• Register
  • Help
Page 1 of 8 1234 ... LastLast
Results 1 to 10 of 74

Topic: why 2496 when we all finally bounce to 4416

Share/Bookmark
  1. #1

    why 2496 when we all finally bounce to 4416

    closed topic

  2. #2

    Re: why 2496 when we all finally bounce to 4416

    'cause we aren't only rendering to 44.1k/16-bit, there are hi-res options like DVD-A, SACD and DSD. Not to mention DSP, which benefits greatly from higher-res audio files when running its processes.
    Alan Lastufka | www.BelaDMedia.com
    Producer/Artistic Design | Content Producer

    20 Things

  3. #3

    Re: why 2496 when we all finally bounce to 4416

    Well, there are a lot of people who know more about this than I do and I'm sure they will add more meaningful insight, but here's my understanding of it. Typically, audio tracks are run through varying degrees of processing, and processing will degrade the sound to some extent. If you start with a recording that is recorded with more detail and at a higher quality than you are going to eventually mix down to, the final result will present a sound that is as good as you can get. If you start out with a sound that is as good as the final media is capable of delivering, when you process it and the quality is slightly diminished, your output will be slightly less than optimal. That is also why it is recomended that when you record something yourself, you should record at 24 or 32 bit.

    This reminds me of an argument I once read about the value of audio DVD and SACD formats that can reproduce high frequencies that are beyond the range of human hearing. It said that these frequencies are present in the recorded sound and as the sound bounces around a room, the frequency gets reduced each time it bounces, and eventually these reflections reach a frequency that can be perceived and therefore they do affect the sound. May be complete bull, but it makes sense to me. Discuss...

  4. #4
    Senior Member newmewzikboy's Avatar
    Join Date
    May 2005
    Location
    San Francisco Bay Area
    Posts
    1,700

    Re: why 2496 when we all finally bounce to 4416

    Instead of 24/96, try recording 24/88.2 and convert to 16/44.1 and see if you get better results. according to theory you should get a better result...

  5. #5

    Re: why 2496 when we all finally bounce to 4416

    There's a very important process that nobody has mentioned.

    DITHERING!! You have to dither a 24 bit wave to squeeze
    it into 16 bits.

    Dithering, aka, Noiseshaping is a very complex subject.
    Best to read up on it on your own when you have alot of time.

    Basically, Much of the higher resolution can and does come through
    when dithering to 16 bits.

    The better your summing and dithering, the better it will sound.

    Sorry, i dont have time to explain summing either...

    TK

  6. #6

    Re: why 2496 when we all finally bounce to 4416

    96 kHz (or 88.2 kHz for that matter) aren't all that important. They allow you to capture and playback frequencies beyond 20 kHz - as if my monitors reproduce them, or I can hear them. There may be some esoteric process that works better with the HF info, but I also know that the HF data can cause problems with non-linear processes, so not having to deal with them may actually be beneficial.

    24-bits is another story. It captures more detailed information, reduces noise and reduces phase errors. More accurate phase leads to more accurate imaging. Most important for us is the ability to process 24-bit information over and over. Each processing pass junks up the signal a bit, so you might end up with 23, 22, 21 or fewer bits of useful information. No problem! It's still better than 16-bits.

    If you start with 16-bits and start trashing it to 15, 14, 13... bits, you'll never get back to 16 good bits.

    I generally work at 24-bits, 44.1 kHz.

    -JF

  7. #7

    Re: why 2496 when we all finally bounce to 4416

    Open the can of worms!!!!

    We've visited this topic several times, do a search, i gave some details as to why capture at higher rates is beneficial.

    Even considering the Nyquist theorum, it is still invalid to say, @ 44.1k the frequency cutoff is at 22.05k and since we cannot hear past 20k, forget about it.' The lowpass filtering on the crystals begins well before 20k, so yes you are eating into valuable sonic realestate.

    Bob Katz and George Massenberg, as well as dozens of other very insighful audio gurus have published several articles on this debate. A few Googles will probably dig them up.

    One caveat - if you have crummy converters, it won't matter

  8. #8

    Re: why 2496 when we all finally bounce to 4416

    Quote Originally Posted by JonFairhurst
    96 kHz (or 88.2 kHz for that matter) aren't all that important. They allow you to capture and playback frequencies beyond 20 kHz - as if my monitors reproduce them, or I can hear them. .

    -JF
    it's my understanding that while this is theoretically correct re: the Nyquil theorem (or is that Nyquist?), the higher sample rates yield better transient response by virtue of the fact that there are more samples, and that's part of why they sound better. also, when it comes time for processing, the more numbers (higher resolution) to work with, the better.

    that being said, i was checking out some preamps a few weeks ago, and did a few passes at 96k, but didn't hear any substantial improvements. this was decidedly unscientific tho, and maybe it's just the converters on my 2408mkiii.

    just chiming in...

    -john
    John DeBorde

    Composer of Music for Film, TV and Interactive Media

  9. #9

    Re: why 2496 when we all finally bounce to 4416

    Quote Originally Posted by John DeBorde
    it's my understanding that while this is theoretically correct re: the Nyquil theorem (or is that Nyquist?), the higher sample rates yield better transient response by virtue of the fact that there are more samples, and that's part of why they sound better. also, when it comes time for processing, the more numbers (higher resolution) to work with, the better.
    Yes, Nyquist. The other is cold medicine.

    While I agree that the transient response is theoretically faster, it's debatable whether you can hear the difference. My monitors don't pass much of anything above 20 kHz, so in my case the transient response won't improve at all.

    A process like a gain or offset change, normailization or compression won't benefit at all from the higher sample rate. Those processes only touch the sample at hand, so more neigboring samples won't help.

    Filtering algorithms may benefit from a higher sample rate. Some don't. Let's say you're doing a boost at 1 kHz. Having data available above 20kHz doesn't affect it, so the high sample rate won't affect it. More complex filtering applications, like impulse convolution, may benefit sonically from more HF info, but even that is debatable.

    Now let's look at the cost: If I use 88.2 kHz samples I'll need twice as much disk space. Disks are cheap, so maybe that's okay.

    Processing is another story. The only benefit (maybe) is for convolution. But 88.2 kHz will eat twice the CPU. That means my five GigaPulse instance machine can only do two instances if I move to 88.2 kHz.

    Cutitng my processing capabilities in half for a benefit that I probably can't hear isn't worth it.

    24-bits is another story. The benefits are tangible and audible. It's the only way to ensure a top 16-bit product.

    I'm sticking with 2444.

    -JF

  10. #10

    Re: why 2496 when we all finally bounce to 4416

    I deliver everything at 48khz so if anything I'm aggravated at all the sample libraries that are sampled at 44.1khz but I realize that the majority of you do work for CD masters and not for AVID/FCP video.

    I just like to know that the option of going to 96khz is available to me, because sooner than later that will become the new standard for video and I like to know I'm ready for it without buying more gear (although I'm sure I will in the meantime!)
    >>Kays
    http://www.musicbykays.com
    Music Composition for Feature Films, Television and Interactive Entertainment

Go Back to forum
Page 1 of 8 1234 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •