I just know someone is going to tell me that this is already there or going to be, but here goes anyway...

I suggest that since v3.0 already has something like 15 EQ preset memories, that it might be nice if we could have memory number one assigned specifically to provide for in-car Equalisation (Auto EQ) and once set that EQ would remain active whilst ever the player is in the car. Then to be able to assign EQ memories 2 through 15 as album / track EQs that would actually be applied as a result of somekind of marker that associates the tracks with that EQ memory, similar to how we have Wendy filters associated now. This would allow us to correct the tonal balance of tracks that are for example light on Bass or overly boomy. To make this work, I imagine that EQ memories 2 - 15 would need to be able to be applied on top of the Car EQ (memory 1) so you are not equalising for both the car and the track in the one memory setting. It also has the advantage that the album / track EQ would only need to be preset once for a track and would be portable from car to car, since EQ memory 1 would take care of the acoustic environment differences between different cars. I would be very interested to hear what you guys have to say about this idea. Thanks..

The "automatically select an EQ in the car" (versus a different EQ when docked) thingie can easily be done in Hijack, if anyone would care to submit a patch for it.

Cheers

George, you're asking for two things here:

1. EQ automatically changes to a certain EQ preset when in the car.

This is already done. Here's the FAQ entry on it.

The idea is that you'd have presets like this:

Preset 01: Flat
Preset 02: Bass boosted and treble rolloff
Preset 03: Bass boost only
Preset 04: Treble rolloff only
Preset 05: Deep cut in the midrange

(etc...)

Each one of those presets has a Home personality and a Car personality. Each of those presets has their own independent car and home settings. You don't have just one car-specific preset, you've got sixteen car-specific presets, and sixteen home-specific presets.

Trying to do "Preset 1=Car" is missing the point and misunderstanding the feature. The idea is to tune the EQ in its native environment and it changes automatically. Limiting yourself to only one car preset would be a pain, and would waste the flexibility of having all those presets.


2. Per-song tone correction.

This one isn't done in the default software. But I always figured that setting up an entire EQ curve per-song would be a pain in the ass. I had a better idea a long time ago, discussed it here on the BBS. It would be a lot easier to implement, and a lot easier to maintain. It involved using the Bass Boost and Treble Boost code in Hijack. Who was it that originally created that code, was that Genixia?

See, setting an entire eq curve just for one particular album is a problem because that's too much detail to go into per-song. All you REALLY need to do is boost or cut the bass or treble on a per-song basis. While letting the EQ's to be your baseline.

Anyway, imagine that a given track has, in the comment field, the text "HIJACKBASS+1" and/or "HIJACKTREBLE-4". Or something like that. Then Hijack could check the comment field (it comes in the Notify text doesn't it?) each time a track plays. If the text isn't there, then it doesn't change the amount of bass/treble boost.

Note that this would be additive, per track, compared to the bass and treble boost settings already there in hijack.

Actually, that reminds me of something that might actually be doable in Hijack, which is turning on voladjust when shuffle is turned on and turning it off when it's not. Obviously, I could do all of that manually, but it'd be cool to have it built in.

Yes, but it only reverts to whichever preset was last active when previously in the car.


OK, but then maybe I am missing something here, how many of us have sixteen cars? Even you Americans don't usually have that many! So why have sixteen car presets? I was suggesting using one of the presets (Preset 01: Flat) as the one that was setup to equalise the car environment, leaving up to fifteen presets in the car to be used for applying correction to tracks / albums. What I envisaged was that when you chose one of the other EQs it would be summed with Preset 01 to achieve a total EQ. This would only need to be done for those tracks or albums that are in need of correction of course, so for those that don't they would be played with only preset 01 active. (I am of course talking in Car here). I should mention also that I figured that the EQ presets would be something like you suggested, ie:
Preset 02: Bass boosted and treble rolloff
Preset 03: Bass boost only
Preset 04: Treble rolloff only
Preset 05: Deep cut in the midrange, etc..
and that you then assign the most appropriate preset to a track or album by means of a selection that I presume would need to be added into emplode, similar to the way you attach Wendy filters to tracks.

Since Preset 01 corrects for the car environment, the other (track/album) presets could be assigned to tracks while listening in the home environment because the degree of coorection required by a track would remain the same no matter where you listen to it (ie: the environment correction part is taken care of seperately in Preset 01 (either for home or car). If your really worried about flexibility here, then there are sixteen presets for in-car use, I am sure we could have Presets 01 and 02 as car environment corrections which you can select from (maybe to correct for car 01 and car 02 if you have 2 cars) and still leave 14 presets available for track/album correction (more than enough).

My purpose here was similar to your idea with using Hijack, namely to correct for badly mixed tracks/albums that would otherwise have you constantly needing to reach for the Bass and Treble adjustments as you move from track to track. I therefore never thought in terms of doing a custom EQ preset for each song, or for that matter each album, but more to merely select the most appropriate preset from a pre-defined selection of EQ corrections that could be setup to best cater for the majority of such tonal balance problems. This would be of particular value if playing in random mode a number of tracks from a variety of albums, avoiding the tendency for some tracks to sound too bass heavy and others to sound too bass light etc. Yes, I expect you could achieve the same result (more or less) by just applying Bass or Treble correction as you suggested, but since we have so many EQ presets available, it seemed ideal to be able to use them to perform a more thorough correction than could be achieved just with Bass and Treble. (If it's there, why not use it?)

I like this idea too, it appeals to my sense of laziness.

Anyway, let's remember this thing is a computer under all that cosmetic beauty, and so it can be made to do all kinds of cool stuff just by doing some programming. This, I think, is it's greatest strength compared to pretty much every other car stereo on the planet. That's why I love it! I believe if you can think of something you'd like it to be able to do (that it doesn't already do) then it probably can be made to do it, which other car players cannot. I am all for anything that makes it safer to drive in your car and not being distracted from driving by having to change EQs, Bass, Treble, Volume, etc, while driving has to be a bonus. We should be able to just sit back and enjoy our music while driving and this machine makes that possibility far more attainable than anything else out there.

Sounds almost like hes wanting a more advanced version of the auto-eq where it changed eq settings based on the songs...

Or an FM strength compressor!

Well, yes, I do want it to change EQ settings based on the track, but not anything like Auto-EQ. That would be impossible to do! however, by just attaching a tag to the track, the player would select the matching EQ preset for the duration of that track.

Eh! How could you come to that conclusion from what I have written?

Just because FM radio stations compress the heck out of every song to try to make them all sound "great". Other than something automated like that, I don't see how one could have EQ settings for all of their music. Unless you limited yourself to 10 or so Genres in your database and had HiJack queue up an EQ setting that you did in advanced for each type of Genre you have.

I have never even remotely implied compressing anything (other than perhaps agreeing with wfaulk that his idea (above) could be useful). All I am suggesting is using a tag in emplode (not there yet but would need to be added) to allow the user to mark a track (or album) to use one of the (existing) EQ presets when that track is played. This is in place of having to manually adjust the Hijack Bass / treble or Empeg's Equaliser while your driving to compensate for the track sounding less than optimum. This is as a result, not of the way we encode our CDs, but the way the recording engineer equalised the recording in the first place. Surely you have had some CDs that don't sound as well balanced tonally as others?

I was just jerking your chain.

No worries, I thought you must be! Could not understand how you could be serious about it.

Anyway, what are your real thoughts on the idea?

This can't really be done by the community for technical reasons. Not as you are asking anyway.

The EQ is implemented in HW, not in SW. The parameters that you set are used to generate DSP coefficients that tell the DSP how to create the EQ. Unfortunately we don't have access to the algorithm necessary to generate the coefficients from the parameters. It is coded into the (closed-source) player application. Obviously the coefficients aren't linear, so you can't just add a couple together to create a new EQ.

I suppose that the empeg player binary in theory could do the math on the parameters before generating the coefficients, but to be honest with you, this feature is never going to get anywhere close to the top of the wish list.

Even if it did, there's some usability issues to tackle anyway. This feature would require that the frequency and Q factors for matched on both the base EQ and the modifying EQ. And applying large changes to the EQ requires an audio mute to prevent potentially dangerous transients (that could otherwise blow a speaker). But both of those issues could be overcome if the desire was there.

But in hijack space, we cannot do this. We could do something similar though.

First, assume that the EQ is being used in 2x10 mode, and not 4x5 mode. Then we only use the bottom five bands to create the base EQ. We use the top five bands to create the modifier EQ. (There is nothing preventing 2 bands having the same frequency and Q factors, in which case the dBs just add).
Then we create all of the desired EQs in this manner with a constant base 5 but differing modifiying 5, and switch between them when the song changes.
There's still a few technical hurdles to leap.

Firstly, the player wouldn't actually know which EQ was really applied, and we don't have a way to tell it. We'd have to deal with that somehow.

Second we'd have to be able to capture and store the required EQ coefficients so that hijack could use them. Hijack only gets to see the cofficients when they are applied by the player. This would require manually setting and capturing the EQs one by one and writing them to disk. Which in turn requires a partition to be mounted read-write during that operation. None of this is insurmountable, just a complete PITA.

Thirdly, we still need to mechanism to decode the tag when the song changes and to trigger the EQ change.

Fourth, we'd have to decide what to do with bass and treble. The existing bass and treble functionality hijacks the top 2 bands. The preferred option would be to rewrite the bass and treble to use the DSP hardware instead, but I've been trying to find time to do that for over a year now. The easy option would be to limit the base and modifier EQs to four bands each (which kinda makes a mockery out of the whole deal). The final option is to make the two features mutually exclusive. Could you live with that?


I personally don't think that this is worth the hassle. Even if this was all done, are you seriously going to sit in your car and set up 10 different EQs and then categorise thousands of tracks to use them? I doubt very much that anyone else is going to!

What could work, and would be _vastly_ easier to implement, would be a bass/treble modifier on a track. We'd just need the tag decoding mechanism for that. And I can really see people using that - it's a quick and easy thing to recognise that a track could use a little more bass or a little less treble.

I'd still want a method of in-car tagging though. I don't know whether there is any space within each fids data in the dynamic data partition that could be hijacked for this. It would add complexity, but might be doable.

It always seemed to me that using EQ to solve poorly-mastered recording problems is the wrong way anyway. You should be editing the "original", since that problem is going to exist no matter what playing device you use. And audio editing tools are better equipped for that sort of problem anyway.

OK, good enough for me. A nice technical reason for not being able to do it is reason enough to give up on my idea, however, I guess that leaves us right back to the Tony Fabris suggestion of using Hijack to do bass / treble corrections, which you aluded to in your answer anyway. That would be a way better solution to the problem than doing nothing at all about this. I for one am all in favour! If the tag edit function can be added to Emplode that would be good too. By the way, I never suggested that a correction should be applied by Genre! That would be very dumb indeed, as if to suggest that all Pop needs a bass boost or something similar! No, I was always intending just to apply a correction to those tracks or albums that require it (the ones where you might be inclined to adjust the Bass and treble in the car anyway) and leave all the other tracks alone (as is). So, how soon can we get this Bass/Treble idea implemented everyone? I would apprecuiate seeing this added sooner than later, and would happily implement it myself if I knew enough about programming to do it.