What sampling rate do you use?

What sampling rate do you record at?

  • 44.1

    Votes: 197 55.8%

  • 48

    Votes: 79 22.4%

  • 88.2

    Votes: 6 1.7%

  • 96

    Votes: 52 14.7%

  • 192

    Votes: 10 2.8%

  • 384

    Votes: 1 0.3%

  • I'm waiting for 768

    Votes: 8 2.3%
  • Total voters
    353
Soundmind?? said:
Recently read an article by one of Apogees designers. Being an expert in digital audio and conversion, he should know what he's talking about. He pointed out that the human ear can't hear much above 18-20khz and that since 48K gives you ability to record up to 24khz
Click to expand...
There are three sides to consider for this debate:

1) The Nyquist theory
The max frequency that you can sample is half the sampling rate.
But this theorem is only valid for ideal converters and filters.
Furthermore, at this max frequency, you only have two samples per complete sine waveform, which is quite not really acceptable in pro audio.

2) The inputs to most soundcards are low-passed anyways.
Look, wether you select 44.1k or 96k, you cannot physically change the input circuit of your soundcard, right ? In order to avoid undersamplig at 44.1k, sound card manufacturers have to low-pass everything below about 20.5kHz. Thus, even when using 96k, you still ain't going to get those 48kHz harmonics many people talk about. Sampling higher only enables you to have "more" fidelity at frequencies higher than about 10kHz (but still lower than 20.5kHz).

3) The higher the sampling frequency, the easier the filtering
Without going too far into sampling theory, a higher sampling frequency reduces the constraints on output filtering, which in return leads to higher fidelity.

Bottom line is that you should use the highest possible/reasonable sampling rate and get over it.
And don't use less than 24 bits for all the good obvious reasons mentionned earlier.
I can actually hear a difference between 44.1k, 48k and 88.2k. Above that, I don't hear a difference.
 
My too sense? 44.1 does goofy stuffy to the high end--some rolloff and some distortion, turning your nice gentle sine waves into sawtooth waves because the 2 or 3 or 4 samples aren't enough to present the full wave as it happened naturally at the source (at least in the high freqs). I personally think you need a few more snapshots of samples up in the higher frequencies to represent them smoothly. If you've ever made a simple sine-wave sweep @ 44.1K in Soundforge or the like, you can see what I am talking about. The samples don't always happen at the peak of the wave, and it gets worse the closer you get to the Nyquist freq.

BUT! Until 96K CDs and 96K Ipods become the standard for consumer music playback or I start recording for DVD, I think I will just stick with 44.1 . Unless people start throwing harddrives and UAD cards in Magma chassis at me...
 
TheDewd said:
1) The Nyquist theory
The max frequency that you can sample is half the sampling rate.
But this theorem is only valid for ideal converters and filters.
Furthermore, at this max frequency, you only have two samples per complete sine waveform, which is quite not really acceptable in pro audio.
Click to expand...

Sorry but this is wrong. The sampling rate must exceed double of the frequency, which means you will always have at least three samples per wave cycle.
 
Restless said:
Sorry but this is wrong. The sampling rate must exceed double of the frequency, which means you will always have at least three samples per wave cycle.
Click to expand...
Noo....!
Let's say you sample a 20kHz signal at 40.00000001kHz, which fits your requirement of "exceeding double the minimal frequency", then you won't get three samples for EACH wave cycle...you're more likely to get TWO most of the time.
 
TheDewd said:
Noo....!
Let's say you sample a 20kHz signal at 40.00000001kHz, which fits your requirement of "exceeding double the minimal frequency", then you won't get three samples for EACH wave cycle...you're more likely to get TWO most of the time.
Click to expand...


Actually it was Nyquist's requirement, not mine.
And sorry, but even thought this argument of yours is irrelevant to this discussion, you are wrong again. You will always get more than two samples per cycle. Like that. At 44.1 Khz you can never have two samples only per cycle, providing the signal is bandwidth limited to 20 Khz.
 
Restless said:
Actually it was Nyquist's requirement, not mine.
And sorry, but even thought this argument of yours is irrelevant to this discussion, you are wrong again. You will always get more than two samples per cycle. Like that. At 44.1 Khz you can never have two samples only per cycle, providing the signal is bandwidth limited to 20 Khz.
Click to expand...
Okay then tell me how much samples (on average) per cycle you will get at 44.1k ?
Much closer to 2 than to 3.
And please explain to me how this is irrelevant in this discussion?
 
Restless said:
Actually it was Nyquist's requirement, not mine.
And sorry, but even thought this argument of yours is irrelevant to this discussion, you are wrong again. You will always get more than two samples per cycle. Like that. At 44.1 Khz you can never have two samples only per cycle, providing the signal is bandwidth limited to 20 Khz.
Click to expand...

No, it has to be at least double, not more than double. Double is the minimum.

It's 44.1 because in reality, the range sampled by 44.1 extends up to 22.05K.
 
TheDewd said:
Okay then tell me how much samples (on average) per cycle you will get at 44.1k ?
Much closer to 2 than to 3.
And please explain to me how this is irrelevant in this discussion?
Click to expand...


iT's irrelevant because 44.1 Khz is the minimum sampling frequency used, not those numbers you're coming with. The math is actually quite complicated, and I can't really say that I understand it all, but reading the articles at http://www.lavryengineering.com/ helps. Anyway, my point is that your statement "The max frequency that you can sample is half the sampling rate." is inaccurate. It's less than half.
 
Last edited:
Restless said:
iT's irrelevant because 44.1 Khz is the minimum sampling frequency used, not those numbers you're coming with. The math is actually quite complicated, and I can't really say that I understand it all, but reading the articles at http://www.lavryengineering.com/ helps. Anyway, my point is that your statement "The max frequency that you can sample is half the sampling rate." is inaccurate. It's less than half.
Click to expand...
Okay first dude, you have to learn how to read.

This is a quote I picked out of the document you liked me to:

"Let us begin by examining a band limited square wave. We set the fundamental frequency to 1KHz and the channel bandwidth to 22.05KHz (as in red book audio CD)."

By this, he means the available bandwidth for a 44.1k sampled system is 22.05kHz, which is half the sampling frequency, as I said.

EDIT: I actually took a look in my EE telecommunications books and the rule is always stated as fs(min) >= fmax. So the maximum frequency that can be recovered is really exactly fs/2. Now what do you have to say against this uh? Tell me dude...

Second, dude, the whole article pretty much says the SAME I've been saying in my first post, which is that no hardware can perfectly and ideally realise the temporal sinc filter function required to retreive the info. Thus, Nyquist is mostly and idealistic theorem, we should sample higher than the minimum sampling rate to ensure easier filtering and better reproduction.

Dude, my point was exactly that....the Nyquist theorem is just a guideline that lets you determine the minimum. But that doesn't mean this minimum is, in practice and in the real world, the best thing do do. This is why sampling higher makes nice sense.

Furthermore, the article states another fact I've been insisting on in my first post, which is that your preamp/mic/signal chain is bandlimited at around 20k anyways, so having a greater sampling rate is NOT to be able to sample 40kHz signals (lol) but rather to ease the filtering process and increase the fidelity of the near 20kHz signals.
 
Last edited:
you are all a bunch of fags.

who cares?

now make me a pie.

always record at 24 bit. if you are going to CD, keep it at 44.1(or 88,2 if you buy the "less math" argument)

if DVD-Audio , go 48 or 96. course if your chain is weak, it is just high-res shit.

192 is good for storage....
 
TheDewd said:
Okay first dude, you have to learn how to read.

This is a quote I picked out of the document you liked me to:

"Let us begin by examining a band limited square wave. We set the fundamental frequency to 1KHz and the channel bandwidth to 22.05KHz (as in red book audio CD)."

By this, he means the available bandwidth for a 44.1k sampled system is 22.05kHz, which is half the sampling frequency, as I said.
Click to expand...

Well, maybe you should read the whole damn thing then:

"Let us examine sampling of a sine wave with frequency above half of the sampling rate. The red wave is the high frequency input. The blue shows the locations and magnitude of the sampling sinc's. Clearly in this case, we do not have enough sample points to track the high frequency. The sum of the sinc's will yield a wrong result. We must sample faster than twice the signal bandwidth. "
 
I'm beginning to think it's all a bunch of hooey. Lately I've been tracking to a two-track reel-to-reel and dumping to my digital recorder....An Alesis Masterlink. Things definitely sound better when recorded at 24/96 vs. 16/44, but GUESS WHAT.....when dithered and downsampled 16/44 sounds better than the dithered and downsampled 24/96. Also sounds better than thanthe dithered and downsampled 24/44, 24/48, 24/88.2.

I know I'm gonna hear now how the Masterlink doesn't do a good job of dither/downsampling. I don't buy it. I think that we've all been sold a bill of goods. Now, if the medium du jour ever changed from CD to something with higher res..i.e. DVD-Audio, now that would be a different story. Then it would make perfect sense to record at 24/96 or higher.........
 
The 44.1/16 CDs days are numbered. Higher resolution capable data storage mediums are readily available, just waiting for economy of scale to reduce costs. Look for MP3 type gear with high resolution to replace players with old fasioned moving parts that consume batteries and have limited dynamic range. In the 70's I collected hundreds of LP records that held a whopping 15-20 minutes of music on each side. Had no idea that they would be virtually junk in 30 years. Their only real value is as collectibles now.

If I record a CD for someone, and that's the only thing they want, I use 44.1/24 (24 for more headroom to prevent accidental clipping). If I'm recording accoustical instruments, horns, etc, I prefer 48/24 for the extra range should it be later useful to transfer it to more advanced medium. Can't wait for 64 bit to take hold....no more clipping maybe....just ruptured eardrums I guess.
 
The minimum I do is 24/48 for the very reason that new CD and DVD formats will be available soon and I don't want my mixes limited to 44.1.


-Stew
 
musicstew said:
The minimum I do is 24/48 for the very reason that new CD and DVD formats will be available soon and I don't want my mixes limited to 44.1.


-Stew
Click to expand...
44.1 isn't really a limitaton when compared to 48k.
 
44.1 purely to save on HD space and the fact I'm more on the rock, metal, synth industrial, everything straight into the DAW, side of things. If I were recording more vocals and acoustic instruments I'd definitley use a highger rate. Mind you I'd probably want to upgrade to Sonar 5 for the 64 bit engine, as well. And get better mikes.
 
44.1 and 16bit on all projects up to now, mostly to avoid having to deal with dithering, but after reading some threads like this one:

http://recforums.prosoundweb.com/index.php/mv/msg/4918/0/144/6837/

I am going to work with 24 bit 44.1 from now on to gain the headroom.

The highest sampling rate my Digi001 can work at is 48000. I would appreciate feedback from anyone who know if Mastering engineers prefer 48000 over 44.1.

Thanks,
bilco
 
bilco said:
44.1 and 16bit on all projects up to now, mostly to avoid having to deal with dithering, but after reading some threads like this one:

http://recforums.prosoundweb.com/index.php/mv/msg/4918/0/144/6837/

I am going to work with 24 bit 44.1 from now on to gain the headroom.

The highest sampling rate my Digi001 can work at is 48000. I would appreciate feedback from anyone who know if Mastering engineers prefer 48000 over 44.1.

Thanks,
bilco
Click to expand...
Even though all mastering engineers can easily do the samplerate conversion, most prefer the target sample rate. (44.1k for CD, 48k for video) Why go through the extra process when you really don't gain anything.
 
You will gain little or nothing, (maybe even do harm to quality during conversion) at 48k over 44.1 if you plan nothing beyond a CD. The 24 bit mode during recording should help prevent clipping (A VERY BAD THING) during tracking if a level is set a little too high or the artist over does it with his/her dynamics. A little more headroom in the digital world is a wonderful thing. If you plan to do a high definition project, u shoul use at least 48/24 or 88.2/24. IMHO
 
You must log in or register to reply here.
Back
Top