Audio through and effects processing

#1

Hi All,
I'm having real trouble using my Axo as an effects processor and I'm sure it's down to stupidity.

I have a Behringer QA1622 mixer with all my synths going into it. My roland piano is my test source. From the aux send I have a short mono (balanced?) cable going to the Axo and the same going back from the Axo to the L channel of the Aux return. All good.

To test I'm using the demo/youtube/multieffects patch but I get the same result with any patches really. If I try to send any more than 50% of my source into the Axo I get distortion coming back out. It appears that the signal is too hot. I've tried the inconfig and outconfig objects, setting gains to -12db or 0db with no results. I simply can't get an audio signal fully into and out of the Axo without it crackling.

I admit I'm pretty much clueless when it comes to signal strengths, db and that sort of thing.

Is there some basic rules around this? I've noticed if I turn down my master AUX Send knob (which is a sum of all channels) and turn up the AUX return I can get more channel % sent without distortion.

#2

The Axo core has a maximum peak to peak input voltage of about 2.4v, quite low, mixing desks can put out 10v peak to peak, driving that sort of level into the Axo will cause distortion and possibly damage it.

The input is an unbalanced stereo input not a balanced mono input, you need a cable with 2 mono plugs on one end and then send an unbalanced line from the desk to one of those.

I had intended to use my axo in a similar way, mostly for guitars though, but without a level matching pre amp for the input and output it's pretty much useless for professional work as an effects processor I'm afraid. I intend to build or buy one but haven't had time or spare funds.

Running a keyboard directly into the axo might work better, most keyboards have unbalanced outputs running around 5v peak to peak.

#3

Thanks for that. That explains why I have to turn it very low to stop the distortion. I also noticed that while it stopped the crackle it still sounded a bit "off". I'll fiddle around a bit and try it on the instrument lines as well. See how I go.

#4

In the patcher there is an input and output module control that enables you to set gain level presets to suit, but it's no substitute for a proper front end, IHMO the analog audio side of the Axo is it's big weakness.

#5

One of the "problems" with using Axoloti as an external effect processor is that Axoloti has a fairly low output. Axoloti can maximum spit out around -14db, before it start saturating the output, which is probably what you are hearing.

So if you use Axoloti as an external effect processor, maybe as a send effect and you send a -5db signal to Axoloti, the Axoloti will saturate thew signal -9db(14-5=9db). I am not sure if the input will have the same effect, as I havent really used it as external effect yet. Which could give even more trouble.....

I think the best solution would probably be lowering the send signal so whatever you send to Axoloti does not exceed -14db, so it is not feed with louder signal than -14db..... And then re-gain the signal AFTER the signal has left Axoloti again, to the level you like.

#6

For my 2c cents worth using decibels for measurement of signal levels is a silly idea, peak to peak voltage tells you much more about what's really going on.

Decibels is a ratio, not a measure of loudness. Useful for levels of air compression by speakers but useless for electrical signals.

#7

I'm "secretly" hoping for a Maxoloti one day. Same basic board but with more CPU power and SRAM and a CODEC with higher levels (as for pro audio gear). In the mean time, mixing desk AUX send and return channels may provide the needed control over gain staging.

#8

With the future facility of inter Axoloti SPI communication, I think that it will be possible to build Axoloti Clusters similar to the Raspberry clusters but dedicated to audio processing :stuck_out_tongue_winking_eye:.

The audio interface of the Axoloti is very good compared to other boards or even to some PC audio interfaces/soudcards.

#9

How dare you suggest Axoloti is not perfect as it is and that the analog audio side of it is not up to professional standard.

This seems to be the reaction I have got when I have suggest the same thing anyway.

#10

Yeah, that is one feeling you get sometimes in here........ I have adressed the same cuplrints several times with same results as you feel that you got. I just gave up and started learning Pure Data instead. Also because I need more DSP than Axoloti offers. So it was kind of natural next step for me anyway.

#11

Well it's an open project so we can take it and make it into what we want it to be.

Having a community that welcomes criticism and ideas for future developments of the project would seem to me to be essential, if the current leaders aren't doing it, let's just ignore them and do it ourselves.

Much of the difficult and time consuming work has already been done, with a few motivated people working effectively together many of the issues could be sorted out reasonably quickly and the project could be amazing.

Before I discovered Axoloti i had imagined something very similar myself and though there are things I would have done differently myself, the core of what could be very close to what I imagined is there and so I think it is worthwhile putting the time in to make it into what I want it to be rather than starting my own hardware project from scratch.

I've spent a fair bit of time with Axo over the last week or two and am enjoying it now I have a handle on how the software works and the object libraries enough to do things. I will try and get more active on the forum as well and see if I can't get help to sort some of the issues I've found out.

I feel I've got a bit of a handle on the code of subpatchs and objects and don't think it's beyond my coding skills to be able to make objects. Getting into the code of the Axoloti app itself however will present more of a challenge as will trying to understand the stm32 architecture and toolchain enough to work on the firmware, but I am in this for the long haul, I have a clear idea of how I can use Axo in my rig with the hardware as it is and it will form a central part of my modular system eventually.

The board should last years if treated well, and they are not so expensive it couldn't be replaced easily if damaged or broken, I'm actually thinking of getting another one anyway to play around with getting them to talk to each other and spread the processing load of the patches I have in mind to build eventually.

I do know a bit about electronics design and again have a very clear idea about a revision of the board that would use basically the same firmware, with whatever modifications were needed to make it run on a different hardware platform, but have a better analog side and better adc/dacs, and then build it into a 19" rack mounted enclosure with balanced lines and also a eurorack format module

A general purpose audio DSP is something I need that would be hellishly expensive to buy but that given the existing code base of axoloti i code conceivably build such board and run axo on it.

The software is where the work is, digital hardware is relatively simple these days because so much is integrated into monolithic chips, developing new firmware and an app to build patches from scratch myself is beyond me I'm afraid but I do think I am capable of being an active useful developer of this project to try and make it into what I need it to be myself.

I'd warmly welcome your comments, ideas and support if you want to work together.

Synergy, it's what music is all about, 2 people working well together can do the work of 4 working alone.
with 3 or even more, the possibilities are astounding.

#12

Oh and btw I don't know much about pure data, can you do axo like things with it?

#13

Yes everything I would say and a whole bunch more. Like FFT processing & phase vocoding, that is more heavy on the DSP.

#14

10 posts were merged into an existing topic: Critter and Guitari - Organelle

#19

I really enjoy the Axoloti,

I can focus on porting my synth and DSP algorithms knowing that @johannes et al. take care of the not that easy implementation of a reliable real time system with very low latency.

I won't underestimate the intricate issues faced when dealing with synchronizing inter board audio signals, implementing support for multiple USB controllers and other cool features while maintaining reliability and low latency.
So, even if I'm waiting like a 8 year old on xmas eve for these features (and spy regularly the ongoing work on github), i understand it will take time.

For me, reliability was the big issue of Clavia's mighty G2. The little Axoloti is much less buggy for a fraction of the price.

About the inputs/outputs, I think it would be useful to test the improvements of some pre amps, amps and direct input boxes with the Axoloti. Shielding and clean analog power supply should also be tested.

Surely, a nice super analog digital / digital analog board/shield including valve pre/post amps would be great.

Inter Axoloti communication (and/or higher CPU) would make polyphonic x2 or x4 oversampling possible, combined with differentiated anti aliasing, it would provide top notch oscillators, filters and distortions :smiley:

#20

Valve preamps might be going too far but high signal level balanced ins and outs with high quality dac/adc's along with more processing grunt and ram would make it close to perfect for me.

Could be built onto a shield I guess, maybe that is a project worth working on.

Do you know anything about circuit design?

#21

Hello strum, I am axoman. You don't know me but I'm generally regarded as one of the more advanced members around here, so if there's anything specific you would like a hand with, please let me know and I'll see what I can do to help :grin:

#22

Yes and that's precisely why I switched to software :sunglasses:

#23

Today you have nice little tube preamps kits for less than 50€.

#24

I'll add a comment for others who might be reading the start of this thread and then leaving with the view that Axoloti can't be used as an effect processor. I've had great success using Axo for FX by using it in line with my synth before going into the mixer. This avoids the level issues when using it as an FX send. Of course, I can only get one synth through it but it works great. Reverb, delays, bitcrush, sidechain compression. All these things are available and work well.