Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Veena Chromatic Tuner

play.google.com

Post reply on HN
Print view View original

41–50 of 50 posts

Re: Show HN: Veena Chromatic Tuner

#41

Post by Ruarl »

v15w wrote: post #3
Ruarl wrote: post #2

This is an interesting project. It isn’t immediately obvious to me how the visualisations aid the tuning process. Please can you say a little more about how you expect a user to interpret those as they perform tuning work?

Pitch is an estimate based on sampled audio. Sometimes, it is not always accurately detected. Here, we have an additional visual cue that produces a stable/stationary waveform when the detected pitch is in tune with the selected reference note. A plucked tone has a slight variation in pitch from start to end; it rises and falls slightly. You can see this in the visual display. This would be very useful for fretting w…

I see. Was there an example of that in the video? Perhaps I missed it.

In what way does the visualisation vary when the note is not in tune? How does the visualisation indicate flat or sharp?

Re: Show HN: Veena Chromatic Tuner

#42

Post by Ruarl »

v15w wrote: post #14
dsego wrote: post #11

Earlier quoted context omitted.

It's just easier to quickly notice a pattern moving or standing still than the needle moving by a few degrees. The needle can also appear twitchy or jump around and lag, making it harder to tune accurately.

Pitch detection is computed and can sometimes be influenced by external noise or sampling limitations.It can jump. However, the visual display is based on the period of the reference note (or set frequency), allowing it to function independently of the detected pitch.

I guess this is a similar idea to the Peterson Strobe tuner? I think they have a rotating object which presents dark and light regions at the chosen frequency, which is a similar idea to building a visualisation on a set of samples whose length is the period of the chosen frequency.

Re: Show HN: Veena Chromatic Tuner

#43

Post by dsego »

Ruarl wrote: post #42
v15w wrote: post #14

Earlier quoted context omitted.

Pitch detection is computed and can sometimes be influenced by external noise or sampling limitations.It can jump. However, the visual display is based on the period of the reference note (or set frequency), allowing it to function independently of the detected pitch.

I guess this is a similar idea to the Peterson Strobe tuner? I think they have a rotating object which presents dark and light regions at the chosen frequency, which is a similar idea to building a visualisation on a set of samples whose length is the period of the chosen frequency.

Same basic idea, like an untriggered oscilloscope tuned to a fixed frequency/period.

Re: Show HN: Veena Chromatic Tuner

#44

Post by dsego »

v15w wrote: post #34
dsego wrote: post #33

Earlier quoted context omitted.

Peterson is the most known for their strobe tuners I believe. There are nowadays many other desktop and smartphone apps and pedals that have a strobe mode, but some are real strobes and some are only simulated. As far as I can tell, a real strobe will recreate the effect based on comparing the input signal frequency to a generated reference. A fake one will just use the estimated frequency (done by a pitch detection…

The visual feedback uses time period (of ref note) and number of cycles to set the scale of waveform display. It doesn't depend on detected pitch. I think strobe tuner's used neat analog electronics based reference and filters to visualise.

A few years ago I made a small tuner that mimicked the strobe effect by resampling the signal with interpolation to match the wave period (similar to an untriggered oscilloscope) and applying an IIR bandpass filter to get a clean sinewave, then showing the sinewave as a color gradient instead of an oscilloscope view. I tried a few different variations but couldn't figure out how to control the speed (sensitivity) of the movement and make it independent of the input frequency/period (apart from showing more or fewer periods in the pattern) and also it required a strong biquad filter to remove harmonics that would bleed into the strobe pattern (a similar effect can be seen in the turbo tuner (*) guitar pedal that uses LEDs to show a spinning pattern and sometimes the pattern smears or flickers), but this introduced unwanted lag. Here is a screen recording of my old app: https://youtu.be/IjYv1fDEopY

(*) I'm not affiliated in any way, I was just researching this a lot.

Re: Show HN: Veena Chromatic Tuner

#45

Post by v15w »

Ruarl wrote: post #41
v15w wrote: post #3

Earlier quoted context omitted.

Pitch is an estimate based on sampled audio. Sometimes, it is not always accurately detected. Here, we have an additional visual cue that produces a stable/stationary waveform when the detected pitch is in tune with the selected reference note. A plucked tone has a slight variation in pitch from start to end; it rises and falls slightly. You can see this in the visual display. This would be very useful for fretting w…

I see. Was there an example of that in the video? Perhaps I missed it. In what way does the visualisation vary when the note is not in tune? How does the visualisation indicate flat or sharp?

No. When sharp even by 0.5Hz the waveform rotates right and left when flat.

Re: Show HN: Veena Chromatic Tuner

#46

Post by v15w »

dsego wrote: post #44
v15w wrote: post #34

Earlier quoted context omitted.

The visual feedback uses time period (of ref note) and number of cycles to set the scale of waveform display. It doesn't depend on detected pitch. I think strobe tuner's used neat analog electronics based reference and filters to visualise.

A few years ago I made a small tuner that mimicked the strobe effect by resampling the signal with interpolation to match the wave period (similar to an untriggered oscilloscope) and applying an IIR bandpass filter to get a clean sinewave, then showing the sinewave as a color gradient instead of an oscilloscope view. I tried a few different variations but couldn't figure out how to control the speed (sensitivity) of…

Interesting. It took time for us to get the waveform stability working. We also have a smoothed wave overlay on top of raw signal. This is more a filtered version of input signal. Tried adding more filtering to get the sine wave output. But, it was too much CPU intensive. We also tried adding few overlays to this stable waveform. They did not work well and hence removed them. I agree, there is no good way to make the speed independent of input frequency. The plucked note pitch variation in flat bridge instruments like South Indian Veena is too wide when compared to pointed bridge instruments. We are planning to add additional cue that shows if a note is in tune based on pitch variation midpoint. This might help tuning the string or setting the frets relatively accurate.

Re: Show HN: Veena Chromatic Tuner

#47

Post by v15w »

mitthrowaway2 wrote: post #37
v15w wrote: post #31

Earlier quoted context omitted.

Wow, impressive stability. In case of violin (bowed instrument), the waveform stability is visualised very well. Closer to sine wave. I tend to see less harmonics than plucked lutes like South Indian Veena. Please select 2 or 4 waveform periods in settings, set the desired reference note and try.

Thanks. I've been looking for a new app to give me immediate feedback on my pitch accuracy while I play, because the violin has no frets and my ear isn't good enough yet to identify whether I'm a little sharp or flat on each note. I think realtime feedback could really help me train my fingers. I'll post more detailed feedback in a day or two when I've had time to experiment. Hopefully you'll still be checking the th…

Post your detailed feedback when you have time or e-mail us. Will be checking the thread. If there is even a slight variation of pitch w.r.t reference note the visual feedback will rotate. But, the feedback would give an idea about the deviation.

Re: Show HN: Veena Chromatic Tuner

#48

Post by v15w »

To wrap up, Thanks, HN community. Your feedback, comments, suggestions, and technical discussions have been truly humbling. We're delighted to offer a complimentary premium version of the app to anyone who contributed to this thread, once it's released. Just email us when it's available.

Re: Show HN: Veena Chromatic Tuner

#49

Post by hcrean »

7bit wrote: post #6
KaiserPro wrote: post #4

Earlier quoted context omitted.

Perhaps it lets you get that last 1-1/4 of a hz that might otherwise get lost? Its also probably really good for beginners trying on their own to train them selves so they can spot when they are over/under tune

You can't tune an instrument to 1/4 of a hz. Temperature, humidity, mechanic impacts all constantly keep working on the instrument, changing it's tune. Instruments are not high precision tools. Whatever features this app has, they're just visual fancy with no value.

Some of us use chromatic tuners to fast (lazy) check our CNC speeds and feeds. 1/4Hz matters to me, go Korg!

Re: Show HN: Veena Chromatic Tuner

#50

Post by v15w »

hcrean wrote: post #49
7bit wrote: post #6

Earlier quoted context omitted.

You can't tune an instrument to 1/4 of a hz. Temperature, humidity, mechanic impacts all constantly keep working on the instrument, changing it's tune. Instruments are not high precision tools. Whatever features this app has, they're just visual fancy with no value.

Some of us use chromatic tuners to fast (lazy) check our CNC speeds and feeds. 1/4Hz matters to me, go Korg!

CNC machines ? Chromatic mode should definitely be useful. The visual feedback requires an accurate reference frequency. You may play with A4 value or ratios to get desired non musical frequency value.
Post reply on HN