The aliasing thread

[intel]

..

[Robert Randolph]

Hum. for a second I almost supported Wusik and wusikstation by releasing a special soundset, joining the affiliate program and purchasing the program...

I guess we all make mistakes... Ill have the soundset done for Rhino 2, who's developer is a little more sane.



developers do little to sell their product by letting their personalities and such run rampant. Wusikstation is great, but Im not gonna support someone who repeatedly lashes out (falsely) like William. Likewise, my favorite VSTi is Discovery, but I wont support discodsp because of George's rude responses to customers in the market place (and I really dont mind george at all, we chat on IRC often). Of course there are other examples... imageline, bombfactory...

You do very little justice for your amazing product by attaching your personal issues and such with them. Ive learned that myself... it's not a pretty thing.

[TomD]

I get the feeling William is having a rough week. Though that doesn't justify the attacks on Rene and George who have both acted honorably in this thread.

Ironically I think Wusikstation sounds better than Zeta, Sytrus, or the DiscoDSP stuff. But it's great because of the design and excellent underlying sample set, not because of any aliasing.

Tom

[george]

I wont support discodsp because of George's rude responses to customers in the market place

I'd kindly ask not to generalize from an isolated case, please. That issue was due two faulty license transfers, unfortunately happening at the same time, which, by the way, have been sorted perfectly. Be aware about license transfer are still free of charge (meaning we deal with the costs).

For a better example of support, you are welcome to visit our forum at KvR anytime, thanks.

[george]

Ironically I think Wusikstation sounds better than Zeta, Sytrus, or the DiscoDSP stuff. But it's great because of the design and excellent underlying sample set, not because of any aliasing.

Congratulations then, you have found a product with a set of features that actually fit your needs, however, regarding this topic focusing on resampling qualities, I'd say it doesn't excel in that area (yet).

[ionized]

I wonder how SampleTank 2 does in this area? Could someone test it also?

(Notice that I'm posting a nice, non-temper flairing question (I hope ) - resisting the urge to say anything about William, George, or Rene . I'll just let these great developers settle this on their own. Arm-wrestling maybe?)

[Innominandum]

William, try not to take things so seriously, and for crap sakes *bitch-slaps* get it together! There is nothing in this discussion that warrants your behaviour. Your product has aliasing, so what? Work in a fix for the next release. I'm always working on to improve my work, and continually find new ways to improve. When something unexpected comes along it can be disappointing, but it's another thing that can be improved.

Some people here aren't particularly friendly, or are geared for technical brass-tacks conversation. Take it with a grain of salt and not personally. At the end of the day, who's on the beach with a blonde goddess under each arm? Nobody can take that away.

Many people here have stated that they like your product, so you already have support. Just keep plugging away, learning and trying to improve. Take a deep breath, thicken up your skin, have some pride, and keep going.

[Z3R0T0N1N]

William:
I just want to say that I own and LOVE W-station. It is an exceptional instrument, but that doesn't surprise me, nor does it surprise anyone. Nobody is going to change their mind about it because of some graphic representation that really has next to nothing to do with what it was designed for. Musik is what it was designed for, and by someone who is pationate about musik, and about designing instruments. A sampler is just a sampler, and really there arent many real samplers in software anyway. W-station is something entirely beyond this stuff, and those who get it, get it. So don't trouble yourself. It's that you were approaching this as a discussion, when really it is more of s lecture. There are some who are good at discussing, and others who are good at, well.... whatever the hell was going on in some of the answers you recieved.

anyway, this isn't meant to be replied to, as I am clicking on the stop watching topic but I just wanted to let you know that things are much clearer than you might be feeling right now.
Cheers, mate.
Thanx for the best virtual instrument I own.

[LlunaSol]

It's clear that WilliamK is a passionate person and had a bad day, and it's also clear that the test comparison that started that thread would had been a great thing coming from an independent 3rd party, which would have add a touch of real fair play, specially being simply one of the details to decide which sampler "is better". Nobody buys a car after reading a breaks comparison. Please, both as musician-user and long years betatester of many products, I would ask the next one to make public that kind of products test, show us the results on SEVERAL areas. Otherwise don't make assumptions on competence products.

Only seeing the place that have Kontakt and Halion on that test coming from companies having "the money" to solve it, can show us how much relative it is to the developer each detail and assumption to the final product goal.

Thanks to Rene for such precise explanations, as usual. For me, how simple I am, the sound coming out my audio system, and the usefulness of the sound libraries is what makes a synth (tool) usable

[Urs]

Hey guys,

sorry for being late, looks like I missed the mess.

Urs is correct in his statement. You are not.
Truth is not achieved thru democracy.

-René

René, what do you refer to?


This is some pages back, dunno if there's already an appropriate answer:

Then you would still need a fir filter to downsample it, as the oversampled sample is at Nx the standard samplerate.

Then you would have certainly some delay.

Doesn't sound like anything you could use in a real-world sampler to me. But perhaps it's just me.

Well, in a sampler, just like in a synth, you can use a trick: If you algorithm has, say, 16 samples latency, you simply calculate 16 samples more at note start and discart the first 16 samples. You simply process with some look ahead

The only thing that screws it up is, the NoteOff will be processed later. (I guess that everybody here lowpasses other control signals anyway, so that 16 samples would be neglectable).

btw. I think I would prefer polyphase filters (IIRs) over FIRs when filtering. Works fine for me, with low latency.


However, I've got a question:

Has anyone ever oversampled a grain (windowed portion of samples) using FFT? - I think one could transform a 128 samples grain to frequency domain (64 real, 64 imaginary values), then map these values to, say a spectrum with 512 bins (512 real, 512 imaginary), do the inverse transform and thus quickly oversample that grain 8x...

Is that a viable way?

(the idea of using a windowed grain is to avoid the effect frequency aliasing, as we can not consider the samples before and after the grain)

Cheers,

Urs

[CypherOne]

ah, a thread by any other name...

[Urs]

Btw, you can officially call me stupid, because at no time I thought of the idea George described, namely oversampling in realtime.



Urs

[george]

I wonder how SampleTank 2 does in this area? Could someone test it also?

Of course. Guidelines are available at the page. If you also want to know the tempo it was made, use 80 BPM. Once someone manages to get the mono wave, PM for details and I will make the rest asap.

it's also clear that the test comparison that started that thread would had been a great thing coming from an independent 3rd party, which would have add a touch of real fair play

What are you talking about? This is a strict audio analysis report. Something like "fair play" doesn't exist, no matter who will make those tests.

I would ask the next one to make public that kind of products test, show us the results on SEVERAL areas

That's as easy as going to each developer page and check for specifications. However, not all of them will reveal resampling quality for obvious purposes, hence this study.

Only seeing the place that have Kontakt and Halion on that test coming from companies having "the money" to solve it, can show us how much relative it is to the developer each detail and assumption to the final product goal.

The sampler comparison study will show a true fact: most of the biggest companies are using ancient algorythms in their interpolation engines, and smaller developers are the ones who are usually pushing the step forward (with several exceptions, of course). If you read the whole thread, you will find several posts where users actually care about this.

Btw, you can officially call me stupid, because at no time I thought of the idea George described, namely oversampling in realtime.

Hey Urs, absolutely no need to call anyone stupid. Humans make mistakes though, and usually wise people is the only who realize about them We aren't robots, are we?

[Sascha Franck]

I don't even care if the samplers I use (EXS and Kontakt) have "poor" quality.
I have done tracks almost exlusively using the EXS, some patches didn't even contain more than one sampler per octave and I did a lot of chord and pitching work too.
Sounds all fine to my ears.
And it doesn't seem as if a lot of "professionals" (whatever that means) would bother either.

I couldn't be arsed using VSampler as the whole thing would just be driving me mad - I'd better take the double amount of aliasing (or whatever you call it) instead of dealing with such an interface.

If it sounds and feels good, it is good.

[bmanic]

If it sounds and feels good, it is good.

See, this is the dilemma. Once you try something that sounds even better, how do you go back?

Try replacing those old EXS tracks (ALL of them) with SFZ doing the playing. On second thought, maybe you shouldn't. You might be upset.

My point: " If it sounds good it is good " but only until you have heard something better, that is, another reference point.

Cheers!
bManic