More About Mix Knobs (phase shift, oversampling and so on)
ฝัง
- เผยแพร่เมื่อ 19 พ.ค. 2024
- In which I investigate the "Softube phase shift issue" that someone commented about, and (spoiler alert) conclude its not really an issue.
That last parallel compression video:
• The Correct Way To Do ...
The Gearspace thread:
gearspace.com/board/music-com...
If you like this type of content and you want to see it more often, consider signing up for Channel Membership: / @danworrall
Or you can just buy me a coffee: ko-fi.com/danworrall - แนวปฏิบัติและการใช้ชีวิต
I'm working in video games and the film industry because of this guy. Thanks for bubble popping and clarity in a very muddled industry, Dan. Your videos, combined with internships and work to be done, have helped push me further than you can imagine. You're a legend.
That’s really cool! Whats the process like for finding work in that field?
@@eriklucasmusicmy answer would be - don't, just don't
He has internships?
@@JosephHenryLuckinbillop just listed what helped them
Lol “do it as a hobby”
But very cool to the OP engineer 👍
Your self described “low effort” videos with you discussing something without a script is still the most information dense and excellent sound videos on TH-cam. Loving the consistent videos
Even the low effort videos have a script. If you sign up for memberships and watch my actually unscripted mix reviews you'll understand why.
@@DanWorrall I think its more how more candid these videos are. Its like one is going to a lecture and this is more talking to the professor after the lecture where you realise there was so much more you didnt know that you didnt know you didnt know. Thats the best analogy I can think of. Very much like both styles of videos though.
i forgot to enable notifications from the best mixing engineer on youtube!!!!!!!!!! enable notifications people! click on the notifications tab after subscribing and select ALL.
This is the only TH-cam channel where I put a like on the video before I watch it. And I never un-liked once I'm done with it.
one of your videos on oversampling convinced me to mix at 48k a few years ago and I have not regretted it one bit
same here lol - IMO it really does make an improvement to the sound quality of just about any mix really, and it doesn't even really affect my CPU headroom or take up too much extra disk space really. Admittedly that quality difference is very subtle and probably wouldn't be noticeable to most people, but it's there and it really doesn't hurt to switch to 48k if you have a decent modern computer, so IMO it's definitely worth doing.
Don’t mix at one bit
if you follow Dan, you know phase shift isn't really a big deal most of the time
I was watching a guy talk about working with bob rock and he was micing the floor for a guitar amp and how nobody would ever do that cause the phase would be al messed up
there’s a difference between time alignment phase shift recording with microphones and phase shift from frequency equalization using a plugin
@@johnviera3884 and either one can sound good or bad depending on the situation
When I did a deep dive into ableton-dubstep-heads, I understood that there's no reason to even think about it. Those guys filter everything so hard, but still the final mixes are loud
As long as it sounds good... right?
The Vocoder! 😄
reminded me of vocoded potion seller
When Dan asks "What sorcery is this?", you know you're about to learn something.
Great analysis. For context, we approach these design issues slightly differently in our multiband compressors, albeit with similar objectives:
+ We process the dry/wet mix separately inside of each frequency band, where phase issues from the crossover filters present less of a concern. Some of our plugins have up to 64 separate dry-wet mixes going on internally - luckily this is fairly cheap in terms of CPU.
+ We avoid using IIR filters for crossovers, especially in the low end. They are not kind to signals with sharp corners, and you can hear it. FIR crossovers however can be made to give "perfect reconstruction" of the multiband signal (absent heavy processing), even at much lower latency than that required for linear phase.
+ We generally use linear phase oversampling on the high-frequency bands, and (cheaper) IIR oversampling for lower frequencies where there's more breathing room between the signal and Nyquist.
10:18 as a plugin developer I can confirm that our voices are just like this, always!
Oh yes! Cableguys seem to know what they're doing with plugins. Some of the smoothest, most well engineered plugins I've come across.
Same as FabFilter plugins. You virtually never see any major issues with them either. Well, except for Twin 3 having its compressor post the output gain! Something I thought they would've designed differently. I caught myself out there, setting the compressor then changing the output gain, which only drove the compressor harder when I increased the output. Had to think for a minute what was going on until I realised the compressor gain reduction was increasing.
Thanks for yet another in-depth, extremely interesting video Dan. Really appreciate them!
seriously shaperbox is the best plugin ever made, ppl need to study that plugin more, it can do sooooooo much
Hey Dan. Would you ever consider doing a mix breakdown of some of the work you've done? Wether that would be on your own music or music that you've mixed for others.. I for one, would love to see that 🙂
Definitely!
I would pay for that. Maybe he should team up with Mix With the Masters or ProMixAcademy. Or sell it on his own site...
I agree, I would love a mix breakdown of Impostor Syndrome
@@alexeytereshonok I'd pay for that too. Perhaps he could do workshops in studio too
You caught the Swedish accent pretty much perfectly.
Dan's making up for lost time! Love it!
Very useful. Thanks Dan.
First video I've seen of yours, Dan, for an age.
Not been getting notifications, as, it appears, TH-cam didn't want me sub'd to you any longer.
Bizarre, never been ghost un-sub'd before. All fixed now though, and ELATED to be back on the DW firm.
these are always SO enlightening lifts those little worries all up!
video number 3 of asking for a clipper video and why it is and it isn't just compression
In your "Fifth Decade" meaning not yet 50?
You're a baby!
"depending on the needs of the algorithm" - ah pure sanity, the most valuable trait human being can have.
as always, great and straight to the point.
Thank You!
I love the way this plugin sounds... adds nice dimension, warm cozy coloration...
and doesnt sound too expensive.😂 grateful to geta copy while they were giving it away!
thanks SOFTUBE, thanks Mr.DanWorrall
It's snappy i like it on drums with a 3:1 ratio
@@TransistorLSD nice.. thanks for sharing that! that actually really inspires me bc im usually ignoring
@@WACkZerden Well, i also ignore
You're the best, Dan
I'd never understand this type of complaints. If the instrument built this way use it as an effect or a part of a sound then. And if the result isn't what you wanna get just use different one. There're plenty of SSL-type compressors on the market. Which is the right/better one? Does it even matters? Just pick one that'll give you the result you want and don't even bother "is it mimicking hardware the right way or not".
Yet I'd never be tired of those "professionals" who started getting concerned about some effects the plugins might bring to the sound only after they have studied it in Plugin Doctor or such thing... It's been very OK before and "impossible to work" now. Lovely!
Thx Dan,
your rhetoric is as skeptical as it is profound !
😂😂
Great demo on that.
As a final take, we now know Softube engineers have been replaced with AI. 😮
Not sure how much of that I understood but as a big Softube user it's nice to know there's some very clever software under the hood. 😎
not really it's basic stuff tbh, well at least this example.
nothing wrong with that tho
I hope you were using the Softube vocoder for that dev’s voice. Brilliant.
Yes he did use that one. You can see it in the background, ignore the text and watch closely 😮
Very interesting that you low pass the overheads at 10k. I may try this. I've heard it mentioned somewhere before as well.
When i do this it almost sounds like i used ribbon mics for overheads during tracking. Warm. Me likey.
@@TransistorLSD What filter do you use? 24db or 12db? I have tried it once or twice before, but only for a second or two. Sounded a bit too dark and I wasn't as low as 10k either. I've heard it mentioned before alright, but thought it might just be internet nonsense until Dan just mentioned it. May try it again so .
@@bugeyedmudafuka2 12 db/oct at exactly 10k. BUT... It depends on the mic used during tracking, actually. Some of them are already dark enough, even if it's not a ribbon. And also your chain down the line... If your mixbus has a 2db highshelf, you might want to filter the highs from your overheads, but if you have a warm Tape plugin which dips the highs... Well... Sometimes you might end up actually boosting them on overheads.
Fascinating video as always, much appreciation for the refined balance between breadth and precision! I have one counter to the notion of the 18kHz notch being unnoticeable, which is simply that stereo information is concentrated towards higher frequencies, and a notch might impact stereo image even if the character of the sound is indiscernably affected. But then, as you mentioned, simply switching to 48kHz sampling or using native mixers solves the issue anyway.
I don't think stereo information is concentrated in the higher frequencies. What makes you think that? Midrange, that's where stereo is important, IMO. Especially lower midrange.
@@DanWorrall I definitely agree that many primary creative decisions regarding stereo placement should be made in the midrange. The thinking goes that lower frequencies have such large wavelengths (e.g., 343Hz has a 1m wavelength) that only the broad pan levels are discernible. My view therefore is that it's actually the harmonic overtones that provide precise stereo information beyond rough placement. Diffuse the top end of any signal with too much reverb, and the stereo placement feels washy, even if the low end is unaffected.
Much of this is obviously subjective, but I followed some of those principles to build a very cool VST that lets you *rotate* a sound in the stereo field. Even with a hard pan, the stereo image is perfectly preserved, rather than flattened into the left or right channels. That must sound strange, but I'll post a comment when I finally publish a demo! JUCE and C++ are doing my head in...
@@sentinelav grab a 100Hz sine wave and pan it around: you can hear that right? You don't need a full wavelength of difference between your ears to get directional information. Your plugin sounds interesting.
@@DanWorrall Oh totally, it's less about needing a full wavelength of difference and more about how high temporal resolution translates to high spatial resolution. As a quick experiment, compare a track with a high cut, and a low cut, at 300Hz. You'll certainly get some broad stereo direction in the < 300Hz range, but almost all the detail lies in the top end. The shorter wavelengths translate into a higher resolution spatial field. Abstractly, this also ties to image processing, where all the detail of an image is in its high frequencies, and the low frequencies define broad shapes.
@@sentinelav I'm suspicious of comparisons to visual processing, our eyes and ears work very differently. But I know that I love stereo width in the lower mid range especially, and if it's mono above 10k or so I don't really mind.
New Dan Worrall video!
First I like it, then I watch it.
Hello Dan! Record a little mixing session for us, please.
Yes that would be awesome.
SOFTUBE!!! GIVE DAN A LICENSE FOR GERMANIUM COMP!!! PLEASE!!!
Germanium has a huge boost in the subs, almost unusable comp for me because of that reason but beside that it sounds great and adds a nice color!
@@huberttorzewski A big boost appears only if you use "Drive" and "Feedback" knobs on the right side, if you turn them down & compensate with clean gain, this boost dissappears completely. And from what i understood, this boost is actually a feedback loop... So it makes the subs not only louder, but longer... Way longer... Booooooom..... Boooooooom.... (it makes the kick sound like that)
@@TransistorLSD Interesting, I couldn't find the proper setting to eliminate that boost when I was testing the plugin 2 years ago. Yea, kick drums were sounding horrible after putting Germanium on the drum bus, very boomy. But anything that doesn't contain any subs sounds great through this plugin
I love the bump and particularly grab that comp when I want it
@@huberttorzewski I just tested, and you need to set Drive and Feedback to 1 and compensate +22.5 db of clean gain, this will give you a flat-ish response with unity gain if no compression is being applied. You don't get the subs, but you get a very gentle highshelf, like +1db at 10k.
And then we have things like Melda's free MFreeFormPhase which allows you to set the phase of individual frequencies...
ppl need to totally chill on phase stuff
Recently dealt with this on the free Voxengo Tube Amp saturator/distortion, kept getting a weird notch/comb on render when using it on a parallel distort send until I found the linear phase os filter option in the settings!
Are all oversampling features created equal?
For example is 4x oversampling on my stock waveshper equivalent to 4x on my Analog Obsession pre amp?
I often want parallel saturation (with minimal aliasing) and have to use Patcher (fl studio) to run 2 copies of say, the TRANSATURATE plugin, one wet and one completely dry to match the oversampling and eliminate the phasing in the high end.
The 4x bit is the same, but the up and down sampling filters can vary a lot. Biggest difference is linear vs minimum phase, but also the steepness, tuning etc. Some developers set the filters lower / steeper to kill all aliasing, others set them higher / gentler to kill enough aliasing with less impact below 20k. The best approach is largely a matter of opinion.
Damn, as a developer I now want to communicate with the plebs solely through a vocoder.
And we will be communicating with you through a Korg MikuStomp pedal.
Effortlessly a genius. The vocoder twist took me out.
Hello DW sir Plz make a video about waht will happend when we use linearity on any insert and master at the same time... Does it make polarity reverted of insert in master?
My ears encounter a lot of times some 'fase shifting' (problems) when I am sending to a parallel Chanel.
I always fear and listen well for those combined " latency / fase " issues.
Thanks
edit: but its not in the upper highs where the difference seems the most, but i hear some wierd fasing or difference in the mids
Interesting idea: sidechain filter in compressors causes phase shift, it means the compressor “sees” the sidechain signal delayed from the main signal. One option is to try to use linear phase sidechain filter and compare the sound. Another option is to apply all-pass phase rotation to the main input matching the phase shift of sidechain filter for sidechain input.
Hi! Sidechain in compressors is a very interesting topic. I dont really get why people say that sidechain oversampling has a dramatic effect on a compressor, it is a little counter-intuitive. Idea for your blog in vk 😉
Phase shift doesn't always mean delay though, right? And if it does, just a tiny amount: effectively it would just mean a minutely slower attack, right? While pre-ringing from linear phase would be like a tiny bit of lookahead?
@@DanWorrall For example 12 dB/Oct high-pass at 50 Hz gives group delay for the lowest frequencies about 6 ms. For fast attack times it matters. Phase non-linearities always give some group delay because mathematically group delay response is the derivative of phase response.
@@DanWorrall Hmm, if minimum phase filter leads to delayed response and linear phase filter gives sort of lookahead, what kind of filter gives perfect time match for the sidechain? I'd investigate 🙂
@@vladgsoundyeah, I don't understand this. If phase shift is delay, how can a zero latency EQ correct the phase shift from another EQ? Is it seeing into the future to apply negative delay?
I was hoping Dan was going to end with "don't fear the Reaper" but that wouldn't have made sense here.
I saw a video of a Mastering engineer saying they turn off over sampling on their mastering clipper and maybe even limiter? Because they find the over sampling filter to be more audible than aliasing from a clipper only kicking in every so often. I tested this myself using Kclip 3 and I could swear I heard a difference when the over sampling was off versus on, even when no clipping was applied. So is it the shift I’m hearing?
If it's linear phase oversampling (quite likely for a 'mastering' type plugin) there won't be any phase shift. The artifacts will be pre-ringing at the very highest frequencies. I can't hear that stuff, but perhaps my ears are too old? If you want to know for sure do a blind ABX test: my guess is you'll find that you can't actually hear a difference, but go ahead and prove me wrong...
These videos oddly remind me that I want to do this stuff for the rest of my life.
You know nothing, Jon Snow.
The Softube Vocode Crystalizer Choir patch video tho…? 🤔🤔🤔
Dan, wasn't it a bit irresponsible to use the engineer's real voice? He has a pretty distinct voice and can surely now be identified. 🤔
Best thing is to use quality samples with layering to avoid using multiband dynamics.
As soon as you changed the samplerate on plugin doctor my tinnitus was triggered.
Hillarious, maybe you’re into some deep science discovery business
Please do an analysis of Slate's VTM. I assert that the bass bump region of his tape emulation is excessive & unrealistic, even after using the bass-bump adjustment slider
I would also love to see that ! But that thing sounds so good on that default preset 🤣🔥
Dan do u have any opinions on har-bal eq?
It is my opinion that I should try it some day.
Totally figures that the Softube dev sounds like a Brit thru a vocoder. Totally apt
Today it was more of an "I and welcome back"
How is your health been Dan?
Recently, a video popped up in my TH-cam recommended feed. The video discussed how a single plug-in could transform this guy‘s mix. When I personally listened to his examples, the only thing I noticed was a slight volume change. What are your thoughts on plug-ins like the one shown in this video? Plug-ins that say they can make your mixes better by messing with the phase are automatically not getting tested by me. Like you have said 1 million times, in most cases, faze shift may sound good and is not something you should be that worried about. Faze shift is not scary, like a lot of people and marketing seem to think it is. You can do some really cool and amazing stuff with faze shifting. I am blind so I cannot see wether a plug-in is bypassed or not, so I cannot be fooled by seeing a green light turn on, or animated wave forms in a plug-in display. I didn’t notice any change at all in the way it sounded except for what could’ve been a 0.1 DB gain boost. The guy in this video mentions that the low mid end of his mix was boosted, but there was no EQ being applied to the sound, and that this plugging was adding more punch to the sound. Like I’ve said before, I just feel like the sound gets ever so slightly louder, which could make someone think that it sounds a bit more punchier. At first, I didn’t even know the plug-in had been enabled, until the guy said “wow”. Then I knew the plug-in was on, and I could not hear any difference at all. Thought I would share this to you. Any plug-in that says it can make your mix better by adjusting the phase shift is not worth my money or my time installing it unless I can noticeably hear a difference that I like between when it’s bypassed or not, with the audio tests being level matched. thought I would send this to you so you could analyze this plug-in and see what it actually does. I wouldn’t be too surprised if it visually makes it look like it’s doing something, but it’s actually just raising the volume up by a slight bit. here’s the link to the video I saw.
th-cam.com/video/LvrQ_GuUvvo/w-d-xo.htmlsi=we-_EeIsNS0w2ukJ
I was expecting some kind of analogue magic snake oil thing, but no, Auto Align is legit, I use it myself. I think this was just a bad demo really: not a scenario where time or phase alignment would help much. A real drum kit would be a much better test, or a live band recorded in one room, where there are multiple mics and correlation between them, that's where this kind of plugin can help.
I feel I should explain what the plugin is actually doing, especially as you can't see the screen. Imagine you recorded a symphony orchestra, with a main stereo pair above the conductor's head plus multiple spot mics for important elements, soloist etc. In the old days you might send the orchestra off for a break while you measure off all the distances from your main array to your spots with a tape measure. You could then convert using the speed of sound (3ms per meter approx) to work out a delay time for each spot mic, to time align them with the main stereo pair. Auto Align v1 did that for you automatically, no need for the tape measure. Version 2 adds to that by also calculating a phase correction filter: a simple delay is usually not enough to match the phase at all frequencies, so that will generally just get the lowest frequencies in phase, and a correction filter is used to fine tune the results and match the phase over the entire spectrum. So Auto Align v1 massively sped up a process that was difficult and tedious to do manually, while version 2 matches phase across multiple channels in a way that would be practically impossible manually. Key point you may have missed without visuals: he's loaded multiple instances of the plugin, one for each channel, but is able to control them all simultaneously from any one of those instances.
@@DanWorrall thanks for the clarifications. I don’t see how this could “ TRANSFORM” a mix, but I do see how it could be useful like how you described. I don’t think an electronic style mix was the right type of demo material for this plug-in. i’m curious how that faze correction filter is supposed to work. I’ve heard of getting delay times right in between mics. Makes sense why there would be a plug-in for that, and I also see how a phase correction filter would be useful in that type of situation such as an orchestral arrangement. I do not see how this plug-in would be useful in a mix like the one shown in the video. I also don’t know how the plug-in is supposed to realize that because two choir patches are in similar octaves, they’re supposed to go together, and be properly aligned. I bet there’s something more to it that I just don’t know. is it possible you could do a better demo of this plug-in, so I can hear what it actually does? Because for an electronic mix, I don’t see how a plug-in like that is going to really change anything of the mix except for maybe a little bit of volume. I would love to see how this plug-in would work in an orchestral arrangement or an acoustic drum kit. Your videos have always been easy to follow, so if I heard you demo it and put it through its paces, maybe I would be able to get a better sense of exactly how this plug-in changes the sound of a mix.
@@levi_gobin it can absolutely transform a mix: I've had night and day differences on drum kits in particular. But it's not going to help at all with an electronic mix like the one in the video, and I don't think he understands what the plugin does and why. Sound Radix have been my clients in the past, so it's not impossible they might ask me to cover it...
Cool. May be a better demo will pop up in my TH-cam recommended feed and I’ll give it a listen. I think the guy in the video was just imagining the differences. Even if it would’ve made any difference on an electronic style mix, his particular mix was all over the place, with Parts that faded in and out, side chaining everywhere, and nothing for me to really notice a change in. The volume differences I noticed could’ve just been from the actual parts of the mix itself, rather than anything the plug-in did to it. I just can’t tell, and that’s why that particular mix was just not that great.
I saw a video posted by Casey Connor on TH-cam talking about an audio illusion. The listener was asked to listen to two recordings of an acoustic guitar. One was made on a cheap imported Chinese mic, while the other one was made on a several thousand dollar calibrated Mike used in laboratories. The catch was, the same audio file was being played for both recordings. When I first listened to it, I actually thought the so-called calibrated Mike sounded a lot better than the cheap one. I even tested it on a friend of mine, same thing. He said the calibrated Mike was a lot better, even though it was just the same audio file played twice. I’m guessing something like this is going on here. he clicks the button, it does its thing, and he sees a bunch of differences that the plug-in has surely made to his mix, and thinks he hears a big difference in the sound. I honestly feel like being blind gives me an advantage, because I don’t get to see when the plug-in is enabled. He doesn’t even say “its now enabled”. He just says wow. I was expecting this to be clickbatey, but I was also aware, and looking out to see if he would mention how faze shift can ruin amix, because it doesn’t. We all know this by now. 😉 Keep up the great work, and thanks a bunch for all the videos you make. I’ve learned a lot from your videos.
as always the defacto statndard for sound engineering knowledge
Iv been meaning to get around to it but i downloaded a free ai mastering plugin and im gona use your i won the loudness war song as a ref
i still want a tutorial how he did it because it sounds amazing an im not joking
Vocoder 🤩😍
👌👌👌
Dan you've missed big chunks of what is important here - the amplitude response still matters. Elliptic filters have wobble in the pass band amplitude response (not good), and you haven't even shown how much rejection is done by each filter, which is important.
Hi Andrew. It wasn't my intention to compare the performance of different OS methods, just the phase / parallel processing aspect. But I wonder: is that stuff important in theoretical striving-for-perfection terms, or is it actually audible in a real mix?
😂😂 man the voice shifter
I have to mix at 48k because my audio interface works at that sample rate, I am nearly in my 5th decade and can’t really hear above 16k so I can’t even hear an 18k notch
You need to release a plugin that just tells me when I'm about to do something irrepairably stupid while mixing in Reaper 🤣
This Reascript should work.
reaper.ShowConsoleMsg("You're doing something stupid.")
All jokes 😛
I need something like this, but for daily life! 😅
It should say "This is bollocks!"
The band-splitting filters having non-perfect phase at the crossover is lazy implementation. Conjugate mirror filters should ideally be used for that, they gaurentee perfect reconstruction after summation
Sidenote.. a small notch for a slight dry/wet mix is likely a non-issue. Regular filters are easier to implement, and multiple crossover bands will require some tricks with multiple filter stages
Hi Dan, a besotted sub here. I'd be incredibly interested to hear your take on a recent youtube post by AP Mastering entitled "Shocking results: 16 EQ plugins tested. Do they all sound the same?!" Thanks!
I think he's trying to grow his channel with some clickbait, which is fair enough I suppose. But his tests are questionable, in many cases he doesn't seem to get much of a null at all, and he should probably be testing them in the high frequency range.
@@DanWorrall I think his point should've been "Different EQ's are much more similar to each other than different compressors and saturators" instead of "All EQ's are exactly the same".
@@DanWorrall Gotcha - thanks Dan.
I can't figure out, why people complain about that issue. Fact is, correct me if I'm wrong or too stupid for this, that you can't get over Nyquist. That's a law, no matter what your sample rate is or where the plugin starts to dip in. Yes, you get a phase shift, if you go beyond the Nyquist level, but that is something, a compressor tries to avoid at any circumstance right?
Oversampling is done to prevent aliasing distortion (distortion which folds back into audible range once it goes above 1/2 of your sample rate in Hz), so in theory you can easily avoid having a phaseshift just by avoiding oversampling altogether. But then you end up with aliasing distortion in the audible range, below Nyquist.
Yeah, I haven't heard 18k in..... awhile, seems like a non issue.
What's left unaccounted for is why you'd ever want to mix the output of a compressor into a parallel dry signal? If it's less of a wall you want, just turn down the compression ratio.
It produces a different sound compared to just turning down the ratio. Some people like it (Andrew Scheps, 90% of his compressors run in parallel), i personally don't.
Conclusion = track with hardware or dsp plugins ^^
i see dan i click
yes!
Si!
Ja!
Да 🗿
He said knobs!
Tell that to whoever produced R.A.M. to filter the cymbals at 10k given that in the hi-res version they pass 30k... And between the two versions the difference is palpable, even for you given that those frequencies You don't listen to them with your ears
What is "R.A.M."?
@@TransistorLSD 🙄... Daft punk last album
@@ziofrenko Yeah i realized a bit later, thanks.
I don't hear 30k with my ears, that's true. And I don't smell bullshit with my nose ;)
@@DanWorrall Various experiments have proven this for at least 40 years... Why do I hear differences between the same drums recorded at 96khz and the same files brought down to 48?! I await your news on frequencies we don't listen to...
This video is the reason why nerds never become good producers.
Yes, perhaps they can develop useful VSTs.
But OCD stressing over things that don't matter in a musical context is time not well-spent.
More on the music, less on the technicalities.
Just have your engineers only use linear phase oversampling methods, duh
This is the stuff I love. :)