MakeACopy uses reproducible OpenCV/ONNX builds and runs Tesseract OCR fully on-device.
The PDF text layer is generated with Apache PDFBox for full searchability.
All binaries are F-Droid–compliant and deterministically reproducible.
Happy to discuss the OpenCV/ONNX integration or how the deterministic build setup works.
Glad to hear you're planning to check out the app! It includes scale training for major, minor pentatonic, blues, and a few more. If you notice any scales missing that you'd like to see added, feel free to reach out — I'm always happy to improve it based on feedback.
Thanks! That sounds like a cool idea too. In Let’s Bend, I also included a scale training mode, where users can practice playing along with different scales. Maybe that could be useful for your project as well. Feel free to dig into the source code, and let me know if you have questions!
Thank you so much for the kind words – I’m really happy to hear you found it helpful, even for violin!
Let’s Bend was originally designed to support harmonica bending practice, but the idea of using the "stay in the square" feedback as a pitch visualizer for other instruments is incredibly exciting. It’s great to see it working in real musical contexts!
Just to address your points:
There’s already a *“Keep screen on”* toggle in the app settings – enabling it should prevent Android from turning off the screen during use.
You can also *change the root key and tuning system* in the settings. While the interface is still harmonica-focused, this makes the app more flexible – including for lower pitches like a violin's low G.
I’m genuinely grateful for your feedback and enthusiasm – and thanks for trying the app in such a creative way!
Thanks — great to hear you liked it, and that’s a really smart suggestion!
Actually, you've inspired a related idea I’ve been thinking about: since the app already shows a magnified pitch view for each target note, I could track how accurately each target is hit — not over time, but per note. That means collecting cent-deviation stats every time a target pitch is active.
For example, if you’re working on bending hole 3 draw down to F#, the app could keep track of:
- how many times that target was attempted,
- your average deviation (in cents),
- and your best attempt so far.
It’s not a full frequency-vs-time graph, but it's very focused feedback — and could motivate practice in a very concrete way. I’d love to make this part of a future version. Thanks again for the spark!
Thanks so much for this detailed and generous comment — it’s incredibly encouraging to hear from someone with your level of experience.
Yes, the email listed on the Imprint page is the best way to reach me — I’d love to hear your suggestions when you get a chance.
And it’s great that you brought up Melody Maker tuning! Even though I don’t currently own a harp in that tuning, I’ve implemented theoretical support for Melody Maker (as well as other Richter variants) in the app already. So users can visualize bends and pitches correctly even when playing in those alternate tunings.
Your Stevie Nicks example is such a beautiful use case. Expressive bending like that — starting just below the pitch and sliding up — is exactly the kind of thing I hope the app can help people hear and understand more consciously.
Also, I really appreciate that you mentioned manual retuning. Let’s Bend was designed to be lightweight and offline, so it’s ideal for checking pitch quickly when tuning by ear or file. That’s one of the practical scenarios I had in mind from the start.
Thanks again — I’m learning a lot from your comment.
Thanks for sharing that — I really appreciate your perspective and experience.
I absolutely agree: bending is a deeply organic thing that you need to feel and hear in your body. No app can replace that. The goal of my app is not to teach you how to bend, but to help you understand what you're actually playing — especially for beginners who are unsure whether their bend reached the target pitch.
It’s more like a "mirror" than a teacher: you still have to do the work with your ears, but the app can help confirm (or challenge) what you think you hear. Some folks find that helpful in the early stages.
Also — yes to Gindick and Jason Ricci! Two amazing resources. I hope my app can complement, not replace, that kind of learning journey.
MakeACopy uses reproducible OpenCV/ONNX builds and runs Tesseract OCR fully on-device. The PDF text layer is generated with Apache PDFBox for full searchability.
All binaries are F-Droid–compliant and deterministically reproducible.
Happy to discuss the OpenCV/ONNX integration or how the deterministic build setup works.