Found a Bug? | Have a suggestion?

What are the benefits of recording with a sample rate greater than 44.1?

Asked 1844 day 4 h | Viewed 37094 times | Updated 1815 day 15 h |

- 4 +

Gravatar photo Gear Guy

Why should I record at any sample rate greater than 44.1k when in the end I'm going to sample back down to 44.1 when I burn a CD?  Over even less if it ends up on an iPod.

Are there any advantages to recording at a higher sample rate, or should I just stick with 44.1k for now?

15 Answers

    Best Answer

  1. Answered: 1832 day 15 h (1) | Permalink

    - 3 +

    Gravatar photoDaniel Kushner

    I used to be all for using the highest sample rate. Doesn't matter so much to me anymore though. The argument goes that harmonics in the highest frequencies, although well beyond the range of human ears, have a phase canceling effect that effects the ultimate tone of a sound. Cutting these off during recording thus changes the accuracy of the recording. This is true.

    The law of diminishing returns applies.  The higher you cut, the less of an effect there is on a tone.

    44 addresses the human hearing spectrum and a bit beyond. In my opinion, it does a fine job. I can't hear any difference between 44 and 96.

    The final thing that comes to mind is that it's better to record at 88 than 96 if your target is 44, because when you down sample, it is an even division, not an approximation (88/2=44  96/2.1818181=44) which makes sense to me. Again though, my ears tell me this is hooey, I've downsampled from 96 and heard no difference.

    There's the argument that with cheap hard drive space it can't hurt. True, I guess. Still not a good reason to do it.

    So my answer is don't stress out over sample rate. In the pantheon of things that will affect your sound, it's way behind microphones, mix process, talent, EQing and compression. Oh, and light years behind "good sounding original source".

    • The math does make sense, and I do know that computers have difficulty rounding floating point numbers at time. I just have this feeling that the sample rate situation ends up being more marketing than necessity. I'd love to record at a higher sample ra

      Gear Guy | Aug 29 at 07:08

    Add a Comment
  2. Answered: 1834 day 8 h (1) | Permalink

    - 6 +

    Gravatar photoMcSmooth

    If you know that you will never need a higher quality master, it is usually safe to go with your target sample rate.  This is a long debated topic that has many arguments for both sides.  Although bigger is usually better, there are many determining factors, and in some cases, a higher sample rate could end up worse than if the target sample rate was used.

    If you want the best possible mix (and a professional is doing the mixdown and your interface doesn't sound worse at higher rates), the engineer might be able to make a slightly better final product with a sample rate of 88.1k or higher.  48k is usually not recommended if you plan to go back to 44.1k.

    You will see a much greater benefit from recording in 24 bit over 16 bit.  Most would also argue that if you are not using high-end preamps and A/D to record with, you will get much more from 44.1k out of high end gear, than 192k out of cheap stuff.

    So although there is a potential for a slightly better sounding final product, it ultimately comes down to your gear, workflow, and if your ears think there is any worthwhile difference.

    • That's a pretty good overview, and it totally agree with your suggestion of recording at 24bit and not at 16bit.

      Justin Vencel | Aug 25 at 01:08

    Add a Comment
  3. Answered: 1831 day 4 h (0) | Permalink

    - 4 +

    Gravatar photoPablo Santiago

    As you may know, bit depth represents resolution in amplitude and the sampling rate (see Nyquist theorem) implies resolution in the frequency range sampled. In 44.1k you'll be taking two samples to digitalize an entire cicle of a sinusoid (or other form) of 22khz, that makes a poor representation of the original wave. When you have 20 tracks with a bad representation of higer frequencies or harmonics mixed in your work you get a thing that just don't sound as you may want to because multiplies the misrepresentation of the original waveforms and multiplies it by 20. By recording at a higer sample rate you try to minimize this effect in your 20 sources so when you mix and edit them don't carry this unwanted effect (that's why usually the audio engines process effects internally at higer bit depth and sampling rate and lately the sound is downsampled). The idea seems to be: "record, edit and mix with precise sound and do damage only once, at the end of work"

    This video should help to explain better the concept, im'not good in english :(

    http://www.youtube.com/watch?v=Fy9dJgGCWZI

  4. Answered: 1793 day 20 h (1) | Permalink

    - 2 +

    Gravatar photoEric Squires

    Check out this article by Craig Anderton: http://www.musiciansfriend.com/document?doc_id=105267&src=3NL9JO1&ZYXSEM=0.

    I agree with him on setting your sample rate to 88 rather than 96. The math is simpler and at least in theory should produce fewer artifacts when reducing it down to 44.1. I  also think that the bit rate is more important than the sample rate.

    One more thing you need to ask yourself before making the switch is "how is this going to be released?" If you're release is solely on CD, then sample rate probably won't make that great a difference. If you're releasing it as the original waves on a DVD or through an online service such as bandcamp.com, then the sample and bit rates become far more important.

    I still hold that even if the majority of people cannot consciously tell the difference in the sample rates their ears can hear it. They may not be able to express the differences, but on some level they can hear it. "Ear training" is not so much training our ears as it is training our brains to recognize and express intelligently the differences we hear.

    • Good point about releasing tracks on new and different mediums. As we move past the CD in years to come there is a great opportunity to get very high quality audio files directly to listeners computers and portable devices.

      Gear Guy | Oct 04 at 03:10

    Add a Comment
  5. Answered: 1788 day 0 h (2) | Permalink

    - 1 +

    Gravatar photoBrian Zieske

    Recordings at 88.2 sound great when converted to 44.1

    • Do you do much of your recording at higher samples rates such as 88.1?

      Gear Guy | Oct 11 at 02:10

    • I really didnt hear a difference at 96kHz and up from 88.2. Memory starts getting crazy when doing a record above 88.2

      Brian Zieske | Oct 18 at 11:10

    Add a Comment
  6. Answered: 1780 day 0 h (0) | Permalink

    - 1 +

    Gravatar photoLiquid Vibe

    There is a Theory called the Nyquist Theory. It states that the sample rate must be at least twice the highest freq. being recorded. The reason is that if the sample rate is not twice as high the A/D converter can not acurately represent the waveform and its freq. in the digital domain. The result is that the actual harmonic content is altered. This is what is known as an alias freq. Most A/D converters contain a fixed low pass filter known as the anti ailas filter. It is set at 20KHz. Even though most mics don't have a freq. response greater that 20k, the anti ailas filter is set at 20k, I like to play things safe. I usualy use 88.2 because it is high enough to be sure there are no alias freq., it is easy math to convert to 44.1, and it takes up less HD space than 96K or 192K.

  7. Answered: 1737 day 1 h (0) | Permalink

    - 1 +

    Gravatar photoSonic Ingenuity

    Unless you are doing live recordings or pre-mixing at a higher rate, its harder to notice.  If used correctly, it tends to add a much deeper detail to the recordings.  Vocals have strong textures, etc.  The problem is the highend waves getting crunched and causing a 'sterile' sound to the mix.  A few ways to fix that too but to answer your question  - Details and textures.

  8. Answered: 1466 day 20 h (1) | Permalink

    - 0 +

    Gravatar photoguest

    Who decides what the best answer is?  The one designated as the best one isn't correct.

    • The person that asks the question can select a "best answer" that they feel helped them out. Any other member can vote on any question or answer by using the + and - buttons. The total number of votes will push an answer up or down in the results.

      Recording Questions | Sep 09 at 02:09

    Add a Comment
  9. Answered: 987 day 13 h (0) | Permalink

    - 0 +

    Gravatar photoAstral Plane Studios

    Higher resolutions and sample rates are best left for recording orchestras and choir groups, otherwise stick to 44.1/88.2 

    Recording at higher rates gives more dynamic headroom and captures more detail of the performance. I tend to record 88.2 as it offers the best balance of sonics and imaging to my ears and is an easy conversion to 44.1 during mastering.

  10. Answered: 751 day 14 h (0) | Permalink

    - 0 +

    Gravatar photoguest

    Wow! That's a really neat anewsr!

  11. Answered: 751 day 9 h (0) | Permalink

    - 0 +

    Gravatar photoguest

    Shiv Shankar Kashyapbahut bahut achi story hai bhagwan ki nioyitn ko samjhne ke liye kash aaj ke insaan bhi in vrkhso ki tarah kaam aate to acha tha, main bhi mehtvkanshi hoon aur duniya ke jhote bandhan main fansa huya hoon pta nahi kab is moh maya se tyag hoga is kahani ke lekhak ko salute karta hoon ..

  12. Answered: 749 day 17 h (0) | Permalink

    - 0 +

    Gravatar photoguest

    zJEOjA onnkeangilkr

  13. Answered: 748 day 23 h (0) | Permalink

    - 0 +

    Gravatar photoguest

    57VGIW , [url=http://mmvuiacpuadi.com/]mmvuiacpuadi[/url], [link=http://jmkkjnkrlxch.com/]jmkkjnkrlxch[/link], http://gwngexkhrkjl.com/

  14. Answered: 736 day 18 h (0) | Permalink

    - 0 +

    Gravatar photoguest

    One reason why a higher sampling rate is better because of phase distortion. In Digital Signal Processing, everytime you use an effect such as eq on a track, it creates a magnitude as well as a phase response for different frequencies. The phase response results in a group delay. Consider a track where you want to cut the higher mids and keep the lower mids as they are. You eq it with a peak filter (usually an IIR filter). what you have done is that you have reduced the volume and added a little delay to the higher mids. To paint a picture, if the two crests in the waveform of the lo mid and hi mid signals were on top of each other before, they might not be on top of each other now, since you have delayed the hi mid signal.Phase distortion is difficult to hear and quantify; comb filtering is a sort of phase distortion. Now for every gain, q and frequency you use for the filter, this delay produced is in number of samples (for digital systems) in discrete time. The higher the sampling rate, the lesser the delay (time delay = number of samples / sampling frequency). The lesser the time delay, the lesser the phase distortion in continuous (natural) time.

  15. Answered: 423 day 17 h (0) | Permalink

    - 0 +

    Gravatar photoguest

    can't you technically use a higher sample rate for a less aliased time stretch ?

Answer this question

(If you are simply responding and not posting an answer, consider leaving a comment instead)

You are currently not signed-in! You can still post this answer but you will not receive credit for it.
Sign In Now!

Not the answers you're looking for? Try asking your own question.