Sampler anti-aliasing comparison updated (including Kontakt)

a.k.a. why Software-Sampler sound quality still needs to improve

http://www.simonv.com/music/quality/

New examples include:
- Native Instruments Kontakt
- Emagic EXS24
- Propellerheads Reason (NN19 & NNXT)
- Steinberg Halion 2
- FL Studio 4

Enjoy listening,
Simon.

 

Very interesting !!! Did you try rgcaudio sfz (http://www.rgcaudio.com/sfzfaq.htm) ? The author pretends to have the highest resample quality...

The Kontakt aliasing is pretty disappointing... I hope to see that improved !!!

 

Just for the record, I'm not entirely sure that guy knows what he's talking about.

 

Looking again at these results I am not sure that the sound he choosed is a good example: given the spectral view of the original note, some of the highs are lows are not artifacts.
Some of the samplers (Fruityloops) seems to have a low pass and a high pass filter on low and high notes. So some low and high harmonics are filtered. Sounds smoother but I don't think this is related to aliasing quality.

 

Use your ears, not your eyes!

In the mix, are you really gonna hear these aliasing artifacts? If it bothers you that much, go and buy the various instruments you use samples of, learn to play them and then record everything live with £1000 mics.

At the end of the day, every step of the recording process is going to colour the sound and create artifacts, and these can often create a great feel to what you're doing, that's why valves are seeing such popularity again.

It may well be true that in a 'scientific' test, the SB live has the best handling of sample aliasing, but if you weigh up the value of all the convienience, creativity and variety offered to you by some of the great software samplers [Kontakt included], the rather limited SB live pales in comparison.

And I hate to point this out, but the majority of the record buying public [like 99%] will be playing back music on cheapo Hi-Fi units with inadequate speakers anyway. They won't hear any difference.

Seriously, all the energy that goes into testing this sort of thing could be better used in actually just sitting down and writing good music. In my experience, a good composition/song should sound good with just about any instrumentation and any quality of technology.

Trust your ears, not some spectragraph!!!

Will you really hear the difference once your sound has been run through compresion, reverb, EQ etc. I hate to think what kind of artifacts all those processes must add! So just listen and if it sounds good, use it.

 

very well said Spud Fully agreeing here

 

Thanks for your comments guys! I use Kontakt everyday to write music. The Tambourine was chosen because it's a musical example, I usually prefer pitching sounds into the right frequency range and less equalizing.

Oskari his aliasing evaluation is more accurate from a scientific point of view: http://jeskola.net/xs1/test

Spud: I agree with you, it's just that playing on a good instrument is more fun.

Cheers, Simon.