I personally prefer high/average/low, although noisy and clean seem equally valid too. Maybe interpolation naming would make more sense, but unfortunately there are some samplers giving no details about this.autloc wrote:Low sounds the same to me as "poor". Why not just say "noisy" and "clean"?
autloc, the part you're missing is that those coefficients (0.3 and 0.7 in your example) change from sample to sample and are very dependent on the relationship between the sampled rate and the playback rate. That part makes them not LTI.autloc wrote:I think we must be talking at cross purposes. The equation y[n] = 0.3*x[n] + 0.7*x[n-1], for example, (a linear interpolation or 1st order FIR) cannot add spectral components.Muon Software Ltd wrote:No, not at all. They're not LTI processes at all and can introduce spectral elements not in the original signal.
Agreed.René wrote:After all, aren't the limits between cathegories subjective?
The great value of those graphs is that they give you a (relatively) objective measurement of interpolation noise. As such, and for political reasons, it's probably best to avoid any subjective labeling on them entirely. Also, you might want to add a brief summary of the better points raised in this thread. Something like "keep in mind that resampling fidelity is only one aspect of synthesizer quality, and any of these instruments is 'good enough for rock'n'roll.'"george wrote:Agreed.René wrote:After all, aren't the limits between cathegories subjective?
I am like manymany other users (owning products by George, WilliamK and the wonderful sfz+ and others from René) just beginning to understand what this all is aboutRené wrote:What about removing the "high, average, low" or "good, medium, poor" or "da bomb, so-so and shitty" cathegorizations completely?
After all, aren't the limits between cathegories subjective?
-René
. Well, actually it's a lie, I downloaded all the pages (mostly before sentences were deleted maybe, I don't know) to understand what could be meant and hope to know it tomorrow .
and musicians round here . It's not only stoicism in philosophy, we learn here . 
) are wonderful for users. Not because of the bashing, of course, but because that's the right words to read and read again and(in my case) maybe understand in some time and then goin' back to your vsti-s and see what you can make of it.right, i agree - that's why there's aliasing with sample rate conversion but not with a constant fractional delay. i was just trying to say that "linear interpolation" or "hermite interpolation" do not necessarily alias. in fact, when used for delay lines (unless there's constant adjustment to delay line length) there is no aliasing at all.autloc, the part you're missing is that those coefficients (0.3 and 0.7 in your example) change from sample to sample and are very dependent on the relationship between the sampled rate and the playback rate. That part makes them not LTI.
Maybe in 50 years we have sinc texture mapping.
In this case the linear interpolation primarily has an impact on the frequency response (specifically lost of high treble). So I guess you could describe this action as being consistant with a filter. Guess we are indeed talking about the same thingwhen used for delay lines (unless there's constant adjustment to delay line length) there is no aliasing at all.
Perhaps you meant tarnce? Only girly rock'n'rollers use samplers.keep in mind that resampling fidelity is only one aspect of synthesizer quality, and any of these instruments is 'good enough for rock'n'roll.'"
That's bullshiet. Everyone and his 6-yo sister does know that psx1 eats n64 alive even on sunday mornings.Adding some fuel to the fire, and to reinforce rene's quoting about 16 vs 12 bits wars time ago (cool), i remember when Nintendo's N64 gaming console was released, it has bilinear interpolation for texture mapping. Half community of the gamers complained that its graphics looked dull compared to sony's psx1, wich just used 'nearest' interpolation for texture mapping...
Or you could invert the relationship:george wrote:I personally prefer high/average/low, although noisy and clean seem equally valid too.
I think it was more because the N64 had no memory, so textures were tiny and very blurred when looked closei remember when Nintendo's N64 gaming console was released, it has bilinear interpolation for texture mapping. Half community of the gamers complained that its graphics looked dull compared to sony's psx1, wich just used 'nearest' interpolation for texture mapping...
Submit: News, Plugins, Hosts & Apps | Advertise @ KVR | Developer Account | About KVR / Contact Us | Privacy Statement
© KVR Audio, Inc. 2000-2026
