Really good project! As a "guitarrist", I can say that this overdrive is a lot better then most of the digital overdrives that we can find in the market, and also sounds very similar to an original Tube Screamer, awesome work and playing, of course (nice Satriani's touch there :) )
You’ve just about convinced me to not miss out on DSP this semester at uni. Thank you again for all your videos. I’ve been getting more and more into PCB design because of them, and now I’m getting all sorts of projects from professors to work on. When any of my fellow students ask me how I learned KiCad and Altium, I point them your way!
Fantastic quality as ever Phil. You carry the torch for audio electronics and DSP videos, hope to see many more like this one! (Maybe an octaver pedal next?)
From what I'm seeing here, this is what pedal folks call a hard clipping distortion algorithm. It is more like a Distortion+, and not the kind of soft clipping that the Tube Screamer you showed actually does. It's fairly easy to get TS style soft clipping, though. Just add in the original dry signal (maybe after some of the filters) and tweak the relative levels a little bit.
Wow. The science of sound manipulation is amazing! So much variety driving decades of music styles from such a simple circuit. I can see why you love making pedal electronics. Thanks for the peek into the world I didn’t know about! Rock on.
Must be real exciting getting it set up and hearing it for the first time. When it comes to personal projects, at that stage you forget about eating and sleeping entirely until you've spent some time playing with it. And your projects are seriously dope.
I just did my undergraduate dissertation project on a hard-clipping digital effect! Your vids really helped me back then to have any idea of what to look out for and where to start. Kinda wish I'd had this video then haha! Still learning lots now... so many questions!
I can't believe that I find always your videos about all of my hobies :) Some day, I design flight controller for long range UAV and quad, then I came across to your video about subject. Now I am planning a DIY effect processor project and bam your video is here. And you are always one step ahead of me :) My idea is copying the behaviour of a 12AX7 tube and amplifier into a microprocessor as a starting point. Basically my method will be based on cascade connected frequency filters and gain stages but definetely this will be not that easy. Edit: Forgot to mention, amazing video.
Looks like our interests align quite well! :) Modelling valves is a great project. Have you checked out 'Digital Audio FX' by Udo Zoelzer? Awesome book on audio-based DSP, including some valve-modelling.
Wow man,this is also my interest ahaha!! I'm a guitarist myself and a electronics engineering student. Until now i implemented in MATLAB a 12Ax7 model and was trying to port the code in c++ ti realize a plugin. But It would be cool to realize this on a microprocessor! If you find some interesting resource please share beacause i love this stuff ahah
Thanks to Phil for the work done, and for bringing up the subject of antialiasing. I hope the topic will continue, you can improve the algorithm with the introduction of lookup-table vacuum tube.
awesome video! i'm watching it with great awe! Just one little thing..The tube screamer has the diodes in the feedback loop of a non inverting opamp, hence, when the diodes are conducting, the stage turns into a buffer. This causes the signal above the diode threshold to get a gain of 1, not zero. This gives the typical "clean blend" overdrive tone of the TS style OD (specially with high output pickups). On the function shown, above 2/3, the signal should have been X, not 1. In fact the function shown is actually hard clipping (diodes to ground, or on the FB of inverting opamp), where the gain above the threshold is zero.
Thank you! While I completely agree with you in the analogue world, given most books and papers in the DSP-world, the soft-clipper is modelled as shown in the video (for example, or similar but cubic) (e.g. Stanford ccrma.stanford.edu/~jos/pasp/Soft_Clipping.html, Udo Zoelzer DAFX book p. 118, DSPrelated www.dsprelated.com/freebooks/pasp/Soft_Clipping.html etc..). The 'softness' comes from the transition between linear and fully saturated (which thereby in that transition region also adds compression). Hard-clipping in contrast, again in the DSP-world, means usually a single linear region - no transition - and hard saturation at the end of the linear region (e.g. Stanford ccrma.stanford.edu/~jos/pasp/Hard_Clipping.html or www.dsprelated.com/freebooks/pasp/Hard_Clipping.html). Lastly, a main difference between soft- and hard-clipping, in terms of frequency domain, is the presence/prominence of higher harmonics. Hard-clipping has a significantly higher harmonic content, whereas soft-clipping - as shown in the video - has (slowly) decaying harmonics. The image I showed with a resistor and back-to-back diodes, while this is indeed more for hard-clipping, I find it far easier as a simplification and illustration for the general clipping process. That's why I included it instead of the op-amp feedback path.
@PhilsLab you're right. The function describes the shunt diode correctly. There's confusion on what "soft clipping" means.. on the analog opamp clipping world, it means "non inverting, diodes on feedback".. but strictly speaking, it means a sigmoid function with soft knees, as you show.
Awesome video and project! Combines a ton of areas of my EE background. I know the video is already packed with content, and you can’t cover everything in detail. But I would love to see more detail about the signal chain on the circuit. I guess because that’s the area I have the least experience in. Specifically how the STM32H7B0 interfaces with the analog signal. A quick look at the data sheet shows it has two ADCs and one DAC. It looks like some slick programming software is provided for it. Maybe I should just build my own pedal and get the experience!
I dont know if i didn't pay enough atention to your explanation, but it would have beed very useful if you mentioned that you are using the absoliute value of the input on the slide with the formula for clipping. I like your DSP series. Do you upload your code and PPTs in some place?
DSP and STM32/ARM stuff still goes way over my head, but cool video! I shall refer back when I have learned a bit more. AVR is my forte and it's aging...
Dude seriously, that was awesome. Are you sure you're an EE or a Guitarista masquerading as an EE. Loving your design videos but super cool to see you play. Thank you. 👍
Hey Phil - Loved this video, thanks for all the really well thought-out insight into such a fun project! Regarding the choice to pre-low-pass-filter the signal (around 12:00 in the video), why not upsample -> nonlinear distort -> low-pass filter -> downsample instead? It's absolutely more computationally intensive, but I think the STM ought to be able to handle it. bandlimited DSP is really tricky lol.
Very interesting indeed, I'm little way behind as I'm looking for the perfect Sine Wave but I'm at the bulbs and Wien Bridge Oscillators stage so a lot of work for me, no need to look over your shoulder just yet LOL !! great video....cheers!!
Since the overdrive is software you can stack 2 overdrive modules to get a better distortion sound, much like an overdrive pedal feeding a clipping amp. High gain amps like your 5150 has several overdrive stages stacked.
I can't understand why the antialiasing filter goes before the nonlinearity. The fault is obviously on my side since I'm not a fluent DSP speaker! ... in your example with the 1 kHz tone, digital converted signal spectrum is just a dirac around 1 kHz (I'm sure there is a lowpass antialias in the analog conditioning stage, so ther isn't aliasing just due to the sampling). Now, if I apply a lowpass Chebyshev filter, I think I end up with the same signal, just time delayed a bit. What am I missing?
Phil, amazing content as always. As an guitarist and electronics engineer these series are my favourite! Maybe a similar DSP project idea is to model a real amp into the STM32. Also what is the delay between the input and output signal on this pedal? Do you notice any delay while playing?
Thank you very much! Amp modelling is a great idea. I've been long planning on making a miniature 20W digital guitar amp. Delay is predominantly due to the audio buffer (variable), so currently 20ms. Can't feel it when playing!
I have the desire to create an IR loader using the stm32 platform. But I'm a beginner in this whole digital world, etc. Would any of your courses help me with this goal of building my own IR loader, and perhaps other projects and pedals?
Hi, I am following this with interest! I am looking to implement similar on a synth system, but using Q15 integer calculations. Do you have a link to your source code as I would like to implement the same design, but change to Q15 integer?
Very good ideas and sound. Has the source code been released? I can't find it in the git link, can you tell me the specific path if there is a release?😆
Thanks for all videos Phil! Keep it up! If you know something about FPGA programming other than xilinx , please concider making tutorial (even paid one) or video about it. Another thing - link to terms and conditions are not avaliable via link on your Fedevel subsite in checkout information. Minor issue but still.
Hi Phil. Amazing work. Do you plan to make delya/reverb videos with stm32h7. Do you think it is capable? I know memory is a problem here, can it be made with external flash?
One thing I've been wondering about with these mixed signal circuits is the reasoning behind the usage of an audio codec for signal input/output: Given that most STM32 MCUs have 12-bit or higher DACs and ADCs, would it be possible to use these for audio I/O, and if so, would it be feasible? Thank you for all these videos, they're really informative and entertaining :)
Thank you! 12-bits isn't that great for this type of audio, in my opinion. However, the ADCs are usually pretty fast in STM32s, so you could potentially oversample and average to gain some resolution. I did try this a couple years ago but the results really weren't great. It's easier and better-performing, although additional hardware is required, to go the external (dedicated audio) codec route.
Great stuff. What does boost do? I assume that's for a future video. Also there's some delay/reverb, is this part of the pedal? Wondering how much headroom there is for other effects i.e. how much % of the processing power is being used right now. I assume it depends on the quality of the filters. Hopefully cabinet modeling/simulation is also coming to this series.
Thanks! The boost adds a constant amplification before the clipping stage (preGain += 50). Can also adjust a DIP switch to add output volume increase for solos. The reverb/delay was added in post on my PC. I haven't actually formally checked load - I'll measure how many cycles are used currently and how many in sleep, and get back to you! Cab sims is a great idea!
How did you do the artwork on pedal? Tayda? or DIY? Can you feel the phase shifting or delay during playing? Obviously off the shelf digital systems like Neural DSP or Axe Fx are really good at imitate the tone but they stiil lack of feeling and dynamic feeling. How about in your desing?
I went with pedalpartsplus for enclosure drilling and UV printing. No noticeable delay while playing (can vary buffer sizes and thus latency) with good dynamics - pretty cool given such a simple algorithm, but it feels very amp-like.
Why not shift the kneepoints on the clipping curve rather than changing the input gain so that zero gain / equal power is (approximately) obtained independent of the overdrive setting ? Or is it the intent that increasing overdrive also increases the output level ? Just asking, I'm no guitarist.
That could definitely be a way of doing it. However, to me it's more straightforward (and based pretty much directly on the analogue prototypes) to amplify the signal before going into the gain stage, as the output level of the gain stage only varies a bit above certain distortion levels. The final level then is tweaked via the volume control, if a big gain adjustment is done.
Please explain why the final "limit the signal to between -1 and 1" would not introduce more distortion and aliasing that would then need more filtering.
It does and I alluded to it around 12:50 (somewhat as 'food for thought'). I said fs/4 was a reasonable starting point. However, I played around experimentally with various cut-offs of the AA filter and running the clipping algorithm, far lower than fs/4. But to me fs/4 yielded completely fine results (as heard in the sound demos), so didn't have to push it any lower in this particular case.
All is well done as usual but I don't see any filter for the neighbor banging on their ceiling from below. 10Hz high-order high-pass filter should do it :P
Nice video can you make one on how sustaniac guitar pickup can be made and another on what the fishman fluence pickups are different than traditional guitar pickup.
@@PhilsLab I'm a PC/MCU enthusiast. I'm not at all proficient as you with CUBEIde, but I know my way around things. ESP-IDF is one of them. Honestly, I'd just like to implement a huge hardware rotary encoder on a FPGA. Can you point me in the way of actual real world solutions?
Excellent breakdown and demo. I'm curious, with 60+ taps on LP at 48k, what's the small time delay through your system? Also, I'd like to see the IMD, intermodulation distortion, FFT with AD generating a 3 note chord. The frequency domain must be full fun peaks. Thanks for the video, well done.
Thank you, Craig! At the moment, delay is about 20ms (predominantly due to the buffer size, which can be varied). Good idea - I'll have a look at that :)
What if your input signal has a DC offset? Then you need the bias (or in the analogue domain, AC couple) for symmetrical clipping. This illustration is for completeness sake and also to show there are adjustments to be made quite easily to achieve different clipping styles.
As someone studying engineering and making DIY pedals as a guitarist, this is totally cool. Hope to see more high gain stuff in the future. Thanks!
That's awesome - sounds like we have the same interests :D More to come! :)
@@PhilsLab
Do you intend to make videos that cover the basics of FET, BJT and OpAmps and how to design actual stuff using them?
@@優さん-n7m Yes, definitely on my list of videos to make!
World's most based guitarist.
Awesome playing and EE work with explanation (as always), keep it up!
Thank you, Jan!
Really good project! As a "guitarrist", I can say that this overdrive is a lot better then most of the digital overdrives that we can find in the market, and also sounds very similar to an original Tube Screamer, awesome work and playing, of course (nice Satriani's touch there :) )
Thank you very much for your kind comment!
You’ve just about convinced me to not miss out on DSP this semester at uni. Thank you again for all your videos. I’ve been getting more and more into PCB design because of them, and now I’m getting all sorts of projects from professors to work on. When any of my fellow students ask me how I learned KiCad and Altium, I point them your way!
That's awesome, glad to hear you'll be doing DSP at uni, Jack :) And thanks also for sharing the videos with your course mates.
Fantastic quality as ever Phil. You carry the torch for audio electronics and DSP videos, hope to see many more like this one! (Maybe an octaver pedal next?)
Thank you very much! Octaver pedal is a great idea!
@@PhilsLabCould be the SLASH OCTAVE FUZZ from MXR? It's an absolutely BEAST !!! KEEP ROCKING !!
From what I'm seeing here, this is what pedal folks call a hard clipping distortion algorithm. It is more like a Distortion+, and not the kind of soft clipping that the Tube Screamer you showed actually does. It's fairly easy to get TS style soft clipping, though. Just add in the original dry signal (maybe after some of the filters) and tweak the relative levels a little bit.
That's not to take anything away from the actual content. It's a great video and project, but...
I love the demo on the guitar, it takes the theoretical and scope-based understanding, and shows you why you should care.
Thank you - yeah, I agree it's good to have a real-world, practical example!
That opening was like the best YT EE video ever. YOU WIN.
Thank you very much!
Wow. The science of sound manipulation is amazing! So much variety driving decades of music styles from such a simple circuit. I can see why you love making pedal electronics.
Thanks for the peek into the world I didn’t know about!
Rock on.
Thanks a lot for watching! Definitely my favourite engineering area with fairly-quick, practical results :)
Must be real exciting getting it set up and hearing it for the first time. When it comes to personal projects, at that stage you forget about eating and sleeping entirely until you've spent some time playing with it. And your projects are seriously dope.
Thank you Jim! It is pretty cool seeing everything come together, both from the hardware and software perspective.
I'm very impressed with the quality of the distortion.
I just did my undergraduate dissertation project on a hard-clipping digital effect! Your vids really helped me back then to have any idea of what to look out for and where to start. Kinda wish I'd had this video then haha! Still learning lots now... so many questions!
That's an awesome undergrad project! If you have any content ideas in this area, let me know :)
I think I just found my new favorite channel!
Great to have you here! :)
I can't believe that I find always your videos about all of my hobies :) Some day, I design flight controller for long range UAV and quad, then I came across to your video about subject. Now I am planning a DIY effect processor project and bam your video is here. And you are always one step ahead of me :) My idea is copying the behaviour of a 12AX7 tube and amplifier into a microprocessor as a starting point. Basically my method will be based on cascade connected frequency filters and gain stages but definetely this will be not that easy.
Edit: Forgot to mention, amazing video.
Looks like our interests align quite well! :) Modelling valves is a great project. Have you checked out 'Digital Audio FX' by Udo Zoelzer? Awesome book on audio-based DSP, including some valve-modelling.
@@PhilsLab didn’t check, definetly will check
Wow man,this is also my interest ahaha!! I'm a guitarist myself and a electronics engineering student. Until now i implemented in MATLAB a 12Ax7 model and was trying to port the code in c++ ti realize a plugin. But It would be cool to realize this on a microprocessor! If you find some interesting resource please share beacause i love this stuff ahah
@@PhilsLab do you think you Will ever make a video about triodes modeling?and maybe some implementation like for this pedal for example
Thanks to Phil for the work done, and for bringing up the subject of antialiasing. I hope the topic will continue, you can improve the algorithm with the introduction of lookup-table vacuum tube.
Thanks, Andy!
Brilliant video as always! Great playing and super-cool pedal artwork 😍
Thank you very much!
You shred on that guitar!!! Nice!!!
Thank you, Jason!
This is such great quality content, we are all massively grateful for what you do Phil!
Thank you for your support!
Fully agreed!
Perfect video again. Every time I get a notification from you is blessing, Phill. Keep it going.
Thank you so much! More to come :)
Would love to see this implemented on FPGA board you've been showing off earlier. Especially since FPGAs prefer fixed point math.
Yeah, that's definitely up for a future video!
Awsome !!!!!
This is the best channel EVER !!!!!
Thank you, Michael!
I found for higher gain distortions, oversampling+antialiasing filters are definitely the way to go.
Like always, you're blowing my mind with each video!
Thank you very much!
awesome video! i'm watching it with great awe! Just one little thing..The tube screamer has the diodes in the feedback loop of a non inverting opamp, hence, when the diodes are conducting, the stage turns into a buffer. This causes the signal above the diode threshold to get a gain of 1, not zero. This gives the typical "clean blend" overdrive tone of the TS style OD (specially with high output pickups). On the function shown, above 2/3, the signal should have been X, not 1. In fact the function shown is actually hard clipping (diodes to ground, or on the FB of inverting opamp), where the gain above the threshold is zero.
Thank you! While I completely agree with you in the analogue world, given most books and papers in the DSP-world, the soft-clipper is modelled as shown in the video (for example, or similar but cubic) (e.g. Stanford ccrma.stanford.edu/~jos/pasp/Soft_Clipping.html, Udo Zoelzer DAFX book p. 118, DSPrelated www.dsprelated.com/freebooks/pasp/Soft_Clipping.html etc..). The 'softness' comes from the transition between linear and fully saturated (which thereby in that transition region also adds compression).
Hard-clipping in contrast, again in the DSP-world, means usually a single linear region - no transition - and hard saturation at the end of the linear region (e.g. Stanford ccrma.stanford.edu/~jos/pasp/Hard_Clipping.html or www.dsprelated.com/freebooks/pasp/Hard_Clipping.html).
Lastly, a main difference between soft- and hard-clipping, in terms of frequency domain, is the presence/prominence of higher harmonics. Hard-clipping has a significantly higher harmonic content, whereas soft-clipping - as shown in the video - has (slowly) decaying harmonics.
The image I showed with a resistor and back-to-back diodes, while this is indeed more for hard-clipping, I find it far easier as a simplification and illustration for the general clipping process. That's why I included it instead of the op-amp feedback path.
@PhilsLab you're right. The function describes the shunt diode correctly. There's confusion on what "soft clipping" means.. on the analog opamp clipping world, it means "non inverting, diodes on feedback".. but strictly speaking, it means a sigmoid function with soft knees, as you show.
What an intro! This dude is smart and a musician
Thank you!
Awesome video and project! Combines a ton of areas of my EE background. I know the video is already packed with content, and you can’t cover everything in detail. But I would love to see more detail about the signal chain on the circuit. I guess because that’s the area I have the least experience in. Specifically how the STM32H7B0 interfaces with the analog signal. A quick look at the data sheet shows it has two ADCs and one DAC. It looks like some slick programming software is provided for it. Maybe I should just build my own pedal and get the experience!
Very cool! I'm pretty new to dsp designs so I'll have to watch your other videos, thanks.
I dont know if i didn't pay enough atention to your explanation, but it would have beed very useful if you mentioned that you are using the absoliute value of the input on the slide with the formula for clipping. I like your DSP series. Do you upload your code and PPTs in some place?
DSP and STM32/ARM stuff still goes way over my head, but cool video! I shall refer back when I have learned a bit more. AVR is my forte and it's aging...
You can’t go wrong with a Charvel when you want a nice bright 80s metal sound, and awesome stable tuning.
Exactly - my favourite guitars!
Dude seriously, that was awesome. Are you sure you're an EE or a Guitarista masquerading as an EE. Loving your design videos but super cool to see you play. Thank you. 👍
Haha thank you very much - glad you liked it! Would like to make more guitar-related videos, but there audience for those is a lot smaller I believe.
Phil, unless I'm mistaken, I'm pretty sure that was Symphony of Destruction. 😊
That's right! Great song and solo :)
Hey Phil -
Loved this video, thanks for all the really well thought-out insight into such a fun project!
Regarding the choice to pre-low-pass-filter the signal (around 12:00 in the video), why not upsample -> nonlinear distort -> low-pass filter -> downsample instead? It's absolutely more computationally intensive, but I think the STM ought to be able to handle it.
bandlimited DSP is really tricky lol.
That's so wunderbar! Great work as always.
Very interesting indeed, I'm little way behind as I'm looking for the perfect Sine Wave but I'm at the bulbs and Wien Bridge Oscillators stage so a lot of work for me, no need to look over your shoulder just yet LOL !! great video....cheers!!
Very pleasing effects!
Since the overdrive is software you can stack 2 overdrive modules to get a better distortion sound, much like an overdrive pedal feeding a clipping amp. High gain amps like your 5150 has several overdrive stages stacked.
That's right. Although I'm pretty happy with the single gain-stage of the pedal for now :)
Even playing some Joe Satriani!! Dig in' it!
this is an incredible level of quality! it's so much easier to do in analog haha
Thanks! Once the hardware + base firmware is in place it's pretty cool and easy to make quick changes, switching algorithms, and so on!
That intro was dope!
Thank you very much!
I can't understand why the antialiasing filter goes before the nonlinearity. The fault is obviously on my side since I'm not a fluent DSP speaker! ... in your example with the 1 kHz tone, digital converted signal spectrum is just a dirac around 1 kHz (I'm sure there is a lowpass antialias in the analog conditioning stage, so ther isn't aliasing just due to the sampling). Now, if I apply a lowpass Chebyshev filter, I think I end up with the same signal, just time delayed a bit. What am I missing?
Phil, amazing content as always. As an guitarist and electronics engineer these series are my favourite! Maybe a similar DSP project idea is to model a real amp into the STM32. Also what is the delay between the input and output signal on this pedal? Do you notice any delay while playing?
Thank you very much! Amp modelling is a great idea. I've been long planning on making a miniature 20W digital guitar amp.
Delay is predominantly due to the audio buffer (variable), so currently 20ms. Can't feel it when playing!
I have the desire to create an IR loader using the stm32 platform. But I'm a beginner in this whole digital world, etc. Would any of your courses help me with this goal of building my own IR loader, and perhaps other projects and pedals?
Hi, I am following this with interest! I am looking to implement similar on a synth system, but using Q15 integer calculations. Do you have a link to your source code as I would like to implement the same design, but change to Q15 integer?
More guitar-playing sections pls!
Very good ideas and sound. Has the source code been released? I can't find it in the git link, can you tell me the specific path if there is a release?😆
Wot!!?? An engineer with decent guitarist mad skillz. 👀
Thank you :)
As the CS4270 is being discontinued, what will you replace it with?
Thanks for all videos Phil! Keep it up! If you know something about FPGA programming other than xilinx , please concider making tutorial (even paid one) or video about it. Another thing - link to terms and conditions are not avaliable via link on your Fedevel subsite in checkout information. Minor issue but still.
Hi Phil. Amazing work. Do you plan to make delya/reverb videos with stm32h7. Do you think it is capable? I know memory is a problem here, can it be made with external flash?
One thing I've been wondering about with these mixed signal circuits is the reasoning behind the usage of an audio codec for signal input/output: Given that most STM32 MCUs have 12-bit or higher DACs and ADCs, would it be possible to use these for audio I/O, and if so, would it be feasible? Thank you for all these videos, they're really informative and entertaining :)
Thank you! 12-bits isn't that great for this type of audio, in my opinion. However, the ADCs are usually pretty fast in STM32s, so you could potentially oversample and average to gain some resolution. I did try this a couple years ago but the results really weren't great.
It's easier and better-performing, although additional hardware is required, to go the external (dedicated audio) codec route.
Great stuff. What does boost do? I assume that's for a future video. Also there's some delay/reverb, is this part of the pedal?
Wondering how much headroom there is for other effects i.e. how much % of the processing power is being used right now. I assume it depends on the quality of the filters.
Hopefully cabinet modeling/simulation is also coming to this series.
Thanks! The boost adds a constant amplification before the clipping stage (preGain += 50). Can also adjust a DIP switch to add output volume increase for solos.
The reverb/delay was added in post on my PC.
I haven't actually formally checked load - I'll measure how many cycles are used currently and how many in sleep, and get back to you!
Cab sims is a great idea!
Man what a sound?!!! Awesome playing (I´m guitar and embedded enthusiastic too!!) and amazing project! This pedal is available to buy??
Thank you very much! I'm afraid the pedal isn't for sale at the moment.
Suddenly my Line6 HX Stomp doesn't quite seem like the little box o' miracles that it once did. Oh, and nice chops, too, mate.
Thanks! Yeah, maybe next step is to try and create a Kemper.. :D
How did you do the artwork on pedal? Tayda? or DIY?
Can you feel the phase shifting or delay during playing? Obviously off the shelf digital systems like Neural DSP or Axe Fx are really good at imitate the tone but they stiil lack of feeling and dynamic feeling. How about in your desing?
I went with pedalpartsplus for enclosure drilling and UV printing.
No noticeable delay while playing (can vary buffer sizes and thus latency) with good dynamics - pretty cool given such a simple algorithm, but it feels very amp-like.
nice work thats really cool.
Thanks a lot!
Why not shift the kneepoints on the clipping curve rather than changing the input gain so that zero gain / equal power is (approximately) obtained independent of the overdrive setting ? Or is it the intent that increasing overdrive also increases the output level ? Just asking, I'm no guitarist.
That could definitely be a way of doing it. However, to me it's more straightforward (and based pretty much directly on the analogue prototypes) to amplify the signal before going into the gain stage, as the output level of the gain stage only varies a bit above certain distortion levels. The final level then is tweaked via the volume control, if a big gain adjustment is done.
Wow... awesome Megadeth intro...
Thanks!
oh waw... not many can play that song.. next is tornado of souls!!
Where did you get the awesome case for the pedal?
I hired a very talented guy on Upwork to do the Tiki-themed artwork. Then had pedalpartsplus drill the enclosures and do the UV prints.
Cool project! Great tiki image on pedal case! Who is artist? :)
Overdriven guitars and programming in the same video? You'd be hard pressed to make me MORE interested. O.O
Awesome, thank you!
Please explain why the final "limit the signal to between -1 and 1" would not introduce more distortion and aliasing that would then need more filtering.
It does and I alluded to it around 12:50 (somewhat as 'food for thought'). I said fs/4 was a reasonable starting point. However, I played around experimentally with various cut-offs of the AA filter and running the clipping algorithm, far lower than fs/4. But to me fs/4 yielded completely fine results (as heard in the sound demos), so didn't have to push it any lower in this particular case.
@@PhilsLab thanks for clarifying!
Great playin’ btw !
Thank you Fred!
All is well done as usual but I don't see any filter for the neighbor banging on their ceiling from below. 10Hz high-order high-pass filter should do it :P
Haha luckily they haven't complained yet!
the Y in Nyquist is silent, the second a in Alias is soft, uh not a, Alius
Nice video can you make one on how sustaniac guitar pickup can be made and another on what the fishman fluence pickups are different than traditional guitar pickup.
Thanks - that's a great idea, I'll look into that!
Yo, great guitar playing man \m/ many thanks for the video, super interesting topic
Thanks a lot!
You're a wizard!!
Thanks!
cool song in the intro :)
Thanks!
this video was just an excuse to flex on the guitar......... AWESOME.
Possibly 😅 Thank you!
Bro is just flexing his hardware, firmware and now even his guitar skills.
Leave some women for us too. 😂
Great video Phil!
Haha thank you.. :D
If you ever start an audio processing company you can count on me for an EE!
Awesome, I'll keep that in mind :)
uhhh, what happened to the crosstalk video, was gonna watch it and when I checked on it later it was gone :(
Marty Friedman would be proud.
Thank you very much!
Phil, I have a mojo v3. Is there any chance for me to get started in the FPGA world?
In terms of DSP algorithms? If you're starting out, I'd recommend going the MCU-route (or even PC) first.
@@PhilsLab I'm a PC/MCU enthusiast. I'm not at all proficient as you with CUBEIde, but I know my way around things. ESP-IDF is one of them. Honestly, I'd just like to implement a huge hardware rotary encoder on a FPGA. Can you point me in the way of actual real world solutions?
Excellent breakdown and demo. I'm curious, with 60+ taps on LP at 48k, what's the small time delay through your system? Also, I'd like to see the IMD, intermodulation distortion, FFT with AD generating a 3 note chord. The frequency domain must be full fun peaks. Thanks for the video, well done.
Thank you, Craig! At the moment, delay is about 20ms (predominantly due to the buffer size, which can be varied). Good idea - I'll have a look at that :)
@@PhilsLab 20mS is mighty fine. Thanks for the fast reply.
fantastic!
Thank you!
what is your PC-SCOPE?
Digilent ADP3450
Ha! Skills Phil! :)
Did I hear a bit of Joe Satriani in there during the test? :D
You did :D
Very nice 👍
Thank you!
It's funny but making guitar effects was my entry ticket into the world of DSP.
Same with me!
me 3!! @@PhilsLab
More effect !!
Is there anything this man can't do??
🔥👍
6:20 bias shown is not correct. The clipping will be not symmetrical in this case. You need a positive and negative biases to work in both periods
What if your input signal has a DC offset? Then you need the bias (or in the analogue domain, AC couple) for symmetrical clipping. This illustration is for completeness sake and also to show there are adjustments to be made quite easily to achieve different clipping styles.
Annnnnnd he plays the guitar.
Where's the hook we hang our collective hats?
👍🙏❤
Leave some women for us!
Haha :D
Ha! Skills Phil! :)