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 Claude Code. The current live products are the NSS Background Remover and Novus Visualizers (both free, in-browser tools). Each product runs on its own subdomain; this hub explains how they fit together.

Yes. The Background Remover and Novus Visualizers are free to use, with no subscription and no paywalled tier. They are supported by non-intrusive display advertising rather than charging users.

No account is required to use the Background Remover, and Novus Visualizers can be used without signing up to test it. The tools are designed to let you get to a result immediately rather than gating the core workflow behind registration.

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, 3D/4D engines, effects, captions, exporting, plus saving projects, albums, the community feed, Creator Studio, the templates library, 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 has a full image editor and video editor, 2×/4× AI upscaling for images and video, and 25+ image and video utility tools (resize, compress, convert, optimize, stabilize, 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 111 engine families, each available in 2D, 3D, 4D, and Advanced modes, with multiple variants per mode — adding up to thousands of distinct presets, plus hundreds of ready-made templates, one-click color themes, and up to four stacked layers per project for custom looks.

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 Album Art Editor, 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.

The free tier includes ten exports per month. Editing, saving projects and albums, and taking part in the community are not metered the same way. For most releases ten exports is plenty, but if you are cutting a single song into many formats in one session it is worth planning your final renders rather than spending exports on test passes.

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, templates, and community presence. Every visualizer is copyright-free, so you keep full ownership of your exports — including any you post.

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 Vercel Analytics); without consent, nothing 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 Claude Code. 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.