Actually, that's made me pause for thought. I wonder if the scratch storage really contains the DSP coefficients, or the 'raw' EQ data in 'human' form (ie f, Q and dB). If the latter, then that would make life significantly easier wrt point 2.

I deciphered most of that, if you can find the thread.

And of course now it works. Hey, all the better.

Where is the number of plays stored? (technical)

Somehow I missed that interpretation the first time around. So part 2 is significantly easier. We can literally write calculated EQs straight into the dynamic data partition and the player will then know about them.

I don't suppose you found any strings corresponding to the EQ preset names? It'd be useful to be able to change that too.

So that just leaves the tricky part of analysing the response and calculating what is needed to correct it.

(And the actual implementation...)

Wow... I am away from my computer for one day and out pops v2.0

In reply to:

---

I just noticed how 2.00 final emplode is less than 1/2 the size of Beta 13 (when zipped). Is that because of the installer?

---

That's because it doesn't include the MSI Engine any more. Installshield 8 (which is what we've started using - despite Rob's best efforts to invent an Installshield 9) has a feature to automatically download the engine if it is required.

So umm Mark is 321 right for the latest Hijack? It seems the Hijack site has gone slightly backwards in time.

v321 is the latest described hijack but 324 exists if you change the link... or did you not want 324 available to everyone? If so, can someone delete my post if I don't get back to the board in time.

Have you seen this thread?

In reply to:

---

despite Rob's best efforts to invent an Installshield 9

---

LOL

And poor thinfourth2 still has 2 and a half months to spend on a boat.

Subject: i'm Off
Poster: thinfourth2
Date: 20/02/03 03:51 PM

Oh well time to go back to work Flying to singapore tomorrow

So see you guys in four months time Hopefully will be back in time for the fun at amerfoort

Oh yeah and if empeg towers follows the norm then they should be realeasing the next update round about the time i leave

If this works I guess we owe Roger some beer

I already sent him one when he said he fixed it. I suppose I'm a little more trusting than you.

Which is not to say that you shouldn't send him one on your own!

In reply to:

---

Hint: Poke around the Player Configuration dialog box...

---

You mean there's a new JEmplode?

It appears that it puts ``play_count=0'' in the [display] section of config.ini.

Hijack v322,v323, and the latest v324 all now "exist" again. Had a minor synchronization issue between multiple machines while I was releasing Hijack versions "on the road" last weekend.

v324 is the latest, and is recommended for everybody.

-ml

Thanks, Bitt. (I subsequently found the same elsewhere in this thread)

I was looking forward to the much-heralded tuner AF improvements. But on my daily drive to work, where normally the tuner re-tunes 3 or 4 times on the way, it didn't re-tune at all, instead just attenuating when the signal level drops to 1 or 2 bars.

The AF indicator light came on as normal (in Radio info), but switching to the AF Debugging info-mode just displayed a "No AF information at present" for the entire (20 minutes) journey. What is going wrong?

Hello tms13,

I had the same here. Is this maybe country dependent?

bye

I'm in the UK, where the Empeg boys do their testing (in fact, very near to Empeg Towers), so I expected it to work here.

This has turned into the all-use 2.00 Final thread...

Q: Have database rebuilds gotten faster? Sure looks like it..

Today, before setting off, I changed frequency down 0.1MHz and back again. The AF Debugging display changed to "Insufficient AF information; WAIT" - and stayed like that for the entire journey. Again, no retuning - it doesn't even fall-back to the 2.0b13 method when the signal gets low. Still not quite what I was looking forward to...

Also, muting the player or switching to MP3 doesn't seem to help - I thought that might give the radio more opportunity to sample the AFs, but it doesn't make any difference AFAICT.

What station are you listening to?

Local stations don't broadcast alternative frequencies in the RDS data - So what you're seeing is expected behaviour for a station like Q103.

Try Radio 1, or another national station. In programming and testing I used Radio 1 and Radio 4, both broadcast about 10 alternative frequencies. In the Cambridge area tune to 98.5 and the AF list should populate within a few seconds (providing your antenna is up to scratch).

Toby

I wish we got Radio 1 over here in the U.S.

I wish we got Radio 1 over here in the U.S

You're better off without it. There's only so much tuneless hip hop any sane person can take.

There's only so much tuneless hip hop any sane person can take.

These days it's all Garage and R&B, rather than classic (even tuneless) hip hop. IMNSHO, Garage sucks almost as badly as jungle.

I've been trying BBC Radio 4; AF switching worked fine on 2.0b13 as I said - 3 or 4 changes between 94.5 and 93.5 in my 20-minute commute. More testing on my way home tonight, I think.

Edit: I'm running Hijack (235), so won't report this as a bug until I've tried it without, just in case.

Okay Toby, please let me know your results.

I've just put hijack on my 2.0 player to see if that changes anything (can't think why it would)

I'll leave it until next week to test sans Hijack, as I can't live without some of Hijack's features over the weekend... Needless to say, I got no further than the WAIT screen last night on Radio 1 and this morning on R4. Presumably, this WAIT message means the player has the full list of frequencies, but hasn't got around to sampling their strengths; is that correct?

But I will test, and if I can reproduce it, then I'll post in Bug Reports.

The real thing I want to check without Hijack is whether it's my IR translations that have broken frequency seeks for me (worked fine in previous beta, but I upgraded Hijack 298 to 235 at the same time).

Regarding the 'WAIT' message:

There are three states the AF debugger screen can be in, and each state relates to the number of frequencies in the alternate frequency list or the range of strength readings.

DORMANT - This state is achieved by there being no alternative frequencies in the list. The message associated with this state is "No AF information currently avilable"

WAITING - This state is achieved by the range of strength readings from the freqencies in the list equaling zero (i.e. strongest reading - weakest reading = 0). The message associated with this state is "Insufficient AF information available, WAIT".

ACTIVE - This state is achieved by both other states not being satisfied - i.e. the AF list has some alternate frequencies in it, and the range of signal strength readings is not zero.

Note that the WAITING state can be achieved by two possible scenarios: If there is only one frequency in the AF list then the range of signal strength readings will be zero - If there are multiple frequencies in the AF list, but they all have the same signal strength (perhaps the initialisation value of zero) then the range will also be zero.

So one of two things is going on with your player - either the signal strength measurements are not happening, or always returning the same value, or the AF list is not getting populated properly. I think the former is the more likely problem.

Silly question, how do you enable the debug mode?

Thanks for the detailed description, Toby. (Are the dev-info and/or FAQ gnomes gathering this?)

I managed to try the tuner without Hijack last night; the good (and bad!) news is that it worked beatifully, going from Dormant into Active within a couple of seconds, and doing everything it should.

Sometimes after changing to a different preset, or on power-up, it remains stuck in Dormant, and I have to change the frequency one notch and back again for it to work. Perhaps there's a subtle race condition - is the AF code getting the frequencies too quickly, before it's reset itself?

I'm surprised that reverting to the stock kernel made a difference (I haven't entirely ruled out coincidence, but it's looking suspicious. I do want to be certain before emabarking on the heresy of suggesting HJ's not perfect, though!) I'm now running the oldest Hijack I have (v221) in the hope that we can identify what's upsetting it.

Silly question, how do you enable the debug mode

>Info>AF Debugger

Somebody's got to come up with "De bugger don't work right", sooner or later, so I might as well do it now!

No doubt about it -- Hijack has bugs. But I've hidden them so cleverly that you aren't supposed to be able to notice them.. my tuner behaves very nicely, but then, we don't have RDS around here.

Cheers