Pioneer HU recognizing most mp3s but not some

[NetfreakBombay]

thanks, that is a good place for technical information.

[DerAlte]

Originally Posted by NetfreakBombay

Can you quote some source for that? If that is indeed the case it would be interesting to read about it.

Since you are a software developer I assume you are familiar with NDAs? Many of us will be violating NDAs if we did!

Originally Posted by NetfreakBombay

Reason I am skeptical is, even 1000 - 2000 Rs devices have more power then that. We are talking about 9000 Rs device here.
Even a 1700 Rs Wireless router that I am using has a 32 bit MIPS processor (BCM3302) running at 200 MHz and runs VxWorks. ...It has 16 MB RAM.
My second PC was pentium 100 MHz with 8 MB RAM and it could do MP3

Perhaps it would be very illuminating if you compare the BoMs of that Router and an average HU. Now compare the BoM cost with the selling cost per unit. If you were an HU designer, you would be hard-pressed to use a micro-controller at all.

Why? In the router, there are 2 chips and a handful of passive components + 8 LEDs for display. In the HU, there are 20 chips, a bushel full of passives, and you wouldn't buy it unless there is a hi-funda DM display. How much cost will you allocate for each function in the HU? And - finally - which micro will you choose in the money left over?

Your second PC would have costed 30K at least - just to play MP3, since when it was playing MP3 it couldn't have done much else!!!

Not very apparent, no?

[RAC]

I vaguely recollect reading somewhere that some Pioneer models may not support Apple Loseless.
Not sure about it, never tried it.

[NetfreakBombay]

Originally Posted by DerAlte

Since you are a software developer I assume you are familiar with NDAs? Many of us will be violating NDAs if we did!

Perhaps it would be very illuminating if you compare the BoMs of that Router and an average HU. Now compare the BoM cost with the selling cost per unit. If you were an HU designer, you would be hard-pressed to use a micro-controller at all.

Agree that it might be economical to build a single chip that does all things all things instead of building HU using generic microprocessor + RAM + Flash etc. Specially if volume is high.

But its hard to digest that integrated chip would be 8 bit.

And reasoning is the same.. it has to do MP3/WMA etc + USB + FAT and many more things.

[greenhorn]

Originally Posted by NetfreakBombay

Agree that it might be economical to build a single chip that does all things all things instead of building HU using generic microprocessor + RAM + Flash etc. Specially if volume is high.

But its hard to digest that integrated chip would be 8 bit.

And reasoning is the same.. it has to do MP3/WMA etc + USB + FAT and many more things.

Yes , eventually if volumes are high, someone will put them all into one LSI chip, but the whole point of using dedicated decoders etc is to do the job with the least processing power and cost .You may need a good processor to playback Mpeg 4 corrrect ? Yet the same can be done fairly easily by most low end GPU's these days. Pretty much the same thing .

The uC in this case doesnt have to be very high performance at all. All it needs to do it sit on the I²C Bus (which doesnt need many bits ) and send commands and/or move small bytes of data ( like the song name etc) around. You certainly dont need horsepower for it. You're thinking of it running on an OS , or something of that variety - I think that's overkill... Some custom C coded ASM should be more than enough for it.

[DerAlte]

Originally Posted by NetfreakBombay

But its hard to digest that integrated chip would be 8 bit.

And reasoning is the same.. it has to do MP3/WMA etc + USB + FAT and many more things.

Sir, why should the core controller run the CODEC? The dedicated silicon CODEC + separate controller is cheaper than higher power micro running a soft CODEC!!!

[NetfreakBombay]

And that CODEC chip is 8 bit?

[greenhorn]

Originally Posted by NetfreakBombay

And that CODEC chip is 8 bit?

The controller may be .

The codec needs to decode 16 bit audio at the least , so obviously not
but the codec is not a 16 bit microprocessor, if thats what you're driving at. Its a digital signal processor

[NetfreakBombay]

I don't think a processor that can handle just 8 bits will be able to act as a controller either.

Yes CODED should so some custom chip with MP3/WMA/AAC etc logic implemented on DSP.

[DerAlte]

To give you some examples of 8-bit micros being currently used, apart from HU controller:
1. The venerable old company Zilog still makes a Z80-derived micro, which is the mainstay of most Remote Controllers for TV, HT, DVD etc., apart from the programmable remotes
2. Most of the engine management ECU for low-end cars are based on 8051 derivatives
3. Most digitally controlled elevators in operation today are controlled by 680x-based control systems
4. Most common PMPs are based on a Philips chipset, which has an 8 bit controller internally (not the CODEC)

Most of us get carried away by Moore's Law and Intel's advertising, forgetting the fact that the bulk of micros and processors working in the field (outside of PCs) are in fact the old clunky slow ones like Z80, 8051, 6809, 68K, 99xx, etc. and their derivatives. They outnumber Pentiums et al 10:1!!! Of course they are not sexy or news-worthy, so you wouldn't have heard of them.

Originally Posted by greenhorn

The codec needs to decode 16 bit audio at the least , so obviously not
but the codec is not a 16 bit microprocessor, if thats what you're driving at.

Why can an 8 bit micro not decode to 16-bit audio? The stream is 8-bit bytes only, no?!

By your logic, a bus-driver should be 3 times the size of an average M800 or Alto driver!!!

[ScarySkulls]

I heard the 6050 wont play VBR mp3 tracks.
Can somebody confirm this?

[greenhorn]

Originally Posted by DerAlte

Why can an 8 bit micro not decode to 16-bit audio? The stream is 8-bit bytes only, no?!

By your logic, a bus-driver should be 3 times the size of an average M800 or Alto driver!!!

Well, I meant that the codec was 16 bit
most 8bit audio codec chips are voice (~48 dB S/N) variety .

[DerAlte]

Originally Posted by greenhorn

Well, I meant that the codec was 16 bit
most 8bit audio codec chips are voice (~48 dB S/N) variety .

We are on parallel tracks! I meant internal Data bus width of the micro, and you were talking about Data width per sample!!! You really need to roll up your sleeves and get your hands dirty, you white collar intellectual

[NetfreakBombay]

Originally Posted by DerAlte

To give you some examples of 8-bit micros being currently used, apart from HU controller:
1. The venerable old company Zilog still makes a Z80-derived micro, which is the mainstay of most Remote Controllers for TV, HT, DVD etc., apart from the programmable remotes
2. Most of the engine management ECU for low-end cars are based on 8051 derivatives
3. Most digitally controlled elevators in operation today are controlled by 680x-based control systems
4. Most common PMPs are based on a Philips chipset, which has an 8 bit controller internally (not the CODEC)

I can understand remote controls using Z80. No idea about elevators but that makes sense as well (not too many parameters to process).

But ECU in today's cars??

I have Alto that is pretty much as low end as you can get (it shares engine with 800). And it has some 32 bit ECU.

Alto's spec page :
Maruti Alto. Know about the specifications of Maruti Alto.

Similarly Maruti 800's spec page talks about 32 bit ECU

[greenhorn]

Originally Posted by NetfreakBombay

And that CODEC chip is 8 bit?

Originally Posted by greenhorn

The controller may be .

The codec needs to decode 16 bit audio at the least , so obviously not

I was replying to Netfreak's question when I said that the codec was not an 8 bit.

Originally Posted by DerAlte

Why can an 8 bit micro not decode to 16-bit audio? The stream is 8-bit bytes only, no?!

I had added that the controller could be an 8 bit in the reply above

Originally Posted by DerAlte

We are on parallel tracks! I meant internal Data bus width of the micro, and you were talking about Data width per sample!!! You really need to roll up your sleeves and get your hands dirty, you white collar intellectual

If You'd taken my replies in context, you'll see that you're the one who's taking me on a parallel path