mmorgan
Max Output Level: -77 dBFS
- Total Posts : 676
- Joined: 2013/02/19 23:39:05
- Location: Bellingham, WA
- Status: offline
Sample rate: 44.1 or 48 that is the question
Recently got to work on a film project which required me to switch my sample rate (SR) from 44.1K to 48K. I also had to use a different DAW. When I returned to Sonar this seemed to wreck all kinds of havoc (okay controlled chaos) with Sonar, particularly existing items (projects and templates).
I'm not sure if there is any inherent advantage to 48K SR although clearly if I were to continue getting film work there seems a requirement there...and, oddly, the anal retentive in me somehow regards 48K as more mathmatically cohesive with a 24bit bit depth (which I always use). Yes I know that's silly...
So here's my question: if you need to work on 48K projects do you adjust the SR on your soundcard before opening projects (i.e. going back and forth between 44.1 and 48) or do you just do everything at 48K. Bear in mind that I have lots of templates - some of which cause a white screen crash during instantiation (Scuffham S-Gear seems to be a particular culprit).
Thanks,
Mike Win8(64), Sonar X3e(64) w/ RME Fireface UFX.
|
emwhy
Max Output Level: -62 dBFS
- Total Posts : 1402
- Joined: 2006/01/03 15:09:02
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/03 11:00:08
(permalink)
In my case it depends on the audio interface. My Edirol UA 101 requires a shutdown before I can change the sample rate. Not so for my M Audio stuff. If I'm opening a project and using the Edirol I have to change the sample rate before opening SONAR.
|
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
- Total Posts : 26036
- Joined: 2006/09/17 11:23:23
- Location: Everett, WA USA
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/03 12:27:20
(permalink)
I'd stick with one sample rate for simplicity. SONAR's sample rate conversion is excellent, so you won't notice any degradation if you record at 48KHz and have to convert to 44.1 for the occasional CD.
If you use a lot of samples, 44.1 might make more sense because most sample libraries are 44.1. But if most of your projects are straight audio and you do a lot of DVD audio, then 48KHz makes more sense.
 All else is in doubt, so this is the truth I cling to. My Stuff
|
mmorgan
Max Output Level: -77 dBFS
- Total Posts : 676
- Joined: 2013/02/19 23:39:05
- Location: Bellingham, WA
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/03 14:05:29
(permalink)
bitflipper: Just curious, if I am at 48K and I pull in some 44.1K samples for a track does the sound card just fill in the extra bits with 0s?
Definitely leaning towards the simplicity thing...I just wonder how many templates I have that might need to be redone - and if I should just bite the bullet (so to speak) and rebuild those templates.
Thanks to both of you for the input.
Regards,
Mike Win8(64), Sonar X3e(64) w/ RME Fireface UFX.
|
Guitarpima
Max Output Level: -34 dBFS
- Total Posts : 4125
- Joined: 2005/11/19 23:53:59
- Location: Terra 3
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/03 14:10:42
(permalink)
If you import them at any sample rate, it should work just fine.
I never knew you could change the sample rate of a project. That would have been useful info a few weeks ago. LOL! Where's that manual. ;-)
Notation, the original DAW. Everything else is just rote. We are who we are and no more than another. Humans, you people are crazy. Win 7 x64 X2 Intel DX58SO, Intel i7 920 2.66ghz 12gb DDR3 ASUS ATI EAH5750 650w PSU 4x WD HDs 320gb DVD, DVD RW Eleven Rack, KRK Rokit 8s and 10s sub
|
jscomposer
Max Output Level: -87 dBFS
- Total Posts : 175
- Joined: 2013/02/03 15:57:01
- Location: Canada, Eh!
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/03 15:51:12
(permalink)
As rule of thumb, most TV and film requires your audio delivered in 48/24 (industry standard). Just set up Sonar using this as the default setting. If you dump in audio that was recorded at 44.1 (loops, etc), shouldn't be a problem.
|
emwhy
Max Output Level: -62 dBFS
- Total Posts : 1402
- Joined: 2006/01/03 15:09:02
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 11:05:33
(permalink)
BTW if you're working at 48, try recording at 96 if your soundcard allows.
|
garrigus
Max Output Level: 0 dBFS
- Total Posts : 8599
- Joined: 2003/11/05 17:23:21
- Location: www.garrigus.com
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 11:20:20
(permalink)
|
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
- Total Posts : 26036
- Joined: 2006/09/17 11:23:23
- Location: Everett, WA USA
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 12:43:40
(permalink)
Something I'd not given any thought to before: SRC for sample libraries. Obviously, SONAR needs to convert a 44.1 sample library to 48 if your project's at 48. Not a big deal. But what if you're not rendering/importing audio, just playing back MIDI-driven samples? Would that not require constant and continuous SRC every time you hit the spacebar?
All else is in doubt, so this is the truth I cling to. My Stuff
|
Paul P
Max Output Level: -48.5 dBFS
- Total Posts : 2685
- Joined: 2012/12/08 17:15:47
- Location: Montreal
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 13:09:37
(permalink)
With Dimension Pro and SD3 samples now as 44,1 flac files,
there has to be a first conversion to wav and then a second conversion to 48.
It would be interesting to know how the various conversions are handled on the fly.
Sonar Platinum [2017.10], Win7U x64 sp1, Xeon E5-1620 3.6 GHz, Asus P9X79WS, 16 GB ECC, 128gb SSD, HD7950, Mackie Blackjack
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 13:35:26
(permalink)
bitflipper
Something I'd not given any thought to before: SRC for sample libraries. Obviously, SONAR needs to convert a 44.1 sample library to 48 if your project's at 48. Not a big deal. But what if you're not rendering/importing audio, just playing back MIDI-driven samples? Would that not require constant and continuous SRC every time you hit the spacebar?
Just do the SRC at the plug's output. Not really any different than a plug that oversamples for processing purposes.
 In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
wogg
Max Output Level: -57 dBFS
- Total Posts : 1819
- Joined: 2003/11/14 16:07:44
- Location: Columbus, OH
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 16:45:21
(permalink)
Just curious, if I am at 48K and I pull in some 44.1K samples for a track does the sound card just fill in the extra bits with 0s? You're confusing sample rate with bit depth. Bit depth is the 16bit / 24bit setting, and yes, if you up 16bit data to 24bit, the least significant bits are simply tacked on as 0's. Sample rate conversion is basically mathematically guessing where the signal lies at the time of the output sample based on the 2 closest input samples. They've gotten really good at that math these days. The best plan used to be to simply use the sample rate of your target medium, 44.1k for audio CD and 48k for video / DVD.
|
Cactus Music
Max Output Level: 0 dBFS
- Total Posts : 8424
- Joined: 2004/02/09 21:34:04
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 18:26:30
(permalink)
I once read a funny analogy to this
Sample rate is how fast the tape speed is.
Bit depth is how wide the tape is.
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/04 19:05:28
(permalink)
wogg
Sample rate conversion is basically mathematically guessing where the signal lies at the time of the output sample based on the 2 closest input samples.
It is NOT guessing and it is NOT based on only the 2 closest input samples. That's how people often assume it works, but it's doesn't work that way. The value of a new sample between two existing samples is calculated and the calculation is based on a series of samples. The accuracy of the result is largely a factor of how much processing power you want to throw at it, but with the power available with modern CPU's there's really no reason for SRC's to cause problems.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
Tom Riggs
Max Output Level: -57.5 dBFS
- Total Posts : 1752
- Joined: 2003/11/08 22:47:26
- Location: Displaced Kansan living in Philippines
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/05 05:08:37
(permalink)
Its my understanding that you can't change the sample rate of a project that already contains audio. Am I wrong?
i7-3770k OC at 4.5Ghz, asus p8z77-m, 16g g.skill aries 1600 c9 ram, Noctua d-14 cooler, RME HDSPe Raydat, Motu FastLane, Nvidea GTX 980 ti 6G, windows 7 and 8.1 pro x64. Sonar Platinum and x3e currently installed My Music My YouTube
|
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
- Total Posts : 26036
- Joined: 2006/09/17 11:23:23
- Location: Everett, WA USA
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/05 10:56:28
(permalink)
drewfx1
bitflipper
Something I'd not given any thought to before: SRC for sample libraries. Obviously, SONAR needs to convert a 44.1 sample library to 48 if your project's at 48. Not a big deal. But what if you're not rendering/importing audio, just playing back MIDI-driven samples? Would that not require constant and continuous SRC every time you hit the spacebar?
Just do the SRC at the plug's output.
Not really any different than a plug that oversamples for processing purposes.
Exactly my point. Oversampling within plugins increases CPU usage, which is why most users only turn it on for rendering. An underpowered machine like mine might not manage 40 or 50 oversampled plugins. If I was using 48KHz as my project samplerate, every Kontakt/SampleTank/Omnisphere/Alchemy sample would require real-time SRC every time I played back the project, with no option to avoid it other than rendering the tracks. That got me wondering if sample-based projects might be significantly more CPU-efficient if the project SR was the same as the sample libraries.
All else is in doubt, so this is the truth I cling to. My Stuff
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/05 21:38:39
(permalink)
bitflipper
drewfx1
bitflipper
Something I'd not given any thought to before: SRC for sample libraries. Obviously, SONAR needs to convert a 44.1 sample library to 48 if your project's at 48. Not a big deal. But what if you're not rendering/importing audio, just playing back MIDI-driven samples? Would that not require constant and continuous SRC every time you hit the spacebar?
Just do the SRC at the plug's output.
Not really any different than a plug that oversamples for processing purposes.
Exactly my point. Oversampling within plugins increases CPU usage, which is why most users only turn it on for rendering. An underpowered machine like mine might not manage 40 or 50 oversampled plugins.
Remember that the oversampled plugins are also doing all of their work at the higher sample rate, which is the major factor in increasing the CPU usage. The SRC part isn't much more than a (hopefully) very good quality LPF, CPU wise.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
losguy
Max Output Level: -20 dBFS
- Total Posts : 5506
- Joined: 2003/12/18 13:40:44
- Location: The Great White North (MN, USA)
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/06 19:04:53
(permalink)
@bitflipper: You raise an interesting question. I scratched my head a couple of times, then remembered that samplers rely on SRC (sort of - read on) as a fundamental operation of their playback engines. It's what allows them to play (for example) a D note based on a sample recording of a C note. The "SRC" is used to play back the recorded sample at a time rate different from the recorded rate (thus shifting the pitch) while preserving the rate at which the samples come out. It's all accomplished with the same type of interpolation process as used on SRC. There, you are playing back a sample at a different sample rate than recorded, while preserving the pitch in the output. In the case of the sampler, changing the sample rate on the output side should simply amount to changing a scale factor in the interpolator. Any increase in computation should then be just the overhead that you get from processing the added samples at the higher rate, which you would get everywhere else in the DAW anyway.
@drew: I concur - though it would be closer in comparison to a good linear-phase lowpass
|
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
- Total Posts : 26036
- Joined: 2006/09/17 11:23:23
- Location: Everett, WA USA
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/06 23:44:41
(permalink)
Good one, losguy. I get it now...SRC isn't a particularly CPU-intensive operation. The reason oversampled plugs gobble CPU cycles is because they're doing calculations on more data, not because of the oversampling itself. SRC overhead while playing back samples at a different project rate is therefore negligible.
All else is in doubt, so this is the truth I cling to. My Stuff
|
jimusic
Max Output Level: -64 dBFS
- Total Posts : 1325
- Joined: 2008/05/21 18:59:52
- Location: Near Vancouver, British Columbia, Canada
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 00:50:52
(permalink)
My head hurts now! 
|
losguy
Max Output Level: -20 dBFS
- Total Posts : 5506
- Joined: 2003/12/18 13:40:44
- Location: The Great White North (MN, USA)
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 17:37:14
(permalink)
Spot on, Oh Esteemed Flipper Of Bits. Incidentally, the interpolator often looks a lot like the truth that you cling to in your sig. In this case, though, your variable "t" is replaced by the new output sample, call it "m", and the parameter "Ts" works as the interpolation factor, call it "K". Oh, and of course, for practical DSP, the summation doesn't go to infinity, but occurs over a finite (moving) time window of samples. Also, a window function is typically applied, usually a symmetric bell shape function "h(m-nK)" over the sample window, to smooth out edge effects (among other things). Since the window is finite and non-uniform, you could say that the "truth" is approximate. But in these modern times, that approximation has been pretty good for audio purposes. (Of course, you may feel free to sum to infinity if you are in need of completely cling-able truth.) 
|
Paul P
Max Output Level: -48.5 dBFS
- Total Posts : 2685
- Joined: 2012/12/08 17:15:47
- Location: Montreal
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 17:45:05
(permalink)
losguy : "Since the window is finite and non-uniform, you could say that the "truth" is approximate. But in these modern times, that approximation has been pretty good for audio purposes. (Of course, you may feel free to sum to infinity if you are in need of completely cling-able truth.) "
That sounds more like religion to me than science :-)
Sonar Platinum [2017.10], Win7U x64 sp1, Xeon E5-1620 3.6 GHz, Asus P9X79WS, 16 GB ECC, 128gb SSD, HD7950, Mackie Blackjack
|
losguy
Max Output Level: -20 dBFS
- Total Posts : 5506
- Joined: 2003/12/18 13:40:44
- Location: The Great White North (MN, USA)
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 18:34:57
(permalink)
Paul P
That sounds more like religion to me than science :-)
"It is difficult to avoid the impression that a miracle confronts us here…or the two miracles of the existence of laws of nature and of the human mind’s capacity to divine them ... The miracle of the appropriateness of the language of mathematics for the formulation of laws of physics is a wonderful gift which we neither understand nor deserve." - Eugene Wigner, The Unreasonable Effectiveness of Mathematics in the Natural Sciences, 1960
|
wogg
Max Output Level: -57 dBFS
- Total Posts : 1819
- Joined: 2003/11/14 16:07:44
- Location: Columbus, OH
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 21:05:47
(permalink)
drewfx1
wogg
Sample rate conversion is basically mathematically guessing where the signal lies at the time of the output sample based on the 2 closest input samples.
It is NOT guessing and it is NOT based on only the 2 closest input samples. That's how people often assume it works, but it's doesn't work that way.
The value of a new sample between two existing samples is calculated and the calculation is based on a series of samples. The accuracy of the result is largely a factor of how much processing power you want to throw at it, but with the power available with modern CPU's there's really no reason for SRC's to cause problems.
Semantics. In the absence of actual data at that time it's still a guess, just a really educated and carefully calculated guess.  Good point on the series of samples, it's probably been a really long time since anyone used two point interpolation.
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 21:23:03
(permalink)
The "actual data at that time" is in fact contained in the series of samples.
That's the very crux of the sampling theorem.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
rabeach
Max Output Level: -48 dBFS
- Total Posts : 2703
- Joined: 2004/01/26 14:56:13
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 22:18:00
(permalink)
drewfx1
The "actual data at that time" is in fact contained in the series of samples.
That's the very crux of the sampling theorem.
IMO the perfect band limitation of an infinite signal/sequence can be viewed as more an absurdity than a reality.
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 22:52:07
(permalink)
rabeach
drewfx1
The "actual data at that time" is in fact contained in the series of samples.
That's the very crux of the sampling theorem.
IMO the perfect band limitation of an infinite signal/sequence can be viewed as more an absurdity than a reality.
But it doesn't have to be perfect - we're dealing with quantized data (likely already containing some noise) and humans. Many of the SRC routines measured here easily surpass both 16bit (or even 24bit) audio and human hearing acuity: http://src.infinitewave.ca/ IMO, if all of the SRC errors are already lost in the quantization error and/or are completely inaudible it's absurd to argue about theoretical imperfections.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
rabeach
Max Output Level: -48 dBFS
- Total Posts : 2703
- Joined: 2004/01/26 14:56:13
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/07 23:14:57
(permalink)
drewfx1
rabeach
drewfx1
The "actual data at that time" is in fact contained in the series of samples.
That's the very crux of the sampling theorem.
IMO the perfect band limitation of an infinite signal/sequence can be viewed as more an absurdity than a reality.
But it doesn't have to be perfect - we're dealing with quantized data (likely already containing some noise) and humans.
Many of the SRC routines measured here easily surpass both 16bit (or even 24bit) audio and human hearing acuity:
http://src.infinitewave.ca/
IMO, if all of the SRC errors are already lost in the quantization error and/or are completely inaudible it's absurd to argue about theoretical imperfections.
My comment was made in jest about the sampling theorem itself being based on perfectly band limiting an infinite sequence derived from an infinite signal; neither of which exist in our reality.
post edited by rabeach - 2013/04/07 23:21:55
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/08 15:52:42
(permalink)
rabeach
My comment was made in jest about the sampling theorem itself being based on perfectly band limiting an infinite sequence derived from an infinite signal; neither of which exist in our reality.
Fair enough! But that wasn't 100% clear, and I wanted to point out that just because theory vs. practice isn't 100% perfect (what you said is of course correct), it doesn't mean it's not "perfect enough". But all this got me thinking. Since I'm a "prove it" kind of guy, I thought it would be fun to test Sonar's SRC vs. an independent calculation and see how good it did. So here's what I did: 1. Calculated the "correct" values for a fairly high frequency 48kHz sine wave (I chose 15,200Hz) in Microsoft Excel. 2. Create a 64bit 44.1kHz 15,200Hz 0dBFS sine wave in Sound Forge 9. 3. Use the SRC algorithms in both Sound Forge 9 and Sonar X2 to convert from 44.1 to 48kHz. 4. Compare the actual SRC outputs with the values Excel said a 48kHz sine should have and express this as dBFS. My results: For Sound Forge 9, the average difference was ~-98dBFS. If you compare this to the results for SF9 given at http://src.infinitewave.ca/, you'll see that (by eye) it roughly matches their results and that SF9 indeed has a fairly poor SRC routine compared to others shown there. For Sonar X2? My average difference was ~-204dBFS (!!!!). The worst error was ~-193dBFS. Those are pretty good "guesses", I'd say.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
rabeach
Max Output Level: -48 dBFS
- Total Posts : 2703
- Joined: 2004/01/26 14:56:13
- Status: offline
Re:Sample rate: 44.1 or 48 that is the question
2013/04/08 17:40:03
(permalink)
|