 Hey, I'm Anfa. I'm an electronic music producer and sound designer, so I exclusively work with open-source software and Linux. In this video I want to show you my favorite DS-ing plugin, Airwindows Deepass. Let's go! Now, for the purpose of demonstrating this plugin, I have prepared a little bit of testing material. Here it is. Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. That's a silly pseudo-sentence I've used to stress this to the answer. Now that I think of it, this sentence would probably do the job too. Anyway, let me play you. How does this sound through the the answer now? Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. And now again without DS-ing. Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. Okay, let me tell you what the heck is going on on the screen. I have prepared a setup where the oscilloscope shows us the difference between dry signal and processed signal. So in red, if I make, for example, the bass only let sibilance through, we have sibilance in yellow and everything else in red, everything else which is missing. And here spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. And here we have intact audio in yellow and the difference in red. So the red peaks are what the bass has removed. If you wonder how does this work, I have basically prepared two audio channels left and right. Right is the dry channel, left is the channel processed by the bass, and they are overlaid on top of each other. Where they overlap, we have yellow. Pretty simple. And I've done this with a bunch of arduous buses and sends. Let's talk about the bass. I'm going to now reset it to its default settings and this is what it is like. By default, the bass does nothing. Spotting subjectively tempting suspicion of turkish sugar trimming. As you can see, there is no DSA going on because there is no red. Now to turn on the bass, all you need to do is move intents to the right. Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. As you can see, it's very subtle. It gently reduces the peaks of asses that it detects or sibilance. There are multiple parameters. Intents is just one of them. Intents is like a threshold, only like a reverse threshold. When intents is at 100%, it's going to react to all the sibilance that it detects, no matter if they are subtle or very strong. If I set intents to 0.5, spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. If I can now overlay the previous version of this oscilloscope display with the current one, I think we'll have a difference that some of the smaller the asses were not touched. Let me move this even smaller, so 0.2. Intents subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. Yeah, so this is like a threshold. We're moving the threshold so only the loudest parts of the asses are affected. If intents is all the way, we affect everything. Now, what is sharp? Sharp determines how the distinguishing algorithm works. As Debass analyzes the sound, at least that's what I understand from Chris's original explanation. I will also link to his original video where he explains how this plugin works and demonstrates it himself. From what I understand, the plugin analyzes the waveforms and looks for chaotic squiggly bits. I think squiggly bits is the technical term that Chris used himself. And sort of analyzes the sound to look for like up-down squiggles that are sufficiently sharp. The sharp parameter determines how squiggly the bits need to be to be considered a sibilant that needs to be acted upon. So if we move sharp all the way to the right, sputting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. It's going to be more picky about the frequency content. So Debass is not looking at frequency content. It's not using like a classic filter to determine what needs to be ducked. It's looking at the shape and the character of the waveform. It's a bit different. So the sharper it is, the less it's going to react to lower frequencies. If I'm going to lower sharp all the way to zero, sputting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. You can see that now it reacts to much more. It kind of like, shaves off a little bit of everything. So if sharp is low, Debass is going to react to less chaotic parts of the waveform. So it's going to be like, we can think of this like a high-pass filter on the side chain signal. So right now, the high-pass filter is all the way down. So the side chain is going to react to everything, shift click. Now it's default. So it's like balanced. And now the high-pass filter is all the way up. So it's only going to react to the highest SS, the most sounds, not to anymore. This is not a filter. This is just an analogy I'm trying to present to explain how this works. The actual algorithms are not using typical filters here. OK, so now we have depth. Depth is like ratio in a compressor. Depth at 0.5 means it, I believe, I didn't test it. But I think it's like it leaves 50% of the energy of the SS intact and removes just 50% of it. If we go with depth 0, it's going to remove all of the SS energy that is present. Let me play this to you. Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. You can see that the red parts are now much deeper. However, there is still yellow bits left over. And that is because we have a filter. I've got to go to filter right away, but depth. If we go with depth all the way up, that means ratio one to one. Again, this is not a compressor. This effect is not a sidechain compressor. It is working differently and it is working excellently. But if depth is at maximum, the best is not going to do anything. Regardless of what other spotting subjective parameters we'll do. Suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. Right, so depth at 0 means maximum reaction. Okay, now filter. If we turn down the filter all the way... Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. Now we are not saving any low-frequency content from the SS. What filter does is allows us to recover lower-frequency content from the removed SS. So if we move the filter all the way up... Spotting subjectively tempting suspicion of turkish... It's gonna do nothing again, but if we move it a little bit down... Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. This makes our de-essing act more like a dynamic shelving EQ that is only dipping the top of our frequency spectrum rather than an attenuator that just dips everything, okay? So filter at 0 means we are attenuating all frequencies equally when SS are detected. And the higher you set the filter the more you preserve the low-frequency content. So if we want maximum removal... Spotting subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. If you want to remove everything, SS 100%, which is a valid use case, because you could copy and paste this same thing and just switch the best to the Sense Vox mode, which is going to give you the opposite of that. So it's going to give you only the SS that it removes in the S-Bode. And then you can, for example, apply reverb to your vocal track only on the de-ess signal and then mix it back with the SS, which are individual. And you can also process the SS separately, like a user compressor, any cue, whatever you want and mix them together. You can do very sophisticated crazy things thanks to this function. But also the Sense on Vox is pretty much like sidechain monitoring. So you can hear what the bass is removing rather than what the result of its processing is. So now you can see we have only the SS. If I move the filter to its default position, which is 0.5 and play again, Well, that clearly is high-passed, right? Let me high-pass this even more. So this is what we are removing. If filter is higher, we are not removing low-frequency content from the SS, which is a good idea, because otherwise it's going to sound pretty weird. That's why by default it's on 0.5. But if you want to completely separate your SS, your Sibilance, and maybe process them separately on a different bus, like I said with a custom compressor or EQ or whatever, you can disable the filter by turning it all the way down and then you're just removing everything. So it's pretty funny. Okay, now the Vox monitoring or the monitoring mode is very useful, because we can hear that the bass is still removing a little bit of the vowels here. So I think what we can do is up the sharp. Sustain, substand. Okay, let's try more. Sustain, substand. Okay, I'm not sure that's what I wanted to do. Let's go with sharp zero. Spotting subjectively tempting suspicion of Turkish sugar truth. Sustain, substantial turbulence of flat length. Yeah, okay, so now if sharp is at zero, it's so reactive to low frequencies that it really just messes up everything. So if we move it up. Interesting. It now does better than when I moved it all the way up maybe. Sustain, substand. It's awesome, like this visualization really helps me figure out that, hey, it really doesn't do what I think it does in the highs. All right, so this seems like the mode that would remove the maximum thing. So let's say sharp at 25, 0.25, depth and filter at zero, intense at one and sense is like what you want. Let's listen to the process signal, not the residue. Spotting subjectively tempting suspicion of Turkish sugar trimming sustained substantial turbulence of flat-lined filth. Interesting, it didn't touch the filth. I wonder if I can make it do it out with... Spotting subjectively tempting suspicion of Turkish sugar trimming sustained substantial turbulence of flat-lined filth. No, I think filth is never gonna happen. Spotting subjectively tempting suspicion of Turkish sugar trimming sustained substantial turbulence of flat-lined filth. It doesn't want to touch the filth even if I'm chewing up everything else. Okay, but that's an interesting thing. I wanted to have the Fs in this sentence as well to see how it will react to those. But they are admittedly much, much quieter than the other stuff. Okay, so if you go with sharp too low, it's going to apply very weird kind of distortion to your spuckle. Subjectively tempting suspicion of Turkish sugar trimming. And I guess if we enable like filter... Spotting subjectively tempting suspicion of Turkish sugar trimming sustained substantial turbulence of flat-lined filth. Oh, with the filter on, the zero sharpness is actually sounding very smooth. Spotting subjectively tempting suspicion of Turkish sugar trimming sustained substantial turbulence of flat-lined filth. It's like it's removing... Oh my goodness, that's so cool. It's like it's removing the sharpness from the vowels as well. Like it's like rounding off the waveforms. I want to actually zoom in. Spotting subjectively tempting system. Spotting... Okay, yeah, so we can see we have assets that are attenuated. Yeah, and then we have a vowel. Yeah, this is really interesting. You see, we are monitoring what we're removing. So we're removing these tiny pulses from every single vocal cycle. Oh, let's try this in reverse and see what the pros... Okay, so we are removing... We are pretty much removing the edge. And this is amazing. It really sounds... Spotting subjectively tempting suspicion of Turkish sugar trimming sustained substantial turbulence of flat-lined filth. It sounds almost like a Lopez filter, but not entirely. Wow, that is fascinating. I... I never tried that. And I've been using this plugin for quite a while. Okay, D-Bass is a huge amount of fun and it's an excellent tool. And now I'm not just using it for vocals. I'm also using it for drums. You know, hi-hats and snares, they also have very prominent highs. Or a tambourine... A tambourine can really make you suffer. So I think D-Bass is a great tool for taming these kinds of sounds. Now, we've been processing a raw recording that was not processed in any way. But S's are becoming a much bigger problem after you apply EQ and compression. So let's do that. I'm going to bypass D-Bass. I'm going to go to the source. And I'm going to apply a parametric equalizer. So our sweet vocal. We want to make it brighter. Sputting subjectively. Sputting subjectively. And yeah, I'm over exaggerating it a bit too. Sputting subject... To make it harder for ourselves. Sputting subjectively. Sputting subjectively. Tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined fill. Sputting subjectively. Tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined fill. Sputting subjectively. Tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined fill. I cut here. Sputting subjectively. Tempting suspicion of turkish... Okay, now we have this kind of hyped radio voice, but need a compressor still. I'm going to use Ace Compressor, which is an Ardor stock plugin, X42EQ is not, but it is a open source plugin that I think I've covered. Oh, maybe I haven't. I need to do that then. Okay, let's compress this, sputting subjectively tempting suspicion of Turkish sugar trimming sustain sip. Now I am overdoing this, sputting subjectively tempting suspicion of Turkish sugar trimming. Okay, so before, sputting subjectively tempting suspicion, and after, sputting subjectively tempting suspicion. Now let's debass this, sputting subjectively tempting suspicion of Turkish sugar trimming sustain substantial turbulence of flat-lined filth. Okay, this is a bit ridiculous because we have sharp so low. I'm going to go with sharp, which is like here, and let's sputting subjectively tempting suspicion of Turkish sugar trimming sustain substantial turbulence of flat-lined filth. Okay, and this versus sputting subjectively tempting suspicion of Turkish sugar trimming sustain substantial turbulence of flat-lined filth. Yeah, so without this de-assing, this is un-listenable. Sputting subjectively tempting suspicion of Turkish sugar trimming sustain substantial turbulence of flat-lined filth. At least un-listenable on my headphones. There is a little bit of noise after the recording ends, or maybe it was my beard like hitting it. I don't know. We can trim that and do it like this. Turbulence of flat-lined filth. That's better. The rest of the video is an explanation on how I made the visualization and some extra stuff. So if you've been here only to learn about debass, that's pretty much it. You can go now. Go. OK, one extra thing that I'm doing that I haven't told you about yet. Is I'm applying DC offset removal. So how does this whole testing thing works? I have the source track, which plays the signal. Now we are applying EQ and compression to it. Then there are two sends created with right click, new oxsend and bus. OK, so this is the bus that the debass is on. And after debassing, I apply DC offset remover. DC offset remover is a special high-pass filter aimed to remove DC offset, which means direct current present as a component in the alternating current or variable current. Who knows? Basically, the waveform is asymmetric. If I disable DC offset remover on the S and on dry signals, I'm going to show you how the analysis works or looks. Sputting subjectively tempting suspicion of turquoise sugar trimming sustained substantial turbulence of flat-lined filth. And it gets a little bit asymmetric at times and also debass, especially if I disable the filter. Sputting subjectively tempting suspicion of turquoise sugar trimming sustained substantial turbulence of flat-lined filth. Look, like here it just like there is we can see a little bit of green here. So it like and he also here. So there is some DC offset as being added as a result of debass. So I think it's safest to add a DC offset remover afterwards. And if I just apply it to debass. Sputting subjectively tempting suspicion of turquoise sugar trimming sustained substantial turbulence of flat-lined filth. You can see that the green waveform is much more centered. But the red one isn't. That's because this recording was not processed. Sputting subjectively tempting suspicion of turquoise sugar trimming sustained substantial turbulence of flat-lined filth. And here both are DC offset removed and they align pretty well. So I did that to make it mostly easier to compare visually what's going on. But I think it's also a good idea to have a DC offset remover after you apply debass. If you're doing something extreme like like this. So I don't know how it would sum up if you like had one track with the S vocal and other track or bus with everything but the S's and then sum them back together. Maybe the DC offset removal would cause some trouble there. I don't know. Let's put it all together. All right. So to recap, debass is an excellent DSR plugin. It's free and open source software created by Air Windows, which is a project by Chris available for Linux, Mac and Windows. The plugin controls are as follows. Intense is like a threshold. Sharp is like a hypers filter on the side chain of a compressor. Depth is like a ratio filter is like a high pass on the actual compressor. So like you're high passing what you're removing and sense is either you listen to what to the result of the processing or you listen to the residual, which is the S's themselves. The plugin comes by default working like that. So it does nothing because the intense is at zero. If you turn intense at one, it's going to work and it's going to do its job. Sputting, subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. But for example, in this piece of vocal, I don't think that is sufficient. So I would first I will lower depth. Sputting, subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. That's better. And then I would lower sharp as well and see what happens. Sputting, subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. Now you may say this sounds weird and unnatural and that there is way too much de-essing going on. Sure, you can back it up with depth. Sputting, subjectively tempting suspicion of turkish sugar. But to me, this is painful already. So I would have more depth. Sputting, subjectively tempting suspicion of turkish sugar trimming sustained substantial turbulence of flat-lined filth. All right. That is all I wanted to show you. Thank you for watching. I hope you've learned something. I hope you will add the best to your plugin arsenal. I really believe every vocal track needs one instance of this plugin. After you compress ADQ, because, yeah, I think even dynamic microphones like the Shure SM57 or SM58, they still can sound pretty, pretty hissy. People who have headphones with more high frequencies will also suffer more. So it's good to be able to, like, you know, monitor on different systems. Yeah. That's all. Thanks for watching. I also want to thank everyone who is supporting my work financially. Thanks to these people, I can justify taking the time to make these videos. Thank you. And if you, dear viewer, would like to join these awesome people, please go to patreon.com.com.anfa or liberapay.com.anfa where you can give me a dollar a month. Every little bit helps. Oh, and if you'd like to meet other people who are using free and open-source software for music production and audio stuff, all of them are using Linux as well, please go to my community chat at chat.anfa.xyz. And you can meet me there and see some of my videos early and hear some of my music early and give feedback and help me fix errors before I publish videos, et cetera. Get inspired, take part in Libre Music Challenge. Learn about new software, learn about new music, tip some tricks, get in touch with the community. All that good stuff. That was a nice pitch, wasn't it? Okay. Oh, I still have my coffee. Oh, and huge thanks to Chris from Airwindows for making this plugin and continuing to make awesome plugins. We love you, Chris. Now go and install the bass.