 Hey, it's Anfa and you're watching Anfa Vlog! Today I'm gonna master an EDM single, not my own track, but something pretty about someone else and sent to me for mastering. I'm no mastering engineer, however I've done some mastering in my life mostly for my own albums, which there was a bunch of them, so I guess I know something about mastering, but mostly this video is there to show you how to master a single using Ardor under Linux using only open source plugins. So, let's get started. Here is my Ardor session. I'm gonna import the source material and this is it. It's in a 44.1 kHz sampling rate. I'm going to upsample it to 48 kHz because this is what my jack is running and this is what I'm using for everything. The sample rate conversion doesn't matter as long as you're using high quality one because otherwise we could have some audible artifacts or even inaudible artifacts, but if I was, for example, to boost the high frequencies a lot in mastering, there could be problems bringing these artifacts up. So the thing is we imported this as two separate mono tracks. I want one track per file as new tracks. Yeah, let's do it one more time and then I'm gonna remove the unnecessary tracks. Alrighty, let's close this dialogue then I can remove these unnecessary tracks. Yup, and what we have is this necks. I haven't heard this track so I'm gonna be listening to it now for the first time along with you. Let's do it. Three, two, one, go. When I first got this track it was a bit too loud and I had to send it back and I asked for a quieter version because it was peaking. To be able to master this track I need to have two things for sure, like first it cannot be clipping and I see that this might be a problem already and second it needs to have dynamic range to work with. I need to return this track and ask for a quieter version and this is what I'm going to do. See you in a while. And it doesn't look like it's peaking right now however it also, I'm not sure, there could be a limiter still that might have been put on the mix bus before exporting this mix. I can't tell for sure the peaks are a little bit too consistent to my eye but well this is what I have and let's try doing something useful with it. So what I'm gonna do is first drop the EQ and play it and take a look. We have some very low bass probably around 25 Hertz. Yep very deep bass. Maybe instead of using a high-pass filter I would normally use a high-pass filter to make sure that no frequencies, no low frequencies like damage or headroom. I might try it, I'll try this use a steep of a high-pass filter as I can get and as flat in the passband so the frequency spectrum where I don't want any filtering as possible. All right that looks good. What I'm gonna do also is analyze this the whole track so I'm selecting a range with the range tool, range bolt and then right clicking spectral analysis and that should give me a nice plot of the frequency of the whole track. So we can see there are some harmonic peaks of chords often used probably but I don't see anything really standing out. It's more or less an even function. This is more or less similar to what a pink noise signal would probably give us. Yeah if I turn on a proportional spectrum which is just using the pink noise slope as a reference you can see it's pretty much flat. There's nothing very deeply happening here. We have kind of a peak around 10-12 kHz and there's a dip around 16-20 but this is pretty natural because we don't really need much information and as you remember we upsampled this from 44.1 kHz so there isn't going to be any high frequency content above 20 kHz probably. I don't see any problems. I could sweep the Q and see if I can hear any and then we can use the measurement to see if it confirms that there might be a problem. I don't see anything wrong with the EQ balance for this mix right now. Let's listen and try. I'm using the headphones because the monitors will only go so low with the bass and also they are more affected with room acoustics when it comes to bass frequencies so listening on headphones especially to the sub bass like below 80 Hz it's just easier. The bass is usually more balanced however you need to know your hardware you need to know equipment. I know how music sounds on these headphones and how much bass I should hear for a balanced mix and actually you can master and mix well them tracks with anything decent when you learn it so you need to listen to a lot of music and consciously analyze how much bass is there, how much highs is there, how does the mid sound and then you can learn how a good mix or a good master will sound on that hardware and then you can reproduce it and when you're mastering or mixing your own stuff. I'm going to first reduce the gain because the loudness now it's going to clip. I'm going to make my mantraing louder. I can do the same for your side. I'm trying to get out a little bit more color out of this so I just added a tiny amount of bass maybe just to mitigate what my hypers could do to just cut off the bass. I don't want to lose too much bass so I'm just adding it back a little bit of that in the higher frequencies range here and also trying to see if there isn't too much too little of the 1k region here the like the mids and I just think that adding just a tiny bit of that might be a good thing and also I added just a little bit of the highs where we have this dip because I think it's a little bit like makes the thing a bit tiny bit brighter but doesn't doesn't hurt. It's not as bright as it would hurt. Okay I'm gonna bring up the level but maybe not as much. Okay so now mastering EPs or albums is much more interesting because you balance tracks between each other. With mastering singles I should probably do some reference tracks. I've added a master limiter. I'm going to make it oversample four times. This is to catch the inter-sample peaks so the limiter is up sampling the signal to a four times higher sampling rate internally analyzing this and then it's down sampling it again after the processing to just make sure that the peaks between samples are not jumping off. They still could so yeah I don't really see anything I don't hear any problems that need to be addressed in this track. Cut the start just to give it a bit less time to roll in. Yeah it's been it's been cut short could be longer. Due to a problem with noise from my microphone and the way I record audio which isn't perfect you can't really hear that tail however I'm gonna paste a clip from the raw audio file so we can hear it. I could try to copy this and add some reverb to make the tail longer however this is already this part is already super quiet if we listen to this okay yeah we can hear it could be better okay I'm gonna duplicate this control little click and I'm gonna try let's see if I can just control right click and now we have region gain I'm gonna make this 12 decibels lower and now okay this is a little bit better and the fade out is a little bit less abrupt let's listen again to make sure it's not broken okay yeah I like it better we managed to mask out the hard cut tail there were there was a a scratch fading out and there was some delay delays bouncing around and they were cut I'm gonna save this I could try to experiment with some master bus compression or maybe saturation I don't want to take the place of a mixing engineer that that works on this you could try you could try doing something something with a compressor and see if we can get anything more exciting out of that all right so I've added just a little bit of compression to maybe a little bit more glue together the main melodic instruments and the drums I'm gonna see what can I do with a little bit of saturation however the cuffs actually I'm going to use a compressor for that because the cuff situator has a problem when you use when you don't use 100 wet processing it introduces some phasing issues and scoops your mids it's not cool so to use a compressor as a saturator um well this compressor has mixed control so we can use it as a parallel compressor but also first thing we can compress the signal so hard and make the attack and release so quick it's gonna basically distort the signal of course in such an amount it sounds just horrible we could actually use our saturating compressor I want to rename this oh I can't why I want to put it before our um more glue compressor I'm actually putting this before the limiter well this is bad however I'm gonna cut it out and paste it on my mix track I'm also gonna change to non-strict IO so we don't have these okay so we are first working on IEQ then we are this saturating the track a bit and then we are compressing all of it I wouldn't do any much more things with this track I don't feel any even what I did is super minimal and actually like I don't hear a huge difference I probably would need to A, B compare before and after a lot to to make sure and actually what if it actually makes a difference I don't want to go overboard with this because well it has to be subtle like if you if you do something so ridiculously obvious in mastering that everybody knows what you did well you're doing it wrong probably unless this is exactly what the music needs and you just had to do it well okay what I'm going to do now is drop in a EBU R128 meter which is a plugin that analyzes the level in loudness units or LU or LUFS and I know that YouTube is currently normalizing to around negative 14 LUFS and well basically a lot of people expect mastering engineers to make things louder and well there usually is a need to make things a bit louder just to manage the peaks to get the maximum of the usable available bandwidth of the distribution medium and this is as true for vinyl as it is for tape and digital files or streaming however sometimes people for mastering deliver mixes they're already too loud and I master things conservatively when it comes to levels I don't like super saturated sound of the modern releases and I if you've listened to my album suppressed or any other albums you probably heard that it has pretty dynamic for EDM and this is what I like and this is what I do with mastering for other people so I'm using this meter to make sure that my levels I can also run the whole thing and let's do host transport integration to host transport can do uh gonna reset and just play the whole thing and let's see uh here the audio didn't record well because I was talking while the music was playing and it's inaudible I was thinking about making subtitles but it sucks so I'm gonna record explanation so I was telling that I'm gonna play whole the whole track through from start to finish with the integrated loudness measurement on on the LUFS meter to tell what is the integrated loudness value and that will tell me how YouTube or other stream platforms will measure this track because they are using the same or analogous methods to determine the overall loudness and for example if YouTube loudness is negative 14 LUFS and it's gonna measure that the track is at negative negative 12 LUFS so it's 2 LU which is roughly two decibels louder it is going to turn it down by two decibels every time it's played back so it will match the rest of the material on YouTube if it is quieter it's not going to be turned up so it is important to make things not too quiet to unless it really makes sense for the material if it's a ballad it can be quiet it's not a problem on the other hand making it too loud will make it just squashed and it's gonna be turned down anyway so there's no point in making it super loud this is why I prefer to make my music dynamic dynamic is the new loud and I stand for that I think it is okay so it looks like this track is actually only one or two decibels lower than I would master it um and this is like this is what I hear and this is also what I see like there's not a lot of transients that actually need limiting or something so I'm going to limit this by three decibels and this is actually doing input gain to a static zero decibel limit so however ebu r128 also recommends like the recommendation is about is for broadcasting and recommends negative 23 lufs integrated level so for a whole program but that's for television and radio uh for music usually we go with like negative 13 14 15 it also depends on the music but you know EDM around negative 14 13 to my ears is good it's not squashed but it's properly loud that you can listen to it on a mobile device and it's not going to be too quiet because that's also a problem like we have these mobile devices have limited power and they are also limited by law how loud can these devices play so we need to consider this when mastering music because are people going to be able to make it as loud as they want on the go in a crowded city in a metro station well it is a consideration and I've been taking many considerations when mastering suppressed like that it's like I think I did the best mastering job for that album so far and I did really lots of research for that one so I'm going to attenuate it by one and a half dp afterwards actually maybe not one a half and now yeah I can get more precise with shift and this probably should give us around negative one decibel true peak and we can do it by rendering the track I'm also going to make the release a little bit shorter let's listen now you see we are not actually touching the limiter yet nothing is happening I wonder what will change when I disable my compression and saturation because that probably that could actually affect it let's see okay we get a lot some peaks okay so I could maybe maybe just back off my saturation to maybe 75 percent so we are letting more through and still the level is nothing really is changing let's see all right so this is getting louder than I would like to go however this is what the numbers say and what your ears tell you is more important usually I don't hear any pumping however I'm going to back off this by half a decibel and that will probably be it I have already controlled the start and ending I could also make sure that there is a clean fade in and the clean fade out yep that's it well for a single track you can export your whole session let's save it so let's just do export to audio files and now the go to format is probably a 16 bit 44.1 kHz wave which is near the CT but you can also do 48 kHz it doesn't hurt however I got the original file in 44.1 so I think I'm gonna do 44.1 because there isn't actually any information that we need to retain and the processing that that is going to happen back and forth it's not really any problem when you have a sufficient noise floor and with 24 bits on the inputs it's basically non-existent so I'm going to use 16 bit I'm going to use shaped noise dither shaped noise dither will make the dithering noise have more high frequency content and low frequency content and the high frequency content it has the same energy but it is less audible however this is really just a non audible difference it's just what I like to do to make sure this is the best thing possible um so I'm gonna yeah I'm gonna use wave 16 bit shape noise for my own purposes I would export flags probably but some people have trouble reading flags so I'm gonna send a wave not now I'm going to normalize the file and this is actually and this is basically a CD audio master I'm gonna call this revision one okay and that's it we're done let's bring more light so here is the export analysis and as you can see we are f***ed because our true peak is about zero decibels we didn't hit the spec we wanted so what we can do now is attenuate our whole track by 1.1 decibel to basically lower these numbers and I'm going to do this I'm going to go to the master limiter and I'm going to first add one more decibel of limiting and then even more than one decibel of output gain reduction or not not a hole and I'm going to export it again as revision two now the good thing about ardor and exporting is that the analysis is also saved as a png file in the folder of your export so you have this and you can send it to someone to prove what are the parameters of your export etc and you can also like have it there for comparison and it's done we have a new export and the true peak is below negative one decibels which is what we want and now we are at negative 12.4 LUFS which is great negative 14 would be better but we don't have we don't have actually transients like in my music there are much more transients I really like to make the drums have a huge spike at the start and usually I don't hear that stuff and this is also why it's easier to limit or crush these things such tracks because the drums don't really jump up and this is also an example of such a track we don't have too much transients the drums we could have more and then we would hear the effects of limiting more here these are pretty hard to hear I don't hear any effects of the limiting which is good I don't like to hear the effects of limiting the the limiter is not there to be a flavor of the sound it's there to make sure that you don't clip so yeah all right this is all for today this video is a bit longer than I thought surprised because I'm not I hope you've learned something master your tracks and tell me what you think if you have any questions or suggestions please leave them in the comments section below and I'll see you in the next video bye wait a minute I forgot about something oh big thanks to all my patreon supporters who are helping me dedicate more time to spend on the videos the process of making these are very time consuming and also as I go I often report bugs and request new features for software that I use because my pipeline is 100 open source and based on Linux and partly why I do that is to help that toolchain develop and actually someone asked me a question why use open source software and there is commercial software that could get your job done faster and I'm going to make a separate on fire rambling indefinitely video about that but for now let's just say that I care about the community and about the software a bit more than I care about getting things done in the fastest possible way which kind of hurts me sometimes so anyway big thanks to all the patreon supporters and everyone who bought my music on bandcamp if you want to find out what do I do head over to bandcamp.com and see for yourself listen if you like that consider subscribing because that lets me know that my content is what you need and yeah it's going to help me out bye by the way it's already May and this thing is rolling along since December