Which audio format sounds better?

This topic was created by Evil Auditor .

  1. Evil Auditor Silver badge

    Which audio format sounds better?

    There's an interesting Gruaniad article on "how much difference is there between MP3, CD and 24-bit audio?" So far so good. Unfortunately, not a blind and let alone a double blind test was applied and hence, as any ful no, the conclusions are utter bollocks.

    It's about data, data transformation, data compression and science. Wouldn't you, El Reg, not be in a much better position to do such research the proper scientific way? (Heck, I might even try to attempt it myself if I found some time...)

    1. Irony Deficient

      Re: Which audio format sounds better?

      Evil Auditor, I’ve found that (nominal) 192 kb/s Vorbis suffices for me from a Rockbox-equipped player — that is, for Vorbis, I can’t distinguish any significant difference between 192 kb/s and a higher bitrate. However, that wasn’t a blind test either, so make of it what you will; people listening to different musical samples, or people who have better hearing, could well draw different conclusions.

  2. cortland

    70-year-old ears +10 KHz tinnitus = I can't tell and don't care which is better.

    And anyway, I buy used CD's. They only have to last as long as I do -- and RIAA gets none of the money.

  3. dan1980

    MP3 is not one quality. 320Kbps sounds pretty good but even on my moderate system I can tell the difference between that and CD. Not on all tracks, of course, but most.

    On my work PC, where I listen through some old earphones in mono I have most of my music in 192Kbps MP3 for sycning to my player but I also have some tracks as FLACs and the difference, even with the lower audio quality inherent in the setup, is instantly noticeable. I have some tracks downloaded in both formats and it's night and day.

    All that aside, the brain is an amazing thing and it tends to adjust the sound you 'hear' until it sounds 'right' - just give it a bit of time. If your system is bass-light then the brain can fill in the missing frequencies.

    One of my sets of earphones got overloaded using a plane adapter and one of the buds end up about 1/4 volume - less maybe. The amazing thing was that after a little while listening to them it sounded relatively normal and it became 'centred' again, where it had initially appeared to be only in the left ear.

    So, after a while listening you adjust and get used to it.

    The only time you REALLY notice the difference in those situations is when you listen to the same track in compressed and uncompressed formats and hear bits you didn't realise where there. After that, you will likely notice the absence when listening to an MP3!

  4. This post has been deleted by its author

  5. Someone_Somewhere

    You certainly notice the difference between lossless and lossy formats when you have to do anything even semi-pro with them - as a DJ, I can tell you that I'd never play an mp3 out on any system, let alone a low-end one with it's own EQ problems to boot, unless I couldn't get my hands on the track in any better format.*

    As a promoter, I insist that the DJs I engage play wav or CD.

    --

    * and even then I'd think long and hard about it

  6. Baldy50

    Wish I had my old equip

    Hi, I guess we'll stick to digital formats!

    No,No,No! Audio note 300b single ended pure class A amp, black box moving coil stylus on any decent turntable (Cast iron platter) preferably English and that goes for the speakers too.

    Just have to let it warm up a bit and don't have the amp or turntable on any surface prone to sympathetic vibrations.

    But to the point I believe WAV's are the best but most people including myself are perfectly happy with any MP3 file at a reasonable Kb's rate.

    But and a big but, many software programs for some reason just extract the data and do nothing else to the file. CD/DVD players have complicated software to equalise the sound built in, you don't know it's doing and what it's doing is making it sound better.

    If you like to store your music collection to a digital media then fine but download a free program called Mp3Gain and it will correct the amplitude of your music, just research the program please if in doubt it just sounds far less distorted and all you need to do is adjust the volume a little higher.

    Many burning programs will 'normalise' the tracks before burning, the same thing really and if making yourself a CD for the car you don't want to be adjusting the volume all the time so normalise the files beforehand.

    As stated above you can't always get the track you want at a good bit rate or a WAV file, I once DL'd some music for someone and one of the tracks when normalised had a gain of 136 Db!

    If I had not checked it would even on a measly 4.8 K rig probably would have left the guests wearing speaker cones.

    Low end PA speaker sensitivity, how much sound it kicks out per watt.

    100 Db output measured with a 1Kh sine wave at 1 Watt input with the microphone 1 meter away in an anechoic room.

    So to get 101 Db I put in 2 Watts, to get 102 Db I put in 4 Watts, to get 103 Db I put in 8 Watts, to get 104 Db I put in 16 Watts, to get 105 Db I put in 32 Watts, to get 106 Db I put in 64 Watts, to get 107 Db I put in 128 Watts, to get 108 Db I put in 256 Watts, to get 109 Db I put in 512 Watts and to get to 110 Db I put in 1024 Watts! It's a Logarithmic scale you see and as you can see very easy to run out of power and thats why your headphones work so well I suppose.

    Regards.

    PS (Db) does not stand for dumb blonde

  7. Neil Barnes Silver badge

    What?

    dB is a measure of relative power. Double the power, increase the dB by 3.

    Years of empirical testing - real people, real ears - have established that a 3dB change in power causes the smallest change in volume that most people can detect. 10dB approximately doubles the perceived volume.

    As an aside, dBv is a measure of the ratio of voltages and is commonly used when the signal is measured in voltage, since it is independent of output impedance. It's numerically twice as large as the equivalent power dB measurement. On a digital representation of such a signal, each bit of the binary value is equivalent to 6dBv; since a CD uses 16-bit linear PCM it has a theoretical maximum dynamic range of 96dB - which cannot be practically achieved for other reasons. Where your 136dB came from I don't know - but you should probably change your software for something that works.

    CD players have internal software to correct for errors and missing bits in the data stream, but they should never perform any other processing; what comes out should be as close a duplicate of what went in as the condition of the medium provides.

    1. jake Silver badge

      Re: What?

      That's nice.

      My McIntosh amps and analog tape & vinyl provide far better sound than anything that the digital world can provide when it comes to recorded audio ...

    2. Baldy50

      Re: What?

      More accurately

      GAIN/power multiplied by

      1DBx1,25

      2DBx1.59

      3DBx2

      4DBx2.5

      5DBx3.125

      6DBx4

      7DBx5

      8DBx6.25

      9DBx8

      10DBx10

      20DBx100

      At 1Kh

      Humans can hear a range of frequencies but we are far more sensitive to sounds between 1 kHz and 4 kHz, most people can detect sounds as low as 0 dB SPL at 3 kHz, but require 40 dB SPL at 100 hertz (an amplitude increase of 100).

      So saying that a person can only detect a 3DB change in power is limited to the frequency that is used, so is not generally correct.

      If you had the same person sit in a quiet place for some time before the hearing test it would have a totally different result to the same person after being somewhere noisy before the same test and having your eyes closed makes a difference too. I've known women that can detect a beer can opening a hundred yards away.

      If you use I tunes just check the gain and clipping of your tracks ripped by their inbuilt ripping software, it's crap. DL the full MP3Gain from sourceforge.net and try it you'll be surprised. You can just analyse a file and not alter the file or copy to another folder and listen to an unmodified track and a processed one.

      All I can say is that's what the program showed me when I checked the file and it sounded far far better after it had been processed and every track that I've used it on that needed it.

      OH BTW The 24-bit DVD disk format has a theoretical 144 dB range and what software was used to rip the track is anyone’s guess.

      Although loudspeakers do consume a large amount of power they are a voltage dependent device not power dependant.

      But in the real world my quote stands as a practicality, we are not just using a 1k sine wave but just play some hard core reggae or dubb on a big rig and see just how much power is required and has always been a good guide to how much power I'm going to need to fill a big room. Shape of room, reflective surfaces and sound absorbent surfaces, packed with people or half full.

      Just by speaker placement alone you can achieve massive gains in DB output, such as folded horns using the side of the wall as part of their horn.

      Ta for reply, Ian.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon