Hi,

In several forums or reviews, the sound quality of synthesizers is often debated. It is not a sterile debate for people listening more to the machines then to music. If you want to make professional sounding music, you need a professional instrument, even if it's just a VST.

Alchemy is often compared to Omnisphere, because both a principally based on samples, even if Alchemy only is opened to third party samples. I must admit (and the Camel team too) that the winner as almost always Ominisphere. Why ?

I don't own Omnisphere, but listened to enough demos to make my own judgment about it (the basic sound quality, not the global behavior). First, it's absolute clear that the huge bank provided with Omnisphere, if not as versatile as Alchemys factory bank, has a better sound quality, because the samples are better: velocity layers, round robin, better ground material… In the factory bank, samples sometimes sound so sad that you can only use them for resynthesis in granular, additive or spectral mode. But the additive and spectral resynthesis, that makes Alchemy to a fantastic sample manipulator (Omnisphere a very far from those capabilities) degrades the original sound so deeply, that it becomes difficult to get something sounding nice. It's original, it's special, it can't be done with any other instrument, but it's not always nice. Very often, I must give it up, and work in basic sample mode. Even for sound designers, it seems to be a challenge to make beautiful sound with pure additive or spectral resynthesis. Very often, an additive or spectral source is mixed with raw samples or VA.

My first wish for V2 is a dramatic improvement of resynthesis and of the factory bank, that should have the quality of the competitors (it's not a problem to download 30 gig of data, we have time!).

There a so many presets based on raw samples (sometimes + VA), that the sample manipulation consists only in a mix of a basic sampler (too often with low quality samples), VA tools (filters) and FX (CamelPhat and CamelSpace built in Alchemy). Of course, Alchemys modulation capabilities are fantastic, but it's not enough. No more.

It may be unnecessary to use 24bits samples for making high compressed MP3 to be listened in a noisy tube or bus. But this argument doesn't work for musicians. When I listen to the superb samples and presets of the new Dream Voices library, or those of Simon Stockhausen (alchemistry), I'm convinced it worth to make the job.

I'm a so big fan of Alchemy (and Camels customer service!) that all my reviews or posts on other forums are always positive. It would be my great pleasure so see the camel grow! I believe that Alchemy is the most creative and inspiring synth that has ever been done. But on this forum, we must also help our friends to go ahead, and sometimes, say what they may not be happy to listen to… if it's true!

Raymond
http://soundcloud.com/marginalray/sets

Hi
I can follow you 100% here.
When i first bought Alchemy i was very disappointed with the factory samples quality.
If i wasn't doing patches for Camel Audio and Alchemy i would wipe out the most of the factory samples because i never use them for my own creations.

Alchemy could have ruled the world now if the factory content was on Spectrasonics level.

Very interesting (and honest) point of view.

I'm a fan of Alchemy as well and I truly love it, but to be honest I've never really used the factory bank nor any default presets. The reason I bought and decided to use Alchemy is because I wanted a powerful tool to manipulate my own field recordings. Sounds that I record myself with the equipement of my choice at the quality and bit depth of my choice as well. So if the sounds sound bad, it's my fault (or choice).

Marginal Ray's point has never really been an issue for me personally, although I totally understand what he means and fully agree. Most of the people use the factory banks coming with their softwares, so having very high quality samples is a must to keep in compitition with the others.

Interesting topic that I will follow.

Judging synths and Plug-Ins by their factory content is of course justified but also misleading. This counts for many synths and Plugs out there but I will not name any or diss any of those. I've never used the Alchemy Factory content as I roll my own anyway, so the quality/non-quality doesn't bother me. I can't compare it to Omnisphere as I have never used that due to the lack of sample import.
What is definitley neccessary in a future Alchemyversion is an improved FX section (amongst other things) and an improved resynthesis quality, although I hardly never use resynthesis to emulate a "real" sound (works alright with simple sounds like e.g. violins and some chromatic percussion...) but for exploring textures I haven't heard before.
/2c

Regarding quality of Factory samples. As we've mentioned before, we simply did not have the budget to license or produce enough high quality content for Alchemy v1, and while a lot of our self produced samples have proved very useful, they certainly don't have the same level of detail as many other sample libs...particularily the samples of real instruments. we were also limited by the necessity to keep each SFZ less than 15mb.

We have already added a lot of much higher quality and more detailed sample content to factory since v1, and I think the library we have planned for v2 should keep most people happy

Col
Last edited by biomechanoid on Mon Jan 16, 2012 10:10 am; edited 1 time in total

As far as pure sonic fidelity goes, Alchemy is, with a handful of others, in a fairly rarified zone. We're talking sound quality here. And you can bet that Alchemy will only get better, while still staying CPU-conscious as time goes on.

Personally, I wouldn't use them for the same thing, so I wouldn't necessarily compare Omnisphere and Alchemy. And again, personally speaking, I think that Alchemy strikes the right balance between quality and quantity with the samples provided. I don't want or need 30GB of factory-preset samples, but I do applaud what they are doing with the reasonably priced sound banks they offer, where they can and do offer higher-quality samples. Way to go.

Alchemy is much more than the sum of its parts. And you can throw in the superlative customer-service into that mix as well.

Preset quality:
The quality of the presets is not important for a sound designer, but It can be for a musician who doesn't want to create what specialists can do better. I bought Alchemy and not just the player, to "turn the knobs" (modulation is an absolute must). But a good preset is often a good start point if you are more concerned about your composition than about the making of sound.

Sample quality:
In Alchemy, the quality of the samples is important, because it's mainly a "sample manipulator". Even a sound designer should not be obliged to use only his personal samples to create good presets.
Camel Audio's idea is simple: you don't need high quality samples, because they will be manipulated with the additive and spectral engines. But these engines doesn't sound so fantastic, and the editors are difficult to use (who is painting in the spectral editor ?!). Alchemy's factory presets uses very often samples in granular mode or even in basic sample mode, and then, the real quality of the original sample is obvious. So I'm happy to read from Col that Camel Audio will provide the V2 with better samples, like the competitors (Omnisphere, Kontakt…) do. And yes, definitively yes, a 30 mega bank has become something absolute normal today. What is not normal is to feed the best sample manipulator of the word with so… average samples (I talk about the factory bank, not the extra libraries, in most of the cases)!

Resynthesis quality:
Sampleconstruct said : "I hardly never use resynthesis to emulate a "real" sound (works alright with simple sounds like e.g. violins and some chromatic percussion...) but for exploring textures I haven't heard before".
For me, the goal is not to emulate, but to start exploring from a high quality ground material. When I record a few words with my microphone, and load them in Alchemys additive engine, it sounds really bad. I'd like to explore new textures, and it would be fantastic to tweak around in the main interface or in the additive editor to create a new voice, different from mine, but that sounds human, and not so noisy. The same with instruments: it's easy to create new textures with additive resynthesis, but difficult to get a really nice sound. So I ask Camel Audio. Will it be dramatically improved in V2 ?

FX quality:
The main problem of the effects section is not the sound quality. I think Alchemy has the best sounding FX section I've ever seen in a virtual synth, especially combined with the modulators. Even the reverb can sound very musical, if you control the high frequencies. The four distortions make Alchemy to a real virtual amp tool, so that you do not need any extra plug to make very fat sound. The Mod FX section may be difficult to adjust, because you have not knobs for chorus, flanger, phaser… but you have the ingredients to make all these effects yourself. I'd like to have presets for the Mod FX section, just to show me the way, and then tweak them. I'd like Camel Audio to transpose the CamelPhat and CamelSpace presets in Alchemy. Of course, they would prefer Alchemy customers to buy these two tools, but if you use only Alchemy, it's not really the money worth, because they are already embedded in Alchemy.
The EQ is very primitive, but "sounds" good. The band reject system is a bit bizarre: It would be so much easier to make the EQ work as a modulator, so you can define a bandpass for every action you want apply on the sound. Example for distortion: right click, add modulation, freq range. On the same level is the FX-Mix knob : It would be so much easier if each of the four sources could be routed to the FX section, or better, if each source would have his own FX section. Another point is the five slot limit for effects, especially if you start from a preset. You often need a third party FX because there is no free slot.

Conclusion:
Don't misunderstand me: Alchemy is magic, but if you have the luck to make music with such a fantastic instrument, you want it to be better. Col said the sample quality is limited by the budget. It's true that a pure VA synth has no sample problem, and that a pure sampler has only this problem!
Marginal Ray

thats not even remotely true. we wouldn't be investing in improving content if we thought like that, and obviously the general principle is better samples = better the presets, regardless of analysis methods.

col

I'm happy to read this. Of course, Camel Audio wants the best for its clients, and if they could sell Alchemy for 200 € with 30 Gig of super samples, they would do it. I just wanted to point out that when the synth engine deeply modifies the sound of the basic sample, its original quality could be less important. It's like puting a maximum of distortion on a guitar : it could become difficult to make a difference between Fender and Gibson ! I want to ear huge difference between the sound generated by each Alchemy engine... but after manipulating, not after loading the sample.

Raymond (hope not so marginal !)

@Raymond: Transforming speech samples in Alchemy does yield incredible results if you manually adjust the spectral content and eliminate the Pitch Var factor - check out this demo I made for my first Alchemistry Bank (if you don't already know it.

...and I think the distortion modules are the weakest parts of the FX section....the Reverb is indeed usable, so is the Mod FX but still both could definitely be improved.

Spectral? Isn't the Pitch Var factor parameter located in the additive engine? Or maybe I'm lost a bit, which is always possible...

Wrong/incomplete terminology from my side - yes, I meant the "spectral content"=the partials in the additive engine- like adjusting the data of the individual Oscillators...

Ray,

In terms of overall sample quality and excellence, have you heard the sample libs by Sampleconstruct (Simon Stockhausen) for Alchemy?

http://www.patchpool.de/alchemy_shop.html

How would you compare those libs to the factory sets by Camel Audio? Better? The same?

I'm just asking because I'm trying to figure our what you consider good samples?

Thanks.

Cheers

Doug

He mentioned my samples in his first post and thought they were alright

I have to concur about the resynthesis quality. I think this should be a major factor in designing 2.0. The ultimate goal should be 100% transparency... although I realize that this might not be realistic for 2.0, perhaps for 3.0 it will be. I can scarcely imagine the possibilities this kind of thing would engender! Imagine an acapella in a pop recording: rich, smooth, tubey, 100% perfect in every way. Nobody realizes it's resynthesized until it starts to subtly shift and go strange, and then not so subtly, and then it's completely different but still intelligible! At no time is there any loss in perceived quality...

That could be like the holy grail for pop music in particular... auto tune?! Pffft...
Last edited by DrWashington on Tue Jan 17, 2012 2:14 am; edited 1 time in total