• Register
  • Help
Page 1 of 2 12 LastLast
Results 1 to 10 of 15

Topic: Sample Rate Question(?)

Share/Bookmark
  1. #1
    Senior Member Styxx's Avatar
    Join Date
    Mar 2004
    Location
    West Seneca, NY
    Posts
    11,075

    Question Sample Rate Question(?)

    On my PreSonus FirePod control for hardware settings there's a choice of sample rates from 44.1 kHz to 96.0 kHz. Will changing the rate to a higher setting help me to obtain better quality playback without clipping?
    I am working on a piece that is driving out on my mind on one beat where there is a tuti fruity. The horns are dominant yet if I back them off too much I loose the whole dynamics not to mention how angry they get and start throwing their mutes at me!
    So, what do you say? Or should I just toss the whole mess in the trash bin and call it a year?

    Now I know there are very intelligent musicians out there that will read this and not say a word SO I AM BEGGING YOU PLEASE PLEASE DON'T TURN YOUR BACK ON ME I DIDN'T DO IT THE PERCUSSIONIST DID THOSE FRUITCAKES!
    Styxx

  2. #2

    Re: Sample Rate Question(?)

    I don't have your particular unit (or GPO for that matter...), but I really don't see how putting it up to 96khz will help with clipping - totally different thing. The best bet is to raise your buffer settings. Have you ever had these sort of clipping problems before?

  3. #3

    Re: Sample Rate Question(?)

    Hi Styxx!

    It will make a difference in the quality. You might be able to do it. You can always give it a try and change it back if it doesn't work. I haven't been able to successfully raise the sample rate on my rig though.

    Steph
    (mm)

  4. #4
    Moderator/Developer Brian2112's Avatar
    Join Date
    Nov 2003
    Location
    Out of my Mind
    Posts
    1,858

    Re: Sample Rate Question(?)

    Typically, I believe that a higher sample rate and bit depth will give you more headroom. When dithered down to 44.1 – 16 bit, it will simply compress the peak a little. A good dither will give the illusion of maintaining the higher dynamic range.



    I always work at 48k 24 bit.

    …2112

    EDIT:
    Location: The man's nuts grab 'em! I'm not going to grab 'em you grab 'em!
    ROTFL!!!!!
    "So what if some parts of life are a crap shoot? Get out there and shoot the crap." -- Neil Peart
    Hint:1.6180339887498948482 Φ

  5. #5

    Re: Sample Rate Question(?)

    Styxx,

    Though i don't have a nice sound card (or even a decent GPO computer - see below in my signature) ( I just use my mac's system one), I do know from my own experiences that RAISING the resolution (In GPO Studio) only makes the clicks happen more (at least for me). GPO Studio is the only place I can really change my resolution size. But in my setup - the computer's CPU handles the full load of the sound card, as well as the other functions of the computer. SO REDUCING for me has helped eliminate the clicks and pops and stuff. It does change the quality a little - but not that bad - my contest piece was done with a sample rate of 19,000. No one seemed to have noticed when listening, so I guess it wasn't that bad.



    Jerry
    MacBook Pro Intel Core 2 Duo 2.5ghz 4GB Ram OSX 10.5.8
    Korg TritonLe & MAudio Oxygen 8
    T3, Logic 9, DP7, K2, GPO4, Strad, Gofriller, C&MB, Finale 2010

    My Website!
    New Film Scores!
    Also, Ever think about having your very own personal documentary? See my new Website!
    http://www.mylegacyfilm.com


  6. #6
    Senior Member Styxx's Avatar
    Join Date
    Mar 2004
    Location
    West Seneca, NY
    Posts
    11,075

    Re: Sample Rate Question(?)

    Quote Originally Posted by Brian2112
    Typically, I believe that a higher sample rate and bit depth will give you more headroom. When dithered down to 44.1 – 16 bit, it will simply compress the peak a little. A good dither will give the illusion of maintaining the higher dynamic range.



    I always work at 48k 24 bit.

    …2112

    EDIT:

    ROTFL!!!!!
    Might have to try tonight.
    Styxx

  7. #7

    Re: Sample Rate Question(?)

    Actually, sample rate has nothing to do with dynamic range. Bit depth does. 24bit has a greater dynamic range than 16bit (144db at 24bit compared to 96db at 16bit; 6db per bit.) 44.1k at 16bit has exactly the same dynamic range as 96k at 16bit or 192k at 16bit. One other thing: The dynamic range gains of higher bit depths are at the *bottom* of the dynamic spectrum - not the top. Digital "0" is digital "0" at any bit depth. The gain with higher bit depths is resolution at the soft end of the spectrum. Also, I personally never record at 48k if the final product is going to be 44.1. There is more loss in the down-conversion from 48 to 44.1 than any supposed gains in the higher sample rate during recording. This is due to the uneven fractional math involved in the down-conversion - unavoidable "rounding" errors in the math with consequent inaccuracies. On the other hand, if you are working in video and the final product is going to be at 48k then it's best to record at that rate. Rule of thumb: Avoid uneven fractional conversions. Recording at 88.2k and down-converting to 44.1k is an even (divide by 2) conversion without the inaccuracies of uneven conversions. On the other hand, 96k to 44.1k is a down-conversion best avoided for the same reasons.

    Now, back to your original question: The correct approach to gain-staging problems is to identify the loudest point in your mix and balance all instruments relative to one another to play without distortion at that point. In other words, if the horns are distorting at measure 53 then use that spot as your reference. First, back off on the horns until the distortion goes away. Then lower all other instrument levels to re-establish your instrumental balances. Or, alternatively, maintain your relative balances and back off on everything in the mix until the distortion disappears. The softer sections of your piece will now be lower in level but the peak at bar 53 will be cleanly reproduced. Under certain circumstances you will need to do some additional detective work to see exactly where in the chain the distortion is being generated (which gain-stage.) Use a process of elimination to narrow your search for the location of the problem. Wherever it occurs that location will necessarily become your reference point for balancing all other elements of your mix. Always remember that instrument voltages are added together as instrument count increases and this can easily lead to later gain-stage overload. The greater the number of instruments, the higher the chance of overload (that is, if you don't reduce the overall volume of all instruments in the mix.) To mitigate the problem, All instruments must be backed off in the mix at a point in the amplification chain prior to the overloaded amplification stage.

    Tom

  8. #8
    Senior Member Styxx's Avatar
    Join Date
    Mar 2004
    Location
    West Seneca, NY
    Posts
    11,075

    Thumbs up Re: Sample Rate Question(?)

    Tom, you are absolutely correct! Changing the bit rate did nothing to solve the problem and on the other side of the coin, neither did changing the mix. I am working with a midi sent to me and produced on some sort of keyboard system. The culprit turned out to be in the bass having been a series of "machine gun" repeated notes at high velocity. After deleting the suspected section the problem was solved. However, investigating the mix proved to be prosperous as well in so much as all is balanced nicely now.

    To all, thanks for the help and professional insight!
    Styxx

  9. #9

    Re: Sample Rate Question(?)

    Quote Originally Posted by Tom Hopkins
    Actually, sample rate has nothing to do with dynamic range. Bit depth does.
    Do you think there is anything to be gained by working at 44K and 24 bit?
    A problem many of us have with computer generated audio,is that the end result is a bit flat.Lack of depth of the sound field, a loss of "fairy dust".
    I seem to remember reading articles on the net,that this is a bit-depth(or is that word-length )issue.I think the article said that if you render audio with effects that change it's level,(e.g. compression) this is better done at a higher bit-depth than 16bit.
    Is there any truth in this in your view?
    regards

  10. #10

    Re: Sample Rate Question(?)

    I've had the same experiences as Jerry. In GPO studio I have to lower the rate down to 3200 to get a good performance. The sound is more muffled than the more pristine 4800, but it still works for me. I notice that when you set a new sample rate in GPO studio all of the instruments have to reload. Does that actually change the instrument's size? With the lower sample rate I could fit more instruments in there, I think, but I don't kow. It could have just been me.

    -Chris

Go Back to forum

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
  •