SoundGate Guitar
AI Music Tutor that gives real-time feedback on your playing
Start new thread

We hit #2! Here’s the top feedback we got for SoundGate and what we’re building next 🎸

by•

Wow! Thank you to everyone who supported SoundGate Guitar! Finishing at #2 exceeded our expectations.

For Vahe and me the best part of the day wasn't the ranking - it was the conversations in the comments. We loved seeing how many of you resonated with the need for a real-time practice companion rather than just another video library.

We took notes on every single comment. Here is the top feedback we received and what it means for our roadmap:

  • New Instruments: A ton of you asked for Bass and Ukulele support so you can jam together. We are focused on perfecting the 6-string engine right now, but Bass is next in line!

  • Advanced Technique Detection: We had great questions about expression detection. We are actively training our models to catch bends, slides, vibrato, and palm muting.

  • Distortion & Amp Effects: Right now our engine is optimized for clean tones. Based on your feedback, we are prioritizing updates to handle distorted tones and heavy effects just as accurately.

  • Android Support: We heard you! Right now we are fully native on Apple (SwiftUI), but we know we need to bring SoundGate to the Android ecosystem.

If you downloaded SoundGate this week, please keep the feedback coming. We are building this for you, so your input directly shapes the product.

For the guitarists here: out of expression detection (bends/slides) or distortion support, which would you want us to ship first?

38 views

Add a comment

Replies

Best

Congrats on the cool launch!
Really happy to see a product built around art and creativity. Wishing you and Vahe lots of success with SoundGate and good luck with everything ahead!

 thanks a lot! And thanks for supporting our launch.

Reading every comment and then publishing what you actually heard is the part most people skip. Respect for that.

Not a guitarist, so I will not pretend to vote on the feature. But the way you framed the choice hides the real difference between the two.

Missing expression detection is a visible gap. The user knows it is not there yet and forgives it.

Getting distortion wrong is invisible. If the engine mishears a distorted note, a beginner does not conclude the model was wrong. They conclude they played it wrong. That is the failure that costs you the user quietly, and you never hear about it, because they just stop opening the app.

So I would ship the one where being wrong is silent. And ship it with the model able to say when it is unsure.

 That’s a great point, and one we think about a lot. User trust is more important than claiming to detect everything. We’d rather be transparent about what the model can confidently recognize than provide incorrect feedback. We’ll continue improving both detection quality and confidence handling as SoundGate evolves. Thanks for the thoughtful feedback!

congrats - keep going, way to go!

Thanks a lot!