Hello all,

I've finally got a 30Gb 5G (video) iPod and, mostly, I'm happy with it. The bundled 'earbuds', as is well known, are utter rubbish and I intend to buy some in-canal earphones for it. In the meantime, I have been using my Sennheiser HD465s, which I have no complaints about.

My music collection is stored in FLAC (with some MPC and MP3 tracks thrown into the mix), and I have used the Nero AAC encoder (with Low Complexity forced) to convert the entire collection, using q=0.35, to AAC for the iPod.

I have noticed, however, a distinct lack in treble (which I've certainly not noticed with these headphones before). I found a very useful website (http://hifiipod.co.uk/) with published RightMark Audio results for multiple iPod versions and, interestingly, different lossy formats. From this it seems the iPod has a near flat frequency response.

However, the comparison between MP3, AAC and OGG highlighted that all these formats have a HF roll-off at around 10-15kHz. While AAC had the flatest frequency response of all, the 15kHz roll-off was quite apparent. This lead me to thinking: can I apply pre-emphasis to the tracks to compensate for the roll-off, so as to level out the frequency response better? I'd much rather do this than use the iPod's EQ (which I am atm - with 'Treble Boost') as I can ensure a higher quality and preserve battery life. I also intend to apply ReplayGain (in Album mode) to the tracks, which I haven't done yet.

This would require converting the collection again (about 35GB in FLAC - around 1500 tracks) which isn't an issue. I realise the problems with lossy -> lossy conversion, so I will avoid this where possible.


Does anyone think this is a realistic venture? How would I go about it? The convolve plugin for foobar2000 looks useful, thought I'm not sure how I'd determine the impluse response to set it up. Perhaps someone can suggest a better solution to my treble-deficit issue?

Thanks,
smiler

Edit: Sorry for double posting! Clicked submit then cancelled page loading. Can the moderators delete the other topic?

Generally, digital formats don't have a classic roll-off. Many of them use a bitrate-dependent lowpass filter, and additionally the psychoacoustic model might decide to leave out some frequency bands from encoding.

If you want to store up to 19 or 20 kHz in an MP3, you have to select a higher bitrate and a good encoder. In the comparison you mentioned I cannot find the used bitrates.

The first question is : are you able to hear a lack of high frequency in encoded files?

Yes, I had noticed a lack of detail. I did see somewhere, however, that all three encoders were are 128kbps (presumiably CBR or closest to CBR, seeing as AAC and OGG don't actually support CBR).

I understand that the encoders use a lowpass filter - I've seen this mentioned innumerable times. However, I've never noticed and issue with lossy encoded files like this before, where I can actually hear a sort of numbness to the sound, which I can only put down to a lack of treble. Certainly, using the 'Treble Boost' EQ setting on the iPod (whatever it might *actually* do) seems to correct the problem. However, it boosts it a little too much, and does increase the load on the battery.