Updated or Pro-grade version of Axoloti?

#21

Food for thought: A friend of mine just got a kyma system and was instantly disappointed with the quality of the audio rate processing, something his kurzweil and expensive eurorack modules does better.

I think a carefully chosen transformer strapped over the axo would do a lot! I'm recording it straight into a nice/transparent and open preamp though good converters (mastering grade, if you will) and I'm getting a lot more bite, punch and more organic high notes then my software synths. It's dirty but it's got a vibe and traditional VA style patches sound surprisingly sophisticated.

More "high quality" modules and code will undoubtedly come along but since something like KYMA is essentially a massive dsp farm I think axoloti will always end up being relatively lofi in comparison. They're pretty much the industry leader in granular and aside from having some very bright people involved they're also rising to meet hollywood style industry demands. So if you want to invest in axoloti sounding more impressive hiring an audio dsp expert/artist to look into making the most of the power available would go a long way.

On the analog side you could do a session at a really well equipped studio and try tons of nice gear out to get a feel for what suits the axoloti best. The best engineers I know have worked out some truly phenomenal chains of units that compliment each other perfectly. A cheaper alternative would be Nebula and impulses of all the classy gear people have sampled that's available for dead cheap.

On that note I don't necessarily agree that there's no optimal colouring stage if you look at it from an instrument design point of view. Truly versatile instruments are rare but they do exist and I reckon that's down to great design. Things like Kurzweil (my old k2000 can process stuff to hell and back and still sound musical), Cwejman and of course Kyma. All systems that are flexible and open sounding.

What are you guys doing to make it sound more "high class" on the patching side?

Stuff I've discovered so far:
- Headroom and gainstaging! The filters seem to be incredible gain sensitive but can sound really smooth when you starve them a bit. Similarly I seem to be rewarded for being attentive to how loud I'm hitting wave shapers in context with the sound I'm going for.
- Designing your own envelopes. There are a few smoothing tools that means you can make curves that suit your ears. The same goes for any sequencing you might be doing. In my experience, this is one of the most effective things you can do to make a digital instrument more interesting and less abrasive sounding.
- Bandlimiting/softening audiorate processing and wave shaping.

#22

you could of course also add an extra analogue filter/tube-amp/distortion stage behind your axoloti..

#23

Putting an axoloti through a Universal Audio interface with their channel strip software might be a nice test. I've been using my axo through the analogue filter of my Shruthi, and then into a guitar amp today. Not hi end in any kind of hi-hi sense, but lots of nice coloration and the filter and amp combo respond nicely to dynamics.

I guess I treat it a bit like Max or PD, where the oscillators and filters don't sound 'sweet' on their own, they need some coaxing.

#24

I can only say I dont agree with your friend. One of my friends have had a Kyma system for years... And kyma sound SUPER smooth and VERY powerful. Thats my opinion.

It has to do with the "consumer stuff" that is used in Axoloti. If it had to do with coding, why dont someone remove this annoying -15 db saturation? For example you cannot use Axoloti as a send effect. If you send in a signal that is louder than -15db it will be saturated.... Which is completely useless in a studio set up. If it cant match the other gears level. You send -8db to the Axoloti and -15 comes out... Not very loud imo.

Yes was thinking about something like that. All though I still would prefer to fix it at the source, on the Axoloti. Would be loverly to have an Axoloti on steroids, with lots of more power and "pro grade" electronics, so we are not stuck with the low level output. Low level really bothers me a lot, compared to most of my other gear Axoloti is just too low in volume. For example my Virus synth. It can EASILY make sounds louder than 0db!!! Goes to +4db, cause it is prograde.

Yes I could do that. But I am moving away form using UAD stuff. Really dont like their marketing approach their sales tactics and how they, in many cases threat their customers. They tell you too much BS, lie and whatever to sell their stuff. Sorry, I understand their business now and dun't want to support it anymore.

Everyone, I KNOW that I could just add an anolog gainstage. But that doesnt remove the fact that Axoloti is made of comsumer grade electronics and NOT pro grade. Which does play a role in how Axoloti sounds. I think this is pretty clear after we got some of the MI objects. They dont sound ANY where close to MI own eurorack modules. ANd mostly the difference is the hardware and also MI gear is mostly 96 khz not 48khz. What I am saying is fix it at the source and not try to polish something later on... Always best to fix it at the source...

I wouldnt mind paying a lot more for a limited version of Axoloti if it was a beast with a lot more dsp etc. It is just too annoying having to work with 3-4 Axolotis at the same time....

Maybe it would be a good idea to integrate more than one Axoloti into the same patcher. A bit like the Nord g2, where you have 4 parts in the same patcher. Maybe Axolotis could use that approach, loading 4 Axolotis into the same software. This would make it easier to work with.

#25

If you need 96kHz oversampling you can try the few oversampled objects in tiar/dev (see the distorted guitar patch in my last contrib).

About the outputs... Well i own a "pro" Clavia G2 Engine and sorry to say that, but the output of the Axoloti is far less noisy (i literally have to disconnect the G2 usb to avoid digital noise to bleed to analog).
If you use a good power supply and shield what needs to be shielded, the Axoloti analog converter is quite good.

The G2 engine - like most "pro" stuff - does not have symetric/balanced outputs.

The Axoloti has a stereo output that can be used as one symetric output for a better S/N ratio and better noise immunity with long cables.

#26

Yes I saw those and all ready commented on them. They are nice :wink: But is is not the same as having a louder better sounding system. Of course it will help a lot, since sound sources will be better quality. But still not the same as overall better system. I also get better quality from using wavetablesm than from using Axolotis own osc. I never use them, only my own wavetables.

Yeah I have got a G2 engine too. G2 is way louder than Axoloti. Can be pushed to around 0db. My G2 does not have very noisy output. I do get some humming from the box itself, but not through the cables. Anyway, I will check up on the noisefloor when I get the chance, just in case my memory is letting me down.

Okay. I dont use long cables, so it is not really an issue. And I dont really think it makes up for anything to be able to use long cables over better quality.. Come on....... WOud you rather have the ability to use long cables than having a better sounding system?

And my post had nothing to do with balanced/unbalanced... Anyway, it is really old by now, but if people still like to debate this, I dont mind either :slight_smile:

#27

lately, I've also created some modules that can be oversampled.
eg. wavetrainOS and LPHQOS (Lp filter)
I've been working on a new filter too, that's kind of similar to the svf, but with a few differences, allowing 2 cutoff points for a wider bandwidth and independent resonance controls for HP, LP and BP (these resonances do influence the other modes too, but they all sound a bit different).

the oversampling process is done simply by putting the necessary audio-processes into a for-loop and taking the avarage of these. (eg. for an oscillator, you also divide the frequency by the oversample-number, same for filters, but with filters you could also leave this, enabling you to reach higher cutoff levels)

#28

@SirSickSik

Interesting with the oversampling. Also @Smashed transistors oversampling stuff is interesting. I just got my Virus C Redback fixed, which has been broken for 10 years, so I am spending a lot of time on that these days, not so much Axoloti. But look forward to get back to Axoloti with all this new stuff :slight_smile:

#29

For the record, MI Clouds and Elements run at 32kHz sample rate, Warps and Rings 48kHz, only Braids uses 96kHz. Quite some "only" have a 16 bit AD/DA. The brds objects in Axoloti use 48kHz, and yes, that is a trade-off.
MI Clouds has a lofi setting for 16kHz, 8-bit compressed (µ-law) recording, to extend the recording time to 8 seconds, no sdram in MI Clouds. Not really a quality decision, still some people will love the grainy sound of that.

@jaffasplaffa I get your point about consumer versus pro level, but for clarity, dB numbers are meaningless without reference, dBu, dBV, dBfs... Axoloti Core uses "consumer line level", I think it is the job of "pro" equipment to be able to adapt input/output gain to whatever you connect.

Discussing sound quality really requires separate topics for gain/level, AD/DA quality, and algorithms, not mixing these all the time.

#30

Hey @johannes Sorry I didnt mean to rip up the old thread, and I didnt expect you to answer. Was a discussion with the other posters :wink: But thanks for the info bout the MI stuff :slight_smile:

But a few had posted new things in the threads, so I just followed up on it.

#31

hi, not to take this too OT but am interested in doing what i can get the best quality out.
tho i also think with a little effort patching it really does sound great as is imo.

will check out the upsampling modules mentioned above and hopefully get into making some modules soon.

@SmashedTransistors could you explain a little further on this please. regarding the 'shield'. what does this mean? also 'a good power supply': as in a powered hub? or are you suggesting that using the dc barrel jack would be ideal for axoloti to perform at it's best?

haven't had any problems with axo output level while recording. if not compressing on the way in then leaving quite a bit of headroom to allow for compression etc afterwards. Usually turning things down, axoloti included. I have noticed the excellent S/N ratio on the axoloti. But does sound like something to take care with when using as a send - thanks for tip OP have not run into that yet.

am also a fan of occasionally adding an analog stage of some sort after axoloti.

#32

Hi,

Low quality switching regulators tend to add noise. High quality linear regulators are better for audio use.
Good shielding avoids electromagnetic pollution (the high freq from a PC or even from the Axoloti itself).

Some "DI units" can also be helpful in case of ground loops (which was the problem with my G2, ground loop through the PC usb port -> pollution by the PC switching power supply).

#33

Averaging is not a very good decimation process. Its frequency response is a sinc, which have sidelobes decaying at -6dB/octave.

#34

thanks for the tip. will do further reading.
had a similar issue with my blofeld and since like to have a DI on hand.

#35

I want to add I am sorry I am not so much into the very low level stuff of Axoloti. And thats why I dont always use the right technical language. know that leads to confusion, once in a while. So I'll work on getting the lingo right, with all the dBu, dBV, dBfs.. etc.

Thanks to anyone contributing to this thread. It all just comes down to wanting to get he best sound out of Axoloti as possible, which I think we all essentially do :slight_smile:

So I apologise if the lack of technical knowledge leads to confusion once in a while.

#36

You can add that many of the contributors are not native English speakers and acronyms and terms differ from a language to another. Thanks to the Axoloti we can share patches, objects and code snippets to illustrate what we are talking about :smiley:

#37

Yes Axoloti patching is a universal language for all of us :slight_smile:

#38

yeah, well, the oversampling was a quick fix that was needed for a patch, but making the right window just asked just a bit too much cpu for the patch to keep working..
it's in the planning though..

#39

If you want to get the best out of axoloti's out you should put it through a DI w/ some nice jensen transformers or a transformer-tube setup. I don't find the axoloti out extremely noisy or problematic, but I have not ever listened critically to it.. there are far more problematic setups like the nord drum 2 for example where the output is ridiculously low for what you'd expect to be a "professional grade" product.

But then again, you start playing with that thing and it sounds so good you don't care anymore about these things :slight_smile:

#40

@bafonso

Yes thanks for the input :slight_smile: I considered that too, the DI option. I am starting off with playing around with all the new oversampled objects. I'll first make it as good as I can from the source and then move on to the output stage. I think I am going to ask a friend to help me build something custom for output stage.