like all my sounds are 44.1, i want to work at 48 or 192 even, upsampling is bad right?

Any word?

Wish more people posted here, because man i hate other gear forums with a passion and i hate being trolled by people i hate in the process as well.

What's the difference actually?
I always wonder when I render my songs in Renoise.
Once I took a sample rate higher than 44100Hz and all it did was make my render time a fuckton longer, didn't really hear a difference.
Same with the bit depth. Is higher better?

your only going to have convert back down to 44.1khz in the end anyway - so that extra conversion step ends up being more degrading than just working at 44.1khz in the first place.

There are pro's and cons to upsampling, if your using a lot of VST's/plug's, it's perhaps worth considering the advantages.

You're not going to be able to "hear" the difference of 96khz, because human hearing peaks at the 20khz area, but you may "feel" the difference in sampling rate and the audio experience.

Some people say VST's shine more at higher sampling frequencies. Important to understand though, is not about opening up a frequency range that can be registered by the human ear, it's just that higher sampling rates filter the audio spectrum in a different way, by working on those frequencies we can't hear, and therefore producing a richer audio experience.

Also it has to be taken into consideration that most audio playback devices are still standardised at 44.1 khz, even CD's I believe (unless it has changed now?).

Soft synths run with out aliasing at 192khz, this vastly improves them for leads and shimery sounds. It also makes fx not alias and render with more resolution, if you use a good khz converter i am told that the benifits are huge for VST soft synths.

"I have done this test myself.

http://dl.dropbox.com/u/15285948/44100Hz.wav
http://dl.dropbox.com/u/15285948/96000Hz_r8b.wav
http://dl.dropbox.com/u/15285948/192000Hz_r8b.wav

Same synth rendered at 3 different samplerates and converted to 44.1. The difference is clearly audible.

EDIT: Converted with the aforementioned R8Brain,it's great."

^ above in quotes is from a kvr site talking about the benefits, this is the proof, listen. the difference is amazing.

Running at high frequencies is not going to make recorded audio better really, but it will reduce aliasing, meaning if anything you record is above 20khz then it will cause aliasing, at higher rates it will not, you can filter it after the fact during conversion and have no aliasing.

The big problem is that all VAs alias past a3 on the keyboard, this is because it is making feqs above nyquist, even though you cannot hear them they get mixed back in as aliasing noise and raise the noise floor of the track, and add unmusical distortion. This is what leads to ear fatigue in digital audio. There is a truth to this and i have heard it myself, i could not stand cubase until i ran it at higher rates, then it opened up and sounded good.

I want to know how bad upsampling will ruin my drums and things.

I wish i knew away to render the software at 192 and convert it back down in my daw sounding good. Then just work at 48 or 44.

This is the reason i got the plugiator, i dont have it yet, tomorrow i will have it. I hope.

All the demos of it i hear do not alias, they high freqs sound great, and the filters sound awesome almost analog, and nothing like the software and VAs i have been using over the years.

Its sound is generated at 176khz then down sampled to 44.1 . This is why, and if i like it, man i will sing its praises.

Yeh you can hear the difference in those clips - but rendering drums at 192 is going to be a waste of GBytes as most drum samples are recorded at 44.1 or 48khz - you wont be adding anything using them at 192khz.

synths, yes if your doing those high rez sweeps then rendering at 192 should help.

Yeah it's clearly audible form those samples, aliasing on the first clip 44100 is really evident, slight improvement on the 96000, and just a lot smoother at 192. The filtering gets smoother the higher the sample rate, so this is interesting. For soft synths this is a must, but as pointed out, rendering drum samples etc at 192 is not going to be beneficial.

heres the way i think of it:

lets say you rip a Cure song off a cd and into a 128k Mp3.

then you take that Mp3 and put it into (insert software here) and up it to, say 320k.

now repeat the same process, only rip the Mp3 at 320 instead of 128, then listen to them both.

you will hear the difference between just ripping at a higher quality compared to ripping at a medium quality, then trying to upsample to a higher bitrate.

hope this makes sense. it does in my head, but dunno if reading this will result in the same effect.

-Sin

@It's all to do with the "source" sound - again if the synths have the ability to run at 192 then by all means use it - but drum samples / other samples wont benefit as you can't "add" anything to it by upsampling as the "source" would be 44.1khz anyway.

I don't use any soft synth but I use lots of digital hardware and I also mix ITB and as per my personal experience, I do think that 192 Khz sounds better, especially when the final mix needs a lot of definition but it will restrict the number / type of plugins per channel and will be a pain in the ass to render / manage. When I use analog synth / drum machine, I tend to track at 44.1 / 48 but if I use digital hardware, I prefer 96 KHz to avoid aliasing issues. The synth himself have also a lot of impact on the digital artifact factor, my JP-8000 alias a way more than my Yamaha AN-200 for example.

I know 192khz sounds better, more air, no aliasing, cleaner sound more ditinct. Still up sampling all my drums would make them not as big ans punchy. That is what i fear, maybe r8brainpro or something would be ideal.

You can upsample them with a converter. Sound Forge has batch conversion scripts that are easy to use.

The thing is that it's pointless to upsample them though. You can leave them as is, use them in a session that's a higher resolution (if your DAW supports that, I know Reaper does)... and only from that point on will any processing you add to it will have that added definition. You can't add it back if the recording wasn't done in that resolution in the first place. Sort of like a snapshot of a picture. It is what it is. If you go trying to do time-stretching/pitch on it the sound will still break apart on some level with the original source recording.

I agree there are huge benefits to working at higher sample rates and bit depths. While I've worked on higher resolution rigs for years... my personal setup at home as it is is just 48/24 max for the time being. Been lusting after a new interface lately.

Oh, besides being a waste of time, it would also make the files unnecessarily large at that point since it will do nothing for the previously existing lower sample rate files.

Bryan, I wouldnt worry too much about sampling rates - just wait till the Plugiator arrives and the quality of it will make you remember it's the source sounds that are much more important than the DAW samplerate - I know you are probably itching for it to arrive.