Luxonix Purity reduced to $49.00!

[fluffy_little_something]

I suppose if you make classical music, the quality is simply not good enough, but for background sounds for pop music it does the trick. I mean, studios will buy better stuff anyway or even hire real musicians.

I use just few sounds of it, mainly the strings which I have modified of course.
Many of the other instruments are not good enough, not even for my amateur needs, for instance the electric bass sounds or the fluegelhorn.

The synth waveforms as such also show some odd behavior. I did a pad the other day and no matter what I did, there always was that slow periodical fattening and thinning effect, probably phase related. So, basically I only use Purity for the few instruments which are impossible to create on my synth.

[HaganeSteel]

Nothing I said has anything to do with that.

OKAY.

Apparently I wasn't clear.

This has nothing to do with the quality of the instruments. Purity is awesome in that regard.

This has to do with the sample timing of this and certain other waveforms as a result of the samples being pitched too far down or stretched too far across the keyboard.

Okay?

It's not too much to ask for better and tighter sample timing.

A real solution to this problem is to keyboard track the sample offset.

The other solution is to edit the samples more tightly, or keep the samples more consistent (as in, don't mix articulations when mapping them to the keyboard).

The problem is that the String ENS 2/Violins waveforms have a Dimension Pro-like problem with their attacks being too slow. It's nowhere near as bad as in Dimension Pro, but Purity does not have an option to adjust the sample offset.

It doesn't even frustrate me anymore. It just makes me laugh now. I know my posts are longwinded, but geeze.

@Fluffy: The above wasn't really directed at you. And yeah, I've noticed that too. It doesn't bother me, though. All romplers have problems like that.

My XP-30's stock french horn is so bad that I'm convinced it's out of tune and completely unusable.

Purity just takes some programming to get right. I mean, what waveforms specifically are giving you a problem? It'd be good info to have.

Like, no one's bashing Purity here. It's okay to say when you have a problem with something. Nothing's perfect, and the more info that's out there, the more we all learn.

I'm kind of saying all this so that Sky Cha knows for any possible future updates, and I think the idea to keyboard track the sample offset is something that'd be easy to implement in the future.

Edit: Trying to be more clear!

Edit #897342347: I'm also going to say this: People presume that because I focus so much on orchestra that I'm out to actually sound like an orchestra.

I'm not.

I like MIDI orchestra.

I want to sound like the Playstation, you know? That's my sound. That's why I look to stuff like Purity and the Hyper Canvas over the more expensive or dedicated instruments.

Okay?

[bailees7irish]

Naoshi Mizuta's mixes sound like they're stitched together with duct tape.

I don't hear that at all, sorry.

[fluffy_little_something]

Both Saw waveforms show that behavior.
Others have a lot of noise, like Vox -> Shine for instance.

But I don't really care, the synth as such is rather limited, I use Tarkus for all synth sounds.

What I find most annoying is the user interface, its colors and size. Not to mention renaming presets.

[HaganeSteel]

Yeah, I just checked.

Shine has too much noise because it's not multi-sampled enough.

It sounds like Shine has one or two samples mapped across the entire keyboard.

That's basically the same problem I was talking about with String ENS 2 and Violins (both use the same samples).

About the saw waves: If you layer two different saw wave samples without detuning them, it'll phase like that. That should happen on any synth.

If you layer two of the same saw wave samples, Purity behaves almost identically to my Wasp XT, except for a loop point that I only notice because I was looking for it.

But there's really no reason to do that without detuning them. Two saw waves layered on top of each other sound pretty much the same as one unless they're detuned or programmed to behave differently from one another.

I'm noticing a pattern of this synth not being multi-sampled enough, but as of this time, there are only three waveforms I legitimately have issues with:

-String ENS 2, Violins (both use the same samples): Attack is inconsistent across the keyboard.

-GM Xylophone: About +20 cents out of tune, and has too slow of an attack the farther down the keyboard you go.

-Shine: Not multi-sampled enough.

So I'd say that is Purity's flaw and that is what people should look out for, that I didn't notice at first. Something like that is very difficult to pinpoint, but has potentially game-breaking ramifications, because, and I'm saying this mostly for Sky Cha:

If a sound is not multi-sampled enough, or pitched too far down, the attack of that sound can become inconsistent, causing it to play slightly out of time. It can also cause issues with tuning for samples that have some sort of modulation in them. <--So far, Purity really has no real issues with tuning, except for the GM Xylophone, but I have noticed this effect with some sounds. It's not ideal, but the other sounds I noticed this in are still mostly usable.

But given that I really only take issue with the String ENS2/Violins and Xylophone waveforms, that's really not that bad.

It's just unfortunate that this happened with the strings.

You're right, though, if you're using these strings in the background, for example, to play chords, they sound fantastic.

The problem is that the way Purity was advertised (and the sound of the samples) made me think it was good at spiccato, but it's really not. But it is decent at legato.

Funnily enough, I think the Emu Proteus 2000's Section patch is the exact same way.

And that's ultimately why I said Purity is hard to recommend. It really does recreate old romplers, warts and all, and if you're not into that sound, or, like me, an idiot who owns GPO but refuses to use it (:lol:), then it's likely to let you down.

[fluffy_little_something]

I am not so sure the noise with some sounds has something to do with sampling as I know that kind of noise from a few synths (not using any samples) as well. Maybe it is simply an audio issue.

With Shine and such samples it is a bit of a shame because normal synths usually don't produce such sounds. For the rare occasion when I need such sounds I use Mulab's internal synth, its sound quality is way better than that of Purity, and it is also more flexible and powerful. The price for that is a higher CPU load of course, and more complex patch programming.

[tapper mike]

Nothing I said has anything to do with that.

OKAY.

Apparently I wasn't clear.

This has nothing to do with the quality of the instruments. Purity is awesome in that regard.

This has to do with the sample timing

Which can be indeed related to the sample quality/oversight during production. I'm not talking about bit rate I'm talking about post sample recording production which one wound find these things and correct them before the product comes to market.

[HaganeSteel]

I am not so sure the noise with some sounds has something to do with sampling as I know that kind of noise from a few synths (not using any samples) as well. Maybe it is simply an audio issue.

With Shine and such samples it is a bit of a shame because normal synths usually don't produce such sounds. For the rare occasion when I need such sounds I use Mulab's internal synth, its sound quality is way better than that of Purity, and it is also more flexible and powerful. The price for that is a higher CPU load of course, and more complex patch programming.

Well, I know for a fact that Purity's engine is ace, but the actual recording process, or the synth that was originally sampled may have had noise in it and probably did, for sure, if the noise is that bad like it is in Shine.

I can see how that'd turn you off because you're right - I'm noticing now a lot of the voxes have very noticeable loop points and/or noise. I think "Choir Ahh" and "GM Halo Pad", although not realistic or anywhere close to being as good as what's in the M1, are still good for synth pad stuff.

The problem is definitely coming from not being multi-sampled enough and having imperfect loop points, which become very noticeable when the samples get stretched so far.

In some cases, like the Brightness waveform, it sounds like a sample is set to C#4, and then stretched all the way up the keyboard, or roughly about 3 octaves.

That's pretty bad.

But I feel that these are actually only minor issues for me.

Purity still gets some awesome sounds with the right programming, it just takes combis.

I wouldn't use Purity for vox-like stuff. It can do it, and I think the "Choir Ens" preset is pretty awesome, but I would go to the M1 for that first, as that Chorus 1 waveform still sounds fresh and fantastic to me.

So in the end, what is Purity really good at?

Well, I still think it's best at doing the K/Jpop-esque Phantasy Star Online stuff.

Here is my main issue with Purity as a whole: The sample content is not that good (Ravity came out during a time when this sample content was actually considered pretty good) and so it may be dated, and thus justifying its price drops.

Because the sample content is not that good (or, as the case may be, inspiring), this creates a higher barrier of entry into programming the synth because you have to learn how to program the synth and also learn what issues affect what sounds, and then learn how to program Purity to overcome these issues.

This means that in order to make Purity sound really good, you have to be really good at using Purity, and you have to know its sounds and samples inside and out, and it's hard to get that good when the base sample content doesn't really inspire you.

So consider what I just said here as the follow-up to my review.

Purity still gets a B+ for me, but I want Luxonix to do better in the future, and I think they have done much better, judging by the Toys in Wonderland and Mini Sketch demos, but I don't own those products because I don't own Kontakt, so I can't say for sure.

Which can be indeed related to the sample quality/oversight during production. I'm not talking about bit rate I'm talking about post sample recording production which one wound find these things and correct them before the product comes to market.

Yeah, I agree. I just need to stress that I wasn't criticizing Purity for the wrong reasons. Sample timing (and now some noise) can have serious consequences for people using the synth.

Edit: No one saw that. Actually, I just changed a word "very" to "pretty" because I want to avoid being taken so literally.

[tapper mike]

Purity only has two osc per sound. Proteus hardware had four and sampled well but they didn't do expecially great post production on "All" of the sounds. Some sounds were awesone and some less then awesome. Although the hardware varieites of Emu were far superior(IMO) to the virtual proteus and dsf's later versions. It was no playstation DAC's make a hell of a difference. The only caveat is... MSO because it was built ground up for protX

Your right, when sampling a single note for the entire range of an instrument one gets a more homogenized sound. One also has to do more correction for the top and bottom end. The aliasing for violins in proteus when played below the usable range of the acoustic instrument is quite atroucious. Same thing for any sampled sound beyond the natural range of the instrument.....And the sounds in early romplers were specifically thin sounding. Layering for velocity is only effective on these types of units is only effective if the osc's were recorded for that velocity. Trying to change the LFO will not bring back the character of the timbre for an instrument that is played hard or soft. If anything you are better off normalizing your levels and putting in some diffusion for your reverb then compress and eq to homogenize the overall sound.

[HaganeSteel]

Purity only has two osc per sound. Proteus hardware had four and sampled well but they didn't do expecially great post production on "All" of the sounds. Some sounds were awesone and some less then awesome. Although the hardware varieites of Emu were far superior(IMO) to the virtual proteus and dsf's later versions. It was no playstation DAC's make a hell of a difference. The only caveat is... MSO because it was built ground up for protX

Your right, when sampling a single note for the entire range of an instrument one gets a more homogenized sound. One also has to do more correction for the top and bottom end. The aliasing for violins in proteus when played below the usable range of the acoustic instrument is quite atroucious. Same thing for any sampled sound beyond the natural range of the instrument.....And the sounds in early romplers were specifically thin sounding. Layering for velocity is only effective on these types of units is only effective if the osc's were recorded for that velocity. Trying to change the LFO will not bring back the character of the timbre for an instrument that is played hard or soft. If anything you are better off normalizing your levels and putting in some diffusion for your reverb then compress and eq to homogenize the overall sound.

Yes! Exactly!

I believe a similar process was used by the guys at Square-Enix back when they used that Proteus 1 string sound a lot. That made me love the Emu sound, only to realize that it didn't sound like that out of the box and actually had a lot of problems.

Emu really didn't process many of their samples very well at all from what I could tell.

But neither did Roland.

It was the Sound Canvases that made me fall in love with Roland's sound, and not the JV/XV's, which sound different despite using the same sample content, and it's because I feel the guys at Edirol cared very much about what they do.

I get this same impression from Luxonix. I feel they care very much about what they do, but I feel that a lack of manpower and experience held back Purity from being stronger.

I would love to work with Luxonix as a beta tester, so I could tell them all these things before they released their material, and we could fix them.

[tapper mike]

I fell out of love with roland before sound canvas and could never return. If nothing more then from overuse. I did a long run with the GR,GI series from the GR1 thru the GR9. I also had a D50 thrown in for good measure.



Here are some odd things about DimPro. I've got to believe the sounds were recorded by DSF and then licensed to Roland/Cakewalk. Because they don't sound like roland hardware to my ears. Whenever I get a new library I let the sounds call to me. If I'm hearing something familiar from the first note it leads me down a path. A few really great sounds in a given library are enough to make it worth my while. I don't load up midi files I just play and listen. I pulled up a piano patch in DSF that rocked my world. I started hearing Carol King and Carly Simon and Todd Rundgren and it guided me to try and figure out the songs even though I hadn't heard them since the 70's. Magical. It all had to do with the entire recording and multi-sampling and refining the samples. That's why I'm so enamored by the PC3K though it's not in my financial realm. I wish they would have treated all the sounds in DimPro with the same level of dedication that they did the piano sounds. When you have something as rich as DimPro the last thing you want to do is layer different instruments like your standard midi stack sounds that come standard in Purity. Purity has it's place in house, electro and of course jpop/kpop but it's not a be all end all.


The other odd part about DimPro is it's future. Roland sold off Cakewalk to Gibson. Gibson has stated that they are going to be hands off of Cakewalk and at this point I don't know if that's a good thing or a bad thing. IMHO Gibson has no place in virtual audioware. Not in recording and not in soundware. Cakewalk on the otherhand has not been the mover and shaker it could have been in the last few years. Let's face it. Sonar isn't Abelton Live or Samplitude by a long shot. Someone needs to light a fire under Cakewalks arse

[HaganeSteel]

Here are some odd things about DimPro. I've got to believe the sounds were recorded by DSF and then licensed to Roland/Cakewalk.

That would make a lot of sense.

Purity has it's place in house, electro and of course jpop/kpop but it's not a be all end all.

Yeah. All of my criticisms towards Purity stem from this fact. I want it to be the be-all and end-all, but there's no way it ever could've been. I think it's possible to find a synth that can do that for me, though. The Hyper Canvas is really close, and this leads me to looking into the SD-50.

I'm now saving up for an SD-50, but I'm still not sure if it's the right move for me.

Dimension Pro is fantastic, but all of my romplers have severe, game-breaking problems.

Dimension Pro is incomplete, even with the Virtuoso.

Hyper Canvas doesn't have enough content, has terrible strings.

M1 takes too much programming, and has so much useless content that it's unwieldy.

Purity could conceivably be a one-stop shop, provided the problems with some of its samples get fixed, but I don't see that happening.

This whole time, I've been wondering if Purity has problems like Dimension Pro, or if I was just hearing things.

I still don't know. To me, it sounds like it does, but to a very minor degree. Is that something that should be criticized, or lived with? I think both.

I think Dimension Pro has made me paranoid, and I go out of my way to find bugs and problems with things now, and sometimes it paralyzes me from using my gear for no reason.

In my review, I mentioned Purity didn't have problems with sample offset or anything, and part of me still believes that, and part of me never did and always expected something to go wrong.

But whereas Dimension Pro is almost completely unusable for me because of its problems, Purity, I think, is still very usable.

I say all this because I also understand I'm holding Luxonix and Purity to an impossibly high standard.

But that's because they've come so close!

The other odd part about DimPro is it's future. Roland sold off Cakewalk to Gibson. Gibson has stated that they are going to be hands off of Cakewalk and at this point I don't know if that's a good thing or a bad thing. IMHO Gibson has no place in virtual audioware. Not in recording and not in soundware. Cakewalk on the otherhand has not been the mover and shaker it could have been in the last few years. Let's face it. Sonar isn't Abelton Live or Samplitude by a long shot. Someone needs to light a fire under Cakewalks arse

Mm, I didn't know Roland sold Cakewalk to Gibson.

I don't believe DimPro has much of a future if its past is anything to go by.

[HaganeSteel]

Okay!

Here's what I did! Maybe you guys can help me with this and decide if this is a problem or not.

I recorded this little test:

https://soundcloud.com/dimensionprotest ... dirol-test

This will decide if what I think is a problem is really a problem. It's Purity, Emu and Edirol all playing an arpeggio (G, Bb, D) against a kick-hihat rhythm that I'm using as a metronome.

The first example is Purity's Violins waveform. The highest note here (D) sounds like its attack is slowed a little, and to me, it throws the arpeggio a little out of sync.

The second example is Purity's GM Strings waveform. This I would say is the second tightest and sounds fine.

The third example is the Emu Proteus 2000 Section patch. This sounds to me like the worst.

The fourth example is the Edirol Hyper Canvas' strings, which are completely tight and have zero problems outside of sounding like crap.

Is what I'm hearing an issue?

Because it sounds like it is when I write music, and it's frustrated me.

[tapper mike]

IMHO DimPro and specific to that DSF has nowhere to go. I've checked out what the newer HD versions have to offer and...not impressed with the realism vs modeled instruments.


While I had great disdain for earlier versions of COSM it's come along way. Modelling synthesis is the way to go. Many midi wind guys I know use the Yamaha VL70 for hardware and Wallender Wivi for vsti.

I kind of wish TonyOstinato would chime into this thread as I bow to him for all things brass and woodwind related. He's the guy who's been there done that both in the acoustic and virtual realm. Intonation, harmonization and articulation (to a lesser degree) are vastly different (imho) on acoustic and virtual representations of instruments. Partly because of the nature of the beast. Trumpets are very very hard to play with exact intonation in upper registers. The character of woodwinds can dramatically change simply do to the way one approaches the mouthpiece.

My experience has been that. Though I too dream of the one huge library that fits all I'm reaching for single well modelled instruments over rompler libraries

[fluffy_little_something]

Number 2 and 4 sound terrible to me, number 3 inconsistent (the A# sounds pretty good), I guess I would take number 1 if I had to take any of those four...