Distrokid's Automatic Mastering Feature 

Hi Dan, I tried that distrokid option for a couple of weeks, when you pay it has 3 loudness options, the lowest being very close to the original and it also has options for the master to be brighter or darker, anyway, I stopped using it because I am not convinced yet. PS: Thank you for your videos, I am not an audio engineer, I am just a musician and thanks to you I have understood many things.

>"ai mastering" >look inside >Ozone 11 auto suggestion

Reminds me of the good old days where HiFi salespeople would make sure the system they wanted to flog you was turned up a fraction louder than the rest.

These mastering tools always seem to me like a one-size-fits-all service directed towards those who may not know any better. I would rather fail with a self-master than process it through a machine algorithm that is geared towards achieving certain set points regardless of how it sounds in the end.

But Dan... If you mastered it, it was already mastered by A.I...

As always, wish you all the strength to fight Crohns!

I don't like distrokid in general, they delete all songs from all streaming-platforms once you unsubscribe, you can't downgrade subscriptions and they are generally horrendously overpriced imo (like Content ID is $5 per song & year + 20% of what you earn with it). And then they also do intentionally misleading stuff like this. Really can't share these "positive feelings" you have towards them.

I agree with you 100%. I’ve noticed that my fully mastered and higher than recommended LUFS STILL SOUNDED QUIET IN THE PREVIEW. I DON’T RECOMMEND ANYONE PAY FOR THIS. THANKS FOR THE COMMENTARY.

When I released my album that was the first time I really mixed and mastered something with the help of numerous articles, videos (Dan's included of course) and memories my late musician dad told me 20 odd years ago. So when I got to this option on Distrokid and I heard that it boosted the highs and lows and it squashed everything I was really conflicted. Am I being deaf and stupid? I worked on this for weeks and this is what it supposed to sound?? Eventually, I decided I didn't care. I liked it the way I made it even if it's subpar. Later I saw a bunch of posts by professionals how awful this feature was so I was relieved. Maybe my mix is bad but at least I ruined it myself 😂

of course they would never gain match it, louder = better for most people = more money for distrokid

I like the ‘conceptual talks’ vids Dan (I have more listening-time available than watching-time).

Wow, you never disappoint... 😆 They probably expect the mix to be @ a certain level and not at 0db so I'm thinking it probably took your finished mix and crushed it to their "pre" master level then re-mastered it... Imagine how brutal that must sound...

Any "AI mastering" service rarely gives results I like, only Ozone seems ok but after an hour or so of tweaking

Distrokid should be sorting out the problem with songs belonging to people being deleted because of "fake streams". There are now hundreds of reports from people who state they have never bought any fake plays or playlist placements but had their songs deleted.

The very top end of your esses are like razor blades. It can be unpleasant to listen to. I found your video about your vocal chain. The SM58 example sounded better in these regards. I can say the same for when you only engaged the high pass filter on the ProQ3 for the condenser mic setup. If that is still the same condenser mic you're using, it might be worth using one of the other mics you have that you said is likely better. As a side note, I also have found the ProDS to sound a little strange. I can't necessarily put my finger on it. The old Waves De-esser (the light grey one) doesn't seem to exhibit any audible strangeness in the way it processes, unless it's in split mode. The crossover seems to augment the signal too much, at least for my liking.

sounds like they just whacked it through Breakaway Audio Enhancer with the settings cranked lol

Hot take: All "Auto Mastering" services are for people who don't know better and it's an absolute last resort if you have no idea how to do things properly. That's why they don't even bother with a proper loudnes-normalized comparison or anything else. The user is supposed to hear that it's louder, maybe have a bit more bass and think "oh, that's so much better than before!"!

If it doesn't pass the long-winded and wonky test, it might be an AI imposter.

Good stuff as always

make a video explaining RU-vid's new "Stable Volume" feature, and what it does.

Up yours!! Dan, you're the man!

I don’t know about the mastering but the track sounds 🔥🔥🔥🔥

WHOO NEW DAN WORRALL 🤙🤙

Have you seen Mixerman last video ? Thanx for kicking some asses PeaceR

Well, there's no algorithm for taste.

Thanks

I fully agree. Even Ozone 11 gets things wrong (although I feel it also gets a lot of things right and is a great tool, judging by my limited experience being a non-pro who’s mastered only one album with it). A few years ago I tried online mastering services and for the most part I felt like I could do a better job myself, even though I wouldn’t claim to know anything about mastering.

you can preview the entire song for $10!

$10 for an automated service that clearly costs them next to nothing, if not nothing is crazy

Landr provides an AI mastering tool 'free' with their distribution platform (and lots of other decent plugins). It has a bunch of options like a gain dial, eq, saturation etc, it's not just a one shot like your preview. I suspect distrokids might offer similar controls too when you buy it. So not sure if this is a fair test without buying it Landrs also has a VST you can drop into a DAW for it, minus some of the options like "warm" or "balanced"

I myself mix and master songs for a group of fellow producers, and each time, we joke around with distrokits ai mastering. Not a single time has it sounded better, and patterns started to emerge. It always sounds like there is a looot more highend. I'm assuming it is using some kind of multiband saturation in combination with EQ. This makes me wonder, however. Is the ai really listening to your mix and trying to improve it based on what it's lacking? Or is it more like a streamlined "mastering present". In case the second option holds true, I guess the fraud argument is even more valid. Then again, I guess this is catered towards newer artists since no credible or even mediocre mixing/mastering engineer would ever consider using a tool like this after having spent time and effort fine tuning everything as close to perfection as possible. A huge improvement might be for them to provide multiple versions, each focused on different frequency rages. One for the lowend, one for the midrange and one for the highend. I see value in that because it might give newer producers insight on common problems within their mix. If the lowend focused versions tend to sound better than the original one every time, then they could focus on the difference and try to improve their ears and skill to process the lows themselves until the AI is no longer needed.

Well it definitely sounded wider with their algorithm on it, for what it’s worth. Not sure if it’s “better” or “mastered” because of that, though.

AGREED !

Great 👍 video.

2:04 truth condensed in 6 seconds.

Compression + a loudness EQ curve, then.

11 bucks is delusional..

Time will go by and automated tools will get better. If you want to be more valuable in the future, learn more today. Pay more attention to the sound of what you do. Just because something sounds good in your expensive studio doesn't mean it will sound good on today's average or mid-range gear.

"Up yours" LOL. Good on ya Dan 🙂 Soooo tired of incontinent do-nothing people who think they can pull down their betters with condescending insults. Live and let live. And whilst doing said living, do something of value to someone (even if only to yourself). Altho you Dan, provide value to all (sane people). :-)

Hi Dan, interesting insight on automatic mastering services as the one from Distrokid... all these should be taken with a pinch of salt though. By the way, it would be very interesting, really helpful and MUCH APPRECIATED for a lot of people if you kindly elaborate on a separate video, a pragmatic vision why high resolution masters converted to lossy format such as MP3, increase their digital levels. And how we can predict the exact value of that increase. I realized myself that these values could vary (move) according to the loudness level of the original master, but also could change according to the source quality wav file, getting different results from a 44.1KHz source file than a 48KHz file, always using 24bit source files. The loudest the original master, the loudest the increase on the converted MP3, but this is not a firm statement at all and not a fixed value, that was only based on empirical experience I observed in a recurrent way. I mean, from a WAV 24bit 48KHz audio file mastered quite hot, around -7 LUFS integrated, and -0.30dB FS peak on the original master, it could reach +1dB FS increase when converted to MP3 320k for instance (all measured within Wavelab Pro 10 "Global Analysis" feature). On a 24/48 master of around -8.5 LUFS integrated the increase value in the MP3 320k is a bit lower, sitting around +0.7dB FS average. I could be completely wrong and could be a fact of the codec used, but would love the have your approach on this please. Fyi, I always convert all my files within Wavaleb Pro 10, using the Fraunhofer v4.1.3 encoder vers, constant bit rate 320k and “highest quality” option. I don’t hear any “distorted” artifact when playing back the converted MP3 file, but of course these over zero peaks can occur in a few samples or fraction of milliseconds, and that would be inaudible for anybody. This bothers me a lot, and want to find a proper answer in order to adjust my hi-res masters peak levels to do not reach over zero values when intended to be converted to MP3. Can’t wait to have your say on this. Keep posting your valuable content. Thank you maestro!

Unsurprising conclusion tbh. What would be much more interesting is the current state of the art of automatic/ai based mastering. Like whats to expect from such services apart from previewing? Soundcloud has a service like that and you have to pick between 3 different styles and can set an intensity level. This makes no sense at all to me, because what they did is just renaming compressor/limiter/eq vocabulary to 'thunder' and 'clouds' and 'chicken' or whatever, but in the end you still need your ears (and at least soundcloud is NOT taking care of the gain, so the mastered version is might simply be just louder. I could properly master myself if I have to rely on my ears anyways, at least benefiting from shared knowledge about established mastering techniques. So is it even possible with current possibilities to have a true one button mastering that is doing more than increasing volume bei 3 db? A video about that would be very intriguing, given that there is enough to say about it.

Yep, hearing it louder is definitely false advertising. Also distrokid keeps taking regular 1€ payments from me and doesn’t explicitly tell why, so I’m biased against them.

More long winded Dan!!!

Does that music contain fart samples? It sounds like farts.

more like automatic OTT, lol

The preview should be level matched you say? Why?? To make an informed decision??? I wonder, who would want that 🤣.

Hi Dan I will be really grateful if you talk about AI music!

seems silly yes but your test is flawed. try it on a non-mastered track, no limiter. a/b/x it - see if you can tell which is what. then try that on a phone or bluetooth speaker

Louder can often sound better on low volume sections ( even though it may not be what the artist wanted). But when you get to the high volume sections it all falls apart and gets truly unpleasant. I dont think this demo got us there...

For orchestral music i am almost sure automatic mastering will not the results hands can do. If they can, this is so surprising as Dan Worral appearing

pretty sure it's intended to be misleading. That's how you sell things you don't need.

Butʻ…. will they listen? :+)

DYNAMICS? WHAT DO YOU MEAN DYNAMICS? SORRY I CANT HEAR YOU OVER THE LOUD!

couldn't they just NOT auto-master songs? i mean artists who don't just upload their tracks to soundcloud but use a distributor like distrokid typically take everything super seriously and because of that already have mastered their music

As stated elsewhere, it's basically a money grabber.

This might be a hot take, but basic mastering techniques aren't very hard to learn. That said, if one is a musician who isn't interested in learning them (which is perfectly valid), I can't recommend they use this. It would be much cheaper in the long run to spend $50 on sonible's pure:limit (or something similar). But whatever you do, DO NOT buy ozone 11 elements (the cheapest version of ozone). I've found that ozone's master assistant is just as likely to ruin the track as it is to improve it, and the elements version doesn't allow the user to adjust any parameters in the chain. At least pure:limit is just a limiter and won't add a nonsensical EQ that can't be fixed. Or, and hear me out, just hire me to do it. 🤗

It makes sense that the incentive is there NOT to level match it.

They don't care. That is all.

This mastering service is way worse than just volume. Firstly, when I upload my tracks, which do need to be loud, then DistroKid makes them sound actually quieter. But aside from that, I tried uploading something with lowered loudness to see how it reacts. It basically changed the EQ and ruined a bunch of stuff, completely removing huge portions of low end. I would rather just use anything else, but not this service. Seriously, it did some awful work. I don't know, maybe there are some genres that it is good for, but for what I'm doing, which is Phytrance and Techno, it is definitely just destroying the mix. And destroying the balance.

AI mastering is old school now ... AI can now generate full songs with just a simple text prompt

Nothing wrong with "long winded and wonky", unlike Distrokid's fraudulent mastering service. I also have had no problems with them but that, sadly, is a pure con.

Here your’s track. And for 9.99 here it is the way I like it. What a cheap scam.

Oh, no. Not this "AI mastering" mumbo-jumbo-dumbo again.

You expect people to honestly trust in a master that has been done by some automatic process which they like to call a.i. power....without a real person in a highly treated room that does this for a living and makes critical decision based upon what impressions or requirements he or she got from either the telephone or real life conversation with the artist.. This might be fun for people who are just starting out recording music, and who are prone to gettting expoited by those who are preying on the weak, or ignorant in the community. I find it disgusting that a service like this is now being used to get more revenue more income to a company that has no high standards when it comes to music obviously, I don't mean the way songs are written and music is played but the quality of the mastering process!!! IF that is the new norm, then it is a wasted caus, making music and getting people to like it....If people are not careful and do not speak out against this a.i. exploitation process, music will indeed lose all of its inspiration and dedication AND human element which is creativity, emotion and message in songs! which non of the so called automated processes that like people to believe they are better than humans can do! IT is the person who programmed this process, who DOES NOT have to be musical in any way it's just fucking math to such a person......unbelievable how stupid and ignorant people are.....

The gain disparity is very obviously done with ill intent, i.e. to mislead customers. I strongly suspect that Distrokid's core business isn't all that lucrative, despite the number of artists they have as customers, because they charge peanuts for their service if their administrative and legal fees are anything like I imagine them to be, and that's on top of the more obvious tech stuff. They're probably in dire need of that new income stream. They might not even have the finances required to develop some proper mastering software

Making money off of laziness and ignorance. Thanks AI, and thanks Capitalism.

If it didn’t make it louder how would they keep up the illusion that they do anything good to your file 🤣🤣🤣. Load of rubbish always use a profession 🙌

Ai mastering, web mastering and robot mastering, applying an profile that been programmed and voila MASTERED! how lame.

First?

I love Dan’s tutorials. But his music… 🤐🤐🤐

You always tend to sound pretty tense. Are you breathing properly when you record? It sounds like you’re holding your breath between pauses, and only breathe out when you’re speaking. Whatever it is you’re doing, I can hear that tension in your voice.

the ai one is a huge improvement over yours, to be fair

Protip, if you take as long as you like, we'll just skip your videos.