| Author
|
SRC comparisons 96 to 44.1 kHz
|
shamantrixx
Started Topics :
7
Posts :
549
Posted : May 3, 2007 04:25
|
Here is a nice page that could shine some light on frequent arguments about samplerate and sequencers. If you're not sure how should "good" graph look like choose ProTools and compare others with it. Users of FLoops don't have to bother viewing because (for the reasons best known to them selfs) FL is not among 30 sequencers and wave editors that research team have tested.
http://src.infinitewave.ca/
  "It occurred to me by intuition, and music was the driving force behind that intuition. My discovery was the result of musical perception"
Albert Einstein, speaking about his theory of relativity |
|
|
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : May 3, 2007 16:26
|
The ProTools SRC isn't really _that_ good. For textbook results check out the iZotope and R8Brain graphs.
One important thing to keep in mind is that these graphs don't tell us how the SRCs actually sound.
UnderTow |
|
|
|
PoM
IsraTrance Full Member
Started Topics :
162
Posts :
8087
Posted : May 3, 2007 17:04
|
|
is it better to use one of these plugs or to record at 44.1 with a rme converter? |
|
|
|
orange
Fat Data
Started Topics :
154
Posts :
3918
Posted : May 3, 2007 17:22
|
|
shamantrixx
Started Topics :
7
Posts :
549
Posted : May 3, 2007 17:57
|
@ atropa: It's just an information. You can take it if you like it or you can forget about it. In it self, pictures have no meaning until you make a meaning out of them.
"It occurred to me by intuition, and music was the driving force behind that intuition. My discovery was the result of musical perception"
Albert Einstein, speaking about his theory of relativity |
|
|
|
bukboy
Hyperboreans
Started Topics :
40
Posts :
803
Posted : May 4, 2007 09:10
|
|
What does this webpage compare? what does SRC stand for. lot of abbreviations of SRC on google. |
|
|
|
PoM
IsraTrance Full Member
Started Topics :
162
Posts :
8087
Posted : May 4, 2007 11:18
|
|
sample rate converter ,96KHz to 44.1KHz for the test in the webpage. |
|
|
|
Spindrift
Spindrift
Started Topics :
33
Posts :
1560
Posted : May 4, 2007 18:04
|
Sure the sound is what counts, but I find these graphs very interesting nevertheless.
First of all I can agree with a lot of processing that the most important is how it sounds, and not the purity of the processing from a technical perspective.
SRC is not one of those kinds of processing...if I input a 1 khz test tone that's all I want to get after conversion.
Many things sound better with a bit of distortion, but I rather let other plugins do that job than an SRC.
Also doing ABX tests on all those SRC's would be very time consuming, and I find it nice to be able to look at those graphs to get a good idea of what SRC'c offer the conversion resulting in the least artifacts.
It for sure encouraged me to give a couple of products a listen.
(``·.¸(``·.¸(``·.¸¸.·`´)¸.·`´)¸.·`´)
« .....www.ResonantEarth.com..... »
(¸.·`´(¸.·`´(¸.·`´``·.¸)``·.¸)``·.¸)
http://www.myspace.com/spindriftsounds
http://www.myspace.com/resonantearth |
|
|
|
shamantrixx
Started Topics :
7
Posts :
549
Posted : May 4, 2007 18:32
|
That's exactly what I had in mind in the first place. I use conversion from 48 to 44,1 kHz but I think that this test would look proportional if conversion was done from 48 to 44,1.
"It occurred to me by intuition, and music was the driving force behind that intuition. My discovery was the result of musical perception"
Albert Einstein, speaking about his theory of relativity |
|
|
|
captain-kirk
Started Topics :
9
Posts :
40
Posted : May 5, 2007 01:00
|
|
also gotta keep in mind that artifacts give character (and that is a taste thing) other wise everything would sound the same |
|
|
|
Mike A
Subra
Started Topics :
185
Posts :
3954
Posted : May 8, 2007 02:46
|
|
why would you work at 96 in the first place?
|
|
|
|
Spindrift
Spindrift
Started Topics :
33
Posts :
1560
Posted : May 8, 2007 02:54
|
Wild guess:
Sound quality.
The difference is indeed very noticeable when rendering synths or doing processing at 96k also after sampling down to 44.1k.
If I had more CPU and DSP power I would not hesitate to work at 96k.
(``·.¸(``·.¸(``·.¸¸.·`´)¸.·`´)¸.·`´)
« .....www.ResonantEarth.com..... »
(¸.·`´(¸.·`´(¸.·`´``·.¸)``·.¸)``·.¸)
http://www.myspace.com/spindriftsounds
http://www.myspace.com/resonantearth |
|
|
|
Upavas
Upavas
Started Topics :
150
Posts :
3315
Posted : May 8, 2007 03:37
|
"(for the reasons best known to them selfs) FL is not among 30 sequencers and wave editors that research team have tested."
Actually I use both Pro Tools and FL7, I see no reason why FL should not be included in that list.
The sampling rate is simply how fast we take samples when we quantize from an analog into a digital signal, or how fast we record digitally. The higher the sampling rate, the better the overall "picture" and the rounder and more precise the Wave, since we are talking digital it can never be fully precise, like analog signals. Since newborn children are the only ones that can hear 20k, once we have a rate of 44.1 k we generally have a fast enough quantization to to avoid audible aliasing. So you can record at 44.1k, or 96k, as long as there are no frequencies above 20khz, simply use a lowband pass filter to take care of the problem and we should generally be fine. This said, I generally go for 44.1 k in sound, except when I edit movies, for that 48k is usually expected. What is more important is the bitrate as it decides our dynamic range.
As far as Pro Tools is concerned, it is leading standart in the movie industry, simply because there is no software more versatile with wav files then Pro Tools. I still don't see an "obvious reason what makes it obvious that FL should not be included in that list. It is not the tool that lacks, it is generally rather the artist that lacks sufficient knowledge of how to use a tool e.g. sequencer.
Upavas - Here And Now (Sangoma Rec.) new EP out Oct.29th, get it here:
http://timecode.bandcamp.com
http://upavas.com
http://soundcloud.com/upavas-1/ |
|
|
|
shamantrixx
Started Topics :
7
Posts :
549
Posted : May 8, 2007 04:06
|
Quote:
|
On 2007-05-08 02:54, Spindrift wrote:
The difference is indeed very noticeable when rendering synths or doing processing at 96k also after sampling down to 44.1k.
If I had more CPU and DSP power I would not hesitate to work at 96k.
|
|
I agree. More complex synth patches (bubbles, hi pitch fx etc) can sound WAY better when exported at 96kHz. It would be great to be able to work at 96kHz but for now it is both CPU and HDD consuming.
@ Upavas: we can't know for sure why they didn't include FL in tests. It's a matter of choice so they are free to test what they like without any rule. So not being tested can mean many things. Beside quality it could be economical or a matter of personal dislikes of people who did a testing. Since this forum has seen an endless arguments about FL vs. SX (or any of the permutations) I have simply noticed that FL is not included in the test.
Since I have very little experience with FL I really can not judge about sound quality. All I can say is that to me it looks like a sequencer for people who don't have any prior experience with sequencers and therefor don't expect any similarity or "logic" that is present in the similar tools. I can not focus on music with interface like that. I could struggle with it and in a while I'd get used to it but with the number of tools I can use by intuition it's just not worth the struggle.
"It occurred to me by intuition, and music was the driving force behind that intuition. My discovery was the result of musical perception"
Albert Einstein, speaking about his theory of relativity |
|
|
|
Upavas
Upavas
Started Topics :
150
Posts :
3315
Posted : May 8, 2007 05:16
|
Shamantrixx,
Bubbles and high pitch fx tend to have sounds we cannot audibly percieve, when the sampling rate is too low we experience a phenomenon called aliasing, based on the Niquist theorem. So what happens is that these higher sounds cannot be expressed properly, the waveform is so high that the sampling rate cannot take an acurate picture. This is beautifully explained here:
http://en.wikipedia.org/wiki/Nyquist_rate#Nyquist_rate_sampling
So all we need to do is add a lowbandpass filter @ 20.000 herz and we get rid of this problem, meaning the human ear cannot possibly percieve any difference. The other way to do it is raising the sampling rate from 44.100 taken samples per second to 96.000 samples per second. I personally prefer the first method as it saves me a lot of cpu and hdd power. As far as FL and Q base is concerned, digital sound is digital sound, no matter where it comes from... the tools in Fl are more or less the same as in Cubase, Logic, Sonar, you name them. I think it has more to do with what you're comfortable with when making music, as far as the soundquality and versatility is concerned, they are pretty much alike.
Upavas - Here And Now (Sangoma Rec.) new EP out Oct.29th, get it here:
http://timecode.bandcamp.com
http://upavas.com
http://soundcloud.com/upavas-1/ |
|
|
|
