Hi Andrej, Thank you for getting back to me. I see what you mean, yeah the content creation community is rather small at the moment. I do have one other idea that I've been thinking about, it'd be an OOT module but it's something that a few communities I'm part of have mentioned interest in and I think might be worthwhile: ft8/WSPR implementations.
Corey had made this a few years ago: https://github.com/ckoval7/automate_ft8 and my idea is in a similar vein. Whereas he is focused on automating QSOs this would be more creating a modular/reusable FT8 OOT module for GNU Radio. Basically let people incorporate FT8 functionalities into their custom GNU Radio applications. There would have to be two main parts: encoding and modulating FT8 messages then demodulating and decoding, an integrated FT8 modem would be the end goal essentially. I'd expect gr-digital and gr-fec to be basically the main parts of GNU radio I'd use right now. But yeah this is the general skeleton of what I'm thinking, I can write up a super detailed proposal like last time if you'd like, let me know. I will need to do some homework if you think this is a good idea though, April 8th is the deadline so hope to talk soon! Daniel P.S. If you think this is too large a project/not the greatest idea, I'm open to the "Revitalize in-tree and out-of-tree (OOT) modules" project idea as well. Not having support for some of these is probably annoying for a lot of people so I see the value of the idea. I'd just appreciate more details when the time comes around. On Tue, Apr 1, 2025 at 4:55 AM Andrej Rode <ar...@gnuradio.org> wrote: > Hi Daniel, > > Sorry for the late response, we have looked at your proposal and discussed > about the idea and implementation. > Generally your idea seems worthwhile, for the GNU Radio project we are > mostly lacking content regardless of the form. > > So if we have a great tool to create videos with GNU Radio we would still > need someone to do that. And generally we are lacking the “someone”. > Another point is that this proposal is quite large and complex in nature > and needs a lot of integration with existing parts of the project. I think > many of those features > are already sufficiently captured with screen recording. Coming back to > the point that we are still lacking content production and not the means of > it. > > Thanks for creating such a detailed idea proposal, maybe there is some > other topic that interests you and you want to implement within GNU Radio? > > Cheers, > Andrej > > > On 13. Mar 2025, at 02:04, Daniel Paul <danielpaul1...@gmail.com> wrote: > > > > Personal Information > > Name: Daniel Paul > > Location: Winnipeg, Manitoba, Canada > > Email: danielpaul1...@gmail.com > > University: University of Manitoba > > Degree Program: B.Sc. in Data Science (minor: Chemistry), expected April > 2025 > > GSoC Project Duration: Full-time, 350 hours over 12 weeks > > GNU Radio Companion Video Creation Toolkit (GSoC idea) > > So the idea I have is to develop a GNU Radio Video Creation Toolkit that > streamlines the process of creating high-quality YouTube videos and > tutorials demonstrating GNU Radio flowgraphs, SDR experiments, and signal > processing concepts. This toolkit will integrate flowgraph exporting > components, visualization, screen recording, dynamic annotations, and video > post-processing to provide a more seamless content creation experience. > > So quick about me: I’m a data science student and I have my HAM advanced > and basic radio licenses. I also have a RTL-SDR and a HackRF One dongle. > Finally, I also have created videos for my youtube channel related to SDR > stuff that you can see below. https://www.youtube.com/@agarauproductions > > Also, one point I’ll make, this isn’t an SRS so if you’re going to ask > about specific implementation details, while I might have ideas, I haven’t > fleshed them out just yet -- but I've given general directions I hope to go > in. > > Now back to the idea. > > Deliverables (to be discussed with mentor on how many to focus – all > would be amazing though): > > 1. High-Quality Flowgraph Exporting > > · Generate formatted screenshots and animated sequences of GNU > Radio flowgraphs. > > · So, imagine if you can extract high quality images of each > component you’re building out, ie. extract each block and arrow. Each part > without the background associated with it, so you can just pick out the > parts you want to export, and you can keep those in the highest quality so > no more of those pesky screenshots if you don’t want them. And give an > option to be more colourful than just white/back when you export. > > 2. Screen Recording with Interactive Overlays > > · Capture screen activity with real-time annotations (e.g., > highlighting, text pop-ups, guided steps). Of course, existing tech allows > you to draw on top of anything on your screen, but if this was embedded in > GNU that would be super useful for tutorials. > > 3. Signal Visualization Recording > > · Capture real-time spectrograms, FFTs, and time-domain plots > directly from GNU Radio. Again just export them directly, in fact I saw the > CyberEther project, maybe they could embed some of those useful features > here too, overall just making the lives of tutorial makers easier. > > 4. Automated Narration & Voice-over Syncing > > · Enable voice recording and auto-align narration with flowgraph > execution steps. Again, this is possible but if you treat this as step 2a > then you don’t have to worry about manual synchronization anymore. > > 5. *Template-Based Scripted Video Creation > > · Offer reusable templates for common tutorial structures, > including recorded steps and visual aids. *I haven’t really worked this one > out yet, but sort of implementation of this would be useful though. I hope > I can talk more about this with someone. > > 6. Export Features > > · A way to export the things I previously mentioned with Youtube > and existing major products that are free to use. The idea being support > the open source community while also making these steps easier overall for > everyone. > > Benefits to the Community > > Creating tutorials and educational content for GNU Radio currently > requires so much manual effort, including taking screenshots, editing > videos, and syncing voiceovers. This toolkit will significantly reduce the > time and complexity here. If we make video creation more accessible, it > will encourage more people to share their expertise and increase the > adoption of GNU Radio and just support the community in general. > > > > Tech stack (most likely): Python, OpenCV, FFmpeg, PyQt. > > What the ideal schedule looks like: > > Phase 1 (Week 3-6) > > Implement flowgraph exporting features and initial screen recording > capabilities. > > Phase 2 (Week 7-10) > > Develop automated narration, voice-over syncing, and visualization > recording. > > Final Phase (Week 11-12) > > Complete template-based video creation, finalize documentation, and test > stability. > > > > > > Background & Coding Experience > > I am a Data Science student at the University of Manitoba with a strong > background in machine learning, software development, and signal > processing. I hold an Advanced Amateur Radio Certification and have > hands-on experience with RTL-SDR and HackRF One. Additionally, I create > SDR-related educational content on my YouTube channel: Agarau Productions. > > Relevant Projects > > • Machine Learning for MRI Classification: Developed a CNN-based > model for early-onset Alzheimer’s detection. > > • Biometric AI (Displaid): Applied signal processing techniques to > analyze raw voltage signals. > > • SSIS Data Pipeline Development: Built an automated data migration > and processing pipeline in SQL and Python. > > I have read and understood the three strikes rule and GNU Radio’s > guidelines > > Secret Code Word: Cyberspectrum is the best spectrum. > > > > Hope to talk with someone about this! > > Thank you, > > Daniel > >
