Decode Mp3 to raw bytes

kubeek

Joined Sep 20, 2005
5,795
It´s open source hardware so you should be able to replicate it, depending on the volume of production you should be able to achieve similar price.
 

kubeek

Joined Sep 20, 2005
5,795
Also mind you that if you want to go with your product into production, there are some license fees you need to pay, though they are quite small if I remember correctly. HW chips usually have the fee incorporated into the price of the chip, but for SW implementation you need to register and pay to stay legal.
 

Thread Starter

m zaid

Joined Jan 9, 2016
46
Most run Linux, and they are not any larger than a board you could make with PIC32 and external RAM.
Bolb
Thanks for pointing out. It'd be great if I could have those or at least the one I found as a functional microcontroller but then what are the support? Got a few pointers though on customization from a recent thread - realized i'm the newbie.
It shouldn't take much of a PIC32 to handle MP3 duty. A DIY board friendly chip like the http://www.microchip.com/wwwproducts/en/en557425 can be used to make a SW mp3 player/decoder.
I think I can appreciate ready development boards now...and yes I'll be doing SW mp3. From here, MIPS devided by sample rate is 590 instructions per sample output - not bad. But with higher processor clock the number of instruction is higher?
...to tell you the truth I already ordered a board pic32-mz instead during post #10. And I didn't tell before, apart from the issue, I'd like to do graphics and maybe external flash is needed. I learned from a thread the pic32mz can do this.
Also mind you that if you want to go with your product into production, there are some license fees you need to pay, though they are quite small if I remember correctly. HW chips usually have the fee incorporated into the price of the chip, but for SW implementation you need to register and pay to stay legal.
gotchya..
http://mp3licensing.com/royalty/

Edit: Nothing wrong with the mp3 duty calculation. I was too hasty and slipped my mind. MIPS requirement was mistakenly took as processor speed correspondence.
 
Last edited:

ci139

Joined Jul 11, 2016
1,898
Mp3 is also most efficient if streaming through the internet.
it may be most efficient in a sense of hardware/software - the codecs supported by (various) - are capable of play back

while say 22050 samples per second (max 11025 Hz !!! - if input phase = sampler phase) 3-bit RAW-WAV has a comparable audio quality and also a similar storage size to a low-quality MP3 e.g. 32kbps to 96kbps (and unlike mp3 doesnot have any parasitic high frequency fancy-ringing anomalies) - considering your project´s probable specific nature ? why not to introduce your own audio codec to the world e.g. a dynamically masked frequency spectra with phase timing info or whatever you find most effective for your needs
 

Thread Starter

m zaid

Joined Jan 9, 2016
46
I can tell a a few more things about the product, it'll have the SW components like in the application in this video:
, and hence also an idea about PIC32's feasibility. In addition to the current issue, I'd like to do maybe composite video.
I'm using mp3 since it is a consumer product. Though, i'm still little for both theory and technical know how - only general and none detailed to be describing better - not to mention other techniques as well. I hoped foruming can help out about these technical problems I encounter not facebook. Maybe later you'll find me posting in those topics.
 
Last edited:

wayneh

Joined Sep 9, 2010
17,498
Yeah, the question was meant to be a bit rhetorical and snarky.

My point was, I'd be reluctant to design something new around such an old format. But some people think vinyl is cool, so what do I know.
 

Thread Starter

m zaid

Joined Jan 9, 2016
46
Last edited:

Thread Starter

m zaid

Joined Jan 9, 2016
46
I realize I've been making you all guess a solution to a problem not well defined. You can't tell a bargain not knowing about the actual needs. But I would really still like to keep the 'thing' a secret. But technical development wise, for the most of it it'll do like in the application in post #27, without the LCD but composite video and It'll have physical buttons instead as input.

Anyway, I have gone by my original 'guess' solution (that I'm more confident now) which was to use the Helix decoder with Microchip Pic32. Here is a picture of the board I got from http://www.microcontroller-board.co...2mz-ef-mcu-starter-kit-pic32mz2048efm100.html. Tell me it isn't the most cost effective PIC32MZ board including shipping cost:D.
pic32mz board.jpg
benta earlier told to handle the Mp3s without the decoder library. Maybe there is enough data for me to directly calculate impulses (using peak amplitude of each frequency domain) before needing to wait for the decoder to produce each sample.. Next is porting the demos (http://www.microchip.com/Developmenttools/ProductDetails.aspx?PartNO=DM320005) ; connecting a DAC and dive into mp3..
 
Last edited:
Top