That's why I prefer libraries sampled at 44.1 from the first time. My audio is already "castrated" (read recorded at 44.1). BTW, castration happens in the lowest area of the body; at the opposite end you simply get a hair cut
Thanks for these examples ! I'm hearing exactly the same effects as in my own tests, so I feel reassured my system's probably alright and I'm not crazy . - "Audio 44k nebula 44k" and "Audio 96k nebula 96k" sound really waaaay better than the converted ones. 96k has more air & resolution, right. Not necessarily my favourite. Liked the light "S*****" of the 44k-version. While the 96k-96k-version is probably closer and truer to the original hardware character, the 44k-44k has a beneficial side effect - in THIS case. - "Audio 44k nebula 96" is slightly better than "Audio 96k nebula 44k", but they both lack mojo. You know, that special Nebula thing we're all crazy for . Try that example with a ridiculous 15dB CLQ boost at 1,5k with narrowest Q and it'll become even more obvious: 96k-96k it sounds intense, but the ears don't hurt. At 44k-96k the ears DO hurt, slightly, and it "only" feels as good as "just" the UAD2-class of plugins
AlexB wrote:Finally, a sine sampled at 44.1kHz and 96kHz. In a world: RESOLUTION !
ouch... "si tacuisses, philosophus mansisses" Explaining the benefits of higher audio resolution with images like that is like explaining sexuality with birds and bees I know you know better than that. Always a good read: http://www.lavryengineering.com/documen ... Theory.pdf - page 24-25
Yes resolution is not sample rate! Resolution relates to bit depth. Sample rate describes the cycles needed to represent frequencies we can hear. More sampling "points" do not equal more resolution. The Lavry papers are indeed a good read to understand this.
I'm quite sure everyone will agree that processing 96kHz audio tracks with 44.1kHz presets will be inferior to 96kHz presets if you are trying to preserve the audio spectrum...
What many are concerned about, is the possible quality loss when using 96kHz presets to process 44.1kHz tracks and the quality of the Nebula src. I have done quite a lot of tests myself and came to the conclusion that there is a definite quality loss when relying on Nebula's internal src. The degree to which this effect is audible depends on several things, such as the source material, the kind of presets used and the number of internal conversions.
Don't forget that even if the Nebula src was as good as izotope, you are still talking about one conversion per preset (which will typically mean many conversions) as opposed to a total of 2 conversion stages if you choose to up/down sample the audio material, and no conversions if you use presets that match your project's sample rate.
If you look at the way that Nebula down-samples 96kHz presets to 44.1kHz you will see phase shifts (as Niklas already pointed out), which are typical for minimum phase src, and often an increase in aliasing noise. More importantly, many presets are simply not converted properly by Nebula, yielding wrong harmonic distortion relations. Below are 2 images of the same preset, first as 44.1kHz (converted internally by Nebula from 96kHz) and then without src:
Yes understood however "visual resolution" is not what it seems when it comes to SR. showing sampling rate in discrete stepped fashion as opposed to less stepped does not demonstrate resolution. Sampling rate, after satisfying Nyquist, does not increase resolution with increased values. With 44.1 KHz you have full resolution of audibility, frequency wise, you don't get more with higher rates. Bit depth is resolution; these increased values allows for better representation of multiple computations. So increasing or decreasing this value at whatever point will affect audio tremendously.
My last two mixes, one of them done in 96kHz with an UAD Quad Omni system with a mixture of other plugins like Waves CLA and Equality, the other one at 44.1 with Nebula (mostly DocFear, Mammoth, APE, Germanos and AlexB filters, preamps and console inputs); the Nebula mix sounds simply sweet, intimate and musical, while the UAD/Waves one sounds punchy but cold and flat.
Both of them mastered through the Phoenix compressor, the Nebula mix sounds even sweeter, while the other one can't sound even close despite being mastered at 96 with some nice AlexB eqs.
Mercado_Negro wrote:At the end of the day Nebula sounds better than most plug-ins out there, even after SRC.
I couldn't agree more. All this SRC buzz falls under the same category as the "aliasing panic" and people dumping their plugins (they actually like) based on some graphs.
Recent example: got to try the Duende S*L bus compressor and WOW, I was floored by it. Even had a comparascent between hardware and this plugin and it's so good. But according to the "alias crew" it would be considered crap.