FAQ & support

Novus Stream Solutions builds free apps that rival paid alternatives — removing the paywalls that block exporting or using a feature, and built with modern AI tooling. The current live products are the NSS Background Remover and Novus Visualizers, and Novus PDF Studio, and Novus Convert, and Novus Learn, and Novus Examples, Novus AI Stats, and early-access Novus Striker (free, in-browser tools). Each product runs on its own subdomain; this hub explains how they fit together.

Yes. The NSS Background Remover, Novus Visualizers, Novus PDF Studio, Novus Convert, Novus Learn, Novus Examples, and Novus AI Stats are free to use, with no subscription blocking the core workflow. Novus Striker is free to play in early access. Advertising helps support the sites rather than forcing users through an upgrade screen.

No account is required to start using the NSS Background Remover, Novus Visualizers, Novus PDF Studio, Novus Convert, Novus Learn, Novus Examples, or Novus Striker. Novus AI Stats is the exception: because it stores a history of your own AI usage, that dashboard is account-based. Optional account features exist elsewhere only where saving work is useful.

Start at the help centre or the contact form for hub-related questions, and the documentation for step-by-step product detail. Each app also has its own help pages on its subdomain.

The tutorials section has 50+ step-by-step walkthroughs across both apps — from a first cutout or first visualizer to deep dives on the image and video editors, edge refinement, upscaling, the utility tools, 3D/depth relief, project saves, custom tuning, and collaborative sessions, and every part of Novus Visualizers (layers, beat-sync and multi-band sync, the audio engine, the visual engines, effects, captions, exporting, plus saving projects, albums, the community feed, Creator Studio, the Studio Workstation, and the companion tools). Each app has its own hub at /tutorials/bg-remover and /tutorials/visualizers. For a complete map of every tool with a short description, see the tool maps, the documentation for the full technical reference, and search the whole hub at any time. The apps themselves also hold deeper, always-current help.

It is a free AI image tool at bgremover.novusstreamsolutions.com that removes image backgrounds entirely in your browser and exports a true straight-alpha PNG. Full detail lives in the Background Remover documentation.

No. There is no backend image-processing server and no API call that sends your image anywhere. The AI model runs entirely on your device, so your images and videos never leave your browser. It is built as a progressive web app whose model weights are cached locally, which is why it can keep working offline after the first visit.

Two models run locally through Transformers.js: a Fast model (RMBG-1.4, about 80 MB) optimized for product shots and portraits with clean edges, and a Best Quality model (RMBG-2.0, the BiRefNet-based bilateral reference network, about 180 MB) for fine hair, fur, transparent objects, and complex edges. Each model downloads once on first use and is cached in your browser.

Most free tools export premultiplied alpha because it is simpler, which produces dark or white halos when you open the file in Photoshop, Figma, or a print layout. The NSS Background Remover writes true straight (non-premultiplied) alpha, so the RGB of transparent pixels is preserved and the cutout composites cleanly with no fringing. There is a deeper explanation in the documentation.

Input formats include PNG, JPG/JPEG, WebP, AVIF, and HEIC (iPhone photos). Output is PNG, WebP, or AVIF with a true straight-alpha channel. Images up to 4096 by 4096 pixels are processed at full resolution; larger images are downscaled for inference and the full-resolution mask is applied back to the original.

Alongside still-image background removal, the app removes backgrounds from video and GIFs, and has a full image editor and video editor, 2×/4× AI upscaling for images and video, and a utility set totalling 35 tools (resize, compress, convert, optimize, stabilize, colorize, inspect metadata, and more) — all running in the browser using the same local-processing approach.

Chrome, Edge, and Opera (recent versions) give the best experience with WebGPU acceleration. Firefox and Safari work through WebAssembly. No special GPU is required — WebGPU makes inference faster where available, but every tool automatically falls back to WebAssembly on the CPU.

No. There are no usage limits, no watermarks, and no premium tier. You can process as many images as you like. The tool is free and supported by non-intrusive display ads.

The models are modest and download once on first use, then cache in your browser: a Fast background-removal model (RMBG-1.4, about 80 MB), a Best Quality model (RMBG-2.0/BiRefNet, about 180 MB), and a separate super-resolution model for AI upscaling. A built-in check probes your device (WebGPU support and available memory) so tools run on the best available path and fall back to WebAssembly on lighter hardware. You can see how much model data is cached and delete it at any time.

Novus Visualizers is a free, in-browser music-visualizer app at visualizers.novusstreamsolutions.com. You upload a track, choose a visual style, customize it, and export a finished video. Full workflow detail is in the Visualizers documentation.

You can upload MP3, WAV, OGG, or M4A files. The app analyzes the track in real time using the Web Audio API — a 32-band FFT plus beat, onset, and loudness detection — so the visuals respond to the actual rhythm and energy of your music.

There are 13 engines — Particles, Trails, Bloom, Character Motion, Radial, Spectrum, Bars, Tunnel, Waveform, Logo, Lattice, Terrain, and Stage Lights — each with 6 modes, for 78 modes in total, that you tune with colors, geometry, direction, and audio bindings. Two editors share one saved document: the Classic Editor for a guided, beginner-friendly setup and the Studio Workstation for direct canvas work with ordered layers, a beat-aware timeline, keyframes, and blend modes.

Closely. On top of real-time 32-band FFT analysis with beat, onset, and loudness detection, multi-band beat sync lets bass, mid, and treble each trigger their own effects — so a heavy element can pulse with the kick while finer details sparkle on the hi-hats, rather than the whole frame strobing on one global beat. The result tracks the arrangement of a track, not just its downbeat.

Export runs entirely in the browser using WebCodecs — no upload and no server render queue. You can export MP4 (H.264) or WebM (VP9) up to 4K at 24, 30, or 60 fps, with platform presets for YouTube, TikTok/Reels, Instagram, Spotify Canvas, X, and Discord.

Yes. The optional AI captions feature uses an on-device Whisper model running through Transformers.js, so your audio never leaves your browser. Transcriptions can be edited per word and styled with preset caption looks.

You do. Exports are copyright-free — you can use them commercially or personally without restriction and without attribution. The app is free and ad-supported.

Yes. Alongside the main visualizer, the platform includes an Cover Art Studio, an Audio Effects Generator, a Lyric Video Creator, Stream Overlays, and an Audio Library.

No — browsing the community, opening the editor, customizing a visualizer, and exporting all work without signing in. A free account is optional and exists so your work can be saved: it adds the Creator Studio dashboard, durable saved drafts and projects, albums, and the ability to take part in the community feed. Signing up is free.

When you are signed in, drafts and projects are saved durably to your account and reopen in the editor with all settings intact — and because they live with your account rather than only in one browser, you can reopen them on another device. Albums group related visualizers (a release, an EP, or a campaign) so they can be managed as a set from the Creator Studio dashboard. The heavy compute (analysis, rendering, and export) still runs locally on your device; only the project data and your community presence are account-backed.

All of them. Export is free with no watermark and no monthly quota — you can render as many videos as you like, as many times as you like. Editing, saving projects and albums, and taking part in the community are free too. Because rendering runs entirely on your own device, there is nothing to meter.

The community feed is a public gallery of visualizers made by creators, with Feed, Trending, Following, and Saved views, genre filters, and sorting by Recent, Most Liked, or Most Viewed. Browsing is free; liking, saving, and posting require a signed-in account. Creator Studio is your dashboard for managing your own visualizers and community presence. Every visualizer is copyright-free, so you keep full ownership of your exports — including any you post.

Novus PDF Studio is a free, in-browser PDF form editor at pdf.novusstreamsolutions.com. Upload a PDF, scan it for likely fields, refine the placement by hand, draw a signature, and export or print the finished document. Alongside the editor it ships twelve page-level tools — merge, split, organize, rotate, page numbers, protect, unlock, compress, flatten, PDF-to-images, images-to-PDF, and OCR. The full map is in the PDF Studio documentation.

No. The editor and every page tool run in your browser. The file you open, its page contents, any password you type, your signature, and the exported result all stay in the tab for the length of the session — there is no server that receives documents. Closing the tab discards the document.

The scan merges three sources: form fields already embedded in the PDF, boxes and underlines detected in the page geometry, and labels detected in the text. The results are de-duplicated and placed as ordinary editable fields, so nothing is locked in. If the scan misses a field, add one from the toolbar or run Rescan to merge newly detected fields; if it adds an extra one, delete it. Every field — scanned or manual — can be moved, resized, restyled, or removed, and you can clear only the AI fields without losing the ones you placed yourself.

Yes, for documents you are authorized to open. Unlock takes the password and verifies that the result reopens without credentials; the other tools ask for a password when a protected input is given and modification is permitted. Protect goes the other way and writes AES-256 encrypted output — note that permission flags inside a PDF are only ever advisory, since enforcing them is up to whichever viewer opens the file.

No — and this is worth being precise about. Highlight snaps a marker stroke to each line of text it covers, and erase covers the text under your selection with a page-coloured box. On export that box is opaque, but the rest of the document keeps its selectable text layer, so erase is a visual cover rather than a removal. If you need content genuinely gone from a file, remove it in the source document before exporting. Redaction, compression, format conversion, and metadata cleanup are not features of PDF Studio today.

Yes. Exports are vector-based, so text stays crisp and selectable, highlights use the same multiply blending you see on screen, and signatures embed exactly as drawn. Text fields use the PDF-safe built-in fonts — Helvetica, Times Roman, and Courier — which is what keeps a finished form rendering the same in every viewer.

Free, with no account, no subscription, and no export limit — the editor and all twelve tools are funded by advertising rather than by metering the work. Because documents are processed on your device, they are never seen by us or by the ad network and never influence which ads appear. It works on current Chrome, Edge, Firefox, and Safari on desktop and mobile; large PDFs render best on desktop, though the editor has touch controls for phones and tablets. Government, benefit, and legal forms are the main use case — always review field placement, values, dates, and signatures before you submit anything official. The getting-started tutorial walks through a full form.

Novus Convert converts images (including camera RAW), video, audio, documents, ebooks, vector/CAD/3D, and fonts — all locally in your browser, with no upload. It also compresses images, video, and audio, and runs a validated mixed batch queue where each file keeps its own output and downloads independently. The live /conversions directory lists every active route.

No. Every conversion currently exposed in the interface runs in browser memory. Temporary result URLs disappear when a job is removed or the page closes.

Images are limited to 100 MB and 80 megapixels, structured text to 25 MB, and archives to 200 MB compressed, 5,000 entries, and 512 MB expanded. Per-format daily download allowances also apply and are shown in the live interface.

It checks the expected signature, container, or decodability before enabling download. Malformed, encrypted, unsupported, over-limit, or invalid files fail without changing the original.

Novus Learn is a free, source-grounded learning platform at learn.novusstreamsolutions.com. Bring it a source through any of nineinput modes — a Wikipedia topic, a URL, a whole indexed site, an open-access paper, a video, a textbook, an upload, a coding path, or a JavaScript codebase — and it builds a summary, study notes, flashcards, a quiz, a concept map, a guided lesson, and visuals out of that source's own sentences, every claim cited. Alongside the study pipeline it also carries a careers layer of 565 occupations, 376 aptitude practice suites, 2,200 cognitive practice questions, 200 puzzles, the Novus Reasoning Challenge, and a 12-module JavaScript course.

No — that was true at the June 2026 launch and has not been true for a while. Wikipedia is one of nine input modes; the others are a URL, a whole indexed site, an open-access paper via Europe PMC, a video, a textbook, an uploaded file, a coding path, and a JavaScript codebase. On top of that sit four layers that have nothing to do with Wikipedia at all: careers, aptitude and cognitive practice, the Novus Reasoning Challenge and puzzles, and an authored JavaScript course.

No. Novus Learn is extractive, not generative. It does not fabricate answers or run a paid AI model — summaries, notes, flashcards, quizzes, concept maps, and visuals are assembled from the sentences of whatever source you gave it, shown with source identity and claim-level citations, so you can trace each claim back to where it came from. When the source cannot answer a question, it says so instead of filling the gap.

565 occupations grouped into 20 families at /careers/families, each with pathway overviews, qualification routes, and links to the official bodies the information came from. The career matcher comes in quick, full, and career-change forms, runs entirely on your device, and suggests families rather than pronouncing a single verdict — no account, and no result leaves your browser.

Yes. The aptitude directory carries 376 practice suites across 24 sectors and 35 constructs — 22,560 keyed items in total — and the cognitive skills section adds 2,200 questions across 22 construct-specific banks, with printable packs. It is practice material with published editorial standards, not a diagnostic, and it does not replace a supervised assessment.

No, and it is deliberately not named like one. It is 9 public forms plus a practice lab of 12 modules, built as an entertainment and educational instrument. It returns no clinical score and is not a substitute for a supervised assessment — the challenge's own methodology and FAQ pages set out exactly how items are built and what a result does and does not mean. Alongside it sit 200 untimed puzzles across 8 categories, each with an explanation that teaches the method rather than just revealing the answer.

Yes — an authored JavaScript course of 12 modules and 77 lessons at /languages/javascript, running from values and types through to a capstone, with exercises graded in a sandbox on your own device. Languages without an authored course get a curated index into their official documentation instead, and each page says which one you are looking at.

No account is required, for any of it. Novus Learn is local-first: study projects, career matcher results, and practice progress are all held in your own browser, so you can open a topic or start a practice suite immediately without signing up.

Yes. Novus Learn is free to use, local-first, and loads no third-party trackers on its content pages. There is no account requirement and no paid tier blocking the study pipeline, the careers layer, the practice suites, the puzzles, or the JavaScript course.

Novus Examples is a free library of sample files at examples.novusstreamsolutions.com. Think of it as a Lorem Ipsum for files — 4,075 ready-made sample files across twenty-four categories that you download to test tools, workflows, and integrations, plus 840 editable visual templates and fifty-two live test targets.

Yes. Novus Examples is free to use, with no account and no paid tier blocking any download. It is browser-first, so you pick a file and download it straight away.

4,075 files across twenty-four categories and 267 extensions — images, video, documents, audio, archives, 3D models, fonts, security fixtures, and structured text such as JSON, CSV, and TSV. Each sample is a valid file generated to an exact specification, so it ships with a spec sheet describing what is inside it.

Yes — 401 of them, making video the second-largest category. They cover fourteen codecs, seven containers stream-copied from one encode so only the wrapper differs, pixel formats, playable HLS and DASH packages, caption files across awkward encodings, chapters and multi-track accessibility fixtures, intentionally-corrupt containers, and ten AI-video suites where each degraded clip is paired with the pristine clip it came from.

Yes. The visual template library holds 840 editable templates across 42 categories, four of which are broadcast categories — live stream overlays, stream alerts and motion frames, stream scene backgrounds, and avatar state sets. Those 80 templates each ship an OBS-compatible scene collection alongside the image downloads, so importing the JSON gives you a scene with the overlay already placed above a capture placeholder. Novus Stream Solutions is not affiliated with or endorsed by the OBS Project.

Anything that takes a file as input — converters, uploaders, parsers, and editors. The samples pair naturally with Novus Convert for trying a conversion and the NSS Background Remover for testing image workflows, so you can check a tool with a known file before using your own.

No account is required. Novus Examples is browser-first, so you can browse the library and download any sample file immediately without signing up.

Novus AI Stats is a privacy-first PWA at aistats.novusstreamsolutions.com. It imports ChatGPT, Claude, Gemini, Claude Code, Codex CLI/Desktop, Gemini CLI, and Cursor history from user-approved exports and files so you can track sessions, tokens, projects, and milestones in one dashboard.

Seven sources are supported today: ChatGPT, Claude, Gemini, Claude Code, Codex CLI/Desktop, Gemini CLI, and Cursor. Each one is documented with its accepted inputs and its honest limitations on the app's integrations page. Anything else you see on the roadmap is listed as planned — a roadmap entry is not an ingestion claim.

It is live and still in progress — imports, metrics, and share cards continue to expand. Free to use, with no payment information required. Viewing your own history needs an account: open the sign-up page to see whether registration is currently accepting new members, and use the contact form if it is closed.

Free to use, with no payment information required. Viewing your own history needs an account: open the sign-up page to see whether registration is currently accepting new members, and use the contact form if it is closed. The sign-up page is the authoritative answer at any given moment.

Exports and logs are parsed transiently in your browser, and only the normalized statistics are stored against your account. Raw prompts, responses, code, attachments, source files, full transcripts, and local file paths are never persisted — not on your device and not on the server.

No. The PWA cannot watch local folders in the background. Every import starts with an explicit file or folder selection.

Yes. No payment information is required, and advertising stays off authenticated analytics pages.

Novus Striker is a fast arcade soccer game at striker.novusstreamsolutions.com. Play in the browser with no download and no install. It is currently in early access / active development.

No. It is in active development. You can play the early build and send feedback while features and feel continue to change.

No. Novus Striker runs in the browser. The game streams and caches as you play.

The in-browser tools process your files on your device and do not transmit them to any server. With your consent, we collect anonymous page-view counts (via Google Analytics 4, with IP anonymization enabled); without consent, no analytics script is loaded at all and nothing is collected beyond standard HTTP request headers. Full detail is in the privacy policy.

The AI models are cached locally so the app can run offline, but your source files are not uploaded anywhere. Any saved projects or history are stored locally in your own browser, not on a Novus server. Nothing you load — image or video — is saved or transmitted off your device.

Where a feature requires an account, the apps support secure email/password authentication. Passwords are stored as hashes (never plain text), and security controls are enforced at the application level. Some integrations may use a provider authorization flow where the external platform requires it.

Dated ecosystem milestones are on the changelog, and deeper write-ups live on the product blog. Each product also publishes its own release notes on its subdomain.

Our filter is simple: build free apps that rival paid alternatives, removing the paywalls that block exporting or using a feature — all built with modern AI tooling. We ship a focused version, watch real usage (activation, not vanity signups), and let that evidence drive what to improve next. The reasoning is in how we ship and test small apps.