WEBVTT

00:00:15.740 --> 00:00:21.820
Have you ever looked at a word on a screen and felt, you know, smarter for a couple of

00:00:21.820 --> 00:00:27.240
seconds, not really learned anything just felt smarter, you know what I mean?

00:00:27.800 --> 00:00:34.860
That's kind of what the word lossless does to people. It's a little badge next to a song title on

00:00:34.860 --> 00:00:42.980
your music app and suddenly you feel like you're hearing the record the way it was actually meant to

00:00:42.980 --> 00:00:52.500
be heard. The way it was made. The real thing. Finally, after years of getting that cheap MP3

00:00:52.500 --> 00:00:59.800
version. Now stick with me for a minute because I'm not making fun of you. I did the same thing.

00:01:01.700 --> 00:01:04.440
I'm Tyler. Thanks for being here today.

00:01:07.800 --> 00:01:11.560
In today's episode is about that little badge.

00:01:12.220 --> 00:01:18.880
Lossless audio, high res audio, all of it finally showing up in every streaming app

00:01:19.840 --> 00:01:29.140
minus YouTube music this year. And the question that nobody seems to be asking,

00:01:29.500 --> 00:01:41.500
which is whether any of us can actually hear that lossless high res version. Can you also,

00:01:41.820 --> 00:01:46.380
first up, standard disclaimer every episode, you know the drill. This is my read on it.

00:01:48.220 --> 00:01:52.000
My opinion. Nobody signs off on this show, but me.

00:01:55.080 --> 00:02:03.200
Here's my confession before we get fully into this, before we dive headfirst into this.

00:02:04.420 --> 00:02:08.580
For about a year, I saw that lossless badge pop up on my phone.

00:02:09.060 --> 00:02:15.160
You know, you see it on Apple Music, you know, especially title if you've got that.

00:02:15.720 --> 00:02:19.080
And I'll be honest, I felt a little smug about it. I'm like, yeah,

00:02:19.800 --> 00:02:26.820
I get the real good version now and never once, you know, did I check whether my headphones

00:02:27.720 --> 00:02:36.200
could actually do anything with that higher quality file signal, whatever. I never checked.

00:02:37.740 --> 00:02:42.760
I don't know. Maybe I didn't want to know. You know, some days I probably still don't,

00:02:42.760 --> 00:02:48.560
but it turns out not looking was doing a lot of work for me. And that's what

00:02:49.500 --> 00:02:55.380
Apple and Spotify and all these other companies are sort of betting on.

00:02:56.680 --> 00:03:00.720
So we've got to define the terms first because there's a lot of jargon

00:03:02.020 --> 00:03:10.040
that I'm going to spew out on this episode. This whole space runs on words that sound like

00:03:10.040 --> 00:03:13.340
they mean the same thing, but they don't stick with me for a minute.

00:03:14.380 --> 00:03:22.480
Lossless has one very boring, one, you know, specific meaning. When you unpack the file,

00:03:22.680 --> 00:03:29.480
you get back an exact copy of the original recording. Not close, not, yeah, you probably

00:03:29.480 --> 00:03:39.100
can't tell. It's identical. Flack, ALAC, Wave, that whole family. They're not even competing

00:03:39.100 --> 00:03:49.380
with each other on sound because, well, they can't. They all decode to the exact same thing that

00:03:50.240 --> 00:03:57.060
20 year old forum argument, you know, the argument I'm talking about, whether Flack

00:03:57.060 --> 00:04:05.100
sounds better than ALAC is two people arguing about which photocopy is more accurate when

00:04:05.100 --> 00:04:13.940
both copies are pixel for pixel, the same damn thing one for one. So don't get caught up in those

00:04:13.940 --> 00:04:21.900
arguments on those audio file forms is what I'm saying. Now, lossy. So there's lossless.

00:04:22.180 --> 00:04:29.900
We just talked about that. Now there's lossy. Lossy is the opposite. Lossy is that MP3, that AAC.

00:04:31.400 --> 00:04:37.920
Basically everything you grew up on. That's the kind of music that you were, you were

00:04:37.920 --> 00:04:42.700
downloading on like Kazan, Napster, and WinMX back in the day, MP3.

00:04:49.400 --> 00:04:53.940
AAC came a little later, but you know what I'm talking about, that watery,

00:04:54.560 --> 00:05:02.500
weird kind of compression sound. That's a lossy file. An algorithm basically looks at the song,

00:05:02.740 --> 00:05:08.520
guesses what you probably won't miss and throws the rest of it out the door forever.

00:05:09.040 --> 00:05:15.420
Because once it's gone, it's gone. That's how you get the smaller file. It's also a

00:05:16.380 --> 00:05:23.920
it's a permanent decision made on your behalf about what you're allowed to hear.

00:05:25.280 --> 00:05:33.840
And one gacha here, one little kind of wrench into the plan

00:05:36.360 --> 00:05:44.100
is another term. It's separate from both of those that we just talked about,

00:05:44.100 --> 00:05:53.220
lossless and lossy. It's called high res, which just means higher resolution than CD.

00:05:55.300 --> 00:06:00.660
Different, different thing altogether. A file can be lossless and not high res.

00:06:01.200 --> 00:06:08.060
A file can call itself high res and not be lossless at all. Hey, just hang on to that last one.

00:06:08.300 --> 00:06:10.180
We'll, we'll, we'll get back to it. I promise.

00:06:12.860 --> 00:06:21.720
Now, by this year, basically everyone has lossless audio now. Apple had it.

00:06:23.180 --> 00:06:27.820
Well, I shouldn't say had it. They have it. I think was Apple one of the first that

00:06:27.820 --> 00:06:30.820
kind of pulled the lever on that? I don't remember, but Apple has it.

00:06:31.400 --> 00:06:35.240
I've got Apple music. It's great. I love it. Sounds awesome.

00:06:36.960 --> 00:06:42.680
Amazon folded it into their regular pricing a few years back.

00:06:43.240 --> 00:06:46.680
Title made it the default for everyone right out the gate.

00:06:47.020 --> 00:06:50.840
And Spotify, after teasing the thing for like four or five years,

00:06:51.080 --> 00:06:55.680
finally shipped it last fall free. No upcharge. Good news, right?

00:06:56.640 --> 00:06:58.820
Case closed. Everybody's ears can rejoice.

00:07:02.240 --> 00:07:07.040
Except almost none of you are actually hearing any of it.

00:07:07.160 --> 00:07:13.040
I know. Sorry to burst the bubble. And you want to know the reason why?

00:07:14.260 --> 00:07:15.660
The reason is Bluetooth.

00:07:22.620 --> 00:07:26.720
Here's the part that doesn't make it into the buzzwords, the ad copy, the,

00:07:27.520 --> 00:07:31.500
the promotions. And it doesn't even mention it in the app.

00:07:32.880 --> 00:07:42.900
A true lossless stream needs way more room to breathe than a standard Bluetooth connection

00:07:42.900 --> 00:07:46.980
can even give it. It's physics. And I'll tell you why.

00:07:48.040 --> 00:07:55.840
It doesn't fit through the pipe. So something has to give in what gives every single time

00:07:56.720 --> 00:07:58.520
is the audio quality.

00:08:01.680 --> 00:08:08.680
Every mainstream Bluetooth codex squeezes that signal down and throws pieces of it away

00:08:09.600 --> 00:08:13.880
before it ever reaches your, your earhole.

00:08:16.560 --> 00:08:23.340
It doesn't matter which one AirPods, you know, that one specifically

00:08:24.220 --> 00:08:29.920
runs on a codec that is not lossless by design. Doesn't matter what tier you're paying for.

00:08:30.340 --> 00:08:36.480
Doesn't matter what that badge on your iPhone is saying. Apple's own audio team basically said

00:08:37.180 --> 00:08:41.420
the, the, you know, the quiet part out loud to a reviewer.

00:08:42.140 --> 00:08:47.740
Their line was that the Bluetooth codec is about reliability,

00:08:48.880 --> 00:08:52.320
not about squeezing out the last bit of quality.

00:08:53.900 --> 00:08:57.100
Fine. That's a defensible engineering answer.

00:08:58.860 --> 00:09:08.940
Reliability over quality. But it also means functionally no AirPods on any current model

00:09:09.660 --> 00:09:16.840
are ever handing you that lossless audio stream over a normal connection.

00:09:17.940 --> 00:09:24.900
That badge in the app describes a file sitting on a server somewhere, probably in California.

00:09:26.120 --> 00:09:29.700
It stops being true the second it leaves your phone.

00:09:32.200 --> 00:09:37.360
However, there's one exception. And it's a small exception.

00:09:37.360 --> 00:09:44.000
A quill com codec exists that genuinely does

00:09:44.640 --> 00:09:52.440
deliver an exact copy over Bluetooth, but it only does CD quality, not high res.

00:09:52.920 --> 00:09:58.060
And it needs a phone and a set of headphones both built specifically for it.

00:09:58.540 --> 00:10:05.440
So if you're on an iPhone, you don't get that option because Apple doesn't support it.

00:10:06.760 --> 00:10:12.220
Now, when I said, you know, you can only fit so much over the pipe, it's physics.

00:10:13.560 --> 00:10:22.680
A CD quality file runs at a bit rate of 1400 kilobytes a second, a little bit over that.

00:10:22.920 --> 00:10:26.780
I'm just rounding it down. Bluetooth can't shove that through there.

00:10:28.060 --> 00:10:32.780
You know, Bluetooth can get somewhere 300. Certain codecs can get 900,

00:10:32.780 --> 00:10:36.880
but you're never quite getting it. Sorry, side tangent.

00:10:39.340 --> 00:10:46.480
Now, let's talk about the part of this that actually makes me a little bit fidgety.

00:10:47.560 --> 00:10:56.120
There was a format out there called MQA. Stood for Master Quality Authenticated. Sounds fancy,

00:10:56.120 --> 00:11:03.320
right? Tidal built its entire premium tier around it, branded it as Masters.

00:11:04.280 --> 00:11:09.640
And the pitch was, this isn't just high quality, it's verified.

00:11:11.640 --> 00:11:22.160
There was a literal light on some hardware that would light up to confirm you were hearing the

00:11:22.160 --> 00:11:29.340
real Studio Master. Guaranteed. Made you feel special.

00:11:31.060 --> 00:11:37.220
But an independent researcher actually tested that claim and a good thing they did because,

00:11:38.060 --> 00:11:43.260
well, when they ran real signals through it, checked what came out on the other end and what

00:11:43.260 --> 00:11:50.880
he found was that a lot of what got sold under that quote, unquote, Masters badge.

00:11:51.820 --> 00:11:58.580
Supposedly straight from the real Studio Masters looked a lot more like a regular file that had

00:11:58.580 --> 00:12:04.460
been dressed up and stamped with the same authenticity seal as the genuine thing.

00:12:05.460 --> 00:12:20.860
The process itself was adding noise into the audio, not even removing it. It was actually

00:12:20.860 --> 00:12:28.040
a good chunk of this. I mean, as they probably would. I mean, it's making them look bad.

00:12:28.840 --> 00:12:37.680
And they said that they never used the word lossless in the strict, you know, technical sense.

00:12:38.060 --> 00:12:44.980
So I'm not going to stand here and call it a closed case, but here's what happened next.

00:12:44.980 --> 00:12:53.160
And nobody disputes this part of it. The company went into administration,

00:12:54.160 --> 00:13:01.720
bankruptcy, whatever. In the following year, title quietly pulled every piece of MQA content off

00:13:01.720 --> 00:13:08.740
of its platform and moved back to plain old flak. I say plain old flak. Flak is still,

00:13:08.740 --> 00:13:18.840
to me, the gold standard, but you know, when the format collapses in the company that built its whole

00:13:18.840 --> 00:13:26.460
premium identity around it, deletes it a year later, you know, the market sort of rendered its

00:13:26.460 --> 00:13:31.380
own verdict. I don't need a lab coat to figure that one out.

00:13:35.910 --> 00:13:43.410
So when do you actually get the real thing? I know it kind of seems a little depressing

00:13:43.410 --> 00:13:50.390
up till now, right? But let's talk about it. A bit perfect, you know, right cable,

00:13:50.790 --> 00:13:55.830
no compression anywhere in the chain. Does it even sound better?

00:13:57.730 --> 00:14:06.610
Well, honestly, the science is still kind of split, which I respect because it means nobody's

00:14:06.610 --> 00:14:13.910
trying to sell you a clean answer. There's an old, well-known study where researchers

00:14:13.910 --> 00:14:20.930
ran blind listening tests for over a year on serious equipment using, you know,

00:14:20.930 --> 00:14:26.070
actual audio engineers and self-described audio files as test subjects.

00:14:26.910 --> 00:14:33.610
Quietly slipping a CD quality bottleneck into an otherwise high res playback chain.

00:14:34.470 --> 00:14:45.890
You know what? Nobody could reliably tell the difference. Not one person did better

00:14:46.610 --> 00:14:56.250
than a simple old coin flip. A newer, much bigger review pulled together close to 20

00:14:57.650 --> 00:15:06.270
separate studies and thousands of individual listening trials.

00:15:08.170 --> 00:15:18.570
And that one did find real measurable ability to tell high res apart from the regular audio.

00:15:19.330 --> 00:15:27.210
But even in that one study that found a genuine effect, trained listeners, people who

00:15:27.870 --> 00:15:31.790
specifically practiced this skill, knew what they were listening for,

00:15:32.510 --> 00:15:38.810
were only getting it right a little better than a coin flip. 50-50.

00:15:40.650 --> 00:15:45.970
That's the winning result. Barely better than a coin flip.

00:15:47.250 --> 00:15:54.030
Public blind tests go even further. One well-known public test had regular people

00:15:54.750 --> 00:16:02.870
try to pick the true lossless track out of a lineup and the results landed right around

00:16:04.430 --> 00:16:09.970
basically random guessing. On one track, more people actually preferred the compressed version

00:16:09.970 --> 00:16:17.710
over the lossless one. So, you know, real differences probably exist in a quiet room on

00:16:17.710 --> 00:16:25.850
really good expensive gear with really trained practice deers on the right kind of recording.

00:16:26.450 --> 00:16:35.270
They're probably small. They're very, very conditional. For how most of us actually listen

00:16:35.270 --> 00:16:41.650
on a bus, in a car, walking the dog, background noise everywhere,

00:16:41.650 --> 00:16:46.530
where you're really not going to tell the difference.

00:16:49.930 --> 00:16:56.190
And here's the stat that, you know, sold me on the whole theory.

00:16:57.410 --> 00:17:06.310
The one major streaming service that still offers no, absolutely no lossless options

00:17:06.310 --> 00:17:12.970
of any kind has also been the fastest growing paid service in the country for

00:17:13.670 --> 00:17:20.570
the last few years running. Not close either. People are picking a service with no lossless

00:17:20.570 --> 00:17:26.850
tier at all over the ones that do have it. In huge, huge numbers.

00:17:28.450 --> 00:17:33.270
Meanwhile, you've got surveys where most people say they'd happily pay more for

00:17:33.270 --> 00:17:41.150
higher quality audio, which tells you people are pretty good at describing what they think they want

00:17:41.670 --> 00:17:46.270
and not that great at actually acting on it once it's sitting right in front of them.

00:17:49.230 --> 00:17:55.350
And that service that has no lossless audio whatsoever, that's YouTube music.

00:17:56.510 --> 00:18:06.010
Now, you know, with YouTube, that's kind of a mixed stat, because yes, it is the fastest growing,

00:18:06.750 --> 00:18:12.890
but YouTube premium subscribers do get access to YouTube music, so the numbers might be a little muddy there.

00:18:19.630 --> 00:18:26.090
Okay, so a couple of things I want to walk you, I want you to walk away with.

00:18:28.230 --> 00:18:33.290
One, lossless streaming became the free standard no extra charge default everywhere.

00:18:33.930 --> 00:18:38.730
Well, I mean, that's a really good thing. There's no downsides to that. I'll take it

00:18:38.730 --> 00:18:43.490
happily. If you've got the gear that supports it, great, you've got the option to hear it.

00:18:44.510 --> 00:18:50.790
Now, another one, if you're, you know, if you're on wireless earbuds, especially AirPods,

00:18:51.230 --> 00:18:57.190
that badge is just a decorative thing on that app. It's describing the file, but not the sound

00:18:57.190 --> 00:19:03.730
you're actually hearing. If you ever actually want what, you know, these companies are

00:19:03.730 --> 00:19:12.910
promising you need a wire or a very specific, you know, matched hardware that most people don't have

00:19:12.910 --> 00:19:22.610
and probably won't ever bother getting. And one more, even when you do get the real thing,

00:19:23.190 --> 00:19:29.070
the honest research says most of us can't reliably hear the difference anyways,

00:19:29.070 --> 00:19:36.910
which doesn't make any of this really a scam. You know, exactly. It just means that the marketing is

00:19:36.910 --> 00:19:45.850
selling a much bigger, it's making it a much bigger deal than it really is. Or your ears can hear,

00:19:46.050 --> 00:19:55.990
you know, any of this can actually back up. That's kind of the whole show in a nut show,

00:19:55.990 --> 00:20:03.350
isn't it? Something gets built well, quietly, correctly, and then it gets sold to you with a lot

00:20:03.350 --> 00:20:12.410
more confidence than the truth can support. That's the episode. If you want to keep up with it,

00:20:12.770 --> 00:20:16.450
subscribe wherever you're listening. And if you're, you know, if you've got a second,

00:20:17.050 --> 00:20:21.630
you're looking for something to do, you got a free moment, drop a five star rating.

00:20:23.570 --> 00:20:29.370
If you, if you would rather just talk to me, you can leave a voicemail. I've got an

00:20:29.370 --> 00:20:39.210
actual phone number set up for that now, 608-668-4702. Again, 608-668-4702. You can text,

00:20:39.210 --> 00:20:47.850
you can leave a voicemail at that number, or you can email me studio at tylerwoodward.mp.

00:20:51.450 --> 00:20:55.090
That's what I got. I'll catch you guys next week.`