Which reverb can rival/test Waves trueverb??

[bduffy]

Holy shit, that's a world of difference.

[jupiter8]

The folder name for the audio files was a particulary nice touch.

[bduffy]

Yes: subtle, yet effective.

[Sascha Franck]

Holy shit, that's a world of difference.

Of-f**king-course it is - I mean, do you see me getting involved into any audioholic discussions very often, such as the best EQ, best compressor and what not? No, you don't, because I don't have golden ears at all, more to the opposite.
But when there's differences like this, which are even clearly standing out for the half-deaf among us, I just can't stand people talking absolute rubbish.
Besides, I'm pretty much pissed off by the fact that our Mr. Shy couldn't even spend the single minute it takes to do such a test himself - instead he wasted several minutes calling people trolls and what not.

Fwiw, bduffy, you can do the same test in Kontakt 2. It will almost sound the same in case it's got to do a lot of internal pitching. But fortunately, with 2.2 there's the HQ interpolation mode, which reduces the aliasing massively. And you can also turn it on/off per patch - most patches with a proper keymapping don't show much aliasing.

And fwiw #2: As said before, I used the EXS to demonstrate the effect because it's really famous for its aliasing, most other plugins aren't even doing halfway as bad. In addition, samplers seem to be particularly affected, properly coded synths are usually doing way better - but this is the very reason why people are full of praise for, say, PolyIblit, which is indeed sounding almost identical at 44.1 and 96.
Yet, in a full mix, these small effects will easily add up.

However, I still see two major problems with working in 44.1 and then switching the samplerate and bouncing to 96kHz for the master:

1) You will have to mass convert all the used audiofiles and have your host "reassign" them, should you use an external sample converter. Not exactly what the doctor ordered in terms of workflow efficiency.

2) Assuming you work on a song with a sound like in the 44.1 example, you will most defenitely adjust both arrangement useage and mixing tweaks to make it sound musically ok. You could even like that aliasing sound (I mean, we're using bitcrushers and the likes as well). Now, in your mixdown, all those "charming" aliasings will be gone and your master will sound extremely different from what you've been mixing.
In a "normal" mix, there should possibly be benefits, but once you used that aliasing creatively, you will just run into problems.

As a solution, I could only imagine using complete track stems, converting them to 96kHz and then importing them into a 96kHz project for mixing (assuming your computer is up to the task). Not exactly an ideal solution, but probably showing quite some benefits.
Working in 96kHz all throughout shouldn't exactly be possible for us mere mortals, due to lack of horsepower, which is especially true once you're using a lot of virtual instruments and other CPU hungry FX. I mean, I have absolutely no problem in bringing whatever machine to its knees in 44.1 already...

[Shy]

Sascha, you expect a "f**king apology"? Ask your grandma for an apology. If you hadn't noticed, I did say you probably meant things that are horribly aliased to begin with, and surprise, you posted just that, an example from a piece of crap piece of software that doesn't represent any form of proper audio tool. How about posting a real example with proper tools that don't have horrible internal aliasing and don't rely on the mercy of your sample rate (which doesn't eliminate any aliasing, only puts some of it elsewhere)?

[bduffy]

Holy shit, that's a world of difference.

Of-f**king-course it is - I mean, do you see me getting involved into any audioholic discussions very often, such as the best EQ, best compressor and what not? No, you don't, because I don't have golden ears at all, more to the opposite.
But when there's differences like this, which are even clearly standing out for the half-deaf among us, I just can't stand people talking absolute rubbish.
Besides, I'm pretty much pissed off by the fact that our Mr. Shy couldn't even spend the single minute it takes to do such a test himself - instead he wasted several minutes calling people trolls and what not.

Fwiw, bduffy, you can do the same test in Kontakt 2. It will almost sound the same in case it's got to do a lot of internal pitching. But fortunately, with 2.2 there's the HQ interpolation mode, which reduces the aliasing massively. And you can also turn it on/off per patch - most patches with a proper keymapping don't show much aliasing.

And fwiw #2: As said before, I used the EXS to demonstrate the effect because it's really famous for its aliasing, most other plugins aren't even doing halfway as bad. In addition, samplers seem to be particularly affected, properly coded synths are usually doing way better - but this is the very reason why people are full of praise for, say, PolyIblit, which is indeed sounding almost identical at 44.1 and 96.
Yet, in a full mix, these small effects will easily add up.

However, I still see two major problems with working in 44.1 and then switching the samplerate and bouncing to 96kHz for the master:

1) You will have to mass convert all the used audiofiles and have your host "reassign" them, should you use an external sample converter. Not exactly what the doctor ordered in terms of workflow efficiency.

2) Assuming you work on a song with a sound like in the 44.1 example, you will most defenitely adjust both arrangement useage and mixing tweaks to make it sound musically ok. You could even like that aliasing sound (I mean, we're using bitcrushers and the likes as well). Now, in your mixdown, all those "charming" aliasings will be gone and your master will sound extremely different from what you've been mixing.
In a "normal" mix, there should possibly be benefits, but once you used that aliasing creatively, you will just run into problems.

As a solution, I could only imagine using complete track stems, converting them to 96kHz and then importing them into a 96kHz project for mixing (assuming your computer is up to the task). Not exactly an ideal solution, but probably showing quite some benefits.
Working in 96kHz all throughout shouldn't exactly be possible for us mere mortals, due to lack of horsepower, which is especially true once you're using a lot of virtual instruments and other CPU hungry FX. I mean, I have absolutely no problem in bringing whatever machine to its knees in 44.1 already...

Yeah, exactly! My 44.1 projects are pushed to the brim, I can't imagine losing one iota of processing power!

So you think stems...there is nothing to gain by merely choosing 96khz in Cubase's export dialogue? You don't have to resample everything in that case; but is it actually creating a 96khz file, or a 44k file, wrapped in a 96khz shell? (if that makes sense, I should go to bed!).

[Sascha Franck]

Sascha, you expect a "f**king apology"? Ask your grandma for an apology. If you hadn't noticed, I did say you probably meant things that are horribly aliased to begin with, and surprise, you posted just that, an example from a piece of crap piece of software that doesn't represent any form of proper audio tool. How about posting a real example with proper tools that don't have horrible internal aliasing and don't rely on the mercy of your sample rate (which doesn't eliminate any aliasing, only puts some of it elsewhere)?

In WHAT WAY hasn't this been a REAL example?
The EXS is used in professional studios all over the f**king planet, some major sample libraries, such as Vienna, are available in EXS format.
Exactly the same goes for Kontakt, by now it's almost THE must have softsampler in a shitload of highly professional studios, be it as a sampler or as a playback engine.
And EXACTLY the same goes for hundreds and thousands of your socalled "proper tools". Yes, sure, many of them are not that horrible to begin with, but they still show similar/related effects.

But I really don't want to argue with you anymore.
This argument was about proving that some (or a lot) plugins would sound better at 96kHz, something you said it wasn't true, so you've been asking for a prove. I delivered you some audio examples that are just EXCELLENTLY proving it, I could as well deliver a shitload of further examples, using what is generally accepted as "proper tools". And while the differences might not be as dramatic as with the EXS, they will still be there in a LOT of cases.
But I can't be arsed to post a single further example as you just seem to be an incredibly unfriendly, rude and ignorant person with a more or less complete lack of reading, understanding and social skills. In addition you fail to come up with counter examples - such as telling us which sampler wouldn't show more aliasing at 44.1, compared to 96, hence qualifying as a "proper tool".
Case closed.

[Bernard Quatermass]

[Sascha Franck]

So you think stems...there is nothing to gain by merely choosing 96khz in Cubase's export dialogue? You don't have to resample everything in that case; but is it actually creating a 96khz file, or a 44k file, wrapped in a 96khz shell? (if that makes sense, I should go to bed!).

I think there is something to gain by exporting to 96kHz, and that's when you plan to apply further mastering plugins, which would benefit from the higher samplerate.
But in that case, I'd probably just export in the project samplerate and up it to 96kHz using the best available samplerate converter (fwiw, Audiomove seems to be doing a nice job), do the mastering on that file and then convert back to 44.1. Of course, perhaps Cubase is doing a better conversion job, I wouldn't happen to know...
Anyways, the problem here probably being the double conversion. It might be counterproductive to the advance of experiencing less aliasing. Again, I wouldn't happen to know, proper mixing and mastering isn't exactly my business.

[bleebsen]

I've done this a couple of times, mostly with oatmeal, which has quite a different sound at 96kHz.

The thread I posted about this ended up in flamewar too, "samplerate" must be one of KVR's powerwords.

I set the project to 96kHz, bounce the single instrument, set back to 44.1, externally convert the bounce to 44.1 and reimport and replace. Kinda manual freeze with oversampling.

Really working at 96kHz is still far out of the question, even with dualcores and two UADs...

[bduffy]

So you think stems...there is nothing to gain by merely choosing 96khz in Cubase's export dialogue? You don't have to resample everything in that case; but is it actually creating a 96khz file, or a 44k file, wrapped in a 96khz shell? (if that makes sense, I should go to bed!).

I think there is something to gain by exporting to 96kHz, and that's when you plan to apply further mastering plugins, which would benefit from the higher samplerate.
But in that case, I'd probably just export in the project samplerate and up it to 96kHz using the best available samplerate converter (fwiw, Audiomove seems to be doing a nice job), do the mastering on that file and then convert back to 44.1. Of course, perhaps Cubase is doing a better conversion job, I wouldn't happen to know...
Anyways, the problem here probably being the double conversion. It might be counterproductive to the advance of experiencing less aliasing. Again, I wouldn't happen to know, proper mixing and mastering isn't exactly my business.

Yeah, the double conversion could be tricky. that was something I was concerned about, way back when. If this is a great solution, then how come it isn't more common? I know my brother told me to boucne down at the highest bitrate when your final mix is ready for mastering; but I haven't heard about up/oversampling your project. But I swear I heard something on the first conversion, now, I'm not so sure.

I do know that the Kingston-Man said that Cubase's SRC is ass, which hasn't been my experience, but I'm in the same boat as you; no golden ears here. I've always found Nuendo/Cubase to do a good job with SRC - but - I've been dealing with 44/48k almost exlusively, so that explains a lot, I'm sure.

But I'm sure we're all in for a serious spanking if Kingston gets wind of this thread. Surprised he hasn't detected ignorance already!

[Sascha Franck]

I've done this a couple of times, mostly with oatmeal, which has quite a different sound at 96kHz.

The thread I posted about this ended up in flamewar too, "samplerate" must be one of KVR's powerwords.

I set the project to 96kHz, bounce the single instrument, set back to 44.1, externally convert the bounce to 44.1 and reimport and replace. Kinda manual freeze with oversampling.

But dude, really, why don't you just use proper tools instead?
I mean, sorry, I wouldn't happen to have a list of them, but I'm absolutely positive our friend Mr. Shy will supply a great amount of links to virtual instruments that just work as well at 44.1.

[bduffy]

I've done this a couple of times, mostly with oatmeal, which has quite a different sound at 96kHz.

The thread I posted about this ended up in flamewar too, "samplerate" must be one of KVR's powerwords.

I set the project to 96kHz, bounce the single instrument, set back to 44.1, externally convert the bounce to 44.1 and reimport and replace. Kinda manual freeze with oversampling.

Really working at 96kHz is still far out of the question, even with dualcores and two UADs...

See, this is what I'm wondering now: when I bounce down instruments, drum submixes, etc, should I do it at a high sample rate and convert it back offline? Do I need to switch the Cubase project to 96k, convert all the wave files, or will Cubase do it on-the-fly? Or should I just stick to the sample rate, after all?

[Kingston]

I seem to recall someone (Kingston?) saying that Audiomove was the best...?

heavens to besty lets hope he confirms it!!!!

here's the shoot out again. no need to troll (again). http://www.sg-acoustics.ch/digital_audi ... arison.pdf



folks, have a look at the weiss saracon graph to see how a resampler *should* perform. r8brain pro at the highest quality settings should look pretty similar, but audiomove is sufficient as well (even if the noise floor isn't such an algorithmic wonder).

[Kingston]

Do I need to switch the Cubase project to 96k, convert all the wave files, or will Cubase do it on-the-fly? Or should I just stick to the sample rate, after all?

it will not do it on the fly. all this must be handled off line, and outside cubase. preferably make a whole new project at the target higher rate, and import only the channel or plugin presets from the older project.