Testing audio shield vs PT8211 vs on-board DAC for audio synthesis

Status
Not open for further replies.

colorado_hick

Well-known member
Hey folks,
I want to get some better data on what the real implications of the different solutions for getting audio signals from a teensy-based synth build.
My idea is to use my first three builds, one which uses the onboard DAC on a 3.6, one that uses a 3.2 with the PT8211, and one that uses a 3.6 with the audio shield. For each one I will upload the exact same synthesis program. I will record the results straight to a DAW and also try and get some visuals with an oscilloscope.

I am thinking some simple sine waves at a variety of extreme frequencies, different levels of polyphony, and some FM synthesis. I also will play around with doing some level adjustments. This is driven by my intended use. I plan to link to the demonstrations here in the forum.

Any feedback on what else I should look for or test? Any other sort of resource monitoring that I should be looking at? Or maybe someone has done this already and can share their findings?

Thanks!
 
anecdotally I have noticed that is the biggest difference with the onboard DAC is that it starts breaking up with more then 3 notes, for the audio shield I can get get up to 8 (have not tried more yet). It could be something in the way I am coding it. I will make sure to include source code for my examples.
 
Yes, but that's because the built in DAC is 12 bits, and the others are 16. That's not really a qualitative issue. Whatever DAC you use, you need to scale the input to not exceed the number of bits.
 
Yeah I thought about trying a different I2S board. But to be honest if I can get the sound quality that works for me out of the PT8211 then it would be hard for me to switch. The great thing about the teensy and the audio libraries is that it all just kinda works.
 
I did a similar test a while ago. Used the same synthesized sounds and played on a teensy 3.6 connected to the audioshield and on a teensy 3.6 using the built in DAC. I didn't really do a structured, scientific analysis, more like a quick comparison using the kind of sounds i usually do with this drum synth. To my ears the difference is not that big, especially if you have sounds that cover the full spectrum and are quite busy anyway. To me the sound of the 12bit DAC sounds a bit like it is always covered with some artefacts or a slight distortion. The sound of the audioshield is definately more clean, and this is most noticable with more quiet sounds or sounds with less high frequency components (in my example when the kick drum is soloed). This is really just me guessing, but i think it might have to do with the lack of antialiasing filters when using the internal DACs.

I uploaded the two versions to dropbox: https://www.dropbox.com/sh/3i1qwnuepks43t3/AACxNp05KdY-yBAIVUqAIo81a?dl=0

I'm interested to hear about your findings, in case you do your comparision.
 
Status
Not open for further replies.
Back
Top