Just heard the radio today. Lets hope that you can enable the algorithms
Get the production line and the sweatshops readyOriginally Posted by rando
-Jesus- King of Kings Lord of Lords
Umm.. is that a good or bad verdict, oleel?
BTW - if you put a beret on that avatar, it could pass for a Libyan dictator.
beretOriginally Posted by tj!2k4
Umm.. it had static if you put your head to the speaker. Also it wouldn't pick up 105.9, a station that was picked up by a stock toyota navigation system. So yeh... Rando said he's working on it. Also said that enableing some algorithms might help. Also said he wasn't sure if he connected it to the right antenna.
Nevertheless, it sounded 100 times better than my pci radio tuner card and I'd pay $100 for it in the current state. Who listens to 105.9 anyways. It's all about 106.7 As long as it gets that, we are ok.
150 watt OPUS
tv/radio tuner card
Dual monitor Radeon
8" Lilliput touch screen
Just a point of correction, it did tune 105.9 a bit later in the afternoon; though certainly not well. It tunes it just fine here in my area so the problem may very well have been related to the antenna system in my truck. The factory Toyota radio uses a diversity tuning system. This means it uses two polarized antennas and selected between them so rapidly that you can't hear the switch. Any time you use only a single antenna from this kind of system, you can expect reduced performance.
It's also the case that the factory radio in my truck was always a poor performer in terms of tuning. So, again, I'm not 100% sure we can cast a verdict on the EM performance just yet.
The noise/hiss that you describe is real. We're still hunting down the source of that. We hope to have more results soon about any improvements in what you all heard today.
Hopefully a few more of you that heard the demo will comment so people can get a more complete idea of how it worked. In all, I believe the consensus was that it was a succesful demo for a non-optimized EM implementation.
Haven't heard such noise from this thread in a while. Sounds like you guys are right on track. Can't wait to see this thing being sold on the site.
How close do you think you are to a completed product?
Progress so far <-------*-> 88%
Kenwood MPV5025 with Aux
LG Slim DVD\CDRW
SYBA USB2.0 4-port Hub
Bufalo 54Mbs Wirelessb/g PCMCIA card
CRAM, at least, is still a ways off. Sorry, we can't be more specific than that. The FAQ will continue to be the most up to date source of information.Originally Posted by Heckle0
I thought that it sounded great at first too, then I put my ear to the speaker It didn't get 106.7 either if IIRC. It just skipped over any station over 100 IIRC. Maybe you guys got that fixed while I was demoing NaviVoiceOriginally Posted by n8scstm
No worries. I know that we'll get that fixed. That's why I didn't want to post the results in the postsOriginally Posted by rando
We shouldn't tested it in my car, but I need to know how to get the damn attenna up first
huh? Definitely not true. I think you're confusing behavior of the demo application (definitely not a production ready app) with that of the actual radio performance. Remember, we just hacked that app together last night.Originally Posted by 0l33l
The seek-scan algorithm (implemented in the code) just looks for stations in .2MHz increments starting at whatever you had entered into the tuning field text box. So if you started with an invalid U.S. tuning frequency (i.e. anything not ending in an odd increment) it would roll all the way around until it got to something greater than 107.9. Then it would jump back to a valid frequency starting at 87.9 and continue seeking; at which point it would find stations again. Sorry, I guess I should have explained that better to avoid this kind of confusion.
The hardware and software you saw can and did tune (and correctly seek-stop) on frequencies all over the FM band. 105.9 didn't tune well even when manually entered. That problem is very likely related to the specific antenna performance in my truck. To a lessor extent, it's possible there are settings in the radio hardware that might have made it possible to lock into the station regardless. Also, the assembly technique (point to point wiring, multiple circuit boards, basic inattention to RF signal crossings, etc) may also limit the end performance of the EM.
We believe the hiss is also related to signal quality because it's not consistent across every station. There are numerous settings in the signal processor that (we think) are designed to compensate for this. The CRAM Demo app is not using any of those settings yet. We'll let you know if we discover something about this.