IIR LPF: you delay the signal and add it to the original, take that, delay and add that to the previous version and so on. That way you "smear away" the high end. FIR LPF: Take the signal, delay it and siphen it off, delay further, siphen that off and so forth, then add all the siphened off ones together. "Infinite Impulse Response" because it's an infinitely recursive feedback loop, "Finite Impulse Response" because there is a finite number of times you're delaying the signal by a sample and storing the results in a buffer. Also good luck hearing linear phase pre-ringing at 24Khz ;)

^what clone me said. The average person is going to have a very difficult time distinguishing between either filtering technique, let alone the differences of over sampling (unless they just love hard clipping).

Very much like expensive hardware really eh?

@@theCloneman5 great explanation! But what does siphen off mean? Can't find it in a dictionary. thanks

@@CyrilleBoucanogh sorry I misspelled, I meant siphon off. You split the signal there and run it in parallel from that point on. After each subsequent delay you add that result to the parallel chain and add them all back together in the end

Yeah I just use my ears. I never worry I just try shit haha

A specific instance where I have always preferred no oversampling is brickwall limiters at the end of the master chain. I just feel it causes some sort of more audible distortion.

Try that being stuck in your head before you press upload 😅 haha

It's interesting cause with the melda stuff I can hear it. When I use their linear phase oversampling I hear a difference in the transients. Just a slight smear

Wait a couple of decades - you won't hear anything over about 16kHz anymore anyway!

@@PaulThirdI doubt it, if your still interested, do a blind ABX (at least 20 times level matched) I very very much doubt you’d hear it

I did do blind AB's at the time. Don't know what it was but I could hear a subtle difference on the snare I was working on. Kept picking the minimum phase oversampling algorithmm. Think it was a clipper or something. Was something like that

@@PaulThird I’d be really tempted to null the two files and see what the result would be because unless your doing pretty extreme filtering with a high Q at low frequencies it’s highly unlikely to hear anything. I’m going to run a test like this for myself and I’ll use a snare drum and do a few where it’s increasingly drastic moves until I can discern it and I’ll null it. If you have the will to do so I’d give it a try too

🤜🤛

🤓🤓

I can dig it

Yes!

Yes, I'm the turn-the-knobs-until-it-sounds-good type!

🤓

🤓

🤜🤛

🤓🤓

@@PaulThird I hope that was a compliment because I had to figure out via trial and error the plugin that can give me that pro analog sound ITB with 2 instances of it on the mastering chain so that's why I'm keeping it a secret. Hint: it's not Slate Digital Fresh Air.

🤓

We're on different pages though. If I'm the mixer then what have I got to do with the arrangement, recording & playing? I'm thinking from a purely mixer orientated perspective. I get the multitracks from the recording engineer / producer and its my job to engineer the record from there. In the box mixing isn't analog and I and other itb mixers don't worry about the limitations and compromises of it, as we don't need to. Because we have other challenges and compromises within the digital realm. Aliasing is a digital problem and unmusical and is something that the mixer adds in via the tools they use and how they use them. If you let it build up it has a sound. The reality is that it is unmusical so why would you build that up throughout an entire mix? That's my viewpoint. Surely to better serve the music you want the pristinely recorded performance to be processed as musically as possible? Its a very subjective thing but in a world of in the box mixers where technology is becoming more powerful and advanced every year, the little things as a mixer that make you better than the next guy or girl becomes narrower. Say a label sends a mix to 4 itb mixers and you are one of them. All using plugins and will imprint their own tastes and techniques but ultimately still mixing the same multitracks in the daw using plugins. In those scenarios you may have to think of ways of getting the most out of what you have. More plugins or a new plugin??.. Top engineers have any plugin they want. So what else can you do.. You can learn them better than those engineers and get a little bit more out of them. Get more out of digital. When your chops are good and you are all on a level playing field it's the micro things that build up that matter. Many people tend to think of what I do as a singular thing where I see it as an overall picture. All of those different decisions made on different sources all building up in the mix. Sometimes it could be just working at 96k which gives you a bit of a clearer top end compared to the other guy battering away with emulations at 48k. Theres many variables but when you are a mixer these days you have the same tools as everybody else but you may undetstand them a bit better which may just give you an edge on them. Small margins but it's a tough world out there these days. Every little thing helps. For many analog guys, that edge meant spending a mortgage on rare gear... For me, its learning more about the digital realm and plugins To each their own

music is art especially how they mix some of those pop music is just beautiful i agree that most of the listeners won't fully experience it but it definitely adds value to the sometimes boring or simplistic pop songs

The harsh reality... There isn't many famous analog upward compressors. Stupid... But true

@@PaulThird hahahaha, I didn't even think of that! I DID think "well yeah, they won't include it in hardware emulations" - but you're right.. the simple fact that most plugins are in fact emulations and most people still don't THINK digitally means they simply don't think of it when doing a purely digital plugin. Also many probably don't even have a code for it yet (because it wasn't necessary for all the analogue emulations), so they can't be bothered. Jesus.. thanks for clearing that one up for me. Sometimes I'm blind when it comes to such things! PS: Don't wanna be a dick.. but seriously: Try out some upwards compression with the mcompressor.. it really is not the same as parallel compression when you use it to an obvious (sound-designy) degree. I just transformed a snare for example - it's a different snare now.. it's like EQing, but not.. you get away with "more change".

I'll try it out but I think I do that with eventide omnipressor. I use an attack preset as a drum preset and it's awesome

@@PaulThird Oh yes, the omnipressor works great on clean scources. On vocals and such you gotta be careful with it's upwards compression though.. if you don't tame it via setting Gain limits you will not get much body, but all the noise. I'm still not "confident" with the Omnipressor.. really takes some time to master that bastard - for me at least. You know I'm slow!

The upsampling is just adding -infinity dB samples between the existing ones before filtering possibly.

The full picture will come to me one day... Haha 😅

Yeah if I do use it I just keep it default which is X8 OS

Hahaha yeah I start to get into it and then after 15 minutes I'm like nah.. F*ck this 🤣

That's a big problem I see. People don't invest enough in their coders. Plugins are only as good as the knowledge and skill of the people making them. Just look at melda, Tokyo Dawn etc.. Guys are dsp wizards. When you speak to some you are blown away by just how much knowledge they have regarding the topic. Unbelievablely clever chaps

Nah just span and the phase align thing

@@PaulThird they are often overlooked like the Melda stuff as the GUI isn't one of those trying to look like hardware type plugins. I'd suggest you demo the Marquis compressor as I think you'll like it

Don't get that one tbh

@@PaulThird Classic Worral s*** eh

@@PaulThird not to be rude, really, but if you didn't get the DW video, you shouldn't speak about this topic....

Tbh without your exact context I can't really say for sure how to interpret your statement. I'm reading this as your implying you must add in a steep lpf at 20k after oversampling, which I don't really see the benefit in. If you are using a lpf after oversampling to roll off needless high end information then that makes sense as you have less high end information to create high order harmonics but that's dependent on the source because the amount of roll off will depend on the source as we don't needlessly roll off high end for reduction in aliasing, we roll off because the source doesn't need it in the context of the mix. And also depends on the saturation as many saturators can cause harmonics made from frequencies as low as 5khz exceed niquist. Depends on the amount of saturation and the level of high order harmonics created Tbh I don't even know if you are implying to roll off high end, saturate it, and then add the high end back in afterwords. Or if you are talking about a parallel saturated track with pre rolled off high end blended in with the full band unsaturated. I undetstand why you would hipass infrasonic frequencies at 96khz & 192khz but at 48k after oversampling I don't get it as the frequency range is already limited to 24khz and most non linear plugins already have anti aliasing filters at 20k anyway which means that the oversampling filter is at 20k..and you want to add another filter after that? The reason why many advise a hpf at 96khz and 192 is because the infrasonic stuff is constantly in the audio and even oversampled still doesn't remove the infrasonic stuff after 20k as niquist in 96 is 48k, and 192 is 96k. This means that the infrasonic stuff can build on top of each other where with oversampling at 48k the audio is up sampled and then down sampled back to 48 which has niquist at 24k. So what's the benefit of adding a lpf at 20k after a lpf at 20k.. Because you've just done it with your oversampling filter.. Doesn't make any logical sense Its a pretty vague comment so without the correct context I don't really get the benefit. However send me the video in question where he discusses this at 48k with a time stamp and I'll re-watch that segment incase I've glanced over it the first time around.

Not in the tests I've done at 48khz. Both are quite audible and can be picked out in a blind test Filtering at 20k only filters out frequencies after 20k, not the harmonics made from the frequencies below 20k. They still surpass nuquist and cause aliasing and that's why we require oversampling

@@PaulThird - All tracks sampled at 48Khz have been filtered with a steep low-pass filter at ~20Khz before being digitized. This imparts a static high-end phase shift that is built into your original source tracks. Adding more static phase shift with further processing does nothing to make this more audible. Human ears are very sensitive to DYNAMIC phase shift, but are incapable of detecting STATIC phase shift once there is no unshifted reference for your ears to compare it to. If your tracks are contaminated with harmonics below 20Khz, you are either processing them with non-linear plug-ins that generate distortion or overdriving the signal somewhere in your gain structure. This is a hazardous practice with digital processing, because those non-linear processes may also produce harmonics above 20Khz that can fold back to produce inharmonic aliasing. If you want to use non-linear processing, it's best to quarantine it in an external effects loop to keep your digital mix clean.

🤓🤓

🤓🤓

All of your questions can be answered by simply applying them and listening for differences. Or render both out and listen to the delta

I see more benefit personally mixing at 96k. Mainly cause it means you have a "more analog starting point". If I had the processing power id mix at 96k without hesitation. Don't care about file size as id just invest in more external storage. No eq cramping at 96k and there is no need for OS to fix it .. However that falls flat on its arse when you use a plugin that already is internally oversampled to stop this but the phase response is more natural at 96k anyway so the phase roll off is shifted up, like what I showed in the kit video and the bettermaker/kirchoff example. On the most part you'll find that aliasing in most analog emulations (bar saturators) is under the audible threshold at 96k so it simply means that you don't need to worry as much as by default there's less aliasing and no eq cramping to worry about.. And you can essentially use any plugin you want within reason with less use of oversampling, or simply none at all. That's it really. Just covers a lot more bases and requires less plugin management. That's just me though.

@@PaulThird Yeah fair enough; that's what I thought ! I just wanted to make sure I hadn't missed any information. The past few months I've been mixing at 96k because my CPU more than manages.

Nah I'll stay in the box haha

Unfortunately we're in a moment in time where we DO need to think about all that shit when working with plugins. For the hobbyist like me - I think strategic hardware purchases make sense - with the rest of the stuff ITB. For example: A good hardware synth or two, a preamp & maybe a mastering compressor < that's what I'm working towards. People who have the luxury of working with gear - awesome for them :D

Tbh Ive got kinda free choice with access analog and mix analog gear and I still dont use them haha Its all about your plugin choice for me. If you pick wisely you won't really have any need in gear for mixing. Takes time though to sort out the wheat from the chaff.. That's what puts people off

@@PaulThird Interesting, I think inconvenience plays a big role there as well. I have a couple of HW synths that I'm not yet using because I want to finish all the tracks I started with VSTs - but in reality, I think I'm procrastinating xD

Try that one kzfaq.info/get/bejne/gLqGfLuA3anFlqc.html

@@PaulThird Interesting, thanks.

🤓

It's all the same tbh. Doesn't change how aliasing, phase, oversampling works. Vsti is still a plugin that saturates & impacts frequency resonsonse

Phase shifting occurs in analog as well though ...

Yup, no such thing as linear phase in the analog domain

I use tdr limiter 6 clipper a lot which I don't think has that problem.. I think cause of the routing as you can have a peak limiter and also an output meter as well at the very end to catch any stray peaks. Tokyo Dawn guys are geniuses so I think they covered all that.. I think!

@@PaulThird Nice I use TDR limiter 6 for all my mixes too but I use kclip before the limiter. Mostly because kclip has an "oversample during render" option so I can save CPU but still render with 32x oversampling. I did not know the oversampling was before the protection limiter in TDR.

I'm taking a wild guess. Only cause its TDR

V-Clip and Standard Clip also have a two stages

That's cool. I've found Aapo from Sixth Sample to be really responsive too. He's adding a feature I requested to the delay.

🤣🤣

Nah. Waste of money when I can just oversample and commit

Tbh SRC is something I generally stray away from. I've seen the many forum arguments over the years haha

@@PaulThird ha ha yes I have been very wary of it too, enjoy your vids by the way :)

Have you upgraded your version of R8 brain? The old version from the mid 2000s is severely outdated and the newer versions make use of multi-core processing among other modern processor calculation features.

🤓

Only looked at his fetish and lala. Tried buster but it bugged out on me. Not looked at anything else

@@PaulThird No worries. All of their plugins feature oversampling (as far as I'm aware), which you enable by clicking on their logo, so I was wondering if you were familiar with them. Thanks 👍

🤓🤓

You can oversample the tracks and still bounce the mix at 48khz

With metaplugin and others that use offline mode yes

Well... Sounds like somebodies got a remortgage coming up 😂

🤓🤓

🤓🤓

Remember to check out my autism channel if you want to learn more about my life 🤓🤓

I can't even get them to work in doctor haha

@@PaulThird use the vst2 in doctor

Doctor only does vst & vst3. Unless they've changed it in v2

@@PaulThird "vst" is vst2 in this case

@@PaulThird it's been vst2 since 1999. "vst1" only lasted for 3 years

I don't even know why I automatically revert to those accents 🤣

@@PaulThird racism? 😂

kzfaq.info/get/bejne/kLxjZpeZlZmrYKs.html

@@PaulThird it makes me sad to think the last 20 years of bbc comedy has been so poor.

Too PC now. They are all afraid to say anything now. As soon as little Britain was getting made an example of i knew comedy was going to become safer and safer. OK a lot of British comedy was just out and out racist back in the day which is not on but a lot of comedy was cleverly thought out. Could you imagine if brass eye came out today haha 😅😅😅

Took myself off Facebook entirely

Fair enough. Each to their own

Very true

Analog shouldn't be laughing as it lives in the modern world which means it has to pass through converters which use anti aliasing filters 😜 haha filtering is unfortunately an unavoidable part of digital audio.

@@PaulThird yeah, once through my mastering engineers stereo converter after mastering. I think i'll live...

🤓🤓

Very well put 🤜🤛

@@BrunodeSouzaLino how would you know they don’t know a lick of it? Are you talking about producers or are you talking about mixing engineers? There’s a distinction. Producers are like managers of the production team and even then, most producers know some engineering or are former audio or mixing engineers themselves.

Idk man, plenty of people have made timeless, quality music without knowing any of this shit, as well as plenty have done the same knowing a lot of it. The bigger danger imo is pretending there are hard and fast rules or 1 road to take. These things exists on a vast spectrum, and this isn’t even considering the whole notion that a mix is secondary to a good song. How many countless songs beloved by many out there sound like they were mixed on a potato? Quite a few lol. Your perception of creativity is also kinda fucked. There are no technical requirements to creativity and no bars that need to be hopped before one can be creative. Can technical skill help further creative expression? Sure can, but there is no threshold where you’ve now magically learned enough to be creative. A lot of folks will tell you in their experience the inverse relationship is true, that the more technical they get the more their creativity is dampened because surprise surprise people are different. What a wild thought. And no, before you start on about it I’m not saying anyone should avoid learning anything, I’m saying things aren’t so black and white. Also your definition of producer from the 1950s doesn’t quite hold up in 2022 when the mantle of “producers” means having 5 different jobs for a massive number of people. The “then and only then you can be creative” thing is the biggest crock of shit I’ve heard in a while. Buncha charlatan word vomit. The funny part is that all you’ve done in response to an extreme viewpoint is take the polar opposite extreme viewpoint (that creativity is somehow locked behind Intimate knowledge of processes) when neither are representative of reality or real people other than rando fringe cases.

@@peen2804 lol okay sure, do whatever you want. You can say that to a painter or an artist like Leonardo da Vinci then because he ain’t gonna be the legendary in creativity if he wasn’t an expert with colours and brushes. Name me legendary producers who don’t know sh*t? Those audio engineers from Abbey road who assisted studio and the Beatles are real engineers who can work for NASA dummy. They had sh*t equipment but were able to manipulate audio because they know how to that. Also, I didn’t say you have to be a master in engineering, I said “proficient enough”. Don’t put words into my mouth. All that garbage of insult just because you’re butthurt about what I said lol. Go on then, what else is f**ked? You wrote all that in YT comment? Go learn music production instead of splashing that long-ass garbage here

@@peen2804 Did I say creativity is locked behind intimate knowledge of processes? Again stop putting words into my mouth. I never said that. Intimate knowledge of science and engineering (the application of science) is not the same as “intimate knowledge of processes”. That statement of yours alone is full of BS and proves you don’t know jack about how engineering and creativity work together. Also, the post was about creativity in the context of mixing and audio manipulations, not creativity in general that toddlers and kids grew up knowing! You are off topic. This channel is about mixing, remember? the problem with you is you think audio engineering is about creativity. If you want to be creative, be a musician then or a producer. The basic job of an engineer is to apply science, and those are facts that I did not invent. It’s in the actual definition of engineering in any field or industry. And in audio and mixing context, it’s applying the science of audio to manipulate it in order to provide a solution which is a clear recording, clear mix, or a master. If you can’t do that basic sh*t properly, then good luck wasting a lot of money on equipment you may not even need and losing clients because you lost control of your mix. If the client wants some creativity then that’s between you and them because at the end of the day the clients are king. If you don’t know jack about engineering, you’ll waste more time trying to get to what the client really wants as a creative touch. Nowadays, you have more competition because of access to gear, good luck trying to standout from the rest. Those ones you said “who don’t know jack” have been interns or apprentices of really good producers, engineers, and mixers. They didn’t just come in the industry suddenly being very good without the science and effort needed to get the application of science right. Any real engineer will laugh at your assumptions. Look at Paul Third himself, he’ll probably laugh at you too. And the guy has certification for audio engineering. Also, you can say your opinions, but you don’t shove them with profanities, buddy. Your credibility went down the moment you said profanities without even explaining anything noteworthy.