<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Mykhailo Kashkarov on Medium]]></title>
        <description><![CDATA[Stories by Mykhailo Kashkarov on Medium]]></description>
        <link>https://medium.com/@mykhailo_kashkarov?source=rss-3986cd562804------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*2xWPkvyvO4yYKWSQ</url>
            <title>Stories by Mykhailo Kashkarov on Medium</title>
            <link>https://medium.com/@mykhailo_kashkarov?source=rss-3986cd562804------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 22:36:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@mykhailo_kashkarov/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[An Ornament That Reads Back]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/an-ornament-that-reads-back-b710da5e16a9?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/b710da5e16a9</guid>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Tue, 25 Aug 2026 11:17:29 GMT</pubDate>
            <atom:updated>2026-08-25T11:17:29.506Z</atom:updated>
            <content:encoded><![CDATA[<p>How text becomes a symmetric bytecode mandala — and why symmetry isn’t decoration, but data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WcdZL66xpWm52qlipfC0ag.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g2gdlbe782GPcFVy_uberA.png" /></figure><p>The idea is provocatively simple. Text is converted into bytes using UTF-8, the international text encoding standard. Bytes become zeros and ones. The bits fill one eighth of a square, and that eighth is mirrored eight times, like in a kaleidoscope. The result is a symmetric mandala. Every pixel of the ornament corresponds to one bit of data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OqUuCWKSV2mLMxRphtQfJA.png" /></figure><p>The symmetry doesn’t decorate the code — it generates it. Decoding runs the same path backwards: the decoder takes one sector, gathers bits into bytes, bytes into text.</p><p>There isn’t a single service byte in the pattern — no markers, no checksums, nothing but the text itself. I removed them not for engineering reasons. I treated this as an art study: I wanted to see the bit-level essence of a character — an honest portrait of a letter, with no crutches. Even if the patterns had turned out unreadable, I would still have an alphabet of ornaments — one I didn’t invent, but revealed.</p><p>Yet without the service bytes, reading became even more reliable: the decoder has to correctly reconstruct the entire data stream, or the text simply won’t come out. A decision made for aesthetic reasons turned out to be sound engineering as well.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XjsUDTQ5bL2SkULJx4-Itg.png" /></figure><p>A few interesting properties follow from the symmetry. The pattern reads in any rotation or reflection — it has no “right side up.” And its size gives away the weight of the character: a Latin letter takes one byte and produces a small snowflake, a Ukrainian letter takes two bytes and a larger one, a Chinese character takes three. The weight of a symbol in the world’s encoding table becomes visible to the naked eye.</p><h3>Beyond Black and White: Color Channels and Layers</h3><p>What started as monochrome octants has now evolved into fully realized multi-layered and multi-channel color bytecodes. By utilizing RGB color space and separating data across distinct physical or spectral layers, the encoding capacity multiplies without breaking the aesthetic structure of the ornament.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BkGPi0Czb4i9Xl6EJbkgyQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YWLPZAB2FB4clAr8JicEzA.png" /></figure><h3>The Most Valuable Part Wasn’t Invented by Me</h3><p>Every day we rely on things created by hundreds of thousands of scientists whose names we’ll never know — behind every font and every emoji stands mathematics nobody thinks about. I didn’t know it either; I learned as I went.</p><p>When I truly understood what Unicode is — a single worldwide catalog holding every character of every writing system — it struck me. People can’t agree on almost anything. But here, they agreed: a Ukrainian letter, a Japanese character, a rune, and an emoji each have a number in one shared table.</p><p>Perhaps that is the most valuable thing about this project. I merely visualized what already exists — gave an aesthetic form to an agreement all of us silently use every day. Each snowflake is a portrait of one entry in that agreement.</p><h3>Try It Out</h3><p>You can generate and scan both monochrome and multi-layered color bytecodes using the working tool here: <a href="https://astra31415926.github.io/Life_10">https://astra31415926.github.io/Life_10</a></p><p>This is where I’m counting on you, readers. If you can think of interesting applications — tell me. And if something like this already exists and I simply failed to find it — please point me to it, I genuinely want to know.</p><p>What would you weave into such an ornament?</p><ul><li><strong>Preprint &amp; Priority (Zenodo):</strong> <a href="https://zenodo.org/records/22092011">10.5281/zenodo.22092011</a></li><li><strong>ORCID:</strong> <a href="https://orcid.org/0009-0006-4060-3465">0009–0006–4060–3465</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b710da5e16a9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[An Ornament That Reads Back]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/an-ornament-that-reads-back-935b807d2c49?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/935b807d2c49</guid>
            <category><![CDATA[creative-coding]]></category>
            <category><![CDATA[ukraine]]></category>
            <category><![CDATA[unicode]]></category>
            <category><![CDATA[generative-art]]></category>
            <category><![CDATA[design]]></category>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Mon, 27 Jul 2026 14:19:06 GMT</pubDate>
            <atom:updated>2026-07-27T14:25:45.870Z</atom:updated>
            <content:encoded><![CDATA[<p>How text becomes a symmetric bytecode mandala — and why symmetry isn’t decoration, but data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/760/1*eWdhTpaLWvhotNHP7WfOAQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/760/1*T8YNEOAc7r0rEIr0eIrldQ.png" /><figcaption><strong>One of these patterns can be read by a program. Try to guess which.</strong></figcaption></figure><p>The idea is provocatively simple. Text is converted into bytes using UTF-8, the international text encoding standard. Bytes become zeros and ones. The bits fill one eighth of a square, and that eighth is mirrored eight times, like in a kaleidoscope. The result is a symmetric mandala. Every pixel of the ornament corresponds to one bit of data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XCEq73WUP0MGL0vj-3oxFw.png" /><figcaption><strong><em>From text to bits, and from bits to a symmetric mandala.</em></strong></figcaption></figure><p>The symmetry doesn’t decorate the code — it generates it. Decoding runs the same path backwards: the decoder takes one sector, gathers bits into bytes, bytes into text.</p><p>There isn’t a single service byte in the pattern — no markers, no checksums, nothing but the text itself. I removed them not for engineering reasons. I treated this as an art study: I wanted to see the bit-level essence of a character — an honest portrait of a letter, with no crutches. Even if the patterns had turned out unreadable, I would still have an alphabet of ornaments — one I didn’t invent, but revealed.</p><p>Yet without the service bytes, reading became even more reliable: the decoder has to correctly reconstruct the entire data stream, or the text simply won’t come out. A decision made for aesthetic reasons turned out to be sound engineering as well.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lsZTKSYipUG34NZUKf85fA.png" /><figcaption><strong><em>Comparison of Latin, Ukrainian, Chinese characters, and emoji in bytecode representation.</em></strong></figcaption></figure><p>A few interesting properties follow from the symmetry. The pattern reads in any rotation or reflection — it has no “right side up.” And its size gives away the weight of the character: a Latin letter takes one byte and produces a small snowflake, a Ukrainian letter takes two bytes and a larger one, a Chinese character takes three. The weight of a symbol in the world’s encoding table becomes visible to the naked eye.</p><h3>The most valuable part wasn’t invented by me</h3><p>Every day we rely on things created by hundreds of thousands of scientists whose names we’ll never know — behind every font and every emoji stands mathematics nobody thinks about. I didn’t know it either; I learned as I went.</p><p>When I truly understood what Unicode is — a single worldwide catalog holding every character of every writing system — it struck me. People can’t agree on almost anything. But here, they agreed: a Ukrainian letter, a Japanese character, a rune, and an emoji each have a number in one shared table.</p><p>Perhaps that is the most valuable thing about this project. I merely visualized what already exists — gave an aesthetic form to an agreement all of us silently use every day. Each snowflake is a portrait of one entry in that agreement.</p><h3>What’s next</h3><p>The next step is teaching the decoder to read bytecode from physical media — from print. That opens the door to exhibition pieces.</p><p>I’m realistic about this: it’s not a replacement for QR codes — those are excellent and useful. Bytecode’s niche is different: art, design, souvenirs — things where a pattern with a hidden message is valuable in itself.</p><p>This is where I’m counting on you, readers. If you can think of interesting applications — tell me. And if something like this already exists and I simply failed to find it — please point me to it, I genuinely want to know.</p><p>What would you weave into such an ornament?</p><p>Try it here: <a href="https://astra31415926.github.io/Life_10">https://astra31415926.github.io/Life_10</a></p><p><strong>Preprint:</strong> <a href="https://doi.org/10.5281/zenodo.19633526">https://doi.org/10.5281/zenodo.19633526</a></p><p><strong>ORCID:</strong> <a href="https://www.google.com/search?q=https://orcid.org/0009-0006-4060-3465">https://orcid.org/0009-0006-4060-3465</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=935b807d2c49" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Метод наскрізного геометричного калібрування колірних просторів на основі спектрального репера SCI]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/%D0%BC%D0%B5%D1%82%D0%BE%D0%B4-%D0%BD%D0%B0%D1%81%D0%BA%D1%80%D1%96%D0%B7%D0%BD%D0%BE%D0%B3%D0%BE-%D0%B3%D0%B5%D0%BE%D0%BC%D0%B5%D1%82%D1%80%D0%B8%D1%87%D0%BD%D0%BE%D0%B3%D0%BE-%D0%BA%D0%B0%D0%BB%D1%96%D0%B1%D1%80%D1%83%D0%B2%D0%B0%D0%BD%D0%BD%D1%8F-%D0%BA%D0%BE%D0%BB%D1%96%D1%80%D0%BD%D0%B8%D1%85-%D0%BF%D1%80%D0%BE%D1%81%D1%82%D0%BE%D1%80%D1%96%D0%B2-%D0%BD%D0%B0-%D0%BE%D1%81%D0%BD%D0%BE%D0%B2%D1%96-%D1%81%D0%BF%D0%B5%D0%BA%D1%82%D1%80%D0%B0%D0%BB%D1%8C%D0%BD%D0%BE%D0%B3%D0%BE-%D1%80%D0%B5%D0%BF%D0%B5%D1%80%D0%B0-sci-3adb1f9416d1?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/3adb1f9416d1</guid>
            <category><![CDATA[cmykrgb]]></category>
            <category><![CDATA[color-calibration]]></category>
            <category><![CDATA[color-science]]></category>
            <category><![CDATA[color-management]]></category>
            <category><![CDATA[3d-printing-industry]]></category>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Sun, 05 Jul 2026 15:33:19 GMT</pubDate>
            <atom:updated>2026-07-05T19:34:11.475Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FfX2IsYLQZ1iLrVyks6fUw.png" /></figure><h3>1. Виділення ортогонального базису</h3><p>Традиційні системи управління кольором (ICC) будують профіль за принципом «ззовні всередину»: спочатку вимірюються граничні точки деформованого простору пристрою у просторі L*a*b*, після чого математично описується поведінка всього внутрішнього об’єму. Усередині цього простору канали залишаються взаємозалежними: присутні паразитне поглинання пігментів, розтискування растрової точки та трепінг (crosstalk).</p><p>Метод SCI (Spectral Channel Integrity) пропонує протилежний підхід — «зсередини назовні». Фізичний відбиток аналізується поканально за допомогою селективної RGB-фільтрації. Після компенсації міжканального проникнення визначається область триканальної лінійності, у якій канали поводяться максимально незалежно один від одного.</p><p>Саме ця область формує внутрішній лінійний RGB-каркас, що використовується як базова система координат. Метод побудови спектрального репера SCI, принцип виділення області триканальної лінійності та фізичне обґрунтування поканальної сепарації детально наведені в попередній роботі автора [1].</p><h3>2. Побудова повного колірного охоплення</h3><p>Після побудови внутрішнього ортогонального базису визначаються координати всіх максимально насичених кольорів, які здатний відтворити конкретний принтер. Ці точки не входять до внутрішнього лінійного куба, а описують його зовнішню деформовану оболонку відносно ідеального цифрового простору. Таким чином формується повна тривимірна карта фізичних можливостей друкарського пристрою.</p><h3>Рис. 1. Геометрична деформація колірного охоплення пристрою відносно еталонного куба RGB</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bcz9mBlWXB40AVXtBvKU_Q.png" /></figure><h3>Рис. 2. Локалізація стабільного ортогонального ядра SCI всередині деформованого простору друку</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d0ufnZSWGYYuRQb3_MDZ7A.png" /></figure><h3>3. Масштабування структури</h3><p>Під час переходу між різними пристроями або матеріалами (наприклад, з якісного крейдованого паперу на газетний офсет) масштабуються не окремі кольори через таблиці LUT, а вся координатна конструкція целиком як єдина шарнірна сітка.</p><p>Змінюються координати граничних кольорів відповідно до фізичних обмежень нового обладнання, тоді як взаємне розташування елементів та пропорції векторів відносно внутрішнього ядра SCI зберігаються. У результаті використовується весь доступний колірний потенціал конкретного принтера (максимальна сочність пігментів), але структура колірних переходів і баланс сірого не руйнуються, повністю виключаючи тональний кліппінг даних.</p><h3>Підсумок</h3><p>У традиційному підході координатна система будується від зовнішнього колірного охоплення до внутрішньої структури. У моделі SCI координатна система будується від внутрішнього ортогонального базису до зовнішньої геометрії колірного простору. Саме цей внутрішній каркас забезпечує єдину детерміновану систему координат для монітора, друкарського пристрою та подальшого масштабування без залучення дорогого спектрофотометричного обладнання.</p><h3>Практичне використання профілю SCI</h3><p>У класичному управлінні кольором (ICC) профілі працюють як громіздкі «словники» (таблиці Look-Up Table), які зіставляють зовнішні координати пристрою з колірним простором. Модель SCI натомість використовує внутрішній <strong>ортогональний каркас</strong> та забезпечує поканальну стабільність без важких таблиць перерахунку.</p><p>Процес застосування профілю SCI складається з трьох простих кроків:</p><ol><li><strong>Задання вихідних координат:</strong> Система бере опорний спектральний репер SCI як внутрішній математичний каркас друкарської системи. Він фіксує незалежні вектори каналів, які не перетинаються між собою.</li><li><strong>Визначення стабільних RGB-точок:</strong> На основі геометрії репера обчислюються точні координати допустимих RGB-адрес. Вони визначаються через аналіз шкали сірого, яка виступає ортогональним вектором для виявлення та компенсації нелінійних спотворень каналів (crosstalk).</li><li><strong>Масштабування та друк:</strong> Під час виведення на друк або перенесення зображення на інший носій/принтер система порівнює внутрішні каркаси пристроїв. Корекція кольору відбувається за рахунок геометричного масштабування самого каркаса, що гарантує збереження тональних переходів та лінійності каналів без перекалібрування кожної окремої точки.</li></ol><h3>Посилання:</h3><h4><a href="https://zenodo.org/records/19633526">SCI Universal Model</a></h4><h4><a href="https://astra31415926.github.io/sci-spec/">SPECTRAL CHANNEL INTEGRITY (SCI)</a></h4><h4><a href="https://www.linkedin.com/in/%D0%BC%D0%B8%D1%85%D0%B0%D0%B9%D0%BB%D0%BE-%D0%BA%D0%B0%D1%88%D0%BA%D0%B0%D1%80%D0%BE%D0%B2-281379162/">linkedin</a></h4><p><strong>ORCID</strong>: 0009-0006-4060-3465</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3adb1f9416d1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SCI Universal Model: Deterministic Spectral Response Normalization]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/sci-universal-model-deterministic-spectral-response-normalization-c331209a7aa4?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/c331209a7aa4</guid>
            <category><![CDATA[spectralchannelintegrity]]></category>
            <category><![CDATA[optics]]></category>
            <category><![CDATA[computational-imaging]]></category>
            <category><![CDATA[color-science]]></category>
            <category><![CDATA[signalprocessin]]></category>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Fri, 24 Apr 2026 19:28:34 GMT</pubDate>
            <atom:updated>2026-05-03T15:01:05.965Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*WUuaAAx4xOeJC3lx2av6eg.png" /></figure><p><strong>Mikhail Kashkarov</strong><br> September 2026</p><h3>Abstract</h3><p>This paper introduces the Spectral Channel Integrity (SCI) universal model — a framework for deterministic spectral response normalization in subtractive media. The inherent limitations of physical pigments — spectral inconsistency, crosstalk, and amplitude imbalance — are addressed through a system of fixed reference levels scaled against a carefully defined spectral limiter.</p><p>By applying linear normalization, the model transforms a physically inconsistent reflective surface into a more stable information carrier with operationally separable spectral channels. This approach opens possibilities for high-fidelity spectral data encoding and multi-layer visualization, previously restricted mainly to additive emissive systems.</p><h3>1. Concept: Adaptive Orthogonalization of a Subtractive Medium</h3><p>The SCI method adapts the fundamental property of additive RGB environments — channel orthogonality and functional independence — for implementation on physical subtractive media.</p><p>It represents a structured configuration where spectral reflection functions of pigments are amplitude-aligned. Essentially, SCI implements additive logic through subtractive physics by using controlled dynamic range compression to achieve deterministic physical separation of information channels.</p><h3>2. Physical Limitations of Subtractive Media</h3><p>The primary constraint when transferring information from emissive to reflective systems lies in the spectral inconsistency of pigments.</p><p>Unlike idealized additive models, where an inactive channel equals absolute zero, subtractive pigments always retain a non-zero residual spectral response due to incomplete absorption. This leads to energy imbalance and variable crosstalk.</p><p>Without a unified reference scale, deterministic channel organization remains unattainable.</p><h3>3. The SCI Method: Formation of Reference Levels</h3><p>To overcome these limitations, the SCI model introduces a unified reference scale defined by two fixed system states:</p><ul><li><strong>Upper Reference Level (Uref):</strong> A single maximum permissible spectral response level for all active channels, limited by the system’s weakest chromatic component.</li><li><strong>Lower Reference Level (Lref):</strong> A unified minimum reflection level corresponding to the residual spectral response of the limiting component, accepted as the logical “0”.</li></ul><h3>4. The Spectral Limiter Component</h3><p>The limiting component is defined functionally: it maximizes reflection at the spectral edges (red and blue regions) while maximizing absorption in the central (green) region.</p><p>This component (typically Magenta in CMY systems) possesses the smallest operational dynamic range and serves as the natural physical boundary for the model’s reference levels.</p><h3>5. Mathematical Formalization</h3><h3>5.1 Amplitude Scaling (Upper Level Unification)</h3><p>To equalize maximum reflectance across all pigments relative to the limiting component:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*YlSPq-8aGDtqujrT8sBZJA.png" /></figure><p>where the peak amplitude of each pigment is scaled relative to the peak amplitude of the limiting component.</p><h3>5.2 Zero-Level Offset (Lower Level Unification)</h3><p>To establish a common baseline and remove parasitic reflectance, mapping the signal into a normalized range:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*-4bWS1F6qXKe0HvJyWQkdA.png" /></figure><p>where the signal is shifted and scaled between the lower and upper reference levels.</p><h3>5.3 Channel Decoupling (Linear Correction)</h3><p>To further reduce residual crosstalk, an inverse matrix transformation is applied:</p><p><strong>[ЗДЕСЬ ВСТАВЬ ФОРМУЛУ 3]</strong></p><p>where the transformation matrix is derived from experimentally measured crosstalk coefficients.</p><h3>6. Applications and Future Development</h3><p>The deterministic structure of the SCI model enables:</p><ul><li><strong>Spectral Data Encoding:</strong> Increasing information density in 2D codes</li><li><strong>Authentication Protection:</strong> Creation of hidden spectral markers</li><li><strong>Multilayer Visualization:</strong> Independent images on a single surface visible under narrow-band illumination</li></ul><h3>References</h3><ol><li>Sharma, G., &amp; Bala, R. (2002). <em>Digital Color Imaging Handbook</em>. CRC Press.</li><li>Kang, H. R. (2010). <em>Computational Color Technology</em>. SPIE Press.</li><li>CIE Publication 15:2018. <em>Colorimetry (4th Edition)</em>. Universal Model: Deterministic Spectral Response Normalization</li></ol><p><strong>Mikhail Kashkarov</strong><br> September 2026</p><h3>Abstract</h3><p>This paper introduces the Spectral Channel Integrity (SCI) universal model — a framework for deterministic spectral response normalization in subtractive media. The inherent limitations of physical pigments — spectral inconsistency, crosstalk, and amplitude imbalance — are addressed through a system of fixed reference levels scaled against a carefully defined spectral limiter.</p><p>By applying linear normalization, the model transforms a physically inconsistent reflective surface into a more stable information carrier with operationally separable spectral channels. This approach opens possibilities for high-fidelity spectral data encoding and multi-layer visualization, previously restricted mainly to additive emissive systems.</p><h3>1. Concept: Adaptive Orthogonalization of a Subtractive Medium</h3><p>The SCI method adapts the fundamental property of additive RGB environments — channel orthogonality and functional independence — for implementation on physical subtractive media.</p><p>It represents a structured configuration where spectral reflection functions of pigments are amplitude-aligned. Essentially, SCI implements additive logic through subtractive physics by using controlled dynamic range compression to achieve deterministic physical separation of information channels.</p><h3>2. Physical Limitations of Subtractive Media</h3><p>The primary constraint when transferring information from emissive to reflective systems lies in the spectral inconsistency of pigments.</p><p>Unlike idealized additive models, where an inactive channel equals absolute zero, subtractive pigments always retain a non-zero residual spectral response due to incomplete absorption. This leads to energy imbalance and variable crosstalk.</p><p>Without a unified reference scale, deterministic channel organization remains unattainable.</p><h3>3. The SCI Method: Formation of Reference Levels</h3><p>To overcome these limitations, the SCI model introduces a unified reference scale defined by two fixed system states:</p><ul><li><strong>Upper Reference Level (Uref):</strong> A single maximum permissible spectral response level for all active channels, limited by the system’s weakest chromatic component.</li><li><strong>Lower Reference Level (Lref):</strong> A unified minimum reflection level corresponding to the residual spectral response of the limiting component, accepted as the logical “0”.</li></ul><h3>4. The Spectral Limiter Component</h3><p>The limiting component is defined functionally: it maximizes reflection at the spectral edges (red and blue regions) while maximizing absorption in the central (green) region.</p><p>This component (typically Magenta in CMY systems) possesses the smallest operational dynamic range and serves as the natural physical boundary for the model’s reference levels.</p><h3>5. Mathematical Formalization</h3><h3>5.1 Amplitude Scaling (Upper Level Unification)</h3><p>To equalize maximum reflectance across all pigments relative to the limiting component:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/663/1*HkqQq5mfpNv8KyMTtBAReg.png" /></figure><p>where the peak amplitude of each pigment is scaled relative to the peak amplitude of the limiting component.</p><h3>5.2 Zero-Level Offset (Lower Level Unification)</h3><p>To establish a common baseline and remove parasitic reflectance, mapping the signal into a normalized range:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/663/1*KRNGunJ6UgvotORYqz_-1Q.png" /></figure><p>where the signal is shifted and scaled between the lower and upper reference levels.</p><h3>5.3 Channel Decoupling (Linear Correction)</h3><p>To further reduce residual crosstalk, an inverse matrix transformation is applied:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/769/1*0sCoTOxBC4CFW9XHvDEIiw.png" /></figure><p>where the transformation matrix is derived from experimentally measured crosstalk coefficients.</p><h3>6. Applications and Future Development</h3><p>The deterministic structure of the SCI model enables:</p><ul><li><strong>Spectral Data Encoding:</strong> Increasing information density in 2D codes</li><li><strong>Authentication Protection:</strong> Creation of hidden spectral markers</li><li><strong>Multilayer Visualization:</strong> Independent images on a single surface visible under narrow-band illumination</li></ul><h3>References</h3><ol><li>Sharma, G., &amp; Bala, R. (2002). <em>Digital Color Imaging Handbook</em>. CRC Press.</li><li>Kang, H. R. (2010). <em>Computational Color Technology</em>. SPIE Press.</li><li>CIE Publication 15:2018. <em>Colorimetry (4th Edition)</em>.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x5nkyRKhS5DfZm3yPcMFWQ.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c331209a7aa4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Universal Model: Deterministic Spectral Response Normalization]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/universal-model-deterministic-spectral-response-normalization-49cc6336e6b0?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/49cc6336e6b0</guid>
            <category><![CDATA[color-science]]></category>
            <category><![CDATA[signal-processing]]></category>
            <category><![CDATA[science]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Fri, 17 Apr 2026 20:33:10 GMT</pubDate>
            <atom:updated>2026-04-17T20:33:10.038Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dAqceqIMUB8-vpYGGZpSyA.jpeg" /></figure><p>Title: SCI Universal Model: Deterministic Spectral Response Normalization</p><p>Subtitle: A deterministic approach to balancing CMYK pigments for optical data encoding, steganography, and multi-layer visualization</p><p>Feature Image:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xV-7snZ1yvaiG5UPoW_5tA.png" /></figure><p>Have you ever wondered why a printed photo never quite matches the vibrant glow of a screen? Or why it’s so difficult to reliably hide digital data inside a color print without it getting corrupted?</p><p>The answer lies in the fundamental physics of how screens and paper create color.</p><p>Screens use additive light (RGB): start with black, add Red, Green, Blue to get white. When a channel is “off,” it emits zero light. This makes RGB channels mathematically clean and orthogonal — perfect for encoding information.</p><p>Paper and ink use subtractive color (CMYK): start with white paper, add pigments that absorb (subtract) certain wavelengths. The problem? Physical pigments are messy. They don’t just absorb their target color; they leave a messy “spectral footprint” everywhere else.</p><p>This paper introduces the Spectral Channel Integrity (SCI) Model — a framework for forcing physical ink on paper to behave more like a clean digital screen.</p><h3>The Core Problem: Why Print is “Noisy”</h3><p>In an ideal world, a Cyan ink would <em>only</em> control the Red channel, and a Magenta ink would <em>only</em> control the Green channel. In reality, pigments suffer from two fatal flaws:</p><ol><li>Amplitude Imbalance: Yellow ink reflects almost 90% of light, while Magenta might only manage 65%. This creates an uneven “signal strength.”</li><li>Crosstalk: Magenta ink reflects some light in the Cyan zone. This is like hearing someone else’s conversation on your phone line. It ruins channel separation.</li></ol><p>Traditional approaches try to fix this mess <em>after</em> scanning (complex algorithms on the decoding side). The SCI Model asks a different question: What if we clean up the signal <em>before</em> we even print it?</p><h3>The SCI Method: Clipping the Wings of Strong Colors</h3><p>The SCI method is a form of deterministic pre-distortion. It forces all colors to obey the rules of the weakest link in the system.</p><p>We define two fixed reference levels for the entire system:</p><ul><li>Uref<em>Uref</em>​ (Upper Reference Level): The maximum reflection allowed for <em>any</em> pigment. This is capped at the peak of the weakest chromatic component (usually Magenta).</li><li>Lref<em>Lref</em>​ (Lower Reference Level): The unified baseline (the “zero” point) for all channels.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xV-7snZ1yvaiG5UPoW_5tA.png" /></figure><p><em>Caption: (Top) Standard CMYK curves: notice the imbalance (Yellow is much taller) and the wide crosstalk areas. (Bottom) SCI Normalized curves: Yellow and Cyan are intentionally “turned down” to match the dynamic range of Magenta.</em></p><h3>The Math Behind the Balance</h3><p>The transformation happens in three distinct stages.</p><p>Step 1: Amplitude Scaling<br>We crush the dynamic range of the strong pigments (Cyan and Yellow) down to the level of the limiter (Magenta).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*YlSPq-8aGDtqujrT8sBZJA.png" /></figure><p>Step 2: Zero-Level Offset<br>We shift the baseline so that the “floor” of the signal is uniform across all channels, removing parasitic reflections.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*-4bWS1F6qXKe0HvJyWQkdA.png" /></figure><p>Step 3: Channel Decoupling<br>Finally, we apply a simple inverse matrix transformation to clean up whatever tiny spectral overlap remains.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*CCmtYOASuHAOiuAsMr4ifQ.png" /></figure><h3>The Trade-Off (Spoiler: The Picture Looks Worse)</h3><p>There is no free lunch in physics.</p><p>If you look at a print made with SCI normalization, the visual image will look dull. The colors will appear flat and low-contrast because we deliberately “turned down” the bright Cyan and Yellow.</p><p>However, to a machine (a scanner or a spectroradiometer), the image is now a high-fidelity data carrier. The channels are separable, deterministic, and robust against the usual noise of paper and ink.</p><p>This makes SCI a powerful tool not for <em>fine art reproduction</em>, but for machine-read data.</p><h3>Why This Matters: Three Applications</h3><ol><li>High-Density Optical Codes: Current QR codes are mostly black and white. Using SCI, we can stack three independent data streams (C, M, Y) into the same physical space without them bleeding into each other. 3x the data in the same square inch.</li><li>Hardcopy Steganography: We can hide a crisp, machine-readable watermark in a printed image that is completely invisible to the human eye (because the SCI layer sits in the narrow dynamic range of the limiter).</li><li>Multi-Layer Visualization: Imagine printing three completely different images on one piece of paper. Under white light, it looks like a gray mess. Under a narrow-band Red LED, you see <em>only</em> Image 1. Under Green LED, <em>only</em> Image 2.</li></ol><h3>Conclusion: Additive Logic, Subtractive Physics</h3><p>The SCI Universal Model demonstrates that inter-channel noise in print is a manageable physical parameter. By moving the normalization step from the <em>decoder</em> to the <em>encoder</em>, we can create a bridge between the clean, orthogonal world of digital screens and the messy, beautiful world of physical ink.</p><p>This is a conceptual framework. The next step is experimental validation with specific pigment sets and spectrophotometers.</p><h3>References &amp; Further Reading</h3><ol><li>Sharma, G., &amp; Bala, R. (2002). <em>Digital Color Imaging Handbook</em>. CRC Press.</li><li>Kang, H. R. (2010). <em>Computational Color Technology</em>. SPIE Press.</li><li>CIE Publication 15:2018. <em>Colorimetry</em>, 4th Edition. CIE Central Bureau.</li><li>Original Preprint: [Link to your <a href="https://preprints.org/">Preprints.org</a> page here]</li></ol><p>Author Bio<br>*Mikhail Kashkarov is a science and art practitioner from Ukraine, specializing in optics and spectral design. He is the inventor of photosensitive painting and has over 25 years of experience in multi-layer encoding on reflective media.*<br>*GitHub: Astra31415926 | YouTube: @Michael-RGB-ART*</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=49cc6336e6b0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A practical model for making RGB channels independent in the real, physical world]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/https-medium-com-new-story-be59cb4c8759?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/be59cb4c8759</guid>
            <category><![CDATA[creative-coding]]></category>
            <category><![CDATA[color-theory]]></category>
            <category><![CDATA[generative-art]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[computer-science]]></category>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Mon, 13 Apr 2026 20:49:47 GMT</pubDate>
            <atom:updated>2026-04-15T11:39:53.163Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*HxzN6JQwUWixb5rYiCqi-g.jpeg" /></figure><p>Most QR codes are black and white for a reason.</p><p>They rely on extreme contrast — black versus white — to remain readable under imperfect conditions. This simplicity makes them robust, universal, and easy to scan.</p><p>But what happens if we try to go beyond that?</p><p>In theory, RGB color channels could triple the data capacity of a QR code. On a digital screen, this actually works quite well. Each pixel emits light independently, and red, green, and blue channels remain cleanly separated.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*0yDpMh4mG6Pd7ggNeGof-Q.png" /></figure><p>But the moment you print that same idea on paper, everything breaks.</p><h3>The Hidden Problem: Crosstalk</h3><p>Unlike screens, printed images don’t emit light — they reflect it.</p><p>Pigments don’t operate in clean, isolated spectral bands. Instead, they have wide “tails” in their spectral response. A red pigment doesn’t just reflect red light — it leaks into green and even blue.</p><p>This creates a fundamental problem:</p><p><strong>color channels are no longer independent.</strong></p><p>Instead of clean separation, we get <strong>crosstalk</strong> — interference between channels that makes reliable decoding extremely difficult.</p><p>Most existing approaches try to fix this at the decoding stage:</p><ul><li>smarter algorithms</li><li>adaptive thresholds</li><li>error correction</li></ul><p>But this approach treats the symptom, not the cause.</p><h3>A Different Idea: Fix the Signal Before It Breaks</h3><p>Instead of trying to decode a corrupted signal, what if we prevent the corruption from happening in the first place?</p><p>This is the core idea behind:</p><p><strong>Spectral Channel Integrity (SCI)</strong></p><p>SCI is a model that normalizes the spectral response of color channels <em>before printing</em>, so that they behave more like independent RGB signals — even in a subtractive medium.</p><h3>The Key Insight: Not All Colors Are Equal</h3><p>Different pigments behave very differently:</p><ul><li>Yellow is bright and reflective</li><li>Blue and violet are much darker</li><li>No pigment is spectrally pure</li></ul><p>This means each channel has a different “strength”.</p><p>SCI introduces a simple but powerful rule:</p><blockquote><strong><em>All channels must be limited by the weakest one.</em></strong></blockquote><h3>The Spectral Limiter</h3><p>We define a <strong>spectral limiter</strong> — the weakest component in the system.</p><p>In practice, this is often a composite violet (magenta + cyan), which:</p><ul><li>strongly absorbs green</li><li>has limited dynamic range</li></ul><p>Instead of trying to boost this weak channel, SCI does the opposite:</p><p>👉 it reduces all other channels to match it</p><p>This creates an artificial balance where all channels operate under the same constraints.</p><h3>Low Contrast / High Fidelity</h3><p>At first glance, this seems counterintuitive.</p><p>We intentionally reduce contrast.</p><p>But in exchange, we gain something more important:</p><p>👉 <strong>clean channel separation</strong></p><p>This leads to a principle I call:</p><p><strong>Low Contrast / High Fidelity</strong></p><p>A weaker but cleaner signal is more reliable than a strong but noisy one.</p><h3>How It Works (Simplified Pipeline)</h3><ol><li>Split data into three QR codes</li><li>Assign each to R, G, B channels</li><li>Normalize all channels using SCI</li><li>Combine into a single RGB image</li><li>Print</li><li>Decode by separating channels again</li></ol><h3>What This Enables</h3><p>This approach is not about replacing traditional QR codes.</p><p>Instead, it opens new possibilities:</p><h4>🎨 Design-first QR codes</h4><p>Color codes that are visually appealing and integrated into artwork</p><h4>🧩 Layered information</h4><p>Different data in different channels</p><h4>🌀 Ornament-based encoding</h4><p>QR codes embedded into complex visual patterns</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*3_Qxg3CyC9ox-JDD_8KTpw.png" /></figure><h3>What It Doesn’t Solve (Yet)</h3><p>Let’s be honest about limitations:</p><ul><li>Requires calibration for printers and inks</li><li>Not universally readable</li><li>Lower contrast than standard QR codes</li><li>Needs a custom decoding approach</li></ul><h3>Why This Matters</h3><p>SCI is not a new encoding format.</p><p>It’s something more fundamental:</p><blockquote><em>a way to make physical color behave like digital color</em></blockquote><p>By controlling the signal before it’s created, we reduce the need for complex correction later.</p><h3>Core Idea</h3><p><strong>SCI is a model of deterministic spectral normalization that restores functional independence of color channels in subtractive media by equalizing them relative to a spectral limiter.</strong></p><h3>Final Thought</h3><p>Sometimes the solution isn’t to build a smarter decoder.</p><p>Sometimes the solution is to stop breaking the signal in the first place.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=be59cb4c8759" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A story of overcoming technical limits in merging physical oil painting with digital data encoding.]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/a-story-of-overcoming-technical-limits-in-merging-physical-oil-painting-with-digital-data-encoding-1c2f5cb216a0?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/1c2f5cb216a0</guid>
            <category><![CDATA[art]]></category>
            <category><![CDATA[computer-vision]]></category>
            <category><![CDATA[research]]></category>
            <category><![CDATA[science]]></category>
            <category><![CDATA[qr-code]]></category>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Sat, 11 Apr 2026 21:21:00 GMT</pubDate>
            <atom:updated>2026-04-11T21:21:00.787Z</atom:updated>
            <content:encoded><![CDATA[<h3>How I Taught Algorithms to See a Color QR Code on Canvas</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HkXL4QD6XI6FE21M3AkJHg.png" /><figcaption>How I Taught Algorithms to See a Color QR Code on Canvas</figcaption></figure><p>Many believe that QR codes are just black-and-white squares. Sometimes they have a logo in the middle.</p><p>But I am working on a different challenge: <strong>creating a color QR code capable of storing more data.</strong></p><p>I should clarify right away — the idea of color QR codes is not new. I am not the inventor of the concept. I am a researcher trying to bring it into physical reality. Because on a screen, everything works perfectly. But in the real world? Not so much.</p><h3>The Beginning: An Illusion of Success</h3><p>It all started with a web generator. On the monitor, everything looked flawless: vibrant colors, instant scanning, and independent RGB channels. But the real challenge began when I tried to transfer this into the physical world.</p><p>I took a canvas, acrylic paints, and manually reproduced one of my generated codes.</p><p><strong>The result? Total disappointment.</strong> The scanner simply couldn’t “see” the code. What worked perfectly in a digital environment turned out to be a blind spot for the camera in reality.</p><h3>Through the Thorns: The Battle for Binarization</h3><p>The following weeks turned into a real investigation. To understand what was happening, I wrote several utility programs — essentially, the “eyes of the algorithm.” They showed me exactly how the machine perceives the image.</p><p>The picture was harsh:</p><ul><li>blown-out highlights</li><li>deep shadows</li><li>everything either pitch black or pure white.</li></ul><p>A complete loss of information. But there was a vital observation: this chaos happened in each channel separately. This meant that the core idea of channel separation actually worked.</p><p>I experimented, tweaked parameters, and came up with workarounds — to no avail. The turning point came when I integrated adaptive binarization — the <strong>Bradley Thresholding algorithm.</strong></p><p>And then, it happened. The code painted on canvas was read. It felt like reaching the summit of Everest.</p><h3>Hunting for Channels</h3><p>But that was only half the battle. The next problem was synchronization. Early versions of the scanner returned a result as soon as it managed to read one or two channels (R, G, or B). But for me, that wasn’t enough — I needed all three.</p><p>I rewrote the logic: now the algorithm doesn’t rush the answer. Instead, it waits for a few seconds, trying to extract information from the third channel. By the way, this isn’t always “Blue,” as many researchers assume. Most likely, they reached those conclusions experimenting with CMYK prints. But I have a completely different model — my own.</p><p>I even started developing an interactive viewfinder that helps the user point the camera correctly.</p><h3>A New Everest</h3><p>Eventually, I realized that while the scanner can be improved infinitely, the real magic lies in the color model used to reproduce the QR code itself.</p><p>Now, I have returned to my long-standing idea of <strong>Ornamental QR codes</strong>. This is more of an Art-concept — merging information with visual form.</p><p>And here, a new problem unexpectedly arose: my new ornamental code, which I also painted with acrylics on canvas, is not yet readable by my own “universal” scanner.</p><p>But this isn’t a failure. This is my new Everest. )))</p><p><strong>Project Links:</strong></p><ul><li><strong>Behance Case Study:</strong> <a href="https://www.behance.net/gallery/247424185">QRGB &amp; Spectral Response Normalization</a></li><li><strong>Web Tool:</strong> <a href="https://astra31415926.github.io/QRGB-Spectral-Encoding/">QRGB Spectral Encoding</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1c2f5cb216a0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Encoding 3 layers of RGB data for Subtractive Media: Beyond the Digital Screen]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/encoding-3-layers-of-rgb-data-for-subtractive-media-beyond-the-digital-screen-b756e8f9f9f7?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/b756e8f9f9f7</guid>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Sat, 04 Apr 2026 16:51:30 GMT</pubDate>
            <atom:updated>2026-04-04T16:51:30.267Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VsLGBH4U94kh1yLdSjngQg.png" /></figure><p><strong>The Problem: Spectral Cross-talk</strong> Standard QR is a binary code. I use three channels (RGB) to store three different datasets in one code. While it works perfectly on screens, physical media (CMYK/pigments) cause the layers to “bleed” due to light absorption by paint.</p><h3>Solution: The Model &amp; The Scanner</h3><h4>1. Spectral Response Normalization (Model)</h4><p>A mathematical algorithm that compensates for pigment distortion. It calculates how specific paint absorbs light and corrects the data so that the subtractive environment doesn’t break the RGB logic.</p><h4>2. Bradley Adaptive Thresholding (Scanner)</h4><p>The technical solution for scanning. The Bradley Thresholding algorithm allows the scanner to ignore surface textures — brushstrokes, acrylic glare, and canvas shadows. Data is extracted correctly even from “live” paintings.</p><h3>Key Facts</h3><ul><li><strong>3-in-1:</strong> Three independent data layers in one physical object.</li><li><strong>Science-Art:</strong> Bridging fine art with deterministic data integrity.</li><li><strong>Open Source:</strong> Technology available for both artists and engineers.</li></ul><h3>Resources</h3><ul><li><strong>Demo/Scanner:</strong> <a href="https://astra31415926.github.io/QR.G.B.-ART/">Web Interface</a></li><li><strong>GitHub:</strong> <a href="https://github.com/Astra31415926/QR.G.B.-ART">Source Code</a></li><li><strong>Paper:</strong> <a href="https://zenodo.org/records/18895764">Zenodo</a></li></ul><p>#Science #Art #Technology #OpenSource #ColorTheory</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b756e8f9f9f7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[QRGB-Spectral-Encoding/: Breaking the Data Density Barrier with Spectral Computing]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/qrgb-spectral-encoding-breaking-the-data-density-barrier-with-spectral-computing-a4fbbad9fa19?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/a4fbbad9fa19</guid>
            <category><![CDATA[innovation]]></category>
            <category><![CDATA[optics]]></category>
            <category><![CDATA[computer-vision]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Tue, 31 Mar 2026 11:46:25 GMT</pubDate>
            <atom:updated>2026-03-31T11:46:25.732Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*khZ6td0jd1Pk7uzmtYR1kg.png" /></figure><p>The dream of multi-layered color encoding (Color QR) has been around for decades. By utilizing independent Red, Green, and Blue channels, we can theoretically triple the data capacity of a standard 2D code. Yet, it remains impractical for physical media.</p><h3>The “Crosstalk” Problem</h3><p>The primary barrier is <strong>spectral cross-talk</strong>. When transitioning from digital RGB to physical CMYK printing, channels interfere with each other. Standard printers cannot perfectly isolate spectral components, leading to data corruption during decoding. Previous attempts to solve this via AI “guessing” at the camera level failed to provide industrial reliability.</p><h3>The Innovation: QRGB Spectral Encoding</h3><p>I propose shifting the solution from the decoder to the encoder. Instead of compensating for poor print quality at the camera level, we must ensure <strong>spectral integrity</strong> during the printing process itself.</p><p>My work introduces the <strong>SCI (Spectral Response Normalization) Model</strong>. This is a deterministic mathematical framework designed to minimize channel interference at the physical level.</p><ul><li><strong>Theory (The SCI Model):</strong> Published on Zenodo (<a href="https://zenodo.org/records/19301486">DOI: 10.5281/zenodo.19301486</a>), it provides the equations to calibrate printing for spectral purity.</li><li><strong>Implementation:</strong> The <strong>QRGB Spectral Encoding</strong> system demonstrates triple data density in digital environments.</li><li><strong>The Shift:</strong> By ensuring high-integrity spectral printing, decoding becomes a simple technical task rather than an impossible algorithmic challenge.</li></ul><h3>Moving Toward Physical Validation</h3><p>The math is ready, but it requires experimental verification in real-world printing environments.</p><p><strong>I am seeking R&amp;D partners with access to:</strong></p><ol><li>High-precision spectrometers.</li><li>Industrial-grade printing equipment.</li><li>Laboratory environments for light-response testing.</li></ol><p>The goal is to move QRGB from a “digital sandbox” to a robust technology for Optical Security, Logistics, and Science-Art.</p><h3>Explore the Project:</h3><ul><li>🔗 <strong>Interactive Demo:</strong> <a href="https://astra31415926.github.io/QRGB-Spectral-Encoding/">https://astra31415926.github.io/QRGB-Spectral-Encoding/</a></li><li>📊 <strong>Full Research Paper:</strong> <a href="https://zenodo.org/records/19301486">https://zenodo.org/records/19301486</a></li></ul><p><strong>Mykhailo Kashkarov</strong> | <em>Technical Inventor &amp; Science-Art Researcher</em> 📩 <a href="/service/mailto:colqrnament@gmail.com">colqrnament@gmail.com</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a4fbbad9fa19" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SCI Universal Model: Deterministic Spectral Response Normalization]]></title>
            <link>https://medium.com/@mykhailo_kashkarov/sci-universal-model-deterministic-spectral-response-normalization-313bd3e67811?source=rss-3986cd562804------2</link>
            <guid isPermaLink="false">https://medium.com/p/313bd3e67811</guid>
            <dc:creator><![CDATA[Mykhailo Kashkarov]]></dc:creator>
            <pubDate>Fri, 06 Mar 2026 20:33:43 GMT</pubDate>
            <atom:updated>2026-03-07T00:16:42.306Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Mikhail Kashkarov</strong></p><h4>1. Definition and Concept: Adaptive Orthogonalization of a Subtractive Medium</h4><p>The SCI method (Spectral Channel Integrity) is a model of deterministic normalization of spectral response that forms a balanced color system based on CMYRGB hues (Cyan, Magenta, Yellow, Red, Green, Blue). The model adapts the fundamental property of the additive RGB environment — the orthogonality of channels (their functional independence) — for implementation on a physical subtractive medium.</p><p>SCI represents a structured configuration of the color wheel in which the spectral reflection functions of pigments are amplitude-aligned, ensuring operational separation of RGB measurement channels during reflected reading. In essence, SCI is additive logic implemented through subtractive physics. The model introduces a controlled compression of the brightness dynamic range in order to achieve deterministic physical separation of information channels.</p><h4>2. Physical Limitation: Spectral Inconsistency of the Subtractive Medium</h4><p>The main limitation when transferring information from an additive environment (light-emitting systems) to a subtractive environment (light-absorbing pigment systems) is the spectral inconsistency of pigments. In the idealized additive model, an inactive channel equals zero, while active channels are formed by independent light emitters.</p><p>In a subtractive system, color is determined by the absorption function of pigments, which forms a complex spectral reflection curve. Suppression of adjacent wavelengths is never complete, so inactive channels retain a non-zero residual spectral response. Furthermore, different pigments have different maximum reflection amplitudes within their target spectral ranges. The absence of a unified amplitude reference creates an energy imbalance between system components.</p><p><strong>As a result:</strong></p><p>The minimum state of the system (“zero”) is not an absolute zero;</p><p>The maximum amplitude (“one”) differs across pigments;</p><p>Cross-talk between channels becomes a variable quantity.</p><p>Such a structure lacks a unified reference scale and therefore does not allow deterministic channel organization.</p><h4>3. SCI Method: Formation of Reference Levels</h4><p>To eliminate amplitude imbalance and inter-channel dependence, the SCI model introduces a unified reference scale defined by two fixed system states:</p><ul><li><strong>Upper Reference Level:</strong> A single maximum permissible spectral response level is established for active channel states. The reflection amplitude of all pigments is limited by this level, which unifies the maximum signal of the system.</li><li><strong>Lower Reference Level:</strong> A unified minimum reflection level is defined corresponding to the residual spectral response of the limiting component. This level is accepted as the normalized baseline state (logical “0”) for all channels.</li></ul><p>As a result, all components operate within a unified amplitude scale, and cross-channel contribution becomes a structurally consistent and predictable quantity.</p><h4>4. Violet as a Spectral Limiter</h4><p>In the SCI system, the role of the limiting component is played by a violet pigment due to its minimal spectral response amplitude among the chromatic elements of the system. Violet simultaneously reflects the edge regions of the spectrum (red and blue) with limited intensity and does not provide full suppression of the central (green) region.</p><p>This combination produces the smallest operational amplitude range among system components. Its spectral efficiency defines the physical limits of the SCI model’s reference levels. The violet component does not introduce an artificial constraint but rather formalizes the natural physical minimum of the reflective medium.</p><h4>5. Areas of Application</h4><p>The deterministic amplitude structure of SCI enables the use of reflective surfaces as normalized information carriers. Potential applications include:</p><ul><li><strong>Spectral data encoding</strong> — increasing the density of information in 2D codes through channel separation;</li><li><strong>Authentication protection</strong> — creation of hidden spectral markers;</li><li><strong>Multilayer visualization</strong> — placing independent images on a single surface with selective spectral reading.</li></ul><h4>6. Future Development of the Model</h4><p>Further research will focus on:</p><ul><li><strong>Dynamic range optimization:</strong> Analysis of two-channel configurations (for example using Cyan as the limiter) to increase contrast.</li><li><strong>Biophysical aspects of perception:</strong> Development of structures that take into account the spectral sensitivity of the human visual system, including visually hidden layers that appear only under narrow-band illumination.</li></ul><p>The SCI model forms a conceptual framework for transitioning from a physically inconsistent reflective medium to a deterministic channel structure. Practical implementation and quantitative optimization of system parameters remain subjects for further experimental research.</p><p><em>© 2026 Mikhail Kashkarov, Ukraine</em> <em>September 2026 — first public version of the document</em> <em>All rights reserved.Full theoretical model available on Zenodo: </em><a href="https://doi.org/10.5281/zenodo.18895764"><em>[посилання]”</em>.</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=313bd3e67811" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>