Pitch Transposer - Range

Here is some other information - I was trying to map out the control input to pitch transpose ratio. When I used the module with the control input directly from an AUX ADC (in a 1701), it was a mess. The pitch was all over the place. I decided to use the slider control with the arrow key to step it - hoping I could get map out a translation - but it is strange. I start at 0 (mid position) - one step (arrrow key press) to the left and I get a negative number of -5.7 (this depend on the MAX RANGE setting). One more step, same direction, gets me a -0.0012. Each step past that is a small decrement from that. The problem is that there is almost no change in ptich and then a big change. Any chance the SW people can throw me a clue? I intend to use an external microcontroller to set the pitich ratio but ...

Hi Bill.

The Pitch Transposer is a new algorithm and we are still in the process of characterizing it. The information on the wiki is not yet complete and I apologize for the confusion. I will put together a better explanation ASAP and get back to you.

Great - I have attached the test file I am using right now. If click on the slider and use the arrow buttons, you can see the values loaded. I monitor the output on an O'scope. Just one step from 0 (mid scale) and you should see the amplitude modulation.

I have done some testing and verified your results. It seems the algorithm is not functioning properly. I am going to bring this to the attention of the software manager, who is out of the office until August 20th. I will let you know the progress of this issue. Again, sorry for the confusion and inconvenience.

Thanks - at least I am not insane This algorithm came out just when we needed it so I am hoping for a quick resolution!

BillJ - did you get a chance to bring this issue to the attention of the software manager?

Bill, He still isn't back unfortunately. I still have this issue on my radar and will address it once he has returned. I can't imagine it being more than another day or two.

Bill, I spoke to the software manager and here is what I learned.

This algorithm was originally developed for a specific customer who needed small values of frequency shift and that shift would remain fixed.The frequency shift curve is not defined yet and relies primarily on experimentation. In the future, this block will probably include more traditional frequency shift controls.

If you are going to use an AUX ADC as your data input, you're going to need to scale down the values you are getting from the ADC before you input them to the Pitch Transposer.

I have attached a sample project with this implemented. I am able to get a linear sweep frequency shift with the max shift determined by the amount the AUX ADC value is scaled down by. You can experiment to see what works for you by manipulating the DC Input Entry value. I created this setup using a EVAL-ADAU1701MINIBZ board. If you have a different setup you may need to change the MP Pins and AUX ADC settings.

I will try this out - does it solve the amplitude modulation problem?

The amplitude modulation is an inherent part of the algorithm. There are two LFOs running that create the AM that you are hearing. The size of the delay buffer and frequency content of your input signal will affect how much or little you hear the AM. In the example I sent you there seems to be a sweet spot in the range of 300 samples in the delay buffer. If you change that value to 1000, for example, you will notice an obvious slow pulsing in the tone. Again, experimentation is key.

The control range is really small but it works. Thanks for the info. I am still playing with the numbers for my applications but it is much better. I wan't planning on using AUX ADC input but program it via the I2C so this setup will help me figure out the numbers.

Thanks,

Bill