I love Mixbustv too. But I agree with you. The issue for creators isn’t the dac conversion so much as the conversion to mp3 or m4a and to streaming because that part we DO have control over. We clearly have tools to check it before release. In logic is a plugin called roundtripaac just to check how Apple will change our file during conversion. Now there’s also adptr streamliner from plugin alliance, to hear how things will sound as well. Anyone who saw only the first half of David’s video will not have received the most important reason we DO care. I wish he said that first instead of later. I’ve done testing too, and it’s clear to me why -1 is a much safer bet not because of a dac but because of digital conversion. I hear lots of crap on KZfaq on my studio system. My dac is good. So, loudness at the expense of more distortion via too much limiting OR ISP is simply unfortunate no matter if we think it’ll be inaudible because it’s buried under iPhone speakers or not. I’ve had big name mastering engineers send me back masters with over a thousand ISP at -8lufs. I honestly don’t call that a master. And Apple digital masters all request -1db peak anyhow, and those even maintain 24bit! Except mastering charges extra for that. So essentially they charge extra to do it right the first time. Anyhow, glad to see your comment. Cheers! (From Barrie north of Toronto)

@@jorykevinberger7047 Nice to meet you fellow Canadian friend! 😁👋 (I live in the suburb city of Mascouche, on the north shore of Montreal). Yes absolutely loving the channel of David and the fact that he shares his incredible knowledge!! I totally agree with what you say here. There is something so simple though, I do not know why nobody does this: master as loud as they want (ask the engineer to push the limiter, even to 0db if they want), and then, before uploading to the streaming platforms, simply reduce the gain of the WAV file by -1.5db, even -2db! But why nobody talks about doing this? This is a very simple solution to keep your loud "sound" without any clipping added by the digital conversions processes done by the platforms and without any effect on the playback loudness as most streaming platforms normalize down to a level to around -14 LUFs or so (with variations on the calculation method though, sometimes with huge differences in gain reduction from one platform to another, I love using the loudnesspenalty website for this), so the solution is simply lowering the gain of the WAV file prior to the upload. Sending a "gain lowered" master to the streaming platforms is a very simple and seamless solution to avoid all these problems and still allow for the same "loud" mastering by the engineer. But now, since nobody does this, it's clipping and distorting all over the place because of the digital conversions combined with a too narrow headroom in the UPLOADED master. Glad to read your comment as well! Always nice to discuss with great passionate people who still think that great sound has its place in music! 😁

@@themagicianofsound that makes so much sense it’s painful I didn’t think of it myself. Thank you for suggesting a simple and elegant solution that doesn’t require over thinking.

@@phadrus Thanks for your comment. I know many engineers say "There must be ONE master and we should only upload this ONE master everywhere", I do not say they are wrong, but since many non-professional artists manage their uploads themselves using distributors like DistroKid, why not simply give them 2 masters, one that goes up to -0.1db if you want, for SoundCloud that does not normalize (because the artist may want his sound louder than all others or at least as loud as the others), and another one for the streaming platforms that normalize the volume, for which you reduce the gain and keep it let say at -12 LUFS with a headroom of at least -2dbTP (by the way, Spotify recommends a -2 dbTP (True Peak) headroom for any loud music, this is not me who tells this, this is Spotify, so why not respecting it, when we can?)). Especially in the non-professional world and this is important because a big percentage of the mastering engineers like me are NOT professionals (although I also work with professional artists too) and their reality dealing with artists is completely different as it's more a "friendly" relationship than a business relationship, so I believe we can adapt the methodology for these non-professional cases, in my opinion. Maybe I'm wrong, but me, I do this, and it works fine, all my artists are happy that I care about their music and that I inform them about the subject. You simply can manage it in 2 uploads: one for the most important streaming platforms from which downloading the digital mp3 or wav or m4a is not possible (ex: Spotify which is a streaming-only platform) and another upload for the other platforms and for that second one, you can use your loud master because your main focus is the loud sound, more than no distortion at all. At least, you will avoid the distortion on the most popular streaming platforms (and more importantly, you will follow their recommendations in terms of headroom in dbTP) and on the other ones, since you absolutely want LOUD music and use all possible headroom, then upload the "non-gain-reduced" version there, thus accepting some level of distortion there only. NOT following the recommendations of the streaming platforms is like accepting not following the recommendations in terms of maintenance of your car. You can say: "Nobody needs to change the oil every 5000km, this is BS, you can wait 8000km or 10,000km" and it will have no long term effect. I do not say it's wrong, but also, is it right really? But since possibly 99% of people do not care about these recommendations and the uploads still get millions and billions views that way, I am guessing that it became "accepted" generally to ignore them and so, who am I as a dummy amateur mastering engineer to claim that they must be followed? 🤣🤣🤣🤣 So... talking to myself.... "Shut up magician, and go back into your magic lamp! " 🤣🤣🤣🤣🤣

It depends on your music! If it's EDM or pop, you probably want ISPs so you can get that extra db or two of loudness.

If the tracks were mp3 that's why

I don't like it

SR conversion shouldn't if done well, if it does, is because they use a shitty software. There's no intersample peak happening from the conversion, not from Wav to Wav, not for SR. From wav to mp3 sure. Otherwise think mastering, I catch at 88 or 96, most of us do, it'd be a nightmare. SR doesn't cause any of that

Just always be careful and use ears 🙂

Well, just because others do it doesn't mean it's right for your particular song. You need to consider the type of music, consider the reproduction/distribution platform (if any), and, as was said, use your ears. You need to do end-to-end checking: encode the file (if it will be converted to, say, MP3), reduce the gain (if, say, spotify is the platform), and play it with something that decodes it. Use a meter if possible so you can see what is happening as well as hear what is happening. Sometimes you might even *want* the ISPs to be in the playback. You might decide you need *more* ISPs.

@@briancase6180 mastering for various platforms is stilly. Normalization is only applied om some reproduction media, not on others, it can be turned on or off and tomorrow it could change, and people are left with weak quiet masters. Master as we've been mastering for decades, loud is not the problem, loud and bad is, loud only scares those who don't know how to do it right.

@@mixbustv Hmm, I wasn't saying to be a slave to the current platforms. I was just saying to have open eyes and ears. If you look at the way many big-time EDM producers, for example (and I would say pop produces too), produce for platforms, they master very loud and they have peaks over 0db. But, just because they do it doesn't make it right for your song. It might be right, though, so you need to be aware. And, I think that was one of the points of your video: be aware that we don't necessarily, and probably don't, care about inter-sample peaks. I agree.

Brian Case I think as mix and mastering engineers we don’t have control over the dac someone uses, but we do over how a file will convert digitally, like to MP3, m4a, streaming etc... I think people who only watch the first half of the video will walk away with the wrong information. The important part was toward the second half/end. And I love how David teaches on a normal day.

It depends, my "final limiter" is usually 3 plugins and a couple of hardware units. Mastering and final limiting is NOT what you see on KZfaq videos 😄 members have access to two mastering courses here that show some of that

You click "upload" 😄 no seriously, this video is exactly about that, first of all, those "limitations" are A) inconsistent platform to platform, B) tomorrow they can change (and you're stuck with the quiet masters) C) users can turn it off, D) they are not even there at all on some media. We don't care about loudness, well, good engineers, care about the sound. And this BS that loud sound bad needs to stop: bad loud is a problem, good loud is not and it's also HARD to achieve, that's why everyone who can't do loud and good scream about loudness, unfortunately there's a LOT of bad loud because everyone with a laptop offers mix and mastering services. Read this kzfaq.infoUgzGS8VPzoan0vQ9Ooh4AaABCQ

You gain exactly that, 1db of level

Streaming platforms don't apply loudness normalization (not penality, that's internet twisting things and getting that wrong too) because of ISP, it applies it based on LUFS, you can have an ISP-free tracks that's loud and it will get turned down anyway. So, it's not related. Some platforms ALSO suggest final output level to be -1 but that's because they are worried about their crappy conversion algos to cause problems. Think about it, nobody had these "problems" before the streaming platforms, the problems are not the professional mastering engineers, is the platforms. Also, loudness normalization is also BS, different from platform to platform, and not even enforced on all media, can be turned off by users.. and most importantly, could change tomorrow. And you'll be left with a song that sounds quieter than anything else. Also, I said this many times, loud is not a problem, loud and BAD is a problem, loudness only scares those who can't achieve it correctly. A good loud master will sound denser, more powerful and perceived louder even if they turn it down to the same level as the quieter one. Bad, will sound bad anyway.

eh.. hold it right there: the slight warmth and roundness "nobody can hear" is BS, sorry. The listener don't technically KNOW what it is, but the reason why some mixes/masters move the listener and some don't are also those details coming from those units and/or plugins. And I think, personal and debatable opinion, engineer who think those things don't make a difference, well, let's say they have the wrong mindset to be an engineer. ISP are objectively negligible, because as said in the video, the media where those would be (would, not necessarily happen) have higher distortion to begin with than any amount of ISP. Warmth and roundness or any other word we want to use, is a FEELING, and that, is what and engineer should focus on, and why many of us still prefer hw to reach that goal.

@@mixbustv Based and warmthpilled

No, -0.2 or -0.1

Yep

It's not a big deal as long as you set the ceiling to anywhere between -0.1 and 0.3

@@mixbustv Okay thanks for the reply. I re-mastered anyhow because I'm paranoid! LOL It does sound cleaner now. I had the "vintage limiter" at the end of the chain not set to true peak, I didn't even see that button.

It depends. Logically speaking no, but if you followed the video, you don't really risk anything because they won't be audible. In the video I differentiate, non classical music, pop, rock, hip hop, metal, EDM and all the various subgenres, you won't hear anything, that doesn't mean, as said in the video, that if you want, you can just release a track ISP free. The tracks analyzed show that no, most people, simply don't care.

Dull sound on transients for the most part

@@mixbustv got it. Thanks man

It's not complicated, just disregard it, like most of us do. Master as loud and as good as possible, like we've been doing for decades. Platforms will change, you can't change your master after, good loud masteres sounded, sound and will sound good forever, is that easy. Everything else is a talking clickbait title for KZfaq. Worry about getting turned down? (Which I explained in other vids that it doesn't matter) print a second version at -14lufs, you won't do that anymore when you hear your master being weaker than the louder ones even when they get turned down

@@mixbustv Thanks, I think you are the best sound tech on youtube. I have made all the music I have up now at -14LUFS, and I have noticed my music is too low. I will try and make some hard house tracks and do some loud masters to see what happens. I have made CD level masters as well, but it is a bit tiering and messy making masters in 3 levels and 3 formats. Do you know if anyone care about broadcast-levels for radio?

From the other hand we don't need to go that high with modern digital audio (16 or even 24 bit depth) where the noise floor of the dither is at the huge distance of -90 - -80 dB (for 16 bit) even with decent amount of noise shaping. Especially with modern standards of loudness normalization in streaming services and broadcast which by default will simply turn down loud stuff. Why just not to use this 1 dB of headroom and simply forget about messy codecs, badly designed DACs etc? This extra 1 dB of gain gives simply nothing in the opportunity but it can lead into unpredictable problems especially with loud and spectrum-heavy stuff.

@@mcdebugger maybe, but commercial releases don't care at all and sounds good to me.

@@digitaltrash_ Partially it's true but we can see more and more releases that have great dynamics and are not peaking as hell!

@@mcdebugger да бохуй чел, это всё условности, просто тести как разные кодеки кодируют мастер файл, я делал -1 пик левел, в итоге мой трек звучал тише чем у других. и самому не нравилось это, сейчас оставляю -0.3дб и этого достаточно, на самых громких моментах всё равно вылетает за 0, но ощутимых искажений я не слышу..

@@mcdebugger но всё же соглашусь -1db на мастере в большинстве играет в плюс, да, децибел заметен, но зато искажений после кодировки разными площадками будет меньше. трек будет чище звучать, особенно электронная музыка, где много лееринга и богатый спектральный диапазон.

Why do you want to use TP?

I know, I do that all the time. I kept calling the cloudlifter "cloudfilter" for like 2 years

@@mixbustv Heh! :)

Inter pencil seeks

Not ok. One thing is ISP but ceiling at zero will most likely result in actual overs. You have no reason to pick zero vs -0.1 or even -0.3 if you're really heavily limited (which I wouldn't advise anyway)

@@mixbustv ok thank you. I experimented with setting my final limiter ceiling at 0 and my Ableton master meter never went above 0, after I bounced I analyzed and it had +0.5dBtP

Why you care about ISP? But also, that's why people hire us to do mastering

you mean like this kzfaq.info/get/bejne/bZ97q7iUzLPNopc.html :D

@@mixbustv I knew it you got it.. star power very talented.

@@willierice2461 😄👍

@@mixbustv Cool, man! Nicely done!

@@mixbustv bro that production is sick AF

That makes no sense, if you're going to clip and distort it, why would you use another tool after it to minimize distortion!?

@@retromograph3893 Because digital to analog conversion leads to peaks that exceed the maximum digital sound level (0 dB). By using a true peak limiter at last in the processing chain, we're making sure that the music that is being played through speakers is free of unwanted artifacts.

If you’re going to distort it on purpose, why would you care about some much more subtle distortion artefacts downstream?

@@retromograph3893 Controlled distortion is OK. Uncontrolled distortion is not OK because it is unpredictible.

Because that's why we have true peak limiters, but most important, because NOBODY cares about overs

Hey Brian, there's a difference between peak and "truePeak". You can have a perfect digital Master which peaks at - 0.1 but has a "truePeak" of +1 or even +2. This is very common and you shouldn't care too much about this.

@@MusokeMSK I’m confused so I shouldn’t be using a true peak limiter because the true peak limiter isn’t allowing my mix to go over ?

@@REALLIFEADON that's up to you. But you don't have to. A lot of professional mastering engineers don't use true peak limiting. That's why those songs have intersample peaks. Like David said: you don't have to worry about those intersample peaks. Professional mastering engineers don't worry neither. Many say they don't like the way truePeak Limiting sounds. Just try both - truePeak Limiting and standard limiting - and decide by ear which you like better. But remember that truePeak is only a calculation on how the waveform behaves on playback systems. It does not occur on your actual mastered file and is not the same as the actual peak of your track.

@@MusokeMSK thank you so much

Nothing to do with the format, this is AD/DA related

You should watch the entire video, dude. He said very few people do that.

@@largepoodle6036 that's because i've actually seen it that i'm asking this question. Spotify for example says that you need to remain under -1 db in order to not distort on their platform. I wanted to know if he master at -0.1 ceiling also for spotify

@@IFeeeeelGood I of course cannot answer for him, but you never master for a particular streaming service nor for streaming services in general! You only do 1 final master, unless your also mastering for vinyl. However streaming does not get its own master like Vinyl does! Their ceilings(streaming sites, Spotify, Pandira, Apple Music, Tidal, ect.)are just something to keep in mind when your mastering. For, instance back in the loudness wars years (say 2005'ish) you'd set all your limiters, clippers and whatnot to the loudest possible without distorting, so if you could get it louder than zero without distorting, then that's what you did.. Like Death Magnetic from Metallica! That was over zero from what I can tell. Now we have LUFS, which is different and some say a better way to measure loudness. In theory say a Pop track (which includes Hip Hop nowadays IMO) Spotify says to be at -14 LUFS and you use 1 limiter (say Ozone) to get your Pop track to - 14 LUFS, your go will be quieter than everyone else's! Now say you use multiple bags compression, multiple limiters (possibly including multiband limiting) and multiple clippers/clipping to let's say reach a general level of somewhere between - 9 LUFS to - 6 LUFS you'll free gravy! Depending on the song as each song is different and will give you different results on how hard you can push it /limit it /how loud you can get it... before distortion or audible negative artifacting occurs! So basically, don't make seperate masters for streaming sites or each individual streaming sites, just shoot for somewhere between - 10 LUFS(to be safe) fed - 6 LUFS (to be super loud and aggressive)! Oh and this is bad too say in laymens terms but on certain tracks - 6 LUFS might be above 0db's! So when going for a hot track you gotta double check all the metering as well! I usually shoot for - 10 LUFS! That way it's not too quiet but not too loud either! Yes the streaming sites will turn it down a few dB's, which might make it sound squashed on certain streaming services, but you can't win them all, and 95% of Top 50 Billboard Tracks are all being turned down on all the streaming sites IMO! But there is also this... If your to quiet they will turn you up! But not your perceived volume! Meaning Your song with 1 Ozone limiter on it, set at - 14 LUFS(Spotify's level lets say) will sound quieter than YoungBloods Billboard track, even though both tracks are being played at - 14 LUFS! And that's why Mastering nowadays isn't as easy as you'd think it would be! You just can't slap a limiter on your master buss and call it a day! But I shouldn't even be answering this, as I'm no mastering engineer, just a recording engineer who sometimes does his own mastering because purple are lazy and cheap and want something for nothing! So you can donate $100 to my PayPal account for this invaluable teaching service @ paymybillsbecausrimdeadfuckingbrokeandjoblesswithnojobprospectsuntilcovidisover!

Not really, quality is crap anyway, ISP are not the problem or A problem

@@mixbustv I guess I'm not following this "crap anyway" argument. If the equipment sounds bad, I want to be the best sounding track on it nevertheless. :)

@@JM_2019 oh please don't get me wrong, that is a given 😄 of course we all want that, mine was an advice, if you hear something wrong, we look at everything else before ISP because is unlikely to be the one causing it.

You're not following mixbustv IG uh? That's every other day mastering, and higher.. (I post clips of actual sessions there)

@@mixbustv I do follow MixbusTV on IG. My concern is what happens to the loudness of the track when turned down by streaming sites? I saw a KZfaq video from sonicscoop on the "Future of Mastering" maybe you should look at it and make a video on your opinion. How do you preserve dynamic range with high LUFS cos my guess is that you'll be hitting up to -2db on RMS too

@@chike1866 i covered my views on streaming platforms levels in other videos and some of the comments in here, it's BS for the most part. As for dynamic range, dynamic range is a number, you can't preserve the number, but you can preserve the punch and the feeling of it, that'd mostly done in mixing, like I explained in my very popular (and copied) video on loudness. Loudness is not in mastering, is in the mix, so that the mastering doesn't need to be anything more than a couple of db of GR to reach -5. And mixing, is a LOT more difficult than mastering.

@@mixbustv I'll go check your video on streaming. I've seen the video on mixing and you are right. Loudness is actually in mixing. Everytime I mix ,I keep that video in mind. You are doing a great job man. 👌

Do you hear distortion when you convert? If you do, they matter, if you don't, they don't

@@mixbustv well it depends on how and who converts it, which is why I like izotopes lossless listening feature. Your 320k mp3 you export might not clip audibly but if some store you use has encoders that clip and sometimes emphasize it though most of these places are getting better and increasing their bitrates. I would never personally convert to 96kbps or 128kbps but there isn't a guarantee that some setting on one store won't. That said you shouldn't need to be that loud tbh nearly all major stores working around -14lufs you shouldn't even need to have any peaks (including ISP) over -1 mind over 0. The worst I've heard is soundcloud DJs that obvious limit to 0 and all the clean loud section are full of stereo distortion (it isn't linear across the panning).

There's no solution, there's no oversampling happening, it's conversion, conversion from D(igital) to A(analog) that is going to be the case until someone makes a device that communicate digital files directly into our brains.

@@mixbustv I was just guessing. :) thanks

@@mixbustv but the majority of Dacs are using an oversampling filter, like brickwall, or apodizing, which alter the sound even further.

IPS are not the cause of distortion, surely not in a club, and not the reason "meters are in the red", meters will be in the red with a perfectly IPS free track if it's loud anyway. If you HEAR IPS's you have a much bigger problem before that

@@mixbustv I've played tracks that I have mastered my self, both having TP limiting and no TP limiting and both where at -0.1dBFS. The non TP limiting tracks, spike into the reds on the DJ mixer as well as the software that's playing them back. Of course its fine for me to see the meter go over 0dBFS on the playback software, but the DJ mixer also spikes momentarily into the reds. This *never* happens with TP limited tracks. And I repeat, both where set to -0.1 on the final limiter.

@@nectariosm so? Of course is that way, you simply proved TP limiters do their job as advertised, what's new? That is NOT the point of the video nor my reply.

@@mixbustv I am replying to the title of the video, "why no body cares". Its not nobody, its case specific and it does matter in some cases, regardless of whether people play back music on their smartphones, to the point where the tiny speakers are breaking up.

@@nectariosm it's a title... And as shown in the analyzed commercial tracks, top charting, most, don't care.

I wouldn't use the true peak function, and I say I wouldn't because I literally got that a week ago, never even used it on a master yet. Just because there is an option doesn't mean I necessarily use it. Ozone has it, Lim#6 has it, Kclip has it, I've been using those for years, never used the TP. Not sure what this is about, maybe you though you can only use the TC with the TP algo, if so no.

@@mixbustv I don't think with the TC it's optional? I still prefer to remove ISP just to save possible issues later with conversions or DA.... Interesting video though and I never realised how many "pro" masters were so loud or even over 0db! .. keep up the great vids!

Minus 0 but what is the sweet spot?.

This makes no sense, ISP distortion is a high frequency phenomena, it has nothing to do with low end punch

@@retromograph3893Maybe it just has something to do with the lookahead aspect of the limiter algorithm itself that can affect the low end? I've heard some pro engineers talk about this downside of true peak before.

I see what you mean, i think i mis-read your comment because i agree with you! A limiter needs to act a lot faster and more drastically to protect from ISPs, so it is indeed less punchy. @@davejohnsonmusic

So you say...

​@@mixbustv So I'm saying that -1 dBTP headroom probably is a good idea and in fact it's recommended on many music platforms today. We simply don't need to go that loud as on the streaming platforms music will be turned down automatically anyway. I agree with you that true-peak meters may act not as good as good old ones PCM peak limiters. For me I found the solution to just use usual limiter with a bit lower ceiling/threshold and the final true peak limiter just for the safety. For example TDR Limiter already has both of them in it's chain by default.