 75% of this forum or this thread is true, and greynol is most times right that doesn't mean the forum or Greynol's word equates The Truth.
Aww, c'mon! Nobody has claimed that the thread title or greynol represents "The Truth". If you make a ridiculous exaggeration, it is your own fault. Most fair-minded people know very well how the thread title is meant to be understood. There's no TOS paragraph that would prevent formulating a provocative thread title.

Quote
.... so this thread is all about how the tool DOES NOT measure Dynamic Range. Now you say it does it right (?). Great.
Huh?

Which meaning of the term "dynamic range" are you presuming now?

The entire problem revolves around the discrepancies arising from mixing up various different meanings of this term. Once you or anybody else start to distinguish them properly and clearly, the problem goes away.

Quote
Either it measures dynamics range and you have to accept filters and playback affects the dynamic range of a signal or what you are measuring is another thing. But you can not maintain both claims according to what you want to prove just to say to some of the audiophiles: f**ck you!
They deserve. ok.. but that's not the right way of proving nothing.
Stop your rambling and start to think! There is absolutely no contradiction in what I wrote once you acknowledge that the tool doesn't measure perceived dynamic range, but measures something that has a technical definition, which does not directly relate to perception. If people interpret the measurement result such that they infer a certain perception, they are mistaken. That's what Ian argues quite correctly. If the tool itself would suggest or foster this misinterpretation, it could conceivably be called "lying", even if the measured result is technically correct.

This is quite similar to a thermometer not giving you a measure of the perceived temperature, since temperature perception varies with a number of other factors. If the thermometer purported to tell you the perceived temperature, it would be lying.

Not hard to grasp, is it?

Quote
Again the "real" definition of Dynamic Range needs not clarification. A CD has less SNR. If the same master is used in vinyl and CD, the CD wins. End. If this thread is here is because there is more to what we understand as Dynamic Range. There is perception and we are trying to measure it. Don't say now you don't want to discuss it.
There's no one "real" definition of dynamic range. There are several meanings to it, depending on the context. Measuring the dynamic range of a production has nothing to do with the dynamic range the recording medium offers. Same word, different meanings. The perceived dynamic range is yet another meaning, and a fuzzy one to boot. It would take psychoacoustic tests to determine it, and it isn't even clear if it would make much sense to do it, since human perception is not constant and uniform.

The TT tool does not measure perception, and if it is trying, it doesn't do a very good job, which is again the point Ian is making. Maybe it would make sense to discuss how it could be improved, but for the moment it seems to me to be a problem with the expectation rather than the implementation.

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

Reply #101 –
What is this business about CDs and SNR?

I also want a link to the post in this discussion where I said my job is to dynamically change the thread title to reflect the discussion in the thread.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

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

Reply #102 –
Quote
What I am wondering is when there is a (non-clipped) peak with a fast rise-time in the signal, it's possible that the cutting head might fail to stop its lateral excursion at the correct point, and continue on for a while (due to a combination of inertia and insufficient damping). This would result in bigger peaks, and hence a greater peak-to-avg ratio. But I must stress that this is pure speculation on my part.

I know this is from the first page. I just saw this thread, which is quite interesting and thanks for putting it up.

A typical cutter head has a fair amount of feedback on it and so is damped. But there can be other factors such as distortion in the power amps for the cutter; my thinking is the 2nd harmonic is playing a role. LP systems make harmonic distortions, digital systems don't (they make aliasing). Between record and playback, I can see the most predominant of the distortion harmonics as adding some energy to the signal.

IOW I would not put this down to any one thing (such as as the LP recorder as opposed to the playback), but if you look at the waveforms that Ian expanded you will see that it is not only higher but also bifurcated, which suggests a 2nd harmonic.

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

Reply #103 –
It would be nice to see a better DR meter. Just brainstorming here but maybe something with a multiband compressor that tries to "brickwall" the music, taking an average of how much compression was applied in each band, and weighing the bands using equal loudness contours to give you a final DR number?

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

Reply #104 –
I don't see how that would work when you can compress a triangle wave just as easily as you can a sine wave.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

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

Reply #105 –
I really doubt that a "better" and "backwards compatible" DR meter is even possible.
At least not with the current problem statement (guesstimate the official DR score after undoing a linear filter with completely unknown characteristics). Well, if it can be solved by some calculus magic, it could be defined as "find a reversible linear filter (which must be relatively close to the identity filter — nothing extreme like long reverb, sharp highpass, etc.) which will fold the signal into one with minimum possible DR score" and then that DR score would be the answer. I guess this must be impossibly hard, but maybe someone who knows math better can add something?

Also perhaps it could work with a completely different metric, but it will measure a different thing which won't necessarily have a perfect correlation with the "official" DR on digital records. It will probably need to eschew peak levels (because they are changed after filtering) and compare some "short-term loudness" with long term loudness, but then it would be very insensitive to clipping (unless it's really extreme) and this is not what we want for digital records.

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

Reply #106 –
Quote from: magicgoose on
but then it would be very insensitive to clipping (unless it's really extreme) and this is not what we want for digital records.

Why? If we want to measure dynamic range, why should we care about unrelated things?

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

Reply #107 –
I'd say that clipping is related (it's a degenerate special case of limiting with zero attack time and zero release time, and limiting is a special case of dynamic compression) but… you might have a valid point here too.
 
SimplePortal 1.0.0 RC1 © 2008-2019