Besides potential advantages of lite version kontakt libraries being CPU and RAM friendly,do tracks exported using lite(16-bit) versions of Kontakt libraries sound worse than 16-bit tracks exported by DAW created using 24-bit original libraries ? Are pro-audio producers(i.e.experienced producers in the music industry) easily able to distinguish these tracks from 24-bit ones?

 

In theory (afaik) the only real difference between the two should be the signal to noise ratio, meaning that the 16 bit will have a bit more noise. So raw and exposed, yeah there's probably a difference but in a mix I bet most of the time there's really not. Just a guess though. I try to stay at a reasonably high quality all the way through so I'd want to stick with the 24-bit samples because that noise adds up- but let your ears be your guide, if the end result sounds okay why sweat about it?

 

As for the poll, I think you need to add an option for "it depends"

 

Thanks for the valuable information dear

Depends on ?

 

Depends on what kind of samples you're using (acoustic instruments vs. electronic for example), depends on how exposed they are on the mix, depends on how many samples layer up as you're playing (one or two percussion sounds versus about five hundred piano layers, say) depends on the type of music and the density of the mix and how upfront they are in that mix, depends on a lot of things.

 

How do you know that a different sample rate is the only difference? You do not.

You do not know if they will sound different to you? Why don't you get one with both versions available and compare them?

Maybe they will sound different. Say they do sound different, is someone listening going to say they know why it doesn't sound the same and it's because of using a Kontakt Lite library at 16bit? With additional fx on them? Yeah,sure.

 

That's exactly how I explain the whole debate about analog emulations of plugins and synths. I'm still looking for someone who can listen to a mix and say they know the SSL Bus comp isn't hardware because it punches less in the low end, etc etc. It's a waste of time if you can't identify anything wrong from your own point of view. What would people do if there were no internet to ask if their Juno emulation sounds legit or not? Not publish any music?

 

If you have the two of them (hw and software version) in front of you, you will notice they sound different in whatever way. You are also feeding them into your DAC and then looking at the results on a computer screen. You have lots of help figuring it out. If someone says they can do the same thing with a fully produced and mixed track, about individual instruments they are pushing it. For stuff on the mixbus and used for mastering, maybe they can; because it is the last bunch of processors it hits. You get a very easy comparison of the two.

Buried inside a mix, like a sampler instrument, plugin, or synth, and their blind listen accuracy percentage would be not much higher than flat out guessing. One of the popular Youtube guys did this exact comparison with an Arturia Poly or Matrixbrute, a decent priced third party virtual instrument plugin, and then something stock from FL; using single oscillator waves. He would have had better luck playing Lotto and it was all his usual gear.

 

Last edited: Jul 10, 2024

Simple answer, of course they sound identical. 16 bit and 24 bit don't sound different at all. There is just more internal dynamic range in 24 bit. Only if the sample company does something very weird, like delivering the samples with a level of -60dbfs and the boosting it with the sampler (Kontakt), you will have an issue. There was one minified library where this was the case and the result was much nosier then the original, but the minifier caught this and raised the level of the samples before making them 16 bit and that fixed it. Can't remember what library....
The fact that a sample is 24 bit does not mean it has a -144db noise floor, it has the noise floor of the recording chain used. The best available AD/DA is in the very low 20s like 20 or 21 bits, but that does not mean the the mic, preamps and mixer deliver this at all. In my humble studio I can not get near 16 bits with mics. Just because of the physical noise.
Ok I found it
Orchestral.Tools.Berlin.Symphonic.Harps.KONTAKT.Minified.FIXED
nfo text:
WHAT WAS FIXED:
The previous release had defective samples; the samples were too noisy. My first guess was that something in the process of 24 >16 bit conversion, perhaps the amount of samples, had caused some trouble and the quality of the conversion got affected. But after some tests, the problem wasn't actually the bit reduction, but the volume of the original samples. The samples in this library are incredibly quiet, so when converted to a lower bit depth, the noise shaping algorithm of dithering replaces the extremely quiet zones into white noise. So, all I had to do was make a batch normalisation of the library, respecting the relative volume difference of the samples, with a resulting increase of 20.3 dBs.
As a result, the files are bigger because there's simply more data.

 

Last edited: Jul 10, 2024

Unless you actually check them and find NI does more to the "lite" free version, than just bitrate conversion. I'm not sure I would guess they would not degrade them some other way also. The whole point of them being free is to sell you the complete, "better copy". You think they are above that sort of thing?

 

The are minified by hacking, not by NI!
If you mean the orchestral suite by NI which has a full and essential version, they are not the same and also not free.