saratoga:
Cheers for the info.
njd:
I have to confess I didn’t consider MP3 HW decoding, however, this would never give a 200% efficiency increase over a decent general processor and SW decoder.
Ffmpeg most certainly does NOT need 300W. The TDP (what it is expected to dissipate at power - AMD, so abs. max if the reports are to be believed) of my CPU is 95W - 4 cores. So, at maximum thrust this baby only uses around 25W on a single core - when at max load. decoding a vorbis file is needing < 1% of one core
Ffmpeg on the Nokia N900 reportedly gets ~18hr playback, the same as MP3. libvorbis (made by the same folks of Tremor fame) gets just 10 hours. So this decoding technique IS 2x (ish) as inefficient as it could be.
I would accept 30% battery loss for it being non mature technology with little time for refinement.
My largest gripe, although not clearly stated before is with the way the product is marketed (so to speak*).
15 hours playback and ogg vorbis support.
Now, although they state that the reference is 128kb/s mp3 you would still expect no more than say 2 hour loss of battery life from other formats. Like I say I’d settle for 5 hours loss.
Now, has anyone (including Sandisk) ever got one of these to play for 15 hours? I’d doubt it. It’s probably a theoretical maximum rounded up.
* I would prefer to argue that if marketing states the product does something then it MUST without fail. If it implies something - which it certainly implies it will be less, but only a little bit for battery life under different formats, then it should be close. not 70% loss in performance. However, I’ve been reliably informed of sales departments making up products to sell, just to get the sale. The company doesn’t sell them, but hey, sales dept. figures they can be made. No consideration for if it is technically possible.
On a more cheerful note, a simple NiMH battery pack should be simple to construct to act as a portable power supply. a set of 4 2100mAH cells should provide ~4 full recharges (assuming typically aged batteries and a 500 mA drain - 5 full recharges should be possible in reality). Fresh NiMH x4 would give 5.6V which is above the max for the USB spec (5.25V), but they will sit at 4.8V (USB min is 4.5V) total once they are drained a little until almost flat.
**NOTE:** Let the cells drain before attaching the player (shove them in a torch for a few minutes or something). Also, don’t use alakalies. We can’t assume the players will be happy with >5.25V (although they probably will go somewhat higher 
Have tried looking at using zener diodes, resistors and regulators, but there is no way you can clamp the output below 5.25V when the batteries are at max voltage, and keep them >4.5V when they are at standard voltage
On a much more cheerful note, looking at other mp3 players the clip+ is possibly one of the best on the market IMO (no, not going to be humble, I’m a demanding customer).
Okay, I digressed big time (with shocking English - I’m tired okay). But, great player especially when compared to others on the market. Shame about the Vorbis non-support. Why great? standard USB and 3.5mm jack. Simple intuitive interface. 16GB internal would have been nice, but you can’t win them all and 8 should last me for now.
Message Edited by dihenydd on 01-15-2010 11:14 PM
Message Edited by dihenydd on 01-17-2010 12:46 AM
Message Edited by dihenydd on 01-17-2010 12:47 AM
