"Does a sampler interpolate even if you play the sample as is, at its root note?"
yes of course, unless the sample's root frequency is equal to the host's output frequency. if you've got a host running at 48khz and a 48khz sample, there _should_ be no filtration or anything done. some real world samplers however will still apply some filter even when the sample's root frequency matches the output sample rate. it depends upon exactly what filters and interpolations are applied.
honestly why do people debate the sound quality of a host?
-
- KVRian
- 507 posts since 14 Mar, 2004
BLING BLING BLING BLING BLING BLING BLING BLING BLINGa bit late but still, my personal list why people like to think that daws have a sound:
- because nobody likes how the whole voodoo and "be a golden ear to hear the subtle differences"-business is taken away from them. that little touch of black magic in our profession is something many people want to preserve at all costs. also, today an engineers job is much simpler than it used to be. they already took away those big mixing desks from us. remember how people came into the studio and first thing they said was "and you KNOW what ALL THESE BUTTONS do???" and you could reply with a fake-modest "well, uh, yeah, no big deal huh". no more. and now all daws are supposed to sound the same? booh!
- because theres a million things that you can do wrong when doing a nulltest, resulting in countless stories of nulltests that didnt cancel out perfectly. yet, if someone made the nulltest correctly and the results canceled out, stories about tests in the same daw and same version where it didnt cancel are just stories of people who werent able to make a proper test.
- because many many people dont understand how pan law works and what it does
- because theres still thousands of engineers (including very good ones) that choose to ignore even the most basic psychoacoustics and make statements like "im not fooled by subtle volume differences, i can hear that file A sounds a bit more spacey than file B" without laughing
- because its VERY cool to say "oh no, digital is not digital, its not that simple my friend" (see also point 1 about the black magic and voodoo aspect)
- because there are people (especially long time pros) that were traumatized long ago when some apps still had stoneage engines, i.e. older protools with 24bit linear resolution and no headroom on the mix bus.
- because those same people bought expensive equipment back then to make up for these shortcomings. no way will they now admit to themselves that they invested such insane amounts of money just to fix a temporary software design problem. so, hosts must have a sound. and since theyre pros, everyone who comes to their studio and hears them rant believes them, cause, you know, theyre the pros are they not? they must know!
- because many people arent able to split the problem into single pieces. they make no difference whether the sound difference might come from the converters, some plugins they used or whatnot. none of it having anything to do with how a daws summing sounds.
- urban legends have a tendency to be immortal.
im sure theres more
You win teh internets!!!!!1!1!!11oneoneone
(or teh KvR, ah cannae remember whuch)
-
- KVRist
- 440 posts since 11 Jul, 2003
Thank-you for the answer Aciddose.aciddose wrote:"Does a sampler interpolate even if you play the sample as is, at its root note?"
yes of course, unless the sample's root frequency is equal to the host's output frequency. if you've got a host running at 48khz and a 48khz sample, there _should_ be no filtration or anything done. some real world samplers however will still apply some filter even when the sample's root frequency matches the output sample rate. it depends upon exactly what filters and interpolations are applied.
-
- KVRAF
- 9600 posts since 17 Sep, 2002 from Gothenburg Sweden
This is just a theory of mine.Not sure how it works in the real world. If you play a sample at the exact same frequency there's no need for filters and interpolation. However if you just move the pitch bend wheel for example you'll need it. So it'll probably sound weird having the interpoaltion/filters/whathaveyous popping in and out again. So it makes more sense having them on all the time.loomchild wrote:Thank-you for the answer Aciddose.aciddose wrote:"Does a sampler interpolate even if you play the sample as is, at its root note?"
yes of course, unless the sample's root frequency is equal to the host's output frequency. if you've got a host running at 48khz and a 48khz sample, there _should_ be no filtration or anything done. some real world samplers however will still apply some filter even when the sample's root frequency matches the output sample rate. it depends upon exactly what filters and interpolations are applied.
For a drumsampler it makes sense to be able to turn that stuff off but for a sampler like Kontakt it makes more sense to have it on all the time.
If that really is the case or not i don't know but it sure looks that way.
-
- KVRAF
- 13446 posts since 14 Nov, 2000 from Hannover / Germany
Hm, I seem to remember doing a test regarding that issue (quite) a while back. I was just playing back some sample in Logics EXS (root key, velocity 127, same sample rate and bit depth of what I was using in my song settings) and compared it to the same sample on an audio track, just that I inverted the phase of the audio track. I think I got *almost* to complete cancellation. The "almost" part might've been because velocity 127 wouldn't exactly represent the audio tracks volume fader at 0dB (even if it theoretically should do so).jupiter8 wrote: For a drumsampler it makes sense to be able to turn that stuff off but for a sampler like Kontakt it makes more sense to have it on all the time.
If that really is the case or not i don't know but it sure looks that way.
And well, considering that sampler interpolation problems are getting way worse when using lots of pitch bending or pitch shifting (by playing notes far away from the root key), I would assume that - at least theoretically - a sample played at root key position with a velocity of 127 shouldn't go through any interpolation process.
Further, with all those several GB large sample libraries, covering each key and many velocity switched zones, you surely wouldn't want your (hopefully pristine) sample quality to be affected by any interpolation processes, regardless of their quality, don't you think? I mean, that's one of the key reasons why people were starting to release those large libraries in the first place - for them not to be affected by any interpolation issues.
So, for me it'd only make sense if (proper) sampler developers would not allow for any interpolation to happen on root key zones.
But what do I know... maybe I should just do another test one day. Should be pretty easy to find out.
As far as the rest of this "debate" goes - ah well, where's John "my audio sounds better with the harddrive placed on a chair" Vestman when you really need him?
Seriously, if a "neutral" summing/nulling test goes wrong, there's just two reasons for this to happen:
- The test wasn't performed properly (master faders going over 0dB so they need to be adjusted, some panning going on, plugins used, dithering algorythms used, automation used etc etc).
A properly performed nulling test requires a bunch of (stereo) audio files, all well below 0dB (so you don't need to touch any fader in order to avoid clipping), preferably a realtime bouncing scenario (offline bounces often happen to be partially broken in quite some hosts, even if things seem to get better these days) and bouncing to the *exact* filetype you used for the imported stereo samples, without any dithering or whatever other file "conversions".
I'm actually wondering how so many people could get an easy procedure like this wrong. But it seems to happen all too often.
- Something's broken in whatever host. And yes, this happened before indeed - with more than one version of more than one host.
There are 3 kinds of people:
Those who can do maths and those who can't.
Those who can do maths and those who can't.
-
- KVRist
- 276 posts since 8 Feb, 2004 from France
Peoples who claim there are differences just don't know how to make good sound out of their host. That's all.
-
- KVRian
- 1144 posts since 26 Sep, 2005 from Germany
why not discuss the sound of digital hosts.
people have discussed the sound of nearly everything.
There doesn't have to be differences. But still it is possible for digital systems to have a different sound, depending on what is build into the system/software behind the curtain. And thats where we are not able to look onto the code and say, ah its done like this or that.
So its unlikely that there are differences but still its not impossible at all.
I mean microsoft tells people that their OS is a secured and well-coded platform. People believe this unless they hear that some hackers found the security hole nr. 2345 or so.
So I think its always good to have some critical people that do not trust a system and what its said to be able to do "perfectly".
The sampler thing above in the thread is a good example. In usual cases you would think, hey they are digital, so they must be all the same. In fact there are differences in the pitch qualities.
people have discussed the sound of nearly everything.
There doesn't have to be differences. But still it is possible for digital systems to have a different sound, depending on what is build into the system/software behind the curtain. And thats where we are not able to look onto the code and say, ah its done like this or that.
So its unlikely that there are differences but still its not impossible at all.
I mean microsoft tells people that their OS is a secured and well-coded platform. People believe this unless they hear that some hackers found the security hole nr. 2345 or so.
So I think its always good to have some critical people that do not trust a system and what its said to be able to do "perfectly".
The sampler thing above in the thread is a good example. In usual cases you would think, hey they are digital, so they must be all the same. In fact there are differences in the pitch qualities.
-
- KVRAF
- 4822 posts since 14 Mar, 2002 from Somewhere else, on principle
WOW... this stupid thread is still alive? 
-
- KVRAF
- 4822 posts since 14 Mar, 2002 from Somewhere else, on principle
What rock have you been living under? It's been discussed countless times already, each time resulting in the same idiotic, meaningless gibberish.hifiboom wrote:why not discuss the sound of digital hosts.
people have discussed the sound of nearly everything.
-
- KVRian
- 1144 posts since 26 Sep, 2005 from Germany
yeah but thats how kvr is.John Vulich wrote:What rock have you been living under? It's been discussed countless times already, each time resulting in the same idiotic, meaningless gibberish.hifiboom wrote:why not discuss the sound of digital hosts.
people have discussed the sound of nearly everything.
Everything has been discussed already countless times. So we could just shut it down and live without.
hehe
-
- KVRist
- 210 posts since 20 Oct, 2003
The most funny part of these discussions is that for far the most people it doesn't even matter what they use. They'll make crap regardless, since at least 90% of all music made by people on whatever DAW is crap. Of course I hope my stuff is in the 10% that isn't, but so does everyone else. Good producers can make anything sound fine, no matter whether it costs $10 or $10000 or how the digital summing is done. It's called talent and it's a rare quality in any case.
-
- KVRian
- 1144 posts since 26 Sep, 2005 from Germany
I wouldn't disagree that talent is the final word if your music ix nice or boring.
But to a certain degree, equipment is important too, IMO.
Without the right stuff you are just limited no matter how good you are.
But to a certain degree, equipment is important too, IMO.
Without the right stuff you are just limited no matter how good you are.