1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

16 or 24bit sampling: real benefit?

Discussion in 'Recording' started by Cresta, Oct 5, 2006.

  1. Cresta

    Cresta Active Member

    hi all :)

    all stuff producers will tell you that recording at 4872394879bit sampling rate is what you truly need to sleep well and have great success with girls...but...who really need high bitrate sampling?
    Which is the REAL difference between recording at 16 or 24?
     
  2. Boswell

    Boswell Moderator Distinguished Member

    You're getting confused, especially about the success with girls bit.

    Sampling rate and wordlength are two separate and independent things. In crude terms, the sampling rate determines the range of frequencies you can digitise without aliasing errors, and wordlength determines how accurate each sample is.

    The audio on standard CDs is set at a wordlength of 16 bits and a sampling rate of 44.1KHz, even though the original studio material may have been recorded at 24 bit and 96 KHz. Newer SACD and DVD-Audio disks have the higher rates and longer wordlengths.

    In the studio, using the extra wordlength gives additional headroom for the signals during recording, and maintains quality through the mixing, adding effects and the editing stages. 16 bits gives a theoretical dynamic range of 96dB and 24 bits gives 144dB. The noise levels in professional audio gear is usually in the 110 - 120dB range, and so when digitising the signals, 24 bit converters are used to avoid the conversion process itself being the limiting factor. Similarly, a sampling rate of 96KHz may be used in the studio to maintain the sound quality and to reduce distortion on transients. The conversion to 16 bit 44.1 KHz is done at the end of the pre-mastering process when all the other processes have been applied, including normalisation of sound levels.

    For live sound, there are engineers who record at 44.1 KHz if they know the final result is to go to CD, since the sampling rate conversion process can have a negative impact on the sound quality, although they will almost certainly use 24 bit wordlength recording. Also, by its nature, live sound tends to go through fewer stages of post-production.

    So the answer to the question about who needs higher sampling rates and longer wordlengths is that recording engineers do, but it's not mandatory to use them. Many great recordings have been done at 16-bit 44.1 KHz.
     
  3. DIGIT

    DIGIT Guest

    Virtually ALL the music produced today ends up being played in a car and/or iPod...so, all those bits, all those expensive microphones, pres, late night tracking & mixing sessions, and it you are lucky your music will be heard through a pair of earbuds from some pod form or another :wink:

    In fact, I am manufacturing and selling a set of turn-key Mixing & Mastering systems precisely tuned for iPod reproduction, I humbly call it the iMaster Sytem :oops:
     
  4. TeddyG

    TeddyG Well-Known Member

    An analog analogy(Some history) if you will...

    When tape was "the way", we used to try to do all of our initial recording at 15 inches per second(IPS) - that is to say we did once even THAT techology became available at the "middle to lower" levels of broadcasting -- took quite awhile... I would say that 15 ips is similar, today, to, say 24 bit/48khz.....? The big boys used to even use 30 ips(Kind've like 24/96)! Anyway, big boys down to us kids ALL recordings, eventually, in radio broadcasting, ended up on a (Correct me on this?) 3 and 3/4 inches per second(?) "endless loop audio tape cartridge", for on-air play. Actually it was more than enough, quality-wise, as radio, certainly AM radio, couldn't pass much frequency-wise or quality-wise. The reason we used 15ips, when we could, is because the "extra" quality of the recording was obvious(Lack of "hiss" was a biggie!), even with inferior playback(Speakers, etc.) once we made another "generation"(Copy) of any recording. Anyway, we stayed with 15ips as long as possible, before "dubbing down" to the slower speeds. IF one intended to use THE ORIGINAL recording, the original tape, as the "end result", one could record at quite low speeds before anything obvious was noticed Same should be true today. Sometimes the mechanics of the thing, like the cost of the tape(!) was a major influence on tape speeds used.

    Today, many of us have the ability to DO super quality recordings(24/96 and better), in our home computers... A good thing? Only if ALL THE REST of the gear is up to it AND if the material is "worth it"(Your call?)!?!? Matter of fact my software(Wavelab) does everything IN the computer at 32bit, then "drops back" to 24 bit at whatever sample rate(I use 48--- 24/48 to make my .wav files.) for any output to anything else.

    Anyway, with hard drives so cheap and processors so fast now(Along with no "wear factor", as tape and it's machines had.), there is almost no reason to even bother with 44.1/16 to do our initial recordings and/or masters, other than the possible dithering down, which can cause problems - but probably won't... Of course it SHOULD be counterproductive to START with anything lower than the "end result" desired(Trying to make an .mp3 INTO a 24/48 .wav - though even here, to do further editing, whatever, on a lower quality piece, "bumping it up" would still be the way to go.)

    Do your recordings at the highest rates you can(Don't get nuts! Unless you NEED 32/192, don't bother, you'll never hear it any better "at home" - if anywhere. 30ips tape machines would have been a huge waste, here at my house, too...).

    I'd stick with 24/48 for I/O, 32/48 for "internal" default on your computer, if available and 44.1/16 for "final" production of CD's or .mp3's...

    TG

    BTW: By the time I got to 3-3/4 I had pretty good success with girls. Actually, 1-7/8(Old audio cassette speed) seemed fine(Never got to use it anyway - don't think I could have?) with my 1st. grade gal pals... Later on, with my personal pet projects, I never even made it to 7.5, to say nothing of 15 or 30 which would have been entirely unnecessary under the circumstances(So I'm told..?)... Use what you have - just use it well......
     
  5. Cresta

    Cresta Active Member

    lol, you're right, I was referring to wordlenght :)

    unfortunately some plugin will not work over 48Khz-24bit but.... now I got two different answers: one is "it is not necessary to work with high wordlenght", one is "work every time you can with high wordlenght" :lol:
     
  6. RemyRAD

    RemyRAD Guest

    Actually the old NAB cartridge system utilized 3 tracks. 2 for music and one for primary, secondary and tertiary tones, which the machine could decode and would instruct it to stop back at the beginning of the song or, let you know it's getting ready to stop by flashing one of its front panel lights and faster flashing of the front panel buttons to thislet you know it's going to stop, or if the vocals were about to begin on the intro. They traveled at 7 1/2 IPS on specially back silicone lubricated recording tape.

    In the original stereo versions, the audio channels were handled as discrete left and right. But in subsequent later versions, a matrix was utilized so as to actually be able to record L+R and L-R as the 2 discrete channels. It would then be recombine in the output matrix back to its original left, right. So they were actually taking a stereo recording and converting it to MS! Why did they do this? Because when you carry the signal as discrete left and right, you're dealing with head azimuth alignment problems and a person listening in mono, would hear a cloudy muffled sound. By encoding to MS, the azimuth problems would simply make the stereo image shift slightly back and forth within the stereo sound field, while maintaining proper high frequency phase alignment. Later, folks even started using DBX noise reduction with the matrixed MS encoding. The best NAB cartridges ever sounded in the early mid-1980s.

    Old NAB cart jockey
    Remy
     
  7. Kev

    Kev Well-Known Member

    thanks Remy

    the L+R, L-R trick and it's variations do get used from time to time in the broadcast and pro-audio world.
    Digital has it's own new tricks and as is often the case,
    what is old becomes new and what is new becomes old
    :wink:
    old is good
     
  8. Cresta

    Cresta Active Member

    I feel A BIT confused... :shock:
     
  9. Mjolniir

    Mjolniir Guest

    You are the ultimate judge on if you "need" to record at 24bit. In my earliest days recording, I was using a pretty basic sound card and some very cheap mics and everything was recorded at 16bit and my sound was horrible. Then picked up a few sm57's and a few decent pre's and everything sounded amazingly better, crystal clear. When I got a better sound card and started recording with the same 57's and pre's, things improved some but not nearly as much as the difference between having cheap-ass mics and pre's and good ones. Now I've bought a few decent condenser mics and some even better pre's and again there has been a big leap.

    The bottom line is that making good records is not just about 16 or 24, it involves every part of the process from the mics, the pre's, the IO and conversion and processing and most of all, YOUR EARS. And none of them will get you chicks, by the way. That is a whole other post.

    If you know and appreciate all of this, then the question you should ask is "Will recording in 16bit limit what I can do with the sound?". The answer to that is yes, and for that reason, it is worth it to record at 24.
     
  10. Cresta

    Cresta Active Member

    hey..thanks, this was really helpful :)

    I mostly create music (for hobby uh) digitally, very few things recorded with microphones, so I wish also to know if there is difference in playing and exporting a (example) synth performance at 16 or 24 bit....
     
  11. FreakStudios

    FreakStudios Guest

    Well the Bits are really mainly for "volume" you might say. I mean thats what youll notice right away. a 16bit recording will be at a lower volume than a 24bit recording. other than that, the 24 bit recording has a better quality because of all the extra headroom + less noise because of the same thing
     

Share This Page