Part 1, Distortion for guitar programmed on a Sigma DSP ADAU1452

This video to shows the differences and sounds of nonlinear devices for guitar distortion in Sigma Studio from Analog Devices on a ADAU1452Eval DSP board.
Differences between hard clipping and softclipping , single or multistage, can be heard with sinewaves and played with guitar too.
You can see the spectra live too.
Cheers
Stephan Moeller

[@freeelectron8261]

Thanks Stephan for showing some of the dsp possibilities of simulating valve amp stages to produce the classic and complex tube distortion and "colour". Bravo!

[@stephanmoller2271]

Thank you very much !

[@hannuhanhi183]

Great job Sir !. One of the most interesting videos this year. I have the same board and have been experimenting with it for some time. Not doing any amp modeling but IR stuff. You could also use the FIR for cab modeling. Your amp emulation sounds fantastic !.

[@stephanmoller2271]

Hannu Hanhi : Yes , Fir are great. But IIR are realtime with no latency

[@hannuhanhi183]

@@stephanmoller2271 Discarding the latency of AD-DA conversion the theoretical latency of 1024 tap FIR with 48kHz sampling rate would be around 10 ms. Would that be too much for real time guitar playing ?.

[@stephanmoller2271]

@@hannuhanhi183 yes. I have done lots of recording with cubase and it was best to have 64 or max 128 samples latency. But what is the advantage of FIR ? IIR is much more realistic.

[@hannuhanhi183]

@@stephanmoller2271 Ok, if that's too much then it is too much. I am not a guitar player so I don't know. What latency does the 64 or 128 buffer latency translate in cubase, a few milliseconds ?. From my point of view the advantage with FIR is the fact that they can be derived easily from an impulse response and loaded to a FIR. So basically I am using the FIR for convolution. Yes I know, it isn't the most effective way. More effective ways might be partitioned convolution, but I cannot find anything regarding the SigmaDSP about the topic. My opinion is that there is no match to convolution in realism. An IIR would be an crude approximation of a FIR in that sense. Also there is the aspect that there is no simple way to convert a FIR response to an equivalent IIR response. There is a good article about cab modeling here. z2dsp.com/2018/07/10/impulse-response-modelling/
Cheers.

[@stephanmoller2271]

@@hannuhanhi183 Yes, 10ms is relativ high. I wouldnt play it. 3ms would be ok

[@mirelchirila]

are you the same Stephan Möller that did the ac30sim ? Opamp ac30 simulation ?

[@stephanmoller2271]

yes, thats me 😀

[@mirelchirila]

@@stephanmoller2271 it was a nice circuit, if you don’t have any more comercial interest in the circuit it would make for a nice video. This way you can have the ac30 implemented in every domain. 😃. Nice videos by the way , following along the parts now .

[@dominiklion3282]

Really amzing, thx for this great "tutorial"
Is it possible to use a adu1701 for this?
What key features are Important for this usecase?

[@stephanmoller2271]

thanks for your like ! It is possible to use any SigmaDSP.

[@kayahansevik428]

Hi Mr.Stephan Can you tell me how dou you connect your guitar into board I cant find 1/4" jack input on these boards Thanks

[@stephanmoller2271]

You can not connect your guitar to the board because the input impedance is too low. You need a high Z impedance preamp

[@kayahansevik428]

@@stephanmoller2271 so how do you test after programming thanks ?

[@TheEaserider]

awesome!

[@stevenburvenich168]

Soundwise this is perfect. Bias shift(s) and sag are yet not simulated I guess. That's where the "sustain" magic happens even at low gain settings.

[@stephanmoller2271]

Have a look on the following videos. There is a section about the nonlinear highpass ( grid current ) and crossover distortion too.