Which one has greater dynamic range, CD or LP?

Discussion in 'Sound Science' started by Pete-FIN, Oct 20, 2017.
  1. gregorio
    1. I've personally never seen a commercial recording with more than 60dB dynamic range. It's entirely possible there maybe some but they would be a real rarity, I'd be quite surprised if any of those mentioned in the link do.
    2. I wouldn't go that far, I've seen quite a few symph orch CDs over 50dB, just none over 60dB.

    G
     
  2. 71 dB
    Let's assume you have a 50 Hz or 60 Hz electric hum in your recording at level -80 dBFS. Such hum is not audible at reasonable listening levels because the hearing threshold of human hearing is about 40 dB at such frequencies. That doesn't mean the dynamic range of the recording has been limited to 80 dB, because at 1 kHz you can have more. For 44.1 kHz sampling rate, quantization noise is spead over frequencies 0-22050 Hz. If the spectral density of the noise is flat, half of the power is on frequencies 0-11025 Hz and the other half on 11025-22050 Hz. On frequency band 0-2205 Hz the noise power is 10 % and so on. What's relevant is how loud broadband noise is to our ears. So, a decaying drum sound at the end of a symphony doesn't limit the need for dynamic range at higher frequencies for a piccolo flute. We should ask what are the dynamic ranges of recordings on critical bands of hearing, because that's more than what the technical dynamic ranges are.

    With dither, the dynamic range of CD is about twice of LP.
     
  3. bigshot
    I know of two... Dorati's 1812 Overture on Mercury Living Presence. It has cannon shots that hit massive peaks- my dogs refuse to listen to it, and Slatkin's Telarc recording of the Carmen suites. It has a kettle drum hit that was completely untrackable on LP- it would make needles pop out of the groove. It's better on CD, but I suspect they applied compression to it to allow the rest of the music to be at a normal listening level. First pressings of Karajan's Parsifal were unlistenably dynamic too. Probably around 60dB. They've fixed it in the CD remaster.
     
    Last edited: Oct 21, 2017
  4. bigshot
    The dB scale is logarithmic, so double the number isn't the same as double the range.
     
    ev13wt likes this.
  5. RRod
    Perhaps a dumb question, but how are we defining the dynamic range of a recording here?
     
  6. 71 dB
    I'm an acoustic engineer, so dB scale is pretty familiar to me, thank you. :wink:

    Double on the dB scale is what I mean, of course. Doubling on linear scale would mean only 6 dB more, but you knew that.
     
  7. bigshot
    The point where the sound falls into the noise floor of the recording or format all the way up to the peak. The noise floor of a recording is considerably higher than the noise floor of the CD itself, because there's a room tone to the place it was recorded. With analogue the noise floor of the format itself may be higher than the noise floor of the recording.
     
    Last edited: Oct 21, 2017
  8. RRod
    So it's an X second RMS value in a (non-digitally-zero) silent section of the disc up to the max post-reconstruction peak? Just wanting something calculable.
     
  9. bigshot
    I'd say that's right. A CD can have a hard and fast theoretical dynamic range like we're talking about here, but analogue formats have more variables. There are multiple noise floors- tape hiss, rumble, surface noise, etc. The sum of all of those is the effective noise floor. It's a little more complicated though, because in analogue formats you can "burn in" a peak above the zero line without incurring the horrendous degree of distortion that doing that in digital creates. It was common in the analogue era to allow for a little bit of overdriving to extend the volume at peak level a little more. That squeezed a little extra dynamic range out of a limited format without causing too much noise. You'd think they wouldn't need to do that with CDs, but they do it anyway and it sounds like a dog's breakfast. Also as we were talking about before, the outer groove has more dynamics than the inner groove. The difference between them is quite pronounced, Because of the difference between outer and inner grooves, hot peaks, and the variable noise floors, it's hard to put an absolute figure to analogue dynamics. But you can certainly put a range to the best of all worlds and a range on typical dynamics in real world LPs.
     
    Last edited: Oct 21, 2017
  10. wnmnkh Contributor
    I highly doubt the number of people who would benefit from higher dynamic range is majority.

    The current environment is just too terrible (played from small bluetooth speakers, etc) that most people just won't notice.... well.. that assumes the music itself is not compressed from the first place.
     
  11. bigshot
    One thing I've found about audiophiles is that they love to go for the absolute extreme and think that is important. For instance, with dynamics, they read that 120dB is the loudest sound we can hear without going deaf. So they demand that a recording format intended to reproduce music go to 120dB. But the ear can't hear quiet and loud at the same time. If someone tries to whisper to you right after you've heard the air horn on a diesel truck, you aren't going to hear it, because your ears have to acclimate to a range of volume. In truth, human ears only have a dynamic range of about 40dB at any given time.

    Audiophiles also read that optimal human hearing is from 20Hz to 20kHz, so they demand that the playback equipment do that with a balanced response too. But the truth is, there are core frequencies that are more important than the ones at the extreme ends. How headphones perform from 100Hz to 10kHz is infinitely more important than that last octave on each end, and the range from 4kHz to 6kHz is the most important of all because that is where our ears are most sensitive. But audiophiles tend to worry about the frequencies from 18kHz up, which don't even exist in musical instruments.
     
    Last edited: Oct 21, 2017
    JaeYoon likes this.
  12. old tech
    [QUOTE="Pete-FIN, post: 13796295, member: 149035"
    So, my questions.

    Which one has greater dynamic range, CD or LP?
    How big is the theoretical maximum dynamic range on CD, and how big is it on LP?[/QUOTE]

    The site page from Hydrogenaudio may be also be of some help. Indeed that forum has many discussions on this topic.

    What confuses some people though is that a record of a particular album may sound more dynamic than its (usually later remastered) CD version. That is a mastering choice based on marketing rather than a reflection of the media capability.

    http://wiki.hydrogenaud.io/index.php?title=Myths_(Vinyl)
     
  13. gregorio
    I didn't know of Dorati's 1812 and have only Karajan's Parsifal on CD. I don't suppose you could measure that recording of the 1812, I'd be interested to know what it's dynamic range is. Not that it's particularly applicable though as cannons are not a standard member of a symph orch! :)

    I suspect you know it wasn't a dumb question, that it's actually quite a complex question because it's partly reliant on perception. Your calculation above would give a poor result. It would for example generally show a lower dynamic range for noise-shaped dither than for a standard triangular dither, when in practise the opposite would is true.

    G
     
  14. RRod
    Yes, one should replace RMS with a perceptual loudness measure. My main concern is how to automate the process to, say, parse an entire album collection without having to pull out the abacus for each album and get: a) a reasonable absolute measure and b) at least a consistent ordering. People are often bandying about about which albums have the most dynamic range, yet they never give a replicable, consistent metric for the concept itself.
     
  15. castleofargh Contributor
    it's something I often wondered when looking at different resolutions for albums. it would probably be worth it to band limit the signal at a fix value within the audible range(maybe even lower?) before looking into the dynamic numbers. at least if the aim is to get something related to an impression of dynamic. even though I suspect a lot of people couldn't tell dynamic from a messy frequency response if they were tested on it. but that's another can of worms.
     

Share This Page