FYI it was brought to my attention that when it comes to inharmonic instruments that have no musical context or harmonic structure ie cymbals and hi hats. There is no noise slope for this. These should be nulled individually on the source itself. This pink noise test will only work with musical sounds that follow a harmonic structure. Test 1 soundcloud.com/paul-third-944739244/sets/6dbfs-kick-test-can-you-hear Test 2 soundcloud.com/paul-third-944739244/sets/test-2 Vote on the community post on my channel which will be up for 24 hours 🤓
one of the neatest tricks I've learned over the years is if you're having trouble with a mix, to throw a pink noise source on the monitor and then level each track individually so that its barely audible over the pink noise, it works surprisingly well and after a while I began to understand how to move towards that result more organically. Doesnt change that my music is shit, but its balanced well now at least !
Gonna try that. Usually if I'm adjusting levels I'll throw a mono plugin on the master bus..set my levels then just turn it off, and it made a huge difference. If not I'm fighting it getting aggravated
We have all been there. I have on more than one occasion reached out for a gain knob on a channel eq and tweaked the gain while clearly hearing the difference until I realized that I was altering the eq on a different channel that was muted.... My brain is surprisingly good at imagining things... :D
The human mind is soooo sensitive to sooo many things 😅 drives me round the bend. That's why I'm glad we have science for most things to stop me going mad 🤣
The thing is we "hear" with our eyes too. There are studies about that. Seeing something that is supposed to make or change sound make us hear that. Check out Eric Valentine video about sound A/B-ing, very interesting and revealing!
Great video, my take on this may help. I come from a live sound engineer background on large analogue consoles. We had one good reverb , one good delay, comps for the busses (rack of 10 dbx comps, 3 man lift!!!) When moving to studio recording I fell over when seeing the amount of plugins available!! I now use 1 rev 1 delay (Comps on each chanel and buses in Mixbus ) Might help a few people get started. Cheers 😀
@@PaulThird It's blown my mind to be honest. I've only just recently purchased Ceilings of sound and the Mix Monolith. Wish I had done it years ago. Keep up the great work you do Paul
Anyone who claims "noise isn't sound" are closed minded, uncreative and in the end just not as musical as they think they are. I saw this awe inspiring video from Abelton with this 30th level Wizard of a sound designer as to how sounds around us can be musical. And then he proceeded to make the sound of Birds from one sine wave and non typical percussion sounds (like not typical snares and high hats stuff) with noise. Also, just taking white noise and throwing down Infiltrator on it makes for some fine beats. Anyway, plz do more geekery like this on your channel. Its fascinating, good sir.
I've made entire EDM tracks with little more than white noise and eq/compression and some effects, and you would never know the difference how the sounds were made, although it takes a lot more work, I just did it as a learning exercise a few years back.
I sampled the sound of a shit hitting the water in the can, then freaked and tweaked it into something musical. Now people who use it have no excuse, their music really is shit.
Yup lol, many audio principles apply to eg. Physics¿ Colour science too. All colours of a spectrum equal what we perceive as white but when broken down we have different colours based on different frequencies.. hence noise IS very much sound LOL
minor correction 2: having perfect ratios of frequencies between pitches is only true for Just Intonation tuning system, and we (most commonly) use Equal Temperament system, where ratios are only "close enough" to mathemathical perfection to fool our ears, but offers a better flexibility in exchange
Yeah in hindsight I could have worded that better and said "when the math deviates a lot from this" instead of being definitive absolute perfection. Makes sense though cause tuning guitars is a pain in the arse at the best of times haha couldn't imagine having to perfectly tune my guitars every time 🤣 Couldn't if I tried so there is slight tolerances we can withstand but how much the math can deviate before we notice it.. 🤷♂️ Tbh that whole bit wasn't even meant to be in the video. It just snowballed the more my autistic brain scrambled trying to tie up loose ends 😅
yes there's definitely wiggle room, microtonal music used in many cultures does not use the exact same intervals but sound sweet in their own context. There's also true temperament guitars with the squiggly frets that do sound slightly sweeter and more consonant than equal temperament but also seem to loose a little flavour and spice as a result. I definitely learned a lot of 'technicals' watching your video thanks!!
Omg I loved this yes! I did want to say for the beginning of the video. Noise and music is the same yes true. Even looking at something like rx dialog de noise. It actually tries to separate the harmonic noise vs the static noise and let's yoy make things sound cleaner but all of it is part of the original audio. Or sound. Beautify done.
I can't believe you needed to make this video. Interesting following you have here 😆 Teasing, I know they're on everyone's channel. Thanks for the video!
Well I'm not going to sit here and say I understood ALL of that, but I did at least finally learn what makes side chaining so important, and it was also quite enjoyable.. Thank you for the work you put into that Paul.
OK I think I get it now. The terminology of this has always annoyed me. Polarity is inverting a positive wave to a negative wave without changing the time domain? Where a 180 degrees phase flip would be altering the time domain of a wave so the wave appeared inverted at that specific plot? Is that right?
@@PaulThird Exactly. Flipping polarity simply flips wave up side down, and phase rotation introduces delay to line up peak with the trough of next wave cycle
I think what's happening is people are confusing static processes with dynamic / nonlinear. Someone who doesn't know the difference won't accept an eq null because they've heard dynamic comparisons as different which were only measured statically. Some tests, noise or otherwise are only appropriate in certain tests and situations. In other words we can't measure psychoacoustic differences over time.
Yup when in reality the null test caters for everything regardless of dynamic or static process. A pink noise null on compression for example would give the same maximum difference amounts but at a fixed gain eduction so to really paint a more consistent picture, in theory, you would have to take snaps shots of the null at different levels of gain reduction but the reality is that in a dynamic test, as long as the max frequency peak is matched whatever the highest gain reduction that source reaches is, the pink noise null will show maximum possible differences at that gain reduction.
Hey Paul, I’m not sure if I’m hearing what you said correctly, but at 5:50 the relationship between the A note and C note is a minor third (not a major third). For the sake of accuracy. Regardless, funny that a comment about 550hz would have happened at exactly 5:50 in the video. 🧐
Yeah it was a slip up. I'm gonna take that specific bit out just so everything is more consistent It was mentioned by somebody in the comments as soon as I released it
@@PaulThird that’s cool. I just figured I would mention it, since there would likely be some wanker who would take that one little bit and then say ‘Paul doesn’t know what he’s talking about’
Holy crap Paul I think you made Dan Worrall yawn and Harrison is currently trying to disprove this video, thanks for the post you put a lot of work into it, I actually learned a thing or two, anyway, would love to chat but I have to get back to downloading today's free plugins. Keep up the good work
Paul, speaking of not purchasing more plugins - there's a company called Direct Approach that just announced all it's plugins (there's 10) are free. They all seem to be unique. Would love your take on these!
@@PaulThird Their metering plugin, SpecTrend is actually what first attracted me. It provides a pink-noise weighted view of the audio spectrum of your track. With pink-noise weighting, all frequency bands will be at the same level when your mix is spectrally balanced. And the +3db low end compensation allows for more low end in modern mixes. I wouldn't necessarily let this tool dictate my mixes, but an interesting referencing tool, nonetheless. It's also completely silent.
Our ears mechanical components have a greater bandwidth than we can hear because there are 33 nerves between our cochlear and our brain , each nerve covers 1/3 octave hence and professional graphic has 30 to 33 bands , one for each 1/3 octave nerve , what this means is our ears mechanical components pick up frequencies up to 40khz and above but we don’t “hear” them as we have no nerves transmitting that info but in tests some people can tell “something “ is there whilst not being able to “hear “ it
This world is becoming too complex and interdisciplinary for even the most intelligent among us to hold a "proper" sheepskin in every field. I think that appeals to credentialism are one of the weakest sort of arguments. It's a form of social "noise", the disclaimers one must make to stave off such attacks. This is how we get to having instructions being printed on a box of matches. I like what you do and how you do it, Paul. Keep it up. I learn something every time I watch your channel.
Really like your videos! Thier fun to watch and completey agree on a lot of your points (in other videos as well). Some questions though, how do you account for if the plugin itself it affecting the transients/dynamics? I suppose that would show up in the null frequency range spectrum or perhaps if you listen to the null signal itself? Also, when it comes to spectral analyzers, what is your favorite? SPAN? was thinking of picking one up to improve upon the base DAW version.
Doctor can show you dynamics. If there's anything dynamic wise there's a whole dynamics section where you can see the transfer curve ramp and the attack and release
The person that wrote “use your eyes, not your eyes” is using their eyes to decide what a plugin sounds like based on the description they read. People also tend to think a plug-in sounds better based on how much money they spent on it. I use Logic Pro, and for the most part their included free plugins. They work great, and when comparing to something like Waves plugins which I demoed, on an equal if not better level of quality. My one concern in the tests is while noise is a valid audio source for null tests, I wondered about transient responses of the plugin. Especially if it’s adding any distortion/harmonics. But overall I find your testing very thorough and well thought out and executed. Null tests don’t lie. Do you think you get “better” audio quality recording at 128k instead of 48k? Null tests show otherwise. You gain aliasing artifacts at high sample rates and gain no better frequency response. But people have bias confirmation and always think their rationalized option on what they just heard proves them right. That’s why we do blind listening tests. To avoid bias.
@@PaulThird By the way, I made the transient comment before I got to that part of the video! So thanks for covering that. I agree with your findings. I’ve been a self taught audio engineer for the past 50 years of recording myself and my bands on all manners of gear. We have so many options these days. Too many! Lol. I say just get a good recording going in and stop thinking some plugin is going to eliminate the skill needed in mixing. That just takes experience and learning how to hear (and see) audio.
If I'm being honest I'd never heard of a Delta Signal before.... I have heard of a Delta Wave though. I think it was mentioned in an episode of Doctor Who.. It was a good episode too, one where the Daleks want to blow up the world (Surprise surprise) ANYWAY.. I digress, and now back to enjoying the rest of the video!
Thanks Paul. It's a sorry fact of life but no amount of showing people the truth will convince those with little skill in critical thinking, but it's courageous of you to try. One note is that polarity inversion is not the same as a 180-degree phase shift. When you invert, the waveforms are still precisely aligned in time. When you phase shift by 180 degrees the two waveforms are out in time by half the wavelength. Thus, with a non-sinusoidal signal, shifting the fundamental by 180 degrees won't move any other frequency by 180. Inversion makes every frequency in the signal opposite phase. 😆
Hi, i'm a hearing aid acoustician and i have small complaint: i'd really like to see the source that claims that frequencies below 250Hz largely bypass the inner ear because they don't and you couldn't hear them if they did. What is true is that we percieve frequencies below 250Hz to be quieter than higher freqencies with the same energy (see "Fletcher.Munson Curve), but we can hear frequencies down to approx. 20Hz.
www.science.org/content/article/sounds-you-cant-hear-can-still-hurt-your-ears www.ncbi.nlm.nih.gov/pmc/articles/PMC4448896/ That's where I got the exact quote from though royalsocietypublishing.org/doi/10.1098/rsos.140166 If I recall I did say that the reason for us perceiving lower frequencies less is due to lower frequencies being inadequate at driving the inner hair cells of the cochlea. That's somewhere in one of these references.
@@PaulThird Ok sorry for the late answer: the term largely bypass the inner ear might be a bit unlucky, because what frame of reference is used here? All energy that has been coming from a certain sound source? But the energy coming from a sound source will mostly bypass the inner ear in all frequencies, only a fraction of the energy will even get there. What is actually true is that a sinus at 250Hz needs to be about 20db louder than a sinus at 1kHz (which equals to 10 times the energy) to generate enough action potentials in the inner hair cells to be conciously perceived. That might sound like a lot, but humans have a hearing dynamic of 100db or more (and since db is logarythmical this equals 100 000 times the energy) so in correlation it is actually not that much. Anyway these frequencies don't "largely bypass" the inner ear, the hearing cells just need more energy to be activated in lower frequencies (interestingly the difference gets smaller in louder levels). Think of it like resonances.
I'm not convinced that any integrated test - especially on a steady state signal - will really show us the effect of slew rate limiting until it gets really extreme. Not up to running the tests right now, though. On a perhaps unrelated point, I find myself wondering if the area under the curve is sometimes more meaningful than the level at any given frequency, but atm am not sure how or why or when that might be true.
I’ve always understood that slew rate limiting manifests as a reduction of high frequencies since the higher harmonics required to produce a fast rise time (according to Fourier) will be necessarily absent. The corollary to this is that digital audio is inevitably slew rate limited at high frequencies (above 1/2 nyquist) as there will be no high harmonics to sharpen up the rising edge of transients. Luckily, most energy in the top octave is from cymbals and the like in which partials are not harmonically related. Aliasing in the top octave (10k-20k) can sometimes be euphonic and improve perceived transient performance. Aliasing in lower octaves is definitely not euphonic though (like any kind of intermodulation distortion).
Well slew rate limiting is dynamic. It’s not the same as a static filter, and I’m just not sure how it translates to windowed FFT analysis of a pink noise source. I believe I can hear it before I can see it on SPAN, but it’s been a while since I did those tests.
Great video, very useful and informative, if not slightly challenging, taking your geekiness to another level lol! Best bit of the video, the silent delta example 😂 Do you think you could do a video or a short on how you do your level matching?
Cause it allows you to understand if a certain process or mixing technique is actually adding anything audible, and if so, how audible that difference actually is and if its worth sweating over This can in turn stop you from hyper analysing small moves and instead focus on more impactful mixing decisions which can result in quicker mixes
I have a theory on why less than 1 db can be heard. It's not that you hear 1 db difference, but that two sound files (wav, mp3, whatever) may have, say, 0.5 db difference, but those files pass through a multiplier (amplifier), that being your sound system or even plugins in the sound chain. In that case, you are not listening to just 1db or less of difference, but that amount multiplied by a certain factor.
SSL have a disclaimer on their page and in their manuals for the new ch strips stating that plugin doctor analysis of their plugins may not be correct due to (this is where I get lost) something about being non linear. Anyway disclaimers about plugin doctor? I feel like they are talking to you directly 😂 do you have plugin companies scared? And what is this they are even claiming? Thanks 🙏🏾
Plugin doctor defaults with a linear test which if fed too much non linearities ie lots of noise and distortion then may warp the linear test reading. Depends on how much non linearity is fed through the system. However there is also a fundamental sweep which caters for non linearities so their point is completely pointless. SSL making a point to discredit plugin doctor in their manual and website about this usually means that they have something you want to hide. So be very aware
Being a fella on the autism spectrum myself, there is one large issue with this video, but it's not with the science or facts: Some people will still argue against any level of correctness, because if what is proven is factually true their core foundation of belief will shatter. People who attach identity with their own reality of facts would rather see the world burn than be proven wrong. TL;DR I too thought I could help more people understand certain things by giving a more thorough presentation of facts. They fought back harder.
Tbh im expecting it. I sent it to a few of my audio educator pals and even they were like.. Its definitely solid work but you'll still have people looking for the slightest crack, technical wording, use of advanced audio theory that most engineers don't have a scooby about, and even people who just plain refuse to take any science on board whatsoever. One of the best quotes I've read is from Hawking "one of the basic rules of the universe is that nothing is perfect. Perfection simply doesn't exist.. Without imperfection, neither you or I would exist" I always remind myself of that when I create anything and just accept that it's not perfect, it's the best I could create with the time and the resources I had during that period of my life. You either spend your lifetime on it or just say that's enough and put it out to the world and hope that it's imperfections don't smear the final result too much
@@PaulThird I certainly appreciate the effort, and I'm positive there is a handful of people out there who will find the massive benefit from watching your work; This alone should justify your hard work. If the goal is to change the minds of the unwilling, I suppose altering the course of the Sun before their eyes wouldn't do much either.
Music is just organised noise. I attended a mastering seminar by Bob Katz a few years ago where he suggested that we can perceive a difference of 0.2dB - louder is better.
I spoke to somebody else who said the same but it was more of a distortion speaker trick you could use at 0.2db but on average sources its much harder. I've found from this test that level difference is easier to hear compared to tonal difference. I'd argue tonal differences of 1db is much harder to identify than a 1db level change. Over 80% failed my double 0.6db test
@@PaulThird - I wouldn’t disagree, a 1dB difference over the whole spectrum is pretty obvious. A 1dB difference in tonality probably depends on the bandwidth of the difference - I’d suggest that a 1dB boost with a Q of 5 is more difficult to discern than a 1dB Boots with a Q of 0.5. An “area under the curve” kind of thing. Narrow cuts are even more difficult to hear.
@@PaulThird It's from So I Married an Axe Murderer with Mike Myers. You have to watch it as it has a Scottish theme. ua-cam.com/video/t-OCjvbV2Z4/v-deo.html
In addition, most people in commercial studios are listening at 80-85dB most of the time (which I find pretty loud)... you can't hear that difference at that level. Not a chance in hell. cheers!
Great videos, but we’d love to see your take on mic emulators. Are they legit? (Examples Lava, ML-1 mic, and a new mic by UA) I personally don’t think they sound the same but they do emulate the mics general vibe.
Lava was pretty much mocked by the industry guys I spoke to about it. They were puting up their own mics and it didn't sound anywhere close. I can't give you any real feedback as I haven't had experience with expensive mics for over a decade
Really cool video! A lot of it is going over my head, but I have one question regarding the idea that a sub 1db difference is inaudible. If a plugin was generating an unintended bump of .3-.6db and it was being instantiated on a large handful of tracks in a mixing session, wouldn't that cause a buildup that would ultimately become audible? I may be way off in understanding what you're getting at in the vid haha but thanks for all the info! was still interesting to watch regardless
Then it must likely wouldnt be a sub dB difference if you added them all together. I'm talking about a difference caused by 1 process on 1 source. You are talking about 1 process over multiple sources. In that instance the bump would accumulate and become audible, well depending on if the source makes a lot of that frequency anyway and other factors. Depends on where the bump is, how wide it is, how much its adding, how many sources its on, the level of those sources in relation to the mix and what frequencies they make. Its all context, are you adding 1 process to 1 source or adding 1 process to multiple sources. There's a very big difference in methods and context but both can still be measured to give you an overall level difference
No. The pink noise is the source nulling a direct copy of itself. One with no processing, another with saturation, and then comparing that null against the null of a musical or acoustic sound using the exact same process and test setup. The only correspondence the pink noise has to the musical source is it's own level in relation to the max frequency peak of the musical source. In other words, the level of the pink noise going into the process is determined by the max frequency peak of the musical/acoustic source. They are still both 2 seperate tests with their own nulls. What the video proves is that regardless of what the musical source is, it will never achieve a louder null on any process as long as the pink noise matches it's loudest frequency peak. So when testing for audible differences in regards of 1 or 2 different processes, pink noise can act as the final arbiter cause if its null shows an inaudible level going in at the maximum level you'd go into that process at.. Then doing ANY musical AB is pointless cause there's not enough audible differences there to decipher which is which
@@PaulThird oh…. Ok, that makes much more sense now. I watched it 3 times trying to figure out what was going on. Lol. But I’m not one of those types that think there would be a difference because ike you said, audio is audio. Thanks for the explanation.
By the way.... There was a video a few years ago where someone pulled a prank at the end of the video ... And ran the exact same clip twice at the end and claimed example A was the hardware and example B was software. Many people actually claimed they could hear how the "hardware" had that subtle extra hardware glue send goodness, making it superior to the "software" example. Blind, level matched tests are the only way to gauge an honest opinion.
I feel like there’s a lot of context outside of this video that I have zero knowledge of to make any sense of what is the debate that it is addressing, and what relevance it has to the video’s title. Does anyone feel like giving me some pointers?
The whole point of the video is showing how you can properly test if a plugin or piece of gear is actually adding anything audible, and if 2 plugins or gear have enough of an audible difference between each other to constitute in a purchase, without being distracted by psycho acoustics and the placebo affect. This can save you money on plugins as you know you can actually measure how much difference a process is adding to your sources/mixes. Measuring the level of any difference proves to you how much a difference that decision is adding to a mix. It can help your mixing cause you understand that it's a pretty low level that the average listener will struggle to pin point. Makes you focus on bigger elements instead of micro details. Understanding the intensity of acoustic sounds gives you an understanding of how your plugins recieve the signal. With the 4.5db slope off you can actually see your fundamental accurately which makes your hpf process make more sense. Puts into consideration how much your hpf sidechain detection in your compressor would need to be in order to bypass most of the low end energy triggering the compression. Understanding the relation to musical notes and frequencies can help you with your eq'ing Theres tons of stuff going on but the video was originally inspired by the waves bb tubes review I did where I showed inaudible pink noise nulls and guys were still commenting that were wasn't enough "actual" audio examples to add context when in reality there were no audible differences added by a lot of the artist presets and I could easily recreate most of it with free plugins..the null proved that again.. And I got tired of explaining to people that Pink noise is the most complex signal you could use compared to any audio example. If the pink noise is a decent level and the nulls inaudible.. Pointless doing any musical AB test
What do you mean by using pink noise with your sounds? Should I layer sounds with a pink noise layer? And how would I do so? Using shaperbox or some stuff like that?
I'm moving to fortnightly videos now the weekly PLAP podcast is back. Remember I have a Mon-Fri day time job as well as continuing my mix portfolio. I can't look at every product now. If there's time I may have a look but surely you'd be able to run a few null tests and it'd prove if it's actually doing something?
@@PaulThird The measurements on mine are -105 and -109. They are early and mid 2000's apogee and lucid designs. You have a great ear Paul, I'm inclined to believe you hear the difference between truncating or dithering from 24 to 16 bit with adequate level. Why have greater than 16 bit precision even for mixing otherwise? I think there's a difference between playback of a -86 dbfs (you can hear that in an adequately noise free environment, it may be buried in home studio conditions regularly)signal and -86 in a mix. The difference is your ear makes out the relative difference. Not only is that audible to me in a home studio, but it's an easy measurement.
@@PaulThird Run a test perhaps. Track 16 bit and 24 bit simultaneously. Use something like waves idr without dither, or an old free program like cool edit 96. Iirc you can run 16 bit mixing on it or similar old software. I bet you measure and hear a difference. I'll reference Bob Katz Mastering Handbook as one of my sources for education on the topic. It's a great read if you haven't seen it. I admire that you're data driven in analysis, it's a great quality of yours.
Tbh I just do 24 bit. 32 bit floating point is a bit extreme for me. I'm not peaking the shit out my stuff. There was a great video on 16 and 24 bit I referenced in a video years ago where it showed the noise differences
@@PaulThird With all respect, you're using 32 and 64 bit. It's in your mix bus, plugins and measurement tools. Every bit in fixed point yields 6db of dynamic range and hopefully SNR. I've seen a common measurement problem where people null a 32bit and 64bit mix bus mix. That's mathematical nonsense, computers are usually great at math. They missed that they're measuring with 32 bit software! It lacks the mathematical precision to measure the difference in math in the first place. I'll offer a second test for you, if only thought experiment. Take a 24bit unprocessed recording and truncate to 16 bit. Do 15 and 14, hell 8 bit. You'll hear it get worse if it's a natural sound. It'll take a lot longer to discern with a sine wave or two tho. I think the reason is those harmonics are perfectly related, unless you happen into a bit of intermodulation distortion or a bad iir anti alias filter. I think we hear the difference faster with less predictable signals like found in nature and music. That makes our measurements much harder to justify, and simultaneously all the more important imo.
Crikey, take it easy Einstein! P.S. I think the Delta wave was popularised by blues musicians riding the wake from paddle steamers on the Mississippi. Great stuff Paul - thanks 🙂
How do you have a degree in audio-engineering yet no qualifications in physics and audio-science?? If that's the case, you don't have a degree; you have a certification. When I was in college, all I was really interested in were the classes I got to work in a nice recording studio for a couple hours everyday, but I also had to pass physics and other science courses.
I have an BA honours degree and an A grade HND. I studied audio engineering in an avid run studio led by lecturers who had worked in the industry. We covered lots of physics and science. Even had tests based around audio science. you don't need a specific degree in science and physics to have an audio engineering degree
That if you use a strong pink noise signal to compare differences between 2 processes ie plugins or gear and it results in a very quiet null (spectrally around -80/85dbfs) then there's absolutely no point in doing any AB tests cause it proves that any musical signal will have completely inaudible differences
Yes, and span show them via the avg + max preset. The max shows your maximum transient level for each frequency. So when I match the pink noise to the loudest frequency peak i.e. The loudest transient, which due to the sound intensity of musical sounds is always going to in the lower frequencies of the signal, it shows you the maximum possible difference of that process because there's no louder transient coming from the acoustic source due to the sloped properties of acoustic sound due to its sound intensity. The pink noise null shows you that by matching the max frequency peak there will be no transients that exceed the pink noise due to the slope of acoustic sounds. The avg and max preset proved that. Pink noise is essentially a "constant transient" which makes no logical sensenin theory but for this discussion it's basically the maximum possible dynamic range that a frequency can make in an acoustic source at the same max frequency peak. So pink noise it's a constant steady dynamic. So it's basically a constant transient across the frequency spectrum. So when the loudest max frequency peak is matched by pink noise no acoustic source will provide a louder null in any frequency.. Because pink noise is sloped in a way to give the maximum possible ceiling for each frequency in relation to the sound intensity of acoustic sounds. So there's no transient of any acoustic source that can surpass it. The kick test proved that
I’m sure people watched completely drawn into the sense, putting down their toys and listening… But sure enough the video will end and all that information will be lost in the clatter of their own thoughts after and the same taking it personally to a debunking of a companies plugin they have nothing to do with and the comments will flow in yet again… and you’d have to make another video having to explain things in even greater depth… because you can’t wake someone up that pretends to sleep, and they can not understand they are pretending to sleep because it’s much easier to fake it by believing it and they then fulfil their own narrative. This is the same as what’s going on over on the Glen’s channel with guitar tone arguments. I myself learned long ago how easily our ears can lie to ourselves when I would have a plugin bypassed without realising it and it would take quite a few seconds to realise the knobs wasn’t doing anything… but I swear I could hear a subtle change as I turned the knob… it was only after a few seconds before I realised the change should be much more before I realised. We have to take a step back from what we think we are hearing constantly with creating audio. I have been making sure I splash myself in the face with theoretical water ever since but some people seem to want to run from the truth in anyway they can… “okay, so you proved this much, but what about this bit of information you left out that might have a little gap for us to run through for you to have to come running after us to save from being ran over”
It's inevitable that many in the audio world will deny hard science but for me anyway I like to have work that I can just copy and paste into a comment and leave it there. Anybody that tries to do the whole "were are the real audio examples?" or "dude we use our ears, not doctor and spectral analysers" in a situation where I've shown an inaudible pink noise null I'll just send them this video and let them argue with science. Truth is they won't argue back, but doesn't mean that they take the science on board either. They just simply don't have the argument to counter against it and that's more than enough for me
I a honestly say: I did not understand this video. 😂 i do know though that a lot of plugins are BS. I also know that saturation for instance is distortion of certain frequencies that although not adding volume they change the sound. I also know that you can get changes in sound that frequency analyzers can't measure.. why are some neumann microphones so good for instance : because they add a certain perceived room/3d "sound" to the recorded audio (and other way more geeky things). point being: I always level match and null test after adding a new plugin to make its not just fooling my ears ps: love your channel.. and your brain.
Sorry to break it to you, but 0,6 dB is very audible. I'd get this one correct 10 out of 10 times and I'm not even trying. Important: When I was mixing on headphones I always got the kicklevels wrong (more so than now). Nowadays I attribute this to not being *able* to judge transient-loudness adequately on headphones. I'll spoil the correct answer as an answer to this comment, so don't click unless you already voted. Wether it still makes sense to use *actually* dynamic music when doing nulltests is something I'll have to think about when I'm not so tired. First thought: One being more complex than the other and thus creating a "bigger impact" to the original, doesn't mean the complexity is of the same kind. (I edited out spelling mistakes, but I did not edit the answer where I spoiled the result. So don't think I cheated!)
Tbh one of these tests can be easily cheated but it's the saturation test that interests me the most as it can't really be cheated. What interests me is if plain level difference is more audible than actual difference 🤔 Suppose it depends where in the frequency spectrum the difference is. Too much science for one day but id disagree that 0.6db is very audible. I think very being the wrong term to use in my opinion. "Very" would tell me that it's clear differences that any Tom dick or Harry could pick out. All of my family couldn't tell me. Not one. 5 people just this afternoon haha
@@PaulThird Very audible to me and plenty other engineers. I don't have great ears. My left one has issues in the low-end. I've always been very self-concious about that. I felt like I picked the wrong profession. Beneath 0,2 dB is where I know it gets messy for me. With transients it's even more audible in context (full mix) because it "pokes out" more or less. Easier to judge it in relation. I commented under your poll which one I picked for the saturation difference - which was a lot more tricky. I had to actually listen closer and my mind might still play tricks on me. The saturation difference is a tonal difference (that also results in a slight difference of perception of loudness - as do ALL tonal differences - of course). So no, 0,6 dB loudness difference is not always audible when the kick is DIFFERENT, but when it's the same (as in no further processing) it's always audible.
So 0.6db in level is very audible in your opinon but 0.6db tonal differences in a signal is less audible? Have I got that right? Tbh it actually makes sense if that's what you are saying
@@PaulThird 0,6 dB change between two otherwise identical signals is very audible (to a half-decently trained ear). 0,6 db tonal difference CAN be very audible, but doesn't need to be. Depends a lot on the ear, the scource and the focus. If you notch down 0,6 dB around 5 kHz in a vocal with a regular Q, you'll certainly hear that (as an audio engineer).
Didn't read the whole comment, but regarding "0,6 dB is very audible" -it varies from person to person. I proved this to myself in a studio full of people.
Your assumption that 0.6dB is inaudible to trained ears is terribly wrong, as any mastering engineer can confirm. In mastering, 1dB is considered a large EQ cut/boost. We tend to EQ in fractions of a decibel changes. We had to go through years and years of ear training, but having those ears is the bar to cross before becoming a mastering engineer. It's also why we obsess over expensive monitors, AD-DA, etc.
0.6db difference between 2 files. Not 0.6db boost or cut which would actually result in a much more audible null Nearly 90% failed the 0.6db test I made while 60% didn't even guess and admitted they couldnt tell which was which .. Explain that
I am very sorry that I did not even understand the essence of the video, unfortunately I am not a native speaker. although I try to understand, I used the translation for the second time when I reviewed it and still did not understand what the video was about.
It's talking about why achieving an inaudible null with pink noise renders any AB test on acoustic sources pointless as an inaudible pink noise null proves that there are no audible differences between 2 processes.
Isn‘t this a kind of a ghost debate? The comment shown at the beginning is actually rather referring to „music“ than to „real audio“. And while noise is real audio, it is not music.
The debate is about whether musical audio examples are needed even though scientifically you can prove that there are no audible differences between 2 processes using just noise as the source. What I've shown is that you don't need musical examples when you have an inaudible pink noise null, musical examples wont give you any more audible differences. So when you have an inaudible pink noise null and somebody says "yeah but try it on 'real' audio".. Its a pointless comment as musical audio and noise are both audio. It's still the same thing. Its like oranges and lemons. They are both different kinds of fruit, but they are still fruit. If you squeeze both they'll give you juice, which will taste different and be perceived different but it's still fruit juice and if you were to say test the acidity levels of fruit and we tested the most theoretical acidic lemon on the planet.. Then why would we want to test a random orange when we scientifically know that oranges have less acidicity than lemons? "cause lemons aren't a real fruit?? Nobody actually drinks lemon juice and orange juice is the most popular fruit juice on the planet" That's essentially what people are debating about but in audio terms
I get that it's just that what I'm saying is that when the science proves there is genuinely no difference to hear, there is genuinely nothing to hear so as much as people prefer to listen to things than do math, the math tells you WHEN you should actually be listening. The human brain is extremely wierd and can make you think you hear things that you actually are not hearing and that's why a combination of math and critical listening, for me anyway, is the best method for discerning differences. No point in listening for stuff you actually can't hear in a blind test
IDK, as an audio engineer, you get so much shit from people that dont know what they are talking about, i just stopped to pay any attention to them. My best one: "I don`t hear a difference. This is all esoteric bullshit. I am sure if you make a double-blind-test, you will see that all those companies just scam you." To wich I replied: "We are engineers. We measure shit." Either you are "too esoteric" or "soulless technical" etc...
Yup, any audio signal can be measured. I think when you understand more about the science (i just scrape the surface) you start to realise just how much the human mind has an effect on your audible perception. All it takes is the idea that a difference may be there to make yourself think you actually hear it. Even last week I hooked up my dx7 pro+ to the apollo x6 via coaxial spdif and I was sure I could hear a lack of high end compared to usb. Done a quick pink noise recording through both and sure enough I was right. Checked the connections and the toppings connection wasn't 100% in 🙈🙈🙈 Pushed it in a bit more and I was like yeah.. This is better but I still felt I could hear the slightest change in the top end. Was going nuts. So done another recording and theres no difference where I thought there was. After that.. I can't hear it anymore haha The human mind is such a temperamental beast and its genuinely the cause of so many arguments in audio. When you stick more to science you start feeling less nuts haha
Songs first... The rest is just a push and pull, tweak and twerk.... But , also it shouldn't start with a drag and drop. Mwhahahahahaha Play it, Own it... and the rest will follow..
Yeah sorry man haha but this doesn't do a single thing to support your theory here. Not gonna waste too much time getting into it this time, but the point you're making here is basically saying that "sound is sound, it doesn't matter what the source is, it all reacts the same way" and that literally could not be further from the truth. If that were true, there would be no difference between subjective and objective mixing. There are things that are "proper" and "correct" (that are objectively correct, they follow the rules) then there's thing that are only correct in THAT specific setting, which in any other case would sound awful. That why engineers have pages of different EQ's and compressors, they all have their different flavors and are suited for their own specific things. In none of these videos have you mentioned anything subjective at all. You're 100%. focusing on the objective science behind it and sorry, but what makes music "human" and relatable is that is is NOT an exact science. Yes, the way sound acts can be measured and treated as an exact science, and that can help us make creative decisions, but its only there to inform us and all rules are meant to be broken, THAT is what makes music listenable. I'm sorry man but it REALLY seems like you're just a negative Nancy haha trash talking every thing people hype up like "oh you fools, I have THE answer." Sorry dude, there is no one answer, and everything your saying (while backed up by science, technically) is still only YOUR opinion and YOUR experience for YOUR music. And honestly it. just seems like you're gaining success from being a trash talked and getting all the other trash talkers on board. Which is DETRIMENTAL to the music community as a whole. You want to show peo[ple alternatives to overhyped plug ins and show how they're basically all the same and there's no "game changers" coming out recently? Well that would be awesome, I'd subscribe and watch every single video. That's what I was hoping for when I first clicked your vid. But all you seem to want to do is bash other companies. In your last video I commented on you didn't offer a solution UNTIL I commented THEN you said "gsat is free." Prooooobably should've just made a video comparing it to Gsat+ in the first place, then talk about how they're essentially the same expect goat may even be better. Instead of just being a hater and bashing things you don't like haha just imagine how successful you could be if you weren't essentially insulting 2/3 of the music world (waves users) with these videos and instead making an unbiased comparison with real world examples. Judging but the fact this video exists at all, I'm one of may who feels this way. Seriously man, stop doing videos just to bash other companies and start doing things in a way that makes people feel accepted no matter their choice Instead of declaring everyone an idiot who doesn't agree with you, and you'll even win ME over. I love the fact that you're diving so deeply into this scientifically, seriously I wish more people would. There are far too many purely subjective videos out there, that aren't quantifiable. But you have the opposite problem haha it's like you're just trying to use science to make others look stupid for their choices. Trust me, I've been there. You'll get a LOT further when you learn to balance the two (subjectivity & objectivity) and when you stop making bash videos. That last video (part one of this) you dint offer a single solution or mention that Gsat+ one that you like a SINGLE time. Literally the entire video was about how stupid waves is. When honestly it's more like their ad campaign was blown out. You're smart & you've got potential dude, it sucks to see it wasted. You could be like THE dude for good comparisons if you dropped the nonsense. I don’t know, maybe I'm literally the only one on earth who seems to like fair comparisons rather than just finding a video with someone agreeing with how much you hate something. But then again, look at the USA 2016 election. It was 100% won by appealing to peoples anger against the party in charge. Seems these days, that's what the masses go for. Seems like you're unfortunately using the same tactic. Sorry man but this is precisely what the world needs less of. I'll keep a look out though, hopefully we see some improvement and positivity (I'm sure it will = money in your pocket as well.) Best wishes man.
Your entitled to your opinon. It's good to know at least that I'm not the right content creator for you so you don't need to be dissapointed in the future, which you will be if you carry on watching my future content as I'll be negative when I feel the need to be negative. Going by your take on this video we'll never see to eye. Tbh I Couldn't really give 2 shits about who gets a sore arse about content I've made. Stopped trying to please everybody a long time ago. I make content I feel like making and stay true to myself. I don't want to the best objective reviewer that can keep everybody happy. If I'm ever negative nowadays it's because I've found something misleading or broken that nobody else is talking about. If I'm overly negative it's because I feel the youtube landscape is overly leaning way too positively on a plugin that has issues that are purposely are being ignored or not properly tested. BB tubes had over 60 positive reviews. Want to know how many critical 'negative' reviews . 2. One was mine and another was a 2 and a half minute video that barely got any views. Just cause I get lots of views doesn't mean I'm not in a tiny minority. UA-cam is awash with videos leading people to spend money. People work hard for their money so in those cases where the landscape is fully tilted, I reign it back hard if I find a lot of overhyping and issues. and that's what gives the objectivity to the landscape. Not objectivity to my video, there's hardly in those cases, but objectivity to all of the community which is over saturated with... BUY THIS. You disagree? Then that's cool, but it won't change how I do things on my channel from time to time. Which is time to time, not every week. I was dissapointed I had to make the waves video cause I hate being called a biased basher when I deal with one tool at a time. Everybody forgets when I did the waves kramer pie test and showed it nailed the hardware. When I made the 5 waves plugins I use video. The community only sees what it wants to see. People blasting tiny amounts of Negativity in a landscape filled with weekly shilling, NFR's, affiliates & overhyping is absolutely crazy. You need the Negativity to counter the completely one sided positivity. If there was actual objectivity in the youtube landscape then I wouldn't feel the need to go so hard. This has turned into my part time job. I study and observe the landscape. I see exactly what's going on. I know the averages, the stats and where a product is leaning on youtube. I've actually made an effort to not be so constantly negative, thus why so much educational content has been on my channel. The reason negativity and criticism is gravitated towards is because there isn't actually that much of it in our field because of money. Viewers see past shilling and respect those who try to save them money and think twice. What do I gain from 'bashing' videos? A lot of hate and a slight rise in ad revenue. I could easily do one sponsored video and earn 4-5x what I did off that waves video. Theres nothing it it for me. I stopped caring about subs after 10k. I do those videos cause I'm genuinely passionate about audio and Im genuinely pissed that people are being misled and not told the full story of what they are buying into. Do you know how many people said they regretted buying the bb tubes and that was before even seeing my video? Loads. Emails, dm's. You don't see what I see. You don't see the shit that goes on behind closed doors. The money that changes hands. The politics. I fight daily from being drawn into the positive world of making money and being positive to constantly sell products. That's why I'm very picky about my sponsored videos. I don't expect you to know or comprehend everything that goes on behind all the lovely positivity out there but I can tell you right now that id rather quit youtube all together rather than lose what I believe in. I'm one guy. One guy who everybody for some reason seems to get their pants in a twist with. I don't get it. I'm not a huge business, nor do I have a huge amount of cash flow behind me. I'm just one guy with science and an opinion vs hundreds, maybe even thousands of youtubers all selling subjective positivity with no real criticsm or hard science behind anything they do. Never forget that. One guy. Just one guy doing this part time while juggling a full time job. I'm not this big threat that loads like to make out. it's just that I see behind the bullshit and stand out by being different from the rest of them. So yeah, don't bother replying or sticking around cause if need be I'll come back and do the same if I feel the landscape dictates it. I'll never stop being who I truly am.
FYI it was brought to my attention that when it comes to inharmonic instruments that have no musical context or harmonic structure ie cymbals and hi hats. There is no noise slope for this. These should be nulled individually on the source itself. This pink noise test will only work with musical sounds that follow a harmonic structure.
Test 1
soundcloud.com/paul-third-944739244/sets/6dbfs-kick-test-can-you-hear
Test 2
soundcloud.com/paul-third-944739244/sets/test-2
Vote on the community post on my channel which will be up for 24 hours 🤓
I didn't fully understand the test, but my guess is that B is louder in terms of dB.
In which test?
@@PaulThird The first test. In the second test the choices are AA and BB.
One test is a 0.6db level difference and another is a 0.6db tonal difference caused by saturation
@@PaulThird Okay, then my guess is that in test 1 B is louder in dB and in test 2 BB is the saturated audio.
one of the neatest tricks I've learned over the years is if you're having trouble with a mix, to throw a pink noise source on the monitor and then level each track individually so that its barely audible over the pink noise, it works surprisingly well and after a while I began to understand how to move towards that result more organically. Doesnt change that my music is shit, but its balanced well now at least !
🤓🤓
lol!
Interesting I have never heard of this. I will have to try this out for fun. Thanks for sharing
Gonna try that. Usually if I'm adjusting levels I'll throw a mono plugin on the master bus..set my levels then just turn it off, and it made a huge difference. If not I'm fighting it getting aggravated
We have all been there. I have on more than one occasion reached out for a gain knob on a channel eq and tweaked the gain while clearly hearing the difference until I realized that I was altering the eq on a different channel that was muted.... My brain is surprisingly good at imagining things... :D
The human mind is soooo sensitive to sooo many things 😅 drives me round the bend. That's why I'm glad we have science for most things to stop me going mad 🤣
I think we all have done that one. 😅
The thing is we "hear" with our eyes too. There are studies about that. Seeing something that is supposed to make or change sound make us hear that. Check out Eric Valentine video about sound A/B-ing, very interesting and revealing!
I have the qualifications mentioned in the opening to confirm that this is “close enough”. Good shit mate.
🤜🤛
This one’s a banger, I can already tell!
It took me weeks so I fkn hope so 🤣😅
Thanks!
🤓🤓
The audio woo woo brigade needs rigorous debunking. You're doing a great job.
🤜🤛
If you make multiple boosts in EQ of white noise you can make beautiful drone sounds, because all sound/noise has frequency
🤓🤓
Cheers Paul, my brain has just melted and come out my bum hole. To be fair you did warn me at the start. 😂
😅😅😅
thank you for another great video, Paul! i really admire the amount of work and knowledge you put into this stuff.
🤜🤛
Great video, my take on this may help.
I come from a live sound engineer background on large analogue consoles.
We had one good reverb , one good delay, comps for the busses (rack of 10 dbx comps, 3 man lift!!!)
When moving to studio recording I fell over when seeing the amount of plugins available!!
I now use
1 rev
1 delay
(Comps on each chanel and buses in Mixbus )
Might help a few people get started. Cheers 😀
🤓🤓
Overall a brilliant video and presentation Paul. Ceilings of sound.
Experimenting with it quite a lot just now 🤓
@@PaulThird It's blown my mind to be honest. I've only just recently purchased Ceilings of sound and the Mix Monolith. Wish I had done it years ago. Keep up the great work you do Paul
my initial thought when watching Paul's video was "COS Pro".
That's how I made the 7.5db slope 🤓
@@PaulThird i almost would have bet on that… it just sounded too familiar after i got to know COS Pro better :-) cheers, mate!🍻
If you take all the "use your ears" comments and null them with white noise you get brown noise. Science.
Hahaha all bout the brown noise
Alot of brown noise online these days 😂
@@bontempo1271 I’m much more worried about the rainbow noise
Anyone who claims "noise isn't sound" are closed minded, uncreative and in the end just not as musical as they think they are. I saw this awe inspiring video from Abelton with this 30th level Wizard of a sound designer as to how sounds around us can be musical. And then he proceeded to make the sound of Birds from one sine wave and non typical percussion sounds (like not typical snares and high hats stuff) with noise. Also, just taking white noise and throwing down Infiltrator on it makes for some fine beats.
Anyway, plz do more geekery like this on your channel. Its fascinating, good sir.
🤜🤛
I've made entire EDM tracks with little more than white noise and eq/compression and some effects, and you would never know the difference how the sounds were made, although it takes a lot more work, I just did it as a learning exercise a few years back.
Instead of asking if noise is sound, ask where sound ends and noise begins.
I sampled the sound of a shit hitting the water in the can, then freaked and tweaked it into something musical. Now people who use it have no excuse, their music really is shit.
Yup lol, many audio principles apply to eg. Physics¿ Colour science too. All colours of a spectrum equal what we perceive as white but when broken down we have different colours based on different frequencies.. hence noise IS very much sound LOL
minor correction 2: having perfect ratios of frequencies between pitches is only true for Just Intonation tuning system, and we (most commonly) use Equal Temperament system, where ratios are only "close enough" to mathemathical perfection to fool our ears, but offers a better flexibility in exchange
Yeah in hindsight I could have worded that better and said "when the math deviates a lot from this" instead of being definitive absolute perfection.
Makes sense though cause tuning guitars is a pain in the arse at the best of times haha couldn't imagine having to perfectly tune my guitars every time 🤣 Couldn't if I tried so there is slight tolerances we can withstand but how much the math can deviate before we notice it.. 🤷♂️
Tbh that whole bit wasn't even meant to be in the video. It just snowballed the more my autistic brain scrambled trying to tie up loose ends 😅
@@PaulThird True. Just picking harder changes the initial pitch.
yes there's definitely wiggle room, microtonal music used in many cultures does not use the exact same intervals but sound sweet in their own context.
There's also true temperament guitars with the squiggly frets that do sound slightly sweeter and more consonant than equal temperament but also seem to loose a little flavour and spice as a result.
I definitely learned a lot of 'technicals' watching your video thanks!!
Omg I loved this yes!
I did want to say for the beginning of the video. Noise and music is the same yes true. Even looking at something like rx dialog de noise. It actually tries to separate the harmonic noise vs the static noise and let's yoy make things sound cleaner but all of it is part of the original audio. Or sound.
Beautify done.
🤜🤛
Thank you for this🙌🏼🙌🏼
🤜🤛
I can't believe you needed to make this video. Interesting following you have here 😆 Teasing, I know they're on everyone's channel. Thanks for the video!
🤓🤓
Well I'm not going to sit here and say I understood ALL of that, but I did at least finally learn what makes side chaining so important, and it was also quite enjoyable.. Thank you for the work you put into that Paul.
🤜🤛
Side chaining is such a useful tool. Dig into it a bit and you'll find massive shortcuts in a lot of mixing areas.
minor correction: rotating phase by 180° ≠ inverting polarity, and with nulling we use the latter
OK I think I get it now. The terminology of this has always annoyed me.
Polarity is inverting a positive wave to a negative wave without changing the time domain?
Where a 180 degrees phase flip would be altering the time domain of a wave so the wave appeared inverted at that specific plot?
Is that right?
@@PaulThird Exactly. Flipping polarity simply flips wave up side down, and phase rotation introduces delay to line up peak with the trough of next wave cycle
Thank f*ck for that cause my brain is done after this video haha 😅
@@PaulThird You think *your* brain is done?! That was like a download from the Matrix. I think I could pilot a helicopter now 🤣
I think what's happening is people are confusing static processes with dynamic / nonlinear.
Someone who doesn't know the difference won't accept an eq null because they've heard dynamic comparisons as different which were only measured statically.
Some tests, noise or otherwise are only appropriate in certain tests and situations. In other words we can't measure psychoacoustic differences over time.
Yup when in reality the null test caters for everything regardless of dynamic or static process.
A pink noise null on compression for example would give the same maximum difference amounts but at a fixed gain eduction so to really paint a more consistent picture, in theory, you would have to take snaps shots of the null at different levels of gain reduction but the reality is that in a dynamic test, as long as the max frequency peak is matched whatever the highest gain reduction that source reaches is, the pink noise null will show maximum possible differences at that gain reduction.
Hey Paul, I’m not sure if I’m hearing what you said correctly, but at 5:50 the relationship between the A note and C note is a minor third (not a major third). For the sake of accuracy.
Regardless, funny that a comment about 550hz would have happened at exactly 5:50 in the video. 🧐
Yeah it was a slip up. I'm gonna take that specific bit out just so everything is more consistent
It was mentioned by somebody in the comments as soon as I released it
@@PaulThird that’s cool. I just figured I would mention it, since there would likely be some wanker who would take that one little bit and then say ‘Paul doesn’t know what he’s talking about’
Holy crap Paul I think you made Dan Worrall yawn and Harrison is currently trying to disprove this video, thanks for the post you put a lot of work into it, I actually learned a thing or two, anyway, would love to chat but I have to get back to downloading today's free plugins. Keep up the good work
🤓🤓
Very good explanation, Paul.
🤓🤓
Paul, speaking of not purchasing more plugins - there's a company called Direct Approach that just announced all it's plugins (there's 10) are free. They all seem to be unique. Would love your take on these!
🤓🤓
@@PaulThird Their metering plugin, SpecTrend is actually what first attracted me. It provides a pink-noise weighted view of the audio spectrum of your track. With pink-noise weighting, all frequency bands will be at the same level when your mix is spectrally balanced. And the +3db low end compensation allows for more low end in modern mixes. I wouldn't necessarily let this tool dictate my mixes, but an interesting referencing tool, nonetheless. It's also completely silent.
Our ears mechanical components have a greater bandwidth than we can hear because there are 33 nerves between our cochlear and our brain , each nerve covers 1/3 octave hence and professional graphic has 30 to 33 bands , one for each 1/3 octave nerve , what this means is our ears mechanical components pick up frequencies up to 40khz and above but we don’t “hear” them as we have no nerves transmitting that info but in tests some people can tell “something “ is there whilst not being able to “hear “ it
Interesting 🤓
as someone who listens to Dubstep, I didn't even consider that someone might consider noise as completely different from other sound xD
🤓🤓
lol same here. That might explain why some people hear dubstep and immediately discount it.
This world is becoming too complex and interdisciplinary for even the most intelligent among us to hold a "proper" sheepskin in every field. I think that appeals to credentialism are one of the weakest sort of arguments. It's a form of social "noise", the disclaimers one must make to stave off such attacks. This is how we get to having instructions being printed on a box of matches. I like what you do and how you do it, Paul. Keep it up. I learn something every time I watch your channel.
🤜🤛
Really like your videos! Thier fun to watch and completey agree on a lot of your points (in other videos as well). Some questions though, how do you account for if the plugin itself it affecting the transients/dynamics? I suppose that would show up in the null frequency range spectrum or perhaps if you listen to the null signal itself? Also, when it comes to spectral analyzers, what is your favorite? SPAN? was thinking of picking one up to improve upon the base DAW version.
Doctor can show you dynamics. If there's anything dynamic wise there's a whole dynamics section where you can see the transfer curve ramp and the attack and release
Great video and great job! 👏🏻🙌🏻👌🏻
🤜🤛
The person that wrote “use your eyes, not your eyes” is using their eyes to decide what a plugin sounds like based on the description they read. People also tend to think a plug-in sounds better based on how much money they spent on it.
I use Logic Pro, and for the most part their included free plugins. They work great, and when comparing to something like Waves plugins which I demoed, on an equal if not better level of quality.
My one concern in the tests is while noise is a valid audio source for null tests, I wondered about transient responses of the plugin. Especially if it’s adding any distortion/harmonics.
But overall I find your testing very thorough and well thought out and executed.
Null tests don’t lie. Do you think you get “better” audio quality recording at 128k instead of 48k? Null tests show otherwise. You gain aliasing artifacts at high sample rates and gain no better frequency response.
But people have bias confirmation and always think their rationalized option on what they just heard proves them right.
That’s why we do blind listening tests. To avoid bias.
🤓🤓
@@PaulThird By the way, I made the transient comment before I got to that part of the video! So thanks for covering that. I agree with your findings. I’ve been a self taught audio engineer for the past 50 years of recording myself and my bands on all manners of gear. We have so many options these days. Too many! Lol.
I say just get a good recording going in and stop thinking some plugin is going to eliminate the skill needed in mixing. That just takes experience and learning how to hear (and see) audio.
If I'm being honest I'd never heard of a Delta Signal before.... I have heard of a Delta Wave though. I think it was mentioned in an episode of Doctor Who.. It was a good episode too, one where the Daleks want to blow up the world (Surprise surprise) ANYWAY.. I digress, and now back to enjoying the rest of the video!
🤓🤓
Thanks Paul. It's a sorry fact of life but no amount of showing people the truth will convince those with little skill in critical thinking, but it's courageous of you to try. One note is that polarity inversion is not the same as a 180-degree phase shift. When you invert, the waveforms are still precisely aligned in time. When you phase shift by 180 degrees the two waveforms are out in time by half the wavelength. Thus, with a non-sinusoidal signal, shifting the fundamental by 180 degrees won't move any other frequency by 180. Inversion makes every frequency in the signal opposite phase. 😆
Yes, I figured that out in another comment haha 😅
Hi, i'm a hearing aid acoustician and i have small complaint: i'd really like to see the source that claims that frequencies below 250Hz largely bypass the inner ear because they don't and you couldn't hear them if they did. What is true is that we percieve frequencies below 250Hz to be quieter than higher freqencies with the same energy (see "Fletcher.Munson Curve), but we can hear frequencies down to approx. 20Hz.
www.science.org/content/article/sounds-you-cant-hear-can-still-hurt-your-ears
www.ncbi.nlm.nih.gov/pmc/articles/PMC4448896/
That's where I got the exact quote from though
royalsocietypublishing.org/doi/10.1098/rsos.140166
If I recall I did say that the reason for us perceiving lower frequencies less is due to lower frequencies being inadequate at driving the inner hair cells of the cochlea. That's somewhere in one of these references.
@@PaulThird Ok sorry for the late answer: the term largely bypass the inner ear might be a bit unlucky, because what frame of reference is used here? All energy that has been coming from a certain sound source? But the energy coming from a sound source will mostly bypass the inner ear in all frequencies, only a fraction of the energy will even get there. What is actually true is that a sinus at 250Hz needs to be about 20db louder than a sinus at 1kHz (which equals to 10 times the energy) to generate enough action potentials in the inner hair cells to be conciously perceived. That might sound like a lot, but humans have a hearing dynamic of 100db or more (and since db is logarythmical this equals 100 000 times the energy) so in correlation it is actually not that much. Anyway these frequencies don't "largely bypass" the inner ear, the hearing cells just need more energy to be activated in lower frequencies (interestingly the difference gets smaller in louder levels). Think of it like resonances.
Just unfortunate wording then. Bypass is probably not the best word
Loving your channel, dude! New subscriber
🤜🤛
I'm not convinced that any integrated test - especially on a steady state signal - will really show us the effect of slew rate limiting until it gets really extreme. Not up to running the tests right now, though.
On a perhaps unrelated point, I find myself wondering if the area under the curve is sometimes more meaningful than the level at any given frequency, but atm am not sure how or why or when that might be true.
🤓🤓
I’ve always understood that slew rate limiting manifests as a reduction of high frequencies since the higher harmonics required to produce a fast rise time (according to Fourier) will be necessarily absent.
The corollary to this is that digital audio is inevitably slew rate limited at high frequencies (above 1/2 nyquist) as there will be no high harmonics to sharpen up the rising edge of transients. Luckily, most energy in the top octave is from cymbals and the like in which partials are not harmonically related. Aliasing in the top octave (10k-20k) can sometimes be euphonic and improve perceived transient performance.
Aliasing in lower octaves is definitely not euphonic though (like any kind of intermodulation distortion).
Well slew rate limiting is dynamic. It’s not the same as a static filter, and I’m just not sure how it translates to windowed FFT analysis of a pink noise source. I believe I can hear it before I can see it on SPAN, but it’s been a while since I did those tests.
Great video, very useful and informative, if not slightly challenging, taking your geekiness to another level lol! Best bit of the video, the silent delta example 😂 Do you think you could do a video or a short on how you do your level matching?
I'll think about it 🤓
Excellent!!!!! 👏👏👏
🤜🤛
the conclusions aren't clear enough for me, how do you transalating in mixing practise as a guideline ?
Cause it allows you to understand if a certain process or mixing technique is actually adding anything audible, and if so, how audible that difference actually is and if its worth sweating over
This can in turn stop you from hyper analysing small moves and instead focus on more impactful mixing decisions which can result in quicker mixes
I have a theory on why less than 1 db can be heard.
It's not that you hear 1 db difference, but that two sound files (wav, mp3, whatever) may have, say, 0.5 db difference, but those files pass through a multiplier (amplifier), that being your sound system or even plugins in the sound chain.
In that case, you are not listening to just 1db or less of difference, but that amount multiplied by a certain factor.
SSL have a disclaimer on their page and in their manuals for the new ch strips stating that plugin doctor analysis of their plugins may not be correct due to (this is where I get lost) something about being non linear. Anyway disclaimers about plugin doctor? I feel like they are talking to you directly 😂 do you have plugin companies scared? And what is this they are even claiming? Thanks 🙏🏾
Plugin doctor defaults with a linear test which if fed too much non linearities ie lots of noise and distortion then may warp the linear test reading. Depends on how much non linearity is fed through the system.
However there is also a fundamental sweep which caters for non linearities so their point is completely pointless.
SSL making a point to discredit plugin doctor in their manual and website about this usually means that they have something you want to hide.
So be very aware
Being a fella on the autism spectrum myself, there is one large issue with this video, but it's not with the science or facts: Some people will still argue against any level of correctness, because if what is proven is factually true their core foundation of belief will shatter. People who attach identity with their own reality of facts would rather see the world burn than be proven wrong.
TL;DR I too thought I could help more people understand certain things by giving a more thorough presentation of facts. They fought back harder.
Tbh im expecting it. I sent it to a few of my audio educator pals and even they were like.. Its definitely solid work but you'll still have people looking for the slightest crack, technical wording, use of advanced audio theory that most engineers don't have a scooby about, and even people who just plain refuse to take any science on board whatsoever.
One of the best quotes I've read is from Hawking
"one of the basic rules of the universe is that nothing is perfect. Perfection simply doesn't exist.. Without imperfection, neither you or I would exist"
I always remind myself of that when I create anything and just accept that it's not perfect, it's the best I could create with the time and the resources I had during that period of my life. You either spend your lifetime on it or just say that's enough and put it out to the world and hope that it's imperfections don't smear the final result too much
@@PaulThird I certainly appreciate the effort, and I'm positive there is a handful of people out there who will find the massive benefit from watching your work; This alone should justify your hard work.
If the goal is to change the minds of the unwilling, I suppose altering the course of the Sun before their eyes wouldn't do much either.
@@PaulThird perfect comment!
Music is just organised noise.
I attended a mastering seminar by Bob Katz a few years ago where he suggested that we can perceive a difference of 0.2dB - louder is better.
I spoke to somebody else who said the same but it was more of a distortion speaker trick you could use at 0.2db but on average sources its much harder. I've found from this test that level difference is easier to hear compared to tonal difference. I'd argue tonal differences of 1db is much harder to identify than a 1db level change.
Over 80% failed my double 0.6db test
@@PaulThird - I wouldn’t disagree, a 1dB difference over the whole spectrum is pretty obvious.
A 1dB difference in tonality probably depends on the bandwidth of the difference - I’d suggest that a 1dB boost with a Q of 5 is more difficult to discern than a 1dB Boots with a Q of 0.5. An “area under the curve” kind of thing.
Narrow cuts are even more difficult to hear.
Plugin devs,,, 'how we gonna get outta this one'
My dude,,, keep makin it harder for the devs to crawl away from their shoddy marketin 🤘
🤜🤛
Heed move!! Great vid as always yer wee sassenach.
🤜🤛 What do you mean when you say heed move? haha i read that as shaking your head like.. Naw..just naw 🤣
@@PaulThird It's from So I Married an Axe Murderer with Mike Myers. You have to watch it as it has a Scottish theme.
ua-cam.com/video/t-OCjvbV2Z4/v-deo.html
In addition, most people in commercial studios are listening at 80-85dB most of the time (which I find pretty loud)... you can't hear that difference at that level. Not a chance in hell. cheers!
🤓🤓
Great videos, but we’d love to see your take on mic emulators. Are they legit? (Examples Lava, ML-1 mic, and a new mic by UA)
I personally don’t think they sound the same but they do emulate the mics general vibe.
Lava was pretty much mocked by the industry guys I spoke to about it. They were puting up their own mics and it didn't sound anywhere close.
I can't give you any real feedback as I haven't had experience with expensive mics for over a decade
Really cool video! A lot of it is going over my head, but I have one question regarding the idea that a sub 1db difference is inaudible. If a plugin was generating an unintended bump of .3-.6db and it was being instantiated on a large handful of tracks in a mixing session, wouldn't that cause a buildup that would ultimately become audible? I may be way off in understanding what you're getting at in the vid haha but thanks for all the info! was still interesting to watch regardless
Then it must likely wouldnt be a sub dB difference if you added them all together. I'm talking about a difference caused by 1 process on 1 source. You are talking about 1 process over multiple sources. In that instance the bump would accumulate and become audible, well depending on if the source makes a lot of that frequency anyway and other factors. Depends on where the bump is, how wide it is, how much its adding, how many sources its on, the level of those sources in relation to the mix and what frequencies they make.
Its all context, are you adding 1 process to 1 source or adding 1 process to multiple sources. There's a very big difference in methods and context but both can still be measured to give you an overall level difference
What are you doing exactly in this video? Are you setting the noise to correspond to the music and doing a null of the two? Or what?
No. The pink noise is the source nulling a direct copy of itself. One with no processing, another with saturation, and then comparing that null against the null of a musical or acoustic sound using the exact same process and test setup.
The only correspondence the pink noise has to the musical source is it's own level in relation to the max frequency peak of the musical source. In other words, the level of the pink noise going into the process is determined by the max frequency peak of the musical/acoustic source.
They are still both 2 seperate tests with their own nulls.
What the video proves is that regardless of what the musical source is, it will never achieve a louder null on any process as long as the pink noise matches it's loudest frequency peak. So when testing for audible differences in regards of 1 or 2 different processes, pink noise can act as the final arbiter cause if its null shows an inaudible level going in at the maximum level you'd go into that process at.. Then doing ANY musical AB is pointless cause there's not enough audible differences there to decipher which is which
@@PaulThird oh…. Ok, that makes much more sense now. I watched it 3 times trying to figure out what was going on. Lol. But I’m not one of those types that think there would be a difference because ike you said, audio is audio. Thanks for the explanation.
By the way.... There was a video a few years ago where someone pulled a prank at the end of the video ... And ran the exact same clip twice at the end and claimed example A was the hardware and example B was software. Many people actually claimed they could hear how the "hardware" had that subtle extra hardware glue send goodness, making it superior to the "software" example. Blind, level matched tests are the only way to gauge an honest opinion.
I genuinely did think about doing that 🤣
I feel like there’s a lot of context outside of this video that I have zero knowledge of to make any sense of what is the debate that it is addressing, and what relevance it has to the video’s title.
Does anyone feel like giving me some pointers?
The whole point of the video is showing how you can properly test if a plugin or piece of gear is actually adding anything audible, and if 2 plugins or gear have enough of an audible difference between each other to constitute in a purchase, without being distracted by psycho acoustics and the placebo affect. This can save you money on plugins as you know you can actually measure how much difference a process is adding to your sources/mixes.
Measuring the level of any difference proves to you how much a difference that decision is adding to a mix. It can help your mixing cause you understand that it's a pretty low level that the average listener will struggle to pin point. Makes you focus on bigger elements instead of micro details.
Understanding the intensity of acoustic sounds gives you an understanding of how your plugins recieve the signal. With the 4.5db slope off you can actually see your fundamental accurately which makes your hpf process make more sense. Puts into consideration how much your hpf sidechain detection in your compressor would need to be in order to bypass most of the low end energy triggering the compression. Understanding the relation to musical notes and frequencies can help you with your eq'ing
Theres tons of stuff going on but the video was originally inspired by the waves bb tubes review I did where I showed inaudible pink noise nulls and guys were still commenting that were wasn't enough "actual" audio examples to add context when in reality there were no audible differences added by a lot of the artist presets and I could easily recreate most of it with free plugins..the null proved that again.. And I got tired of explaining to people that Pink noise is the most complex signal you could use compared to any audio example. If the pink noise is a decent level and the nulls inaudible.. Pointless doing any musical AB test
@@PaulThird OK got it. I'll give the video a rewatch with this context.
Very nice of you to teach the noobs (or trolls). Not necessary…but nice.
🤓🤓
What do you mean by using pink noise with your sounds? Should I layer sounds with a pink noise layer? And how would I do so? Using shaperbox or some stuff like that?
I only use pink noise as a source for determining the difference between 2 processes. It's just testing differences
@@PaulThird Aight Bet! Thanks. I was a bit confused at one point in the video😅 thanks again❤️
Well, Prof.Third, that was very geeky.
It was. And I now need a week off to recover 🤣
the GameChanger!! lol !! epic... sorry had to bring that back #GOLD
😅😅
Spectrally, fucking, perfect...
🤜🤛
5:51 *Minor third
Fuck sake, trust it to be one of the only things I copied and pasted from an article 🙈 haha
@@PaulThird Do you refer to your kids as minor thirds when they make you sad and major thirds when they make you happy? Because in my head you now do.
In my house I'm the major third and they are the minor thirds 🤣
Gordito...still waiting for your review of "Make Believe" Plugins
You'll be waiting a while cause my diary is pretty full for this year already 😅
@@PaulThird mmmmmm you at least should check them ... may change your mind
I'm moving to fortnightly videos now the weekly PLAP podcast is back. Remember I have a Mon-Fri day time job as well as continuing my mix portfolio. I can't look at every product now. If there's time I may have a look but surely you'd be able to run a few null tests and it'd prove if it's actually doing something?
@@PaulThird gordito.... i know what I am telling you this.... many say is the biggest snake oil in history and if is true... you will be in Disneyland
i forgot that headphones are in the jack and thought the first 2 minutes its no audio on purpose.. duh
Hahaha
Just record at 16 bit if it's not audible under -96. I'm willing to bet you absolutely hear the difference.
-96db noise floor? Cause that's all going from 24 bit down to 16 bit results in. A higher noise floor
@@PaulThird The measurements on mine are -105 and -109. They are early and mid 2000's apogee and lucid designs. You have a great ear Paul, I'm inclined to believe you hear the difference between truncating or dithering from 24 to 16 bit with adequate level. Why have greater than 16 bit precision even for mixing otherwise? I think there's a difference between playback of a -86 dbfs (you can hear that in an adequately noise free environment, it may be buried in home studio conditions regularly)signal and -86 in a mix. The difference is your ear makes out the relative difference. Not only is that audible to me in a home studio, but it's an easy measurement.
@@PaulThird Run a test perhaps. Track 16 bit and 24 bit simultaneously. Use something like waves idr without dither, or an old free program like cool edit 96. Iirc you can run 16 bit mixing on it or similar old software. I bet you measure and hear a difference. I'll reference Bob Katz Mastering Handbook as one of my sources for education on the topic. It's a great read if you haven't seen it. I admire that you're data driven in analysis, it's a great quality of yours.
Tbh I just do 24 bit. 32 bit floating point is a bit extreme for me. I'm not peaking the shit out my stuff.
There was a great video on 16 and 24 bit I referenced in a video years ago where it showed the noise differences
@@PaulThird With all respect, you're using 32 and 64 bit. It's in your mix bus, plugins and measurement tools. Every bit in fixed point yields 6db of dynamic range and hopefully SNR. I've seen a common measurement problem where people null a 32bit and 64bit mix bus mix. That's mathematical nonsense, computers are usually great at math. They missed that they're measuring with 32 bit software! It lacks the mathematical precision to measure the difference in math in the first place. I'll offer a second test for you, if only thought experiment. Take a 24bit unprocessed recording and truncate to 16 bit. Do 15 and 14, hell 8 bit. You'll hear it get worse if it's a natural sound. It'll take a lot longer to discern with a sine wave or two tho. I think the reason is those harmonics are perfectly related, unless you happen into a bit of intermodulation distortion or a bad iir anti alias filter. I think we hear the difference faster with less predictable signals like found in nature and music. That makes our measurements much harder to justify, and simultaneously all the more important imo.
he smart...that is a lot of science for me brain *shuts down*
🤓🤓
Doing God's work again I see👼
Dunno if I'd call it that haha
wow. Thanks.
🤓🤓
is it true that every audio effect is just a variation of a delay?
Lol!!!!! So funny that you actually needed to post a video about this! 😎😅🤪😂🤣🙃🙃😁
Not had any backlash yet so hopefully it's enough to keep a lot of the trolls at bay.. For now 😂😅
Crikey, take it easy Einstein! P.S. I think the Delta wave was popularised by blues musicians riding the wake from paddle steamers on the Mississippi. Great stuff Paul - thanks 🙂
🤓🤓
👍👍
🤓🤓
Mixing with your eyes. A real DAW problem. Train your self to look away or close your eyes 👀
That's why I'm looking into the presonus faderport
Well done Paul. It's frustrating when people whine about this or that when it's usually 'this'.. moan the Aspies...👍
🤜🤛
How do you have a degree in audio-engineering yet no qualifications in physics and audio-science?? If that's the case, you don't have a degree; you have a certification. When I was in college, all I was really interested in were the classes I got to work in a nice recording studio for a couple hours everyday, but I also had to pass physics and other science courses.
I have an BA honours degree and an A grade HND. I studied audio engineering in an avid run studio led by lecturers who had worked in the industry. We covered lots of physics and science. Even had tests based around audio science. you don't need a specific degree in science and physics to have an audio engineering degree
👌🏻👌🏻
🤓🤓
I’m pretty much dumb to this kind of stuff. But in simple terms what does this mean to audio engineers? Sorry lol
That if you use a strong pink noise signal to compare differences between 2 processes ie plugins or gear and it results in a very quiet null (spectrally around -80/85dbfs) then there's absolutely no point in doing any AB tests cause it proves that any musical signal will have completely inaudible differences
Hmm I don’t trust myself with these type of test. But I put my confidence in you. Have you done this test with Tube-tech Cl1b plugin vs hardware?
No not yet
Some of the music played on the radio sounds like noise ¯\_(ツ)_/¯
Now I've got an image of somebody listening to 'noise FM' in the car sitting absolutely loving life... "6.5db slope... Beautiful 👌"
🤣
But transients at different frequencies though ...
What do you mean
@@PaulThird pink noise doesn't have transients, so how are you seeing how it's reacting to them?
Yes, and span show them via the avg + max preset.
The max shows your maximum transient level for each frequency.
So when I match the pink noise to the loudest frequency peak i.e. The loudest transient, which due to the sound intensity of musical sounds is always going to in the lower frequencies of the signal, it shows you the maximum possible difference of that process because there's no louder transient coming from the acoustic source due to the sloped properties of acoustic sound due to its sound intensity.
The pink noise null shows you that by matching the max frequency peak there will be no transients that exceed the pink noise due to the slope of acoustic sounds. The avg and max preset proved that.
Pink noise is essentially a "constant transient" which makes no logical sensenin theory but for this discussion it's basically the maximum possible dynamic range that a frequency can make in an acoustic source at the same max frequency peak. So pink noise it's a constant steady dynamic. So it's basically a constant transient across the frequency spectrum.
So when the loudest max frequency peak is matched by pink noise no acoustic source will provide a louder null in any frequency.. Because pink noise is sloped in a way to give the maximum possible ceiling for each frequency in relation to the sound intensity of acoustic sounds. So there's no transient of any acoustic source that can surpass it. The kick test proved that
I’m sure people watched completely drawn into the sense, putting down their toys and listening…
But sure enough the video will end and all that information will be lost in the clatter of their own thoughts after and the same taking it personally to a debunking of a companies plugin they have nothing to do with and the comments will flow in yet again… and you’d have to make another video having to explain things in even greater depth… because you can’t wake someone up that pretends to sleep, and they can not understand they are pretending to sleep because it’s much easier to fake it by believing it and they then fulfil their own narrative.
This is the same as what’s going on over on the Glen’s channel with guitar tone arguments.
I myself learned long ago how easily our ears can lie to ourselves when I would have a plugin bypassed without realising it and it would take quite a few seconds to realise the knobs wasn’t doing anything… but I swear I could hear a subtle change as I turned the knob… it was only after a few seconds before I realised the change should be much more before I realised.
We have to take a step back from what we think we are hearing constantly with creating audio.
I have been making sure I splash myself in the face with theoretical water ever since but some people seem to want to run from the truth in anyway they can… “okay, so you proved this much, but what about this bit of information you left out that might have a little gap for us to run through for you to have to come running after us to save from being ran over”
This is why Religion rules the world but never seems to have to progress like phycology and science does
It's inevitable that many in the audio world will deny hard science but for me anyway I like to have work that I can just copy and paste into a comment and leave it there.
Anybody that tries to do the whole "were are the real audio examples?" or "dude we use our ears, not doctor and spectral analysers" in a situation where I've shown an inaudible pink noise null
I'll just send them this video and let them argue with science.
Truth is they won't argue back, but doesn't mean that they take the science on board either. They just simply don't have the argument to counter against it and that's more than enough for me
Music is subjective, I might think that walking on a gravel path as music as I listen to the rhythmic crunching under My feet.
🤓🤓
I a honestly say: I did not understand this video. 😂
i do know though that a lot of plugins are BS. I also know that saturation for instance is distortion of certain frequencies that although not adding volume they change the sound.
I also know that you can get changes in sound that frequency analyzers can't measure.. why are some neumann microphones so good for instance : because they add a certain perceived room/3d "sound" to the recorded audio (and other way more geeky things).
point being: I always level match and null test after adding a new plugin to make its not just fooling my ears
ps: love your channel.. and your brain.
🤓🤓
🤓🤓🤓
Remember to check out my autism channel if you want to learn more about my life 🤓🤓
Does that mean Justin Bieber is music as well? 🥺
Yes Justin Bieber music is sound as well haha
@@PaulThird Ok, 'sound'.. just 'sound' for this one 👍
Sorry to break it to you, but 0,6 dB is very audible. I'd get this one correct 10 out of 10 times and I'm not even trying. Important: When I was mixing on headphones I always got the kicklevels wrong (more so than now). Nowadays I attribute this to not being *able* to judge transient-loudness adequately on headphones. I'll spoil the correct answer as an answer to this comment, so don't click unless you already voted.
Wether it still makes sense to use *actually* dynamic music when doing nulltests is something I'll have to think about when I'm not so tired. First thought: One being more complex than the other and thus creating a "bigger impact" to the original, doesn't mean the complexity is of the same kind.
(I edited out spelling mistakes, but I did not edit the answer where I spoiled the result. So don't think I cheated!)
Tbh one of these tests can be easily cheated but it's the saturation test that interests me the most as it can't really be cheated.
What interests me is if plain level difference is more audible than actual difference 🤔
Suppose it depends where in the frequency spectrum the difference is. Too much science for one day but id disagree that 0.6db is very audible. I think very being the wrong term to use in my opinion.
"Very" would tell me that it's clear differences that any Tom dick or Harry could pick out.
All of my family couldn't tell me. Not one. 5 people just this afternoon haha
@@PaulThird Very audible to me and plenty other engineers. I don't have great ears. My left one has issues in the low-end. I've always been very self-concious about that. I felt like I picked the wrong profession. Beneath 0,2 dB is where I know it gets messy for me. With transients it's even more audible in context (full mix) because it "pokes out" more or less. Easier to judge it in relation.
I commented under your poll which one I picked for the saturation difference - which was a lot more tricky. I had to actually listen closer and my mind might still play tricks on me. The saturation difference is a tonal difference (that also results in a slight difference of perception of loudness - as do ALL tonal differences - of course). So no, 0,6 dB loudness difference is not always audible when the kick is DIFFERENT, but when it's the same (as in no further processing) it's always audible.
So 0.6db in level is very audible in your opinon but 0.6db tonal differences in a signal is less audible? Have I got that right? Tbh it actually makes sense if that's what you are saying
@@PaulThird 0,6 dB change between two otherwise identical signals is very audible (to a half-decently trained ear). 0,6 db tonal difference CAN be very audible, but doesn't need to be. Depends a lot on the ear, the scource and the focus. If you notch down 0,6 dB around 5 kHz in a vocal with a regular Q, you'll certainly hear that (as an audio engineer).
Didn't read the whole comment, but regarding "0,6 dB is very audible" -it varies from person to person. I proved this to myself in a studio full of people.
I just realized your name isn't 'Paul Turd' as they say on other web sites I visit, learn something new every day.
Yeah definitely Paul Third 😂
Your assumption that 0.6dB is inaudible to trained ears is terribly wrong, as any mastering engineer can confirm. In mastering, 1dB is considered a large EQ cut/boost. We tend to EQ in fractions of a decibel changes. We had to go through years and years of ear training, but having those ears is the bar to cross before becoming a mastering engineer. It's also why we obsess over expensive monitors, AD-DA, etc.
0.6db difference between 2 files. Not 0.6db boost or cut which would actually result in a much more audible null
Nearly 90% failed the 0.6db test I made while 60% didn't even guess and admitted they couldnt tell which was which .. Explain that
I am very sorry that I did not even understand the essence of the video, unfortunately I am not a native speaker. although I try to understand, I used the translation for the second time when I reviewed it and still did not understand what the video was about.
It's talking about why achieving an inaudible null with pink noise renders any AB test on acoustic sources pointless as an inaudible pink noise null proves that there are no audible differences between 2 processes.
Isn‘t this a kind of a ghost debate? The comment shown at the beginning is actually rather referring to „music“ than to „real audio“. And while noise is real audio, it is not music.
The debate is about whether musical audio examples are needed even though scientifically you can prove that there are no audible differences between 2 processes using just noise as the source.
What I've shown is that you don't need musical examples when you have an inaudible pink noise null, musical examples wont give you any more audible differences.
So when you have an inaudible pink noise null and somebody says "yeah but try it on 'real' audio".. Its a pointless comment as musical audio and noise are both audio. It's still the same thing.
Its like oranges and lemons. They are both different kinds of fruit, but they are still fruit. If you squeeze both they'll give you juice, which will taste different and be perceived different but it's still fruit juice and if you were to say test the acidity levels of fruit and we tested the most theoretical acidic lemon on the planet.. Then why would we want to test a random orange when we scientifically know that oranges have less acidicity than lemons?
"cause lemons aren't a real fruit?? Nobody actually drinks lemon juice and orange juice is the most popular fruit juice on the planet"
That's essentially what people are debating about but in audio terms
@@PaulThird I'm not saying anything is wrong with your explanation, but some musicians just want to hear things instead of doing math, That's all. :)
I get that it's just that what I'm saying is that when the science proves there is genuinely no difference to hear, there is genuinely nothing to hear so as much as people prefer to listen to things than do math, the math tells you WHEN you should actually be listening.
The human brain is extremely wierd and can make you think you hear things that you actually are not hearing and that's why a combination of math and critical listening, for me anyway, is the best method for discerning differences.
No point in listening for stuff you actually can't hear in a blind test
The only people that complain about the effectiveness of audio tools are the ones who don't know how the tools work.
Noise is not music?
IDK, as an audio engineer, you get so much shit from people that dont know what they are talking about, i just stopped to pay any attention to them. My best one: "I don`t hear a difference. This is all esoteric bullshit. I am sure if you make a double-blind-test, you will see that all those companies just scam you." To wich I replied: "We are engineers. We measure shit." Either you are "too esoteric" or "soulless technical" etc...
Yup, any audio signal can be measured. I think when you understand more about the science (i just scrape the surface) you start to realise just how much the human mind has an effect on your audible perception.
All it takes is the idea that a difference may be there to make yourself think you actually hear it.
Even last week I hooked up my dx7 pro+ to the apollo x6 via coaxial spdif and I was sure I could hear a lack of high end compared to usb. Done a quick pink noise recording through both and sure enough I was right. Checked the connections and the toppings connection wasn't 100% in 🙈🙈🙈
Pushed it in a bit more and I was like yeah.. This is better but I still felt I could hear the slightest change in the top end. Was going nuts. So done another recording and theres no difference where I thought there was. After that.. I can't hear it anymore haha
The human mind is such a temperamental beast and its genuinely the cause of so many arguments in audio.
When you stick more to science you start feeling less nuts haha
thanks for making this and for your informative videos. careful with spelling: perceive, theoretical…
Tbh it was all scripted without spell check. Re-done quite a few times. Spelling doesn't really bother me
did you just say "math"
How is noise not real audio? It sounds just like what comes out of the mouth of the teenagers at home. :P
My house is constant noise with 2 young girls and a border terrier haha
@@PaulThird Lol! Same here but, we skipped the terrier.
comlexity.
🫳🏻🎤
Songs first... The rest is just a push and pull, tweak and twerk.... But , also it shouldn't start with a drag and drop. Mwhahahahahaha
Play it, Own it... and the rest will follow..
Sounds like a guide for picking up chicks at a nightclub 🤣
@@PaulThird Worked for me...... Mwhahahahaha
@@PaulThird To Add.. Strangely, Sometimes its better to think that way.. Because Maths/Code/Daw's dont make a song.
Yeah sorry man haha but this doesn't do a single thing to support your theory here. Not gonna waste too much time getting into it this time, but the point you're making here is basically saying that "sound is sound, it doesn't matter what the source is, it all reacts the same way" and that literally could not be further from the truth. If that were true, there would be no difference between subjective and objective mixing. There are things that are "proper" and "correct" (that are objectively correct, they follow the rules) then there's thing that are only correct in THAT specific setting, which in any other case would sound awful. That why engineers have pages of different EQ's and compressors, they all have their different flavors and are suited for their own specific things. In none of these videos have you mentioned anything subjective at all. You're 100%. focusing on the objective science behind it and sorry, but what makes music "human" and relatable is that is is NOT an exact science. Yes, the way sound acts can be measured and treated as an exact science, and that can help us make creative decisions, but its only there to inform us and all rules are meant to be broken, THAT is what makes music listenable. I'm sorry man but it REALLY seems like you're just a negative Nancy haha trash talking every thing people hype up like "oh you fools, I have THE answer." Sorry dude, there is no one answer, and everything your saying (while backed up by science, technically) is still only YOUR opinion and YOUR experience for YOUR music. And honestly it. just seems like you're gaining success from being a trash talked and getting all the other trash talkers on board. Which is DETRIMENTAL to the music community as a whole. You want to show peo[ple alternatives to overhyped plug ins and show how they're basically all the same and there's no "game changers" coming out recently? Well that would be awesome, I'd subscribe and watch every single video. That's what I was hoping for when I first clicked your vid. But all you seem to want to do is bash other companies. In your last video I commented on you didn't offer a solution UNTIL I commented THEN you said "gsat is free." Prooooobably should've just made a video comparing it to Gsat+ in the first place, then talk about how they're essentially the same expect goat may even be better. Instead of just being a hater and bashing things you don't like haha just imagine how successful you could be if you weren't essentially insulting 2/3 of the music world (waves users) with these videos and instead making an unbiased comparison with real world examples. Judging but the fact this video exists at all, I'm one of may who feels this way. Seriously man, stop doing videos just to bash other companies and start doing things in a way that makes people feel accepted no matter their choice Instead of declaring everyone an idiot who doesn't agree with you, and you'll even win ME over. I love the fact that you're diving so deeply into this scientifically, seriously I wish more people would. There are far too many purely subjective videos out there, that aren't quantifiable. But you have the opposite problem haha it's like you're just trying to use science to make others look stupid for their choices. Trust me, I've been there. You'll get a LOT further when you learn to balance the two (subjectivity & objectivity) and when you stop making bash videos. That last video (part one of this) you dint offer a single solution or mention that Gsat+ one that you like a SINGLE time. Literally the entire video was about how stupid waves is. When honestly it's more like their ad campaign was blown out. You're smart & you've got potential dude, it sucks to see it wasted. You could be like THE dude for good comparisons if you dropped the nonsense. I don’t know, maybe I'm literally the only one on earth who seems to like fair comparisons rather than just finding a video with someone agreeing with how much you hate something. But then again, look at the USA 2016 election. It was 100% won by appealing to peoples anger against the party in charge. Seems these days, that's what the masses go for. Seems like you're unfortunately using the same tactic. Sorry man but this is precisely what the world needs less of. I'll keep a look out though, hopefully we see some improvement and positivity (I'm sure it will = money in your pocket as well.) Best wishes man.
Your entitled to your opinon. It's good to know at least that I'm not the right content creator for you so you don't need to be dissapointed in the future, which you will be if you carry on watching my future content as I'll be negative when I feel the need to be negative. Going by your take on this video we'll never see to eye.
Tbh I Couldn't really give 2 shits about who gets a sore arse about content I've made. Stopped trying to please everybody a long time ago. I make content I feel like making and stay true to myself. I don't want to the best objective reviewer that can keep everybody happy.
If I'm ever negative nowadays it's because I've found something misleading or broken that nobody else is talking about. If I'm overly negative it's because I feel the youtube landscape is overly leaning way too positively on a plugin that has issues that are purposely are being ignored or not properly tested.
BB tubes had over 60 positive reviews. Want to know how many critical 'negative' reviews . 2. One was mine and another was a 2 and a half minute video that barely got any views.
Just cause I get lots of views doesn't mean I'm not in a tiny minority. UA-cam is awash with videos leading people to spend money. People work hard for their money so in those cases where the landscape is fully tilted, I reign it back hard if I find a lot of overhyping and issues. and that's what gives the objectivity to the landscape. Not objectivity to my video, there's hardly in those cases, but objectivity to all of the community which is over saturated with... BUY THIS.
You disagree? Then that's cool, but it won't change how I do things on my channel from time to time. Which is time to time, not every week.
I was dissapointed I had to make the waves video cause I hate being called a biased basher when I deal with one tool at a time. Everybody forgets when I did the waves kramer pie test and showed it nailed the hardware. When I made the 5 waves plugins I use video.
The community only sees what it wants to see. People blasting tiny amounts of Negativity in a landscape filled with weekly shilling, NFR's, affiliates & overhyping is absolutely crazy.
You need the Negativity to counter the completely one sided positivity. If there was actual objectivity in the youtube landscape then I wouldn't feel the need to go so hard. This has turned into my part time job. I study and observe the landscape. I see exactly what's going on. I know the averages, the stats and where a product is leaning on youtube.
I've actually made an effort to not be so constantly negative, thus why so much educational content has been on my channel.
The reason negativity and criticism is gravitated towards is because there isn't actually that much of it in our field because of money. Viewers see past shilling and respect those who try to save them money and think twice.
What do I gain from 'bashing' videos? A lot of hate and a slight rise in ad revenue. I could easily do one sponsored video and earn 4-5x what I did off that waves video.
Theres nothing it it for me. I stopped caring about subs after 10k. I do those videos cause I'm genuinely passionate about audio and Im genuinely pissed that people are being misled and not told the full story of what they are buying into.
Do you know how many people said they regretted buying the bb tubes and that was before even seeing my video? Loads. Emails, dm's. You don't see what I see. You don't see the shit that goes on behind closed doors. The money that changes hands. The politics.
I fight daily from being drawn into the positive world of making money and being positive to constantly sell products.
That's why I'm very picky about my sponsored videos.
I don't expect you to know or comprehend everything that goes on behind all the lovely positivity out there but I can tell you right now that id rather quit youtube all together rather than lose what I believe in.
I'm one guy. One guy who everybody for some reason seems to get their pants in a twist with. I don't get it. I'm not a huge business, nor do I have a huge amount of cash flow behind me. I'm just one guy with science and an opinion vs hundreds, maybe even thousands of youtubers all selling subjective positivity with no real criticsm or hard science behind anything they do.
Never forget that. One guy. Just one guy doing this part time while juggling a full time job. I'm not this big threat that loads like to make out. it's just that I see behind the bullshit and stand out by being different from the rest of them.
So yeah, don't bother replying or sticking around cause if need be I'll come back and do the same if I feel the landscape dictates it. I'll never stop being who I truly am.