WEBVTT 00:00:00.000 --> 00:00:02.859 OK, let's just let's unpack this. For years, 00:00:02.940 --> 00:00:06.320 there have been these two seemingly impossible 00:00:06.320 --> 00:00:10.160 problems haunting AI image generators. The first 00:00:10.160 --> 00:00:13.839 one, trying to get them to write complex, readable 00:00:13.839 --> 00:00:16.960 text inside of visual. It was always just a garbled 00:00:16.960 --> 00:00:19.420 mess like digital spaghetti. Right. And that 00:00:19.420 --> 00:00:21.539 was the industry failure point. But the second 00:00:21.539 --> 00:00:24.559 problem was, you know, just as critical for businesses 00:00:24.559 --> 00:00:27.100 and storytellers. Character consistency. Exactly. 00:00:27.420 --> 00:00:30.190 Generating the same mascot across. dozens of 00:00:30.190 --> 00:00:33.530 scenes without it changing its face or its proportions. 00:00:34.369 --> 00:00:36.890 Google's new system, Nano Banana Pro, it just 00:00:36.890 --> 00:00:39.950 changes the rules entirely by solving both. Welcome 00:00:39.950 --> 00:00:42.829 back to the Deep Dive. Today we are tearing into 00:00:42.829 --> 00:00:44.850 the technical specs of what a lot of reviewers 00:00:44.850 --> 00:00:47.250 are calling a generational leap in generative 00:00:47.250 --> 00:00:49.210 AI. Yeah. We're talking about the Nano Banana 00:00:49.210 --> 00:00:51.909 Pro model, which is powered by the incredible 00:00:51.909 --> 00:00:54.890 reasoning capabilities of Gemini 3. We've sourced 00:00:54.890 --> 00:00:57.070 a brand new technical review to guide us through 00:00:57.070 --> 00:00:59.189 this. And our mission today is pretty simple, 00:00:59.229 --> 00:01:01.869 but it's also a challenge. We really want to 00:01:01.869 --> 00:01:04.329 understand the foundational mechanisms here. 00:01:04.689 --> 00:01:08.359 How does this system move past... just basic 00:01:08.359 --> 00:01:10.120 pattern matching, you know, making things look 00:01:10.120 --> 00:01:13.760 good to a genuine conceptual understanding that 00:01:13.760 --> 00:01:16.359 allows for real accuracy. So we'll cover the 00:01:16.359 --> 00:01:19.060 core tech that enables this. We'll dive deep 00:01:19.060 --> 00:01:22.280 into the revolution this creates for, say, infographics 00:01:22.280 --> 00:01:24.519 and typography. And then we'll look at the elegant 00:01:24.519 --> 00:01:27.140 solution for character and brand consistency. 00:01:27.340 --> 00:01:30.719 And critically, we will examine the honest limitations, 00:01:30.920 --> 00:01:34.079 what it still can't do. Yeah. so let's start 00:01:34.079 --> 00:01:36.180 with the leap itself this isn't just a bigger 00:01:36.180 --> 00:01:38.459 version running faster it's a totally different 00:01:38.459 --> 00:01:41.120 approach the claim is that this is a generational 00:01:41.120 --> 00:01:44.379 shift not just an update why it really just boils 00:01:44.379 --> 00:01:47.420 down to process control before you'd hand a model 00:01:47.420 --> 00:01:49.560 a prompt and it would just immediately start 00:01:49.560 --> 00:01:51.620 generating pixels right the rendering the whole 00:01:51.620 --> 00:01:53.340 system is focused on the look of the prompt yeah 00:01:53.739 --> 00:01:56.780 But NanoBanana Pro, it introduces an advanced 00:01:56.780 --> 00:01:59.620 large language model, Gemini 3, as a reasoning 00:01:59.620 --> 00:02:01.599 engine. So it's at the beginning of the workflow. 00:02:01.840 --> 00:02:04.780 Right at the start. The model thinks first. Okay, 00:02:04.819 --> 00:02:07.540 so if I type a prompt, what is Gemini 3 actually 00:02:07.540 --> 00:02:11.340 doing before any pixels even start to form? It's 00:02:11.340 --> 00:02:14.099 like an internal architect. It translates your 00:02:14.099 --> 00:02:18.659 concept into rigid structural constraints. It 00:02:18.659 --> 00:02:20.939 breaks down the prompts semantically. So if you 00:02:20.939 --> 00:02:23.180 ask for an image comparing two historical events, 00:02:23.379 --> 00:02:26.000 it doesn't just look for pictures of those events. 00:02:26.080 --> 00:02:28.840 It reasons. It reasons. It asks, what facts are 00:02:28.840 --> 00:02:30.860 needed here? What's the best format to communicate 00:02:30.860 --> 00:02:33.599 this? And crucially, that reasoning step is tied 00:02:33.599 --> 00:02:36.120 to real -time verification, isn't it? Exactly. 00:02:36.319 --> 00:02:38.840 Gemini 3 can use web search to gather data in 00:02:38.840 --> 00:02:41.620 real time. or verify facts that are needed for 00:02:41.620 --> 00:02:44.449 the image. Only after that whole reasoning and 00:02:44.449 --> 00:02:46.870 verification phase does it build a detailed plan. 00:02:47.050 --> 00:02:49.090 A constraint map. A constraint map. Yeah, the 00:02:49.090 --> 00:02:52.069 Nano Banana Pro, the image generator part, has 00:02:52.069 --> 00:02:54.090 to obey. It's like a digital architect planning 00:02:54.090 --> 00:02:56.550 the precise load -bearing structure before laying 00:02:56.550 --> 00:02:59.669 a single visual break. Reviewers even saw a thinking 00:02:59.669 --> 00:03:02.169 drop -down feature that let them audit the process. 00:03:02.490 --> 00:03:04.849 You could literally watch Gemini 3 tracing its 00:03:04.849 --> 00:03:06.990 history and verifying facts before the image 00:03:06.990 --> 00:03:08.930 was even rendered. It's the difference between 00:03:08.930 --> 00:03:11.530 asking an artist to paint a historical scene 00:03:11.530 --> 00:03:14.919 from memory. And giving that artist a few hours 00:03:14.919 --> 00:03:17.400 to research the exact clothing, the setting, 00:03:17.500 --> 00:03:19.960 the factual context before they even pick up 00:03:19.960 --> 00:03:23.599 a brush. So how does that? Thinking first step 00:03:23.599 --> 00:03:27.240 translate to practical, factual accuracy in the 00:03:27.240 --> 00:03:29.639 finished image. It uses web search to verify 00:03:29.639 --> 00:03:33.060 facts, ensuring accuracy before rendering the 00:03:33.060 --> 00:03:35.240 visual. That deep planning leads us straight 00:03:35.240 --> 00:03:38.300 to the text revolution because text was the industry's 00:03:38.300 --> 00:03:41.259 big, big bottleneck. Oh, yeah. When we talk about 00:03:41.259 --> 00:03:43.800 text failing, we mean the AI just saw letters 00:03:43.800 --> 00:03:47.159 as visual textures, as abstract squiggles, not 00:03:47.159 --> 00:03:49.539 as symbols with meaning. The severity of that 00:03:49.539 --> 00:03:51.319 problem was crippling. I mean, you could spend 00:03:51.319 --> 00:03:53.500 10... minutes crafting the perfect prompt and 00:03:53.500 --> 00:03:56.120 the AI would still spell basic words wrong. All 00:03:56.120 --> 00:03:58.800 the time. Nano Banana Pro has achieved one -shot 00:03:58.800 --> 00:04:01.580 perfect results. The first generation is the 00:04:01.580 --> 00:04:03.560 perfect result, even with complex paragraphs 00:04:03.560 --> 00:04:06.979 of text. I have to admit, I still wrestle with 00:04:06.979 --> 00:04:09.560 prompt drift myself, especially when I'm just 00:04:09.560 --> 00:04:11.400 trying to get a simple sign right, trying to 00:04:11.400 --> 00:04:14.099 generate a product label or a book cover with 00:04:14.099 --> 00:04:16.819 accurate titles. It was just a prompt credit 00:04:16.819 --> 00:04:19.160 graveyard before this. And this is where that 00:04:19.160 --> 00:04:21.680 reasoning model really shines. It treats the 00:04:21.680 --> 00:04:25.279 text not as a visual thing, but as a conceptual 00:04:25.279 --> 00:04:28.420 requirement dictated by Gemini 3's initial plan. 00:04:28.600 --> 00:04:30.660 Right. I mean, look at the specific tests they 00:04:30.660 --> 00:04:33.480 ran. They created this highly technical, clean, 00:04:33.639 --> 00:04:37.660 medical style, infographic, breaking down REM 00:04:37.660 --> 00:04:40.000 versus deep sleep. And it had perfectly readable 00:04:40.000 --> 00:04:42.279 labels, correct terminology, none of that AI 00:04:42.279 --> 00:04:45.139 weirdness, which is so crucial for any kind of 00:04:45.139 --> 00:04:47.279 technical content. Then they pushed the factual 00:04:47.279 --> 00:04:50.199 boundary. They asked the system to research the 00:04:50.199 --> 00:04:52.860 top five budget coffee machines and to accurately 00:04:52.860 --> 00:04:55.180 pull pros, cons, and real ratings from different 00:04:55.180 --> 00:04:57.279 sources. And lay it out in a chart. And lay it 00:04:57.279 --> 00:04:59.480 out in a neat comparison chart. That requires 00:04:59.480 --> 00:05:02.399 data ingestion, reasoning, and then very precise 00:05:02.399 --> 00:05:04.980 layout. That shifts the tool from being just 00:05:04.980 --> 00:05:07.319 a purely aesthetic engine into a serious design 00:05:07.319 --> 00:05:09.600 and data business assistant all in one process. 00:05:09.860 --> 00:05:12.360 The ultimate stress test, though, for typography 00:05:12.360 --> 00:05:14.560 and layout had to be that comic encyclopedia 00:05:14.560 --> 00:05:18.319 page. That test is just... absurdly difficult 00:05:18.319 --> 00:05:21.319 for a generative model. It demands placing large 00:05:21.319 --> 00:05:24.439 blocks of verbatim text nested in speech bubbles. 00:05:24.600 --> 00:05:27.759 It needs dynamic formatting, dramatic typography, 00:05:27.939 --> 00:05:30.920 color -coded sections, and precise power -level 00:05:30.920 --> 00:05:34.699 stats, all in one complex layout. That's a challenge 00:05:34.699 --> 00:05:37.120 that needs a semantic understanding of the text 00:05:37.120 --> 00:05:40.060 blocks, not just placing a few letters. The system 00:05:40.060 --> 00:05:42.379 had to handle paragraph fitting, font changes, 00:05:42.639 --> 00:05:45.209 stat boxes, and keep it all accurate. So beyond 00:05:45.209 --> 00:05:47.110 simple captions, what was the most difficult 00:05:47.110 --> 00:05:50.170 text formatting test it handled? The model flawlessly 00:05:50.170 --> 00:05:52.930 formatted a dense, multi -section comic book 00:05:52.930 --> 00:05:55.850 encyclopedia page. Wow. Now we move to the second 00:05:55.850 --> 00:05:59.589 major solution, consistency, or the lack of it. 00:05:59.709 --> 00:06:02.250 Character drift was the real commercial and narrative 00:06:02.250 --> 00:06:04.610 barrier for all the previous AI systems. You 00:06:04.610 --> 00:06:07.129 could generate a beautiful mascot one time, but 00:06:07.129 --> 00:06:09.290 the second time, the line weight on its ear was 00:06:09.290 --> 00:06:12.050 slightly off, or its proportions changed in a 00:06:12.050 --> 00:06:15.170 subtle way. And that's a disaster for any agency 00:06:15.170 --> 00:06:18.230 or brand manager trying to scale content. You 00:06:18.230 --> 00:06:21.110 can't build brand trust if your mascot's face 00:06:21.110 --> 00:06:24.370 changes in every single ad. No. So NanoBanana 00:06:24.370 --> 00:06:27.449 Pro addresses this using a technique the reviewer 00:06:27.449 --> 00:06:30.490 called the Pro Workflow. Okay, let's detail that 00:06:30.490 --> 00:06:32.329 workflow. This is really critical for anyone 00:06:32.329 --> 00:06:34.689 who's involved in brand identity. It relies on 00:06:34.689 --> 00:06:37.990 using Gemini 3's analytical power. So step one, 00:06:38.110 --> 00:06:41.209 you ask Gemini to analyze an existing asset, 00:06:41.449 --> 00:06:44.790 a logo, a previous character render, and then 00:06:44.790 --> 00:06:46.930 formalize the brand guidelines for that. The 00:06:46.930 --> 00:06:49.750 vibe, the colors. The vibe, the specific colors, 00:06:49.870 --> 00:06:52.189 hex codes even, and the typography constraints. 00:06:52.569 --> 00:06:55.170 So we're using the LLM to codify the style. What's 00:06:55.170 --> 00:06:57.470 step two? Well, since those guidelines can be 00:06:57.470 --> 00:07:00.290 long and text prompts have limits, you convert 00:07:00.290 --> 00:07:02.569 them into specific visual assets. So you're basically 00:07:02.569 --> 00:07:04.509 just taking screenshots of the style guide. I 00:07:04.509 --> 00:07:06.569 see. Step three, you upload those screenshots 00:07:06.569 --> 00:07:08.970 as dedicated reference images to Nano Banana 00:07:08.970 --> 00:07:11.740 Pro. This process gets around the limits of text 00:07:11.740 --> 00:07:13.519 -only prompts, and it locks in the aesthetic 00:07:13.519 --> 00:07:15.759 for every new asset you generate. That means 00:07:15.759 --> 00:07:18.500 the model is treating the brand style as a single, 00:07:18.660 --> 00:07:21.480 immutable visual token instead of just a list 00:07:21.480 --> 00:07:23.540 of suggestions. Exactly. And the character tests 00:07:23.540 --> 00:07:25.620 proved it. Right. They tested the mascot across 00:07:25.620 --> 00:07:28.240 wildly different scenarios. Holding a latte, 00:07:28.519 --> 00:07:31.220 driving a scooter, and the mascot was perfectly 00:07:31.220 --> 00:07:33.899 reproducible. Identical line weight, identical 00:07:33.899 --> 00:07:36.600 proportions, no matter what the action was. They 00:07:36.600 --> 00:07:39.500 also ran an even more demanding test. The emotion 00:07:39.500 --> 00:07:42.800 panel test. This generated a six panel sheet 00:07:42.800 --> 00:07:46.879 for emotions, cheerful, sleepy, annoyed, where 00:07:46.879 --> 00:07:48.740 the character had to shift its expression without 00:07:48.740 --> 00:07:51.139 any structural distortion. And the fidelity was 00:07:51.139 --> 00:07:53.639 unwavering. It was. And maybe the most revealing 00:07:53.639 --> 00:07:57.180 was the storyboard camera test. Shifting perspective. 00:07:57.839 --> 00:08:00.379 like from a mid shot to a full body shot, that 00:08:00.379 --> 00:08:03.379 usually causes tiny details to drift, a facial 00:08:03.379 --> 00:08:05.759 structure, an accessory. But here the reviewer 00:08:05.759 --> 00:08:07.819 found that every small detail was preserved, 00:08:08.000 --> 00:08:11.160 even across dramatic perspective shifts. So what's 00:08:11.160 --> 00:08:13.439 the secret to ensuring the brand style stays 00:08:13.439 --> 00:08:16.240 locked when generating new assets? Uploading 00:08:16.240 --> 00:08:18.600 converted brand guideline screenshots as dedicated 00:08:18.600 --> 00:08:22.199 reference images is the best technique. Mid -roll 00:08:22.199 --> 00:08:26.560 sponsor read. Okay, shifting focus a little. 00:08:26.639 --> 00:08:29.600 The model's performance in both consistency and 00:08:29.600 --> 00:08:32.379 text, it points to its deeper and I think most 00:08:32.379 --> 00:08:35.980 exciting power. Genuine conceptual understanding. 00:08:36.509 --> 00:08:40.090 It understands ideas and relationships, not just, 00:08:40.090 --> 00:08:41.889 you know, collections of pixels. That conceptual 00:08:41.889 --> 00:08:44.289 power can sound a bit abstract, but the practical 00:08:44.289 --> 00:08:46.129 examples are really stunning. Take the reverse 00:08:46.129 --> 00:08:48.830 engineering or recipe test. A reviewer uploaded 00:08:48.830 --> 00:08:52.009 a photo of a finished kind of complex steak dish. 00:08:52.230 --> 00:08:54.990 Right. Then they asked the AI for a photo of 00:08:54.990 --> 00:08:57.090 all the ingredients labeled with their quantities. 00:08:57.389 --> 00:08:59.830 And the result was that the model correctly identified 00:08:59.830 --> 00:09:02.610 and visualized the meat, the butter, the garlic, 00:09:02.730 --> 00:09:04.769 the heavy cream, everything necessary for that 00:09:04.769 --> 00:09:07.970 dish. required understanding the visual inputs 00:09:07.970 --> 00:09:10.190 and the underlying chemistry that created the 00:09:10.190 --> 00:09:12.649 meal. It's not just matching steak to a database 00:09:12.649 --> 00:09:15.169 of ingredients. It's inferring the whole process. 00:09:15.409 --> 00:09:17.549 Then there was the geographic intelligence test, 00:09:17.830 --> 00:09:20.809 zooming in on Vatican City and maintaining accurate 00:09:20.809 --> 00:09:23.610 spatial relationships. The position of the trees, 00:09:23.850 --> 00:09:27.970 the obelisk, even at a 67x zoom. Which suggests 00:09:27.970 --> 00:09:31.309 the model is not relying on flat 2D approximations. 00:09:31.409 --> 00:09:34.070 It seems to maintain some kind of internal synthetic 00:09:34.070 --> 00:09:37.240 coordinate system. projecting constraints onto 00:09:37.240 --> 00:09:39.159 what's essentially a 3D map of the location. 00:09:39.440 --> 00:09:41.600 And the final piece of evidence for that conceptual 00:09:41.600 --> 00:09:44.299 linkage was the translation accuracy test. It 00:09:44.299 --> 00:09:46.899 coherently translated English text on a cereal 00:09:46.899 --> 00:09:50.039 box into French. No made -up words. No made -up 00:09:50.039 --> 00:09:51.980 words. It showed genuine language processing 00:09:51.980 --> 00:09:55.679 tied directly into the visual output. Whoa. Two 00:09:55.679 --> 00:09:58.639 sec silence. Just imagine scaling that conceptual 00:09:58.639 --> 00:10:01.299 understanding, that ability to reason and translate 00:10:01.299 --> 00:10:04.820 ideas into structural maps to a billion queries 00:10:04.820 --> 00:10:07.679 a day, handling things far beyond just images, 00:10:07.779 --> 00:10:10.659 integrating data streams across media. That is 00:10:10.659 --> 00:10:12.820 the actual generational leap. That conceptual 00:10:12.820 --> 00:10:15.440 power is undeniable, but the reviewer did offer 00:10:15.440 --> 00:10:17.480 an honest assessment of limitations. It is not 00:10:17.480 --> 00:10:19.220 infallible yet, and this is important for users 00:10:19.220 --> 00:10:21.460 to understand. Right. The biggest frustration 00:10:21.460 --> 00:10:24.740 was around specific geometric instruction, specifically 00:10:24.740 --> 00:10:28.399 pose control. If the reviewer asked the model 00:10:28.399 --> 00:10:31.139 to make a character adopt a very specific, detailed 00:10:31.139 --> 00:10:34.620 pose from a reference drawing, say a complex 00:10:34.620 --> 00:10:37.480 martial arts stance, Nano Banana Pro consistently 00:10:37.480 --> 00:10:40.230 ignored it. And it just substituted its own pose 00:10:40.230 --> 00:10:42.990 instead. Why? I mean, if it understands complex 00:10:42.990 --> 00:10:45.049 concepts and spatial relationships, why does 00:10:45.049 --> 00:10:47.110 it struggle with specific input like a reference 00:10:47.110 --> 00:10:50.070 pose? It ignores pose reference drawings, preferring 00:10:50.070 --> 00:10:51.970 to generate its own character positions instead. 00:10:52.169 --> 00:10:54.309 It seems to prioritize the character's identity 00:10:54.309 --> 00:10:57.669 and its action, what the character is doing over 00:10:57.669 --> 00:11:00.730 the precise form, the exact geometry of the pose. 00:11:00.970 --> 00:11:03.230 Right. And there are minor flaws, too. Small 00:11:03.230 --> 00:11:05.850 text on products, tiny text or fine print on 00:11:05.850 --> 00:11:08.850 product labels or, say, watch faces, often fails 00:11:08.850 --> 00:11:10.889 to render crisply when you zoom in. Even though 00:11:10.889 --> 00:11:13.889 the large logos are perfect. Perfect. And finally, 00:11:14.009 --> 00:11:17.250 the reality of any AI rollout. Inconsistent community 00:11:17.250 --> 00:11:20.389 results. Variation is expected. There are constant 00:11:20.389 --> 00:11:22.350 model updates. There's randomness in the generation. 00:11:22.809 --> 00:11:25.149 Your results might vary a bit from the peak examples. 00:11:25.529 --> 00:11:28.190 After analyzing all the evidence, though. The 00:11:28.190 --> 00:11:30.750 verdict is pretty compelling. It's best in class. 00:11:31.070 --> 00:11:34.350 Nano Banana Pro absolutely blows all other image 00:11:34.350 --> 00:11:37.269 generation models out of the water for versatility, 00:11:37.490 --> 00:11:40.629 accuracy, and just practical application. We 00:11:40.629 --> 00:11:42.350 can still look at the competitive landscape. 00:11:42.710 --> 00:11:45.289 Mid -journey is still incredibly strong for raw, 00:11:45.429 --> 00:11:49.029 pure, aesthetic beauty. It can maybe create a 00:11:49.029 --> 00:11:52.450 subjectively prettier image, but it's still fundamentally 00:11:52.450 --> 00:11:56.110 weak on text and consistency. Nanobanan Pro takes 00:11:56.110 --> 00:11:58.269 the crown because it delivers both high quality 00:11:58.269 --> 00:12:01.049 and high utility, especially for commercial and 00:12:01.049 --> 00:12:02.929 technical use cases. So what does this actually 00:12:02.929 --> 00:12:05.190 mean for you, the listener, in your day -to -day 00:12:05.190 --> 00:12:07.350 work? Well, for marketers, it means generating 00:12:07.350 --> 00:12:09.830 campaign -ready assets that automatically adhere 00:12:09.830 --> 00:12:14.220 to brand guidelines in a single prompt. No constant 00:12:14.220 --> 00:12:16.799 human oversight. For educators, you can now describe 00:12:16.799 --> 00:12:20.000 a complex concept like photosynthesis and get 00:12:20.000 --> 00:12:22.200 a publication -ready accurate visual explanation 00:12:22.200 --> 00:12:25.559 or a labeled chart instantly. Yeah. And for professional 00:12:25.559 --> 00:12:27.659 creators, you finally have consistent characters 00:12:27.659 --> 00:12:30.139 across unlimited scenarios and camera angles. 00:12:30.340 --> 00:12:32.679 It unlocks some serious world -building potential. 00:12:33.230 --> 00:12:34.850 So we should probably just reiterate the best 00:12:34.850 --> 00:12:38.830 practices here. Use Gemini 3 to formalize and 00:12:38.830 --> 00:12:40.789 create those guidelines. Right. Use screenshots 00:12:40.789 --> 00:12:42.870 of those guidelines as your visual references 00:12:42.870 --> 00:12:45.649 to liken the style. And always use Nano Banana 00:12:45.649 --> 00:12:48.649 Pro to generate text from scratch. That is its 00:12:48.649 --> 00:12:52.129 strongest use case. But manually verify any highly 00:12:52.129 --> 00:12:55.570 specialized technical text just to be safe. The 00:12:55.570 --> 00:12:57.649 accessibility message here is just so powerful. 00:12:57.909 --> 00:12:59.889 These aren't capabilities that are promised for 00:12:59.889 --> 00:13:01.929 the future. They are proven and available today. 00:13:02.669 --> 00:13:05.070 It represents years of research paying off in 00:13:05.070 --> 00:13:07.830 a tangible, practical leap. The fact that this 00:13:07.830 --> 00:13:10.509 system solved consistency and complex text at 00:13:10.509 --> 00:13:13.149 the same time is proof that the technology has 00:13:13.149 --> 00:13:15.210 moved beyond treating the world as just pixels. 00:13:15.590 --> 00:13:18.309 This conceptual understanding moves past simple 00:13:18.309 --> 00:13:20.590 pattern matching. It's moving toward genuine 00:13:20.590 --> 00:13:23.049 comprehension of intent and structure. The tools 00:13:23.049 --> 00:13:25.470 exist. The capabilities are proven. The only 00:13:25.470 --> 00:13:27.909 question left is... What will you create now 00:13:27.909 --> 00:13:29.690 that this level of fidelity and accessibility 00:13:29.690 --> 00:13:32.230 is available to everyone? Out to your own music.