Thread: An Article on Sample Rates

Results 1 to 20 of 35

  1. #1
    Join Date Apr 2010
    Location Second largest French City
    Posts 1,267
    Rep Power 536871041

    Default An Article on Sample Rates

    Womb experts; I am happy working in 44.1 kHz, however I hear a lot of noise from people telling me that I am somehow giving my clients less than they deserve.

    I found this article today that seems to help justify my position; I would love some expert input on whether it has merit.

    What say, you, experts?
    The emotionally and tonally deaf cannot hear the important things and never will. They will scream to the heavens the rest of us are wrong and produce useless papers to prove their point.
    Anybody claiming to be the first probably wasn't!
    www.timvandeven.com
  2. #2
    Join Date Oct 2011
    Location caprica
    Posts 1,558
    Rep Power 536871031

    Default Re: An Article on Sample Rates

    The article starts off on the whole "measured differences" thing which I think is bullshit.

    Then it goes on to list every exception and concludes that what you do with your EQ is more important than your sample rate. I agree with that.

    I work at 44.1 unless I am doing some bare bones stuff and then I will work at a higher rate, because I can afford the processing power. But when you've got 30 some odd tracks and heavy processing going on, higher sample rates can introduce headaches on older machines.

    weedy says he can hear a difference in the quality of plugin processing at 96khz. I tend to believe him.

    Also, dithering (or lack thereof) may factor into this, as well.
  3. #3
    Join Date Nov 2006
    Posts 5,361
    Rep Power 536871219

    Default Re: An Article on Sample Rates

    Well, the article seem to outline most practical points on the samplerate effects. There are some details, though.
    One point is regarding oversampling for DSP.
    The author writes
    But not so fast: most plugin designers, knowing this full well, have written oversampling into their code. Even in a 44.1kHz session, plugins that benefit from oversampling automatically increase their internal sampling rate. To gain the full benefits of this, itís important to note that the audio doesnít have to be recorded at this higher sample rate, itís just the processing that must happen at the higher rate.
    While this is true, imagine that you have a several plugins in a chain each using oversampling. Firstly, this would require more CPU than giving them oversampled data stream right away.
    But what's more, oversampling as everything in life is not perfect. It may have some ill-effects in the audible range (different ones depending on implementation).
    In my experience, which is mixing 100% in the box, 96k offers substantial sonic benefits.

    Next, another HUGE point is how your converters sound, and as author points out it is up to you to decide. But I am surprised that he mentions the resources needed to get 96k session running as something dramatic. In my opinion it is not.

    As for 192k, well, yeah, I'm with Lavry, there should be a point of diminishing returns.

    Anyway, why just not give it a try and see for yourself?
    When in doubt, mumble!

    EVERYTHING SOUNDS LIKE SHIT IF YA LISTEN LONG AND HARD ENOUGH.
  4. #4
    Join Date Apr 2010
    Location Second largest French City
    Posts 1,267
    Rep Power 536871041

    Default Re: An Article on Sample Rates

    I don't use plugs in my tracking/mixing. Does this negate the need for 96kHz?

    I was just curious about this article and sample rates in general because there is a lot of noise on the internet about the various rates.
    The emotionally and tonally deaf cannot hear the important things and never will. They will scream to the heavens the rest of us are wrong and produce useless papers to prove their point.
    Anybody claiming to be the first probably wasn't!
    www.timvandeven.com
  5. #5
    Join Date Sep 2007
    Location East Glenbeckistan
    Posts 3,796
    Rep Power 536871157

    Default Re: An Article on Sample Rates

    While I have no opinion on sample rates, bit depth is a different story. In my experience, more bits definitely result in better sound and going from 16 bit to 24 bit can be startling. Even when working with 16-bit source material, processing and D/A at 24 bit can make a huge difference, especially with high-dynamic range material on quiet passages. And reverb tails that don't abruptly cut off.
    "Far too many people suffer from the anal-cranial inversion. They got their heads stuck up their asses!" - Tom Dowd
    "A great, timeless album is more than just an entertainment app." - Don Was
    "Success does not mean success. You are only as successful as your last hit record." - Patrick Adams (record producer)
    "The tech sector is like a guy in a crowded prison cell, who finds a hand grenade, and sets it off, so that he'll have the entire cell to himself." - Dwoz
  6. #6
    Join Date Apr 2010
    Location Second largest French City
    Posts 1,267
    Rep Power 536871041

    Default Re: An Article on Sample Rates

    While I have no opinion on sample rates, bit depth is a different story. In my experience, more bits definitely result in better sound and going from 16 bit to 24 bit can be startling. Even when working with 16-bit source material, processing and D/A at 24 bit can make a huge difference, especially with high-dynamic range material on quiet passages. And reverb tails that don't abruptly cut off.
    I always use 24 bit.
    The emotionally and tonally deaf cannot hear the important things and never will. They will scream to the heavens the rest of us are wrong and produce useless papers to prove their point.
    Anybody claiming to be the first probably wasn't!
    www.timvandeven.com
  7. #7
    Join Date Nov 2006
    Posts 5,361
    Rep Power 536871219

    Default Re: An Article on Sample Rates

    I don't use plugs in my tracking/mixing. Does this negate the need for 96kHz?
    Partially, but that also puts more emphasis on how your converters sound (as it is printed if you use them as output for processing/summing)(not to say that the performance of the converters is irrelevant if you mix ITB), and that also makes it easier to run 96k files too!

    I was just curious about this article and sample rates in general because there is a lot of noise on the internet about the various rates.
    Well, there's noise about anything on those internets, no?
    Anyway, up to a point it's always no worse to run higher rates soundwise.
    When in doubt, mumble!

    EVERYTHING SOUNDS LIKE SHIT IF YA LISTEN LONG AND HARD ENOUGH.
  8. #8
    Has Many Personal Intergritys Expensive Boat Anchor
    Join Date Nov 2006
    Location New York
    Posts 10,666
    Rep Power 536871395

    Default Re: An Article on Sample Rates

    not only is there "noise' but the prevailing GS attitude tends to be always in favour of the "you don't need anything" meme, even as it worships blindly at the feet of one piece of fetish gear or another.

    so "all you need is a 57"... and "all you need is 44.1" , even though you "need" 26 different mic pres

    underlying all this is, of course, the desperately, but fragilely, held belief that 'you don't need a professional studio or gear or people to do anything. you can do it all yourself at home'

    so anything that suggest you don't 'need' the higher end (or perceived as more 'pro') option is going to be eagerly jumped on as validation every time.


    in this case, I hear 96k as sounding hugely better.
    Is it about 'implementation'? I don't actually care.

    I have heard some (well, one) A-D convertor that sounded so great that it beat, at 44.1, others working at 96.
    BUT... all it did was 44.1. it didn't HAVE a 96 option.

    I have said this before, and some others said they had different experience, but I have NEVER heard an A-D convertor that went up to 96k not sound better at 96k.
    They ALL do.

    and this is leaving out the issue of plug ins sounding better at higher sample rates.


    so for me, it's simple.
    is there a REASON (such as track count) to not do it at 96?
    if there isn't, I do.

    really the question is why WOULDN'T you?

    UNLESS, you're just trying to prove how DIY, indie, GS and rebellious you are.

    the first guy to play a guitar solo with gloves on is a youtube phenomenon.
    The 25th or 200th guy to do it is just a sheep.
    with gloves on.
  9. #9
    Join Date Apr 2010
    Location Second largest French City
    Posts 1,267
    Rep Power 536871041

    Default Re: An Article on Sample Rates

    not only is there "noise' but the prevailing GS attitude tends to be always in favour of the "you don't need anything" meme, even as it worships blindly at the feet of one piece of fetish gear or another.

    so "all you need is a 57"... and "all you need is 44.1" , even though you "need" 26 different mic pres

    underlying all this is, of course, the desperately, but fragilely, held belief that 'you don't need a professional studio or gear or people to do anything. you can do it all yourself at home'

    so anything that suggest you don't 'need' the higher end (or perceived as more 'pro') option is going to be eagerly jumped on as validation every time.


    in this case, I hear 96k as sounding hugely better.
    Is it about 'implementation'? I don't actually care.

    I have heard some (well, one) A-D convertor that sounded so great that it beat, at 44.1, others working at 96.
    BUT... all it did was 44.1. it didn't HAVE a 96 option.

    I have said this before, and some others said they had different experience, but I have NEVER heard an A-D convertor that went up to 96k not sound better at 96k.
    They ALL do.

    and this is leaving out the issue of plug ins sounding better at higher sample rates.


    so for me, it's simple.
    is there a REASON (such as track count) to not do it at 96?
    if there isn't, I do.

    really the question is why WOULDN'T you?

    UNLESS, you're just trying to prove how DIY, indie, GS and rebellious you are.

    the first guy to play a guitar solo with gloves on is a youtube phenomenon.
    The 25th or 200th guy to do it is just a sheep.
    with gloves on.
    Thanks, Weedy; I was mainly curious because the gear that I own only goes as high as 48kHz. I haven't the funds for what I'd like to have as converters at the moment, but I was curious to hear actual professional responses (unlike at GS or even some of the local dopes that I know that have no ears).

    I assure you; it's a hardware limitation that prevents me from trying the other sample rates, I despise "indie-cred" as a reason for trying to justify a shitty-sounding recording.
    The emotionally and tonally deaf cannot hear the important things and never will. They will scream to the heavens the rest of us are wrong and produce useless papers to prove their point.
    Anybody claiming to be the first probably wasn't!
    www.timvandeven.com
  10. #10
    Has Many Personal Intergritys Expensive Boat Anchor
    Join Date Nov 2006
    Location New York
    Posts 10,666
    Rep Power 536871395

    Default Re: An Article on Sample Rates

    fair enough

    but then I have to say, I find 48k is a massive jump up from 44.1 as well

    maybe even bigger than the one from 48 to 96

    why WOULDN'T you do 48 all the time?

    it doesn't even have a track or CPU usage penalty relatively speaking.


    we're stuck with 44.1 for now in CONSUMER formats... but there is simply no reason to use it on the multitrack recording end.
  11. #11
    Join Date Apr 2010
    Location Second largest French City
    Posts 1,267
    Rep Power 536871041

    Default Re: An Article on Sample Rates

    fair enough

    but then I have to say, I find 48k is a massive jump up from 44.1 as well

    maybe even bigger than the one from 48 to 96

    why WOULDN'T you do 48 all the time?

    it doesn't even have a track or CPU usage penalty relatively speaking.


    we're stuck with 44.1 for now in CONSUMER formats... but there is simply no reason to use it on the multitrack recording end.
    I didn't think there would be a major difference, but I'll give 48 kHz a shot. Thanks for the advice!

    This is another reason why I like this place.... the signal to noise ratio is mostly signal.
    The emotionally and tonally deaf cannot hear the important things and never will. They will scream to the heavens the rest of us are wrong and produce useless papers to prove their point.
    Anybody claiming to be the first probably wasn't!
    www.timvandeven.com
  12. #12
    Join Date Nov 2006
    Posts 6,812
    Rep Power 2500000000

    Default Re: An Article on Sample Rates

    I found that article to be complete and utter bullshit.

    The dialectic process the author follows is based on bait-and-switch, half truths, arguing from specific to general and vice versa, and all kinds of other problems.

    Basically, here's my short list of problems with his points:

    1) "cheap" converter implementation these days is indeed better than it was 10 years ago, but is still much worse than he allows for.

    2) Nyquist theory discusses sines. it does not discuss music. Therefore if you're recording sine waves, you needn't bother with any of this farting around. If you're recording music, you will find that limiting bandwidth to 20kHz (or actually more like 18k or 19k in a cheap converter) WILL result in issues. All the credible evidence I've seen points to us needing to go up to PERHAPS 30kHz to capture all signal that will matter in a production system.

    3) in spite of what the ethan winers of the world say, there is a difference between a production system and a RE-production system. Signal that is irrelevant in a REPRODUCTION system MAY BE relevant in a production system.

    4) Microphones do pick up audio past 20k. In their official specs, they typically list, "20Hz-20kHz" which are the -3dB down points, not the brick wall end-of-all-signal.

    5) The argument today is that 44.1k sample rates place too much faith in the design of the brick-wall filters required to support that rate AND satisfy the human hearing range. Also that a high-frequency transient has some information above 20kHz that may not be perceptible in and of itself, but that is significant in some situations in how signals are summed and processed downstream of the converters. i.e. it is NOT a self-evident assumption that 20k is the highest frequency that MATTERS, just because that's the highest frequency that many people HEAR. In a PRODUCTION system.

    6) The author wanders dangerously into the weeds claiming that higher sample rates are less accurate. This would be true, but for the vastly improved clocks that he mentions earlier. Higher sample rates (and higher bit rates) significantly change the nature of the problem of quantization error, making it a much simpler problem.

    In conclusion: the author is confused.
    "...but ma, audio engineering IS gainful employment!..."

    "...If I wuz at that club where Miles played one note I would have bounced ONE BOTTLE off his shiny fucking coconut. What? He's Phil Glass now?..." -Slipperman

    "...never attribute to magic, that which can be explained by conspiracy."
  13. #13
    Martini Drinker Van Morrison's enunciation coach
    Join Date Nov 2006
    Posts 3,209
    Rep Power 2500000000

    Default Re: An Article on Sample Rates

    there is a difference between a production system and a RE-production system. Signal that is irrelevant in a REPRODUCTION system MAY BE relevant in a production system.
    Stop making fun of my Rane mixer.
  14. #14
    Join Date Sep 2007
    Location East Glenbeckistan
    Posts 3,796
    Rep Power 536871157

    Default Re: An Article on Sample Rates

    I found that article to be complete and utter bullshit. SNIP

    In conclusion: the author is confused.
    On re-read I find the article to be HUGELY political. Like the author has a bone to pick with someone.
    "Far too many people suffer from the anal-cranial inversion. They got their heads stuck up their asses!" - Tom Dowd
    "A great, timeless album is more than just an entertainment app." - Don Was
    "Success does not mean success. You are only as successful as your last hit record." - Patrick Adams (record producer)
    "The tech sector is like a guy in a crowded prison cell, who finds a hand grenade, and sets it off, so that he'll have the entire cell to himself." - Dwoz
  15. #15
    Join Date Oct 2011
    Location caprica
    Posts 1,558
    Rep Power 536871031

    Default Re: An Article on Sample Rates

    It sounds to me like he was trying to say "both sides are right" when in fact, one side is right because they are the ones making the actual decisions with the audio, and the other side's opinion totally does not matter whatsoever.
  16. #16
    Join Date Sep 2009
    Location Quake City
    Posts 11,777
    Rep Power 536871398

    Default Re: An Article on Sample Rates

    There was one assertion in the article that really jumped out at me as bullshit (in spite of all the other bullshit surrounding it). That assertion was that analog circuits become non-linear at the "ultra high" frequency of 192kHz.

    WTF????

    Has this guy never heard of analog radio?

    Total poppycock and flapdoodle!
    http://www.johnnyoklahoma.com/

    Originally Posted by Bob Ohlsson
    Everything is some mixture of awesome and suck. We simply want the awesome to be highlighted sufficiently that it distracts listeners from the suck.
    Originally Posted by Bob Ohlsson
    The appropriate role for science is the study of observed phenomena to gain an understanding. It is not dictating what people ought or ought not to be observing.
    Hey, if I'm Grumpy, where the hell is Snow White????
  17. #17
    Join Date Nov 2006
    Location Songwriter Gulch
    Posts 10,141
    Rep Power 536871379

    Default Re: An Article on Sample Rates

    ..., I find 48k is a massive jump up from 44.1 as well

    maybe even bigger than the one from 48 to 96

    why WOULDN'T you do 48 all the time?

    it doesn't even have a track or CPU usage penalty relatively speaking.


    we're stuck with 44.1 for now in CONSUMER formats... but there is simply no reason to use it on the multitrack recording end.
    We aren't even stuck at it in consumer formats other than Apple demanding 44.1 so people can burn CDs from i-tunes files.

    I sent a friend some 96 x 24 AAC files of his jazz album and he flipped out over how great they sounded. The encoding reduces the file size a lot so I got away with not even telling him they were 96 k. until he called back wondering what I'd done.

    Unfortunately they won't play on an i-pad or i-phone but we certainly can do 48x24 and they sound better than 44.1. as do 48k MP3s.
    Bob's room 615 562-4346
    Georgetown Masters 615 254-3233
    Interview
    Artists are the gatekeepers of truth!- Paul Robeson
  18. #18
    Join Date Nov 2006
    Location Sweden
    Posts 14,276
    Rep Power 536871516

    Default Re: An Article on Sample Rates

    The article?

    Regurgitations, and in some cases mild misrepresentations of the opinions of authorities like Dan Lavry.

    Authored by a guy with ONE mix assistant credit on ONE established artist's record.

    Not for nothing, but I think you can do far better right here on this forum.


    otek
    "Tube color is not the 'thing'. Why would the most linear amplifying device have a color?" - Jonte Knif
  19. #19
    Join Date Nov 2006
    Location Songwriter Gulch
    Posts 10,141
    Rep Power 536871379

    Default Re: An Article on Sample Rates

    My experience with digital audio has been that it's really all about implementation.

    Theory tells us what is potentially best or potentially sufficient but real world implementation will often lead to different conclusions that are specific to one piece of gear and certain specific software. This is why using your ears is so important.

    That said, thus far I agree with William. I've always found the biggest improvement to be going to 48 and ironically this has often sounded even better than recording at 88.2. 96k is even better. Dan says we ought to be recording at 60 but the gods of Sony and Panasonic dictated our current sample rates so we are stuck with them.

    Also just like analog tape, the biggest improvement is gained while recording off the floor.
    Bob's room 615 562-4346
    Georgetown Masters 615 254-3233
    Interview
    Artists are the gatekeepers of truth!- Paul Robeson
  20. #20
    Join Date Apr 2010
    Location Second largest French City
    Posts 1,267
    Rep Power 536871041

    Default Re: An Article on Sample Rates

    My experience with digital audio has been that it's really all about implementation.

    Theory tells us what is potentially best or potentially sufficient but real world implementation will often lead to different conclusions that are specific to one piece of gear and certain specific software. This is why using your ears is so important.

    That said, thus far I agree with William. I've always found the biggest improvement to be going to 48 and ironically this has often sounded even better than recording at 88.2. 96k is even better. Dan says we ought to be recording at 60 but the gods of Sony and Panasonic dictated our current sample rates so we are stuck with them.

    Also just like analog tape, the biggest improvement is gained while recording off the floor.
    This is great; thanks very much, Bob.

    This place rocks.
    The emotionally and tonally deaf cannot hear the important things and never will. They will scream to the heavens the rest of us are wrong and produce useless papers to prove their point.
    Anybody claiming to be the first probably wasn't!
    www.timvandeven.com

Posting Permissions

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