Sample Rates Overrated?

Discussion in 'DAW' started by Davey Jones, Apr 21, 2016.

  1. Andrew

    Andrew AudioSEX Maestro Staff Member

    Joined:
    Oct 24, 2011
    Messages:
    1,981
    Likes Received:
    1,201
    Location:
    Between worlds
    24bit depth is more than enough for that.
    My comment was aimed more towards high sample rate, which is for the most part useless in audio chain.
     
  2. stevitch

    stevitch Audiosexual

    Joined:
    Aug 27, 2014
    Messages:
    869
    Likes Received:
    582
    Location:
    Here
    Take a an audio file that's a rip from a vinyl record at "24/96" and downsample it to "16/44.1,": and zoom in on the respective waveforms to, say, the same peak or valley in the timeline. You will see a difference, with the former's waveform appearing smoother, or less "stairstepped," than the latter's.

    Then comparatively listen to both, and see if you can tell a difference. Have someone else do the play back for you, so you don't know which is which. Will you be able to tell a difference?
     
  3. bluerover

    bluerover Audiosexual

    Joined:
    May 3, 2013
    Messages:
    1,249
    Likes Received:
    1,124
  4. fiction

    fiction Audiosexual

    Joined:
    Jun 21, 2011
    Messages:
    1,889
    Likes Received:
    687
    Yep, that's what I meant: Better use 24bit than 16bit resolution.

    What I find very interesting is the fact that quite a number of Audio enthusiasts swear by the surprising sound improvement using their own upsampling device which in fact is an SRC upsampling 44k1 or 48k to 96k of the digital output of any audio player, followed by D/A conversion.
    And that's all after production! :mad:
     
  5. Iggy

    Iggy Rock Star

    Joined:
    Jun 21, 2011
    Messages:
    1,092
    Likes Received:
    434
    Location:
    The stage, man
    While I understand and respect your opinion on why you prefer lower bit-depths, you haven't given any reasons why you're against mixing and mastering at higher bit-depths and sample rates. Most machines and hardware are more than capable of mixing and mastering at 96 kHz/24-bits and even higher. If you have the capability, why not utilize it? I could understand if the reasoning was "My processor can't handle that kind of load" or "I only have a DAT deck from 1995" or even "my interface sucks", but you keep saying it's a "waste". What, exactly, is anyone "wasting" by mixing and mastering at a higher bit-depth and sample rate? Hard drive space is cheap and plentiful. Internet speeds are fast enough that it takes virtually no time to download the 200 MBs or so of a standard four-minute song rendered at 96 kHz/24-(or even 32-) bit. Isn't it more of a waste to buy a fast computer, a decent audio interface and cutting edge audio software just so you can mix and master at CD quality or lower? You might as well be doing all your audio work on an el cheapo Zoom digital multitrack unit … except those go up to 96 kHz/24-bit nowadays, too. The issue is clearly a divided one; some people, like me, can hear a clear difference between CD quality and higher resolution, while others can't. There's plenty of data and personal experience out there that seems to support both points-of-view … yet, I still don't understand why you or anyone else would be against higher bit-depths and sample rates. Would it be better for music and audio if we only did everything at CD quality? Music was mastered at 48 kHz/24-bits since the dawn of digital audio and CDs, then, went up from there, so you can't even really go backwards, technology-wise. It's been the standard from Day One. I can understand you having a personal preference, but to be actively railing against higher resolution like it's hurting music somehow is just confusing, to say the least. It's also potentially damaging to advise anyone not to go above CD quality, considering the standard for HD broadcast and DVD (on the verge of going extinct) is 48 kHz/16-bit and uncompressed Blu-ray/UHD audio is 96 kHz/24-bit … meaning that someone who was scoring a movie or planning on licensing their music for television (currently the highest-paying outlet for new music) will be totally screwed when they can't provide a broadcast quality master.
     
    Last edited: Apr 24, 2016
    • Agree Agree x 1
    • Interesting Interesting x 1
    • List
  6. fiction

    fiction Audiosexual

    Joined:
    Jun 21, 2011
    Messages:
    1,889
    Likes Received:
    687
    The problem with these theoretical verdicts is that they don't judge by the ear.
    What if blind listening tests prove exactly the opposite? That happens so often...

    My conclusion is: The human ear and perception of sound is a much more subjective matter than most think.
    My latest revelation was the KVR thread on U-He's new free alpha emulation of a Sequential Circuit Pro-One.
    see here
    The idea was to find the most "analog" sounding filter out of 5 anonymous filter models.
    The fact that the filter with the most ugly digital artefacts has amongst the highest rating for "analog sounding" is close to ridiculous.
     
    Last edited: Apr 24, 2016
  7. Matt777

    Matt777 Rock Star

    Joined:
    Oct 17, 2015
    Messages:
    602
    Likes Received:
    410
    While I agree with you in most points, I only have one question: at the moment I'm at the laptop + novation nio 2/4 ext audio card which handles 48/24.
    If this was my only setup (fortunately it's not..) would your advice be to immediately stop making any music until I get a "proper" interface - because everything that I produce will be useless? If this was my only setup would you advise to first invest in a new interface - no other gear/sw? (let's suppose everything produced in the box, so preamps and stuff don't matter..)
     
    Last edited: Apr 24, 2016
  8. Von_Steyr

    Von_Steyr Guest

    @Iggy
    I agree with most you have said,though saying you will be screwed....there are several factors,you are generalizing things.
    Yes some may indeed complicate,most wont.Its not as black as white as you state here.
     
  9. Rasputin

    Rasputin Platinum Record

    Joined:
    Jun 29, 2012
    Messages:
    362
    Likes Received:
    233
    You're right in terms of metering expressions, I suppose, but remember that bit-depth expresses dynamic range, and thereby also the ratio of signal-to-noise. Since noise tends to stay at the same level (or can't be diminished past a certain point) the most advantageous thing is to add the ability to have a stronger signal. Since we're adding the ability to be "louder" then we're raising the point at which the top of our signal won't be lopped off like a guillotine. Extra headroom.

    Of course all these metaphors fall apart on some level when the actual math is involved, so there's no "correct" terminology. Sometimes it just comes down inheriting things from the analog world.

    I mean, every time you add a bit then you're adding headroom, footroom, kneeroom, neckroom, etc. simultaneously. It's like blowing up a balloon, not adding cargo space in your vehicle.
     
  10. Rasputin

    Rasputin Platinum Record

    Joined:
    Jun 29, 2012
    Messages:
    362
    Likes Received:
    233
    Because you can't just take a file and re-save it at a higher sample-rate/bit-depth? Who is going to call your bluff? The dogs, bats, and dolphins that can hear the ultrasound up past 20kHz?

    Remember, as soon as your WAV files hit your DAW mixer or a VST/AAX/whatever, you've already converted to a higher bit-depth anyway. All you have to do is save it. The initial bit-depth is only relevant in terms of capturing your source above an acceptable noise floor. You either hit your target there or you don't.
     
    Last edited: Apr 24, 2016
  11. Rasputin

    Rasputin Platinum Record

    Joined:
    Jun 29, 2012
    Messages:
    362
    Likes Received:
    233
    Subjective: In other words, skewed from reality by the biases of the observer. Seems like what we want to avoid here.

    Now, creative choices and what is "appealing" or desirable, yes, those are subjective. If people like warm, aliased, distorted, hiss, etc. and want to associate that with analog or whatever then great. By all means, throw some artifacts and non-linear distortions in your audio, just don't argue that those are more true-to-life than the mathematically correct model.

    I think people confuse what may sound better to them with what is technically correct. I like "crappy" Ensoniq/Emu converters in my songs. Doesn't mean that should be the universal standard for capturing digital audio though.
     
    Last edited: Apr 24, 2016
    • Agree Agree x 2
    • Like Like x 1
    • List
  12. oisinn

    oisinn Ultrasonic

    Joined:
    Jul 6, 2011
    Messages:
    205
    Likes Received:
    39
    Location:
    Wild West
    Bob Katz always upsamples and ups the bit depth of any material before he masters it. It doesn't matter if its ultimately going to be dithered down for say CD delivery at 16bit and 44.1 because the finished product sounds better when he uses his processing and plugins at high resolution - less aliasing, artifacts and phase issues:like:
     
  13. Zuluw

    Zuluw Noisemaker

    Joined:
    Apr 24, 2016
    Messages:
    1
    Likes Received:
    3
    Location:
    France
    Wrong.
    What you see is a graphical representation of digital information.
    The stairsteps don't exist.

    see here:
     
    • Like Like x 3
    • Winner Winner x 1
    • List
  14. spacetime

    spacetime Platinum Record

    Joined:
    Jan 28, 2016
    Messages:
    418
    Likes Received:
    171
    There are excellent focal monitors, alpha 50 and 80 for 250-350ea
     
  15. bluerover

    bluerover Audiosexual

    Joined:
    May 3, 2013
    Messages:
    1,249
    Likes Received:
    1,124
    Dan Lavry makes some of the finest converters in the business.

    He does say that using 88.2 and 96 is fine, but the optimal frequency, according to his experience, is 60. Anything above 96 could add anomalies into your signal.

    Again, it's the quality of your converter and clock, along with other factors that count; not the sampling frequency. Weiss vs Focusrite @ 16/44.1? Weiss - night and day.
     
  16. Matt777

    Matt777 Rock Star

    Joined:
    Oct 17, 2015
    Messages:
    602
    Likes Received:
    410
    There is no spoon..!
    [​IMG]
     
  17. Blue

    Blue Audiosexual

    Joined:
    Sep 6, 2015
    Messages:
    1,807
    Likes Received:
    954
    Yeah exactly!! For me all my projects are in 44k. 48K is the DVD standard format for the audio.
     
  18. Iggy

    Iggy Rock Star

    Joined:
    Jun 21, 2011
    Messages:
    1,092
    Likes Received:
    434
    Location:
    The stage, man
    I think you misunderstood my point. First, I'm not even sure what constitutes a "proper" audio interface, I was merely using that as an example of someone giving a reason why they would prefer to record at 44.1 kHz/16-bit and slam anything else as "wrong". If you had a 48 kHz/24-bit interface (and to give you context, my first interface was a MOTU 828 mk1 FW with no preamps and only 48 kHz/24-bit resolution. That's what I immediately started recording all my audio at, even though I'm sure those first-gen converters were probably crap) and it sounded good, why wouldn't you record and mix at that resolution? Kwiss was claiming that anything over 16-bits was a "waste", without explaining why it was a waste.

    Okay, this is just confusing. You're arguing that you can just save a 44.1 kHz/16-bit audio file at a higher bit-rate and sample depth and it would be just the same as if you'd recorded, mixed and mastered at 96 kHz/24-bit?!? How the hell is that possible?!? If you up-coverted a CD-quality audio file, it's still only CD quality. You can't add what isn't already there. That would be like taking an old VHS tape and digitizing it to HD video. Sure, you can burn it to Blu-ray, but the quality of the picture isn't going to improve -- it's still the same resolution as that VHS tape was. If you're the type of person who believes nothing exists beyond 16-bits and 44.1 kHz, I guess this isn't a problem, but anyone who believes there is actually a difference in quality. Plus, the irony is, you wound up up-converting your CD-quality sound file to the exact same format you were shunning by mixing and mastering at 44.1 kHz/16-bit in the first place. If you're prepared to up-covert your files … why not just mix and master at that higher resolution and down-convert to CD quality?!?

    Again, I'm reading a lot of arguments about not being able to hear the difference between 44.1 kHz and 96 kHz, but so far, no one has provided a reason why you shouldn't mix and master at a higher resolution. Will it blow up your computer or something if you mix at anything higher than 44.1 kHz/16-bit?
     
  19. oisinn

    oisinn Ultrasonic

    Joined:
    Jul 6, 2011
    Messages:
    205
    Likes Received:
    39
    Location:
    Wild West
    Some of you guys must have learning difficulties. The reason to up the sample rate and bit depth for mixing and mastering (even if you recorded at 16bit and 44.1 Khz and intend to dither back down again to release the finished product at that resolution) is to get better response from your plugins. WHY DO SO MANY OF THE BEST PLUGINS OFFER x2 x4 x8 OVERSAMPLING THEN?!?!?!
     
  20. Andrew

    Andrew AudioSEX Maestro Staff Member

    Joined:
    Oct 24, 2011
    Messages:
    1,981
    Likes Received:
    1,201
    Location:
    Between worlds
    Sample rate and bit depth do not necessarily go together.
    It's best to use at least 24bit for production. No problem there, in fact 32bif FP WavPack is both convenient and small solution.
    FP also allows extra headroom to avoid clipping during production (apart from some vintage plugs)

    You got a point there, oversampling does enhance precision at a cost of very high CPU usage, but the result might not be audible. Sure there's no harm in doing so as long as your HW can handle it. Oversampling is also commonly used in DAC chips to improve precision of Digital-to-Analog conversion. Usually on a much higher scale, like 256x or 192x.
    That's why most of 90s CD playback hardware sound inferior to modern devices, as the amount of filters used and processing power wasn't great at that time. Most early discmans truncated the bit depth to just 12bits.
     
Loading...
Similar Threads - Sample Rates Overrated Forum Date
UVI Workstation question about Sample Rates Software Feb 15, 2024
arturia CMI V lockup on oddbal sample rates Software May 18, 2023
Kontakt Libraries - Sample Rates Kontakt Oct 7, 2021
Gregory Scott (The House of Kush) on high sample rates Mixing and Mastering Oct 3, 2020
Help with sample rates.. Working with Sound May 25, 2016
Loading...