1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

24 vs 32-bit sound

Discussion in 'Sound Science' started by Vilhelm, Aug 22, 2018.
1 2
  1. bigshot
    No, the problem is the peaks. To raise the noise floor up, you have to also raise up the peaks. 80dB is as loud as most people will tolerate. Above that, you can't take it for long periods of time. A live rock concert might have louder peaks but those are mistakes by the mixer. Also, in order to raise your noise floor up to audible levels, you need to raise it an additional amount to get above your environmental noise floor. This applies to headphones too. Add 20dB for your headphones or IEMs on top of 96dB for a CD and you are getting into the territory of threshold of pain, where you can't tolerate it for even a few seconds. Above that is hearing damage. And yes, it is very easy to damage your hearing with IEMs. The fact is, most commercially recorded music is designed for a 40dB dynamic range. The most dynamic music only exceeds that by 15dB or so. You just don't need 96dB.

    If you have a SPL meter, try turning a speaker system up above 80dB. You'll find out what the numbers actually mean.
    Last edited: Dec 18, 2018
  2. jagwap
    I would rephrase your first statement as "Yes, the problem is the peaks". Turn a system down 20dB to cope with them with only 96dB dynamic range at the output, when the volume control is digital, and you get 76dB signal to noise. This is for a normal system. Make it sensitive and and you lose more. This is a reason to have 20bit dynamic range or more. It allows the 16bit through at

    Sometimes if you read what I say you may find I am agreeing with you. Automatically disagreeing with someone who has argued with you in the past on an internet forum can give people the wrong impression.
  3. bigshot
    76dB dynamic range is more than you need. Take a track that is normalized and export it without dithering. It still sounds fine for normal listening, even at loud levels. There’s a demonstration of this in the AES link in my sig file with downloadable examples.
  4. jagwap
    Probably why vinyl sounds fine then.

    That wasn't what I said. With sensitive speakers or headphones, and sensitive amplifiers, you lose more than that. Maybe another 20dB in extreme cases of mismatching. So this is why 20bit capable DACs are useful. There is metrit for analogue level controls.

    You don't reply to people's points. You find something to disagree with.
  5. bigshot
    Vinyl rarely exceeds 40dB dynamic range, even less at the inner grooves.

    Transducers and amps have nothing to do with the dynamic range of CDs. If you have mismatched impedance, you should address that problem so your amp and speakers are functioning to spec. You don't need a deeper noise floor on the recording medium. The noise floor is so low already, worrying about a level pot isn't necessary. You have lots of room. Most music is under 50dB. Most rooms in normal homes have a noise floor much higher than your music. I just can't see how in normal listening conditions with properly set up equipment this would even be an issue.
  6. jagwap
    OK, you seem to be confusing terms like sensitivity and impedance. You are going to find it difficult to make a valid point if you cannot follow the engineering.

    Many amplifiers have much higher gain than others. Also raising the noise floor as they do so. This may be because they need to because they have a very high power output.

    Some speakers have much higher sensitivity than others, such as horn loudspeakers. This creates more SPL per volt, so raising the noise floor.

    If the volume control is digital, when you turn it down the DAC noise floor does not reduce with the signal.

    Simple really.

    So more than 16bit is useful in these cases. Not for the programme material, but for lowering the levels. All your comments that it doesn't matter in your system do not apply when dealing with kW amps on compression drivers. This is my job.
  7. Don Hills
    May I suggest that you're an ideal person to write a brief explanation of gain staging and how people could achieve optimum staging in their systems?
  8. bigshot
    Why would you use an amp that doesn't work properly with your speakers? To me, that is like blaming an iPod not sounding good with headphones through the headphone out that really should be amped. I've had all kinds of speakers all my life, and I have both high sensitivity speakers and low sensitivity speakers in my current system, and I've never heard any noise floor. I just hear music really loud if I turn the volume up.

    Would this raise the noise floor above -50dB?
    Last edited: Dec 19, 2018
  9. TheSonicTruth
    In theory, audio gain-staging is where each link(component) in the audio chain contributes neither too much or too little gain to the signal before it reaches the speakers. Too high, distortion is introduced, too low, noise is introduced - and potentially amplified by the next component.

    For example, if, in a live audio setup, you notice that in a rack of processors one component is really cranked and another is turned way down, the gain staging might not be optimized.

    In my personal listening system, I employed rudimentary gain staging by inserting RCA input attenuators on a couple of inputs on my receiver: CD, Video, etc. This allows me to operate the receiver volume control in a more useful range than without the attens. 6dB attens. should be fine for most situations, and 3 and 12dB are also available, sold in pairs from places like Parts Express. Insert on the back inputs of the receiver, not in back of the playback components. THey are sonically transparent as far as I can hear.

    If you have a more traditional pre-amp > amp setup, you can achieve this more flexibly with each input on some pre-amps.
    Last edited: Dec 19, 2018
  10. jagwap
    More to the point is that it could raise the background noise to more than 50dB SPL when using speakers like Klipsh horns.
  11. bigshot
    I have a Klipsch center channel that is horn loaded, and it has no more noise than any of my other speakers. I have JBL bullet tweeters too.

    Is this just theory, or have you run across this problem with your own system? I'm being genuine here. I've used horn loaded tweeters and Klipsch speakers for a very long time, and I've never had problems with noise. Why would that be?
    Last edited: Dec 20, 2018
  12. jagwap
    You may have an analogue volume control in your system after the DAC. You probably have a >96dB S/N DAC. A Klipsch centre is not the same as a Klipsh horn at >105dB/2.828V rms. Any one of these mitigates the situation.

    Your system does not represent all possiblities. It is only one case. When you design equipment you have to consider other cases.
  13. castleofargh Contributor
    I'm pretty much never full scale on my digital volume, but if I needed something like -50dB, I would definitely take action and change some gears to solve the nonsensical gain mismatch.
  14. bigshot
    When you buy equipment to put together a system, I would think that you would choose components that work together. I don't see any reason to abandon the CD format because some person out there made a very poor choice of equipment and might be not getting his full allotment of -96dB. This can't be a common thing, because I've never heard anyone say that they have an unnaturally audible noise floor on their speaker system.

    My AVR has a normal volume control that raises or lowers in half dB increments. Is that the kind of volume pot we're talking about?

    I'm still curious though... what kind of noise floor boost are we talking about in a type? Can this raise the noise floor to the point where it would be audible at normal listening volumes? That would require a bump up of about +40dB or more. That seems unlikely to exist in the real world to me.
  15. sonitus mirus
    I think it is possible to have issues with noise, but I do take steps to avoid problems where I can reasonably make them. Like you, I also use horn-loaded Klipsch speakers. In my case, 2 RP-280F floor speakers in a relatively small room. I have no issues at all that I can hear with noise as the system is dead silent. I did make sure to get an amp that paired well with the speakers from the design specifications I could find and reviews that were available. I've been using balanced XLR connections from my DAC to the amp, and it is a requirement I now look for in any DAC. While the DAC that is paired with the Klipsch speakers uses a switching power supply and is connected to a laptop via USB, the DAC measures spectacularly and I have had no issues at all with interference or other noise. (RME ADI-2 DAC)

    That said, it doesn't mean that the Klipsch speakers can't have issues with noise under certain circumstances, just not in my setup. I hate hiss/noise, and I would find a solution to remove it. I picked up the speakers because they were on sale for a great price and I knew that I would be able to EQ them to great effect without compromising too much. They will never be able to compete with some of the best speakers available, but for the room I had, these are very nice and extremely capable with a little assistance with room correction using a calibrated mic and my DAC's built-in DSP.
1 2

Share This Page