i used ogg aotuv at quality 7 a couple of years back (maybe last year) and tried flac.. noticed a big difference, it's not as transparent as you think it is.

now as long as it's not 128kbps i don't really care

Quote:



mp3 has been even longer.
Quote:



It's a pure speculation, and indeed false. First of all, when you find a good album, say Diana Krall's, you'll hear sight treble rolloff even in -Q9, this is what makes it worse than Q10/lossless.
Quote:



Guitar timbre is incomplete. If you don't hear it, it means you have problems with amplification because this is mostly hardware what corrupts sound quality and makes people state the mp3 -v0 is indistinguishable from lossless. The amplification kills the subtelties inducing THD/IMD/TIM.
It's even easily explainable why midrange differentiation is lost in lower bitrates of mp3 and ogg Vorbis - both codecs use the MDCT transform which being cosine-based does not like even harmonics - the cosine function is symmetrical itself, that's why. The higher the bitrate, the more data to encode additional sonic information lost in the transform, and more data to lower the psychoacoustic cut-off threshold.
Quote:



No, but I'm able to request an audition of a system based on von Schweikhert monitors but how does it matter? You don't need that much - see above.

Vorbis has developed considerably in the past two years, largely due to Aoyumi and his custom tunings... I'd encourage anyone to listen to the results from more recent encoders, you might be surprised!

Oggenc2.85using aoTuVb5.5

RareWares

@lexnasa, isn't the Lancer SSE3 (based on aoTuV b5) good enough?

majkel - I'll try and keep an eye out for those albums and report back here with my conclusions. At the moment I'm still not convinced - I've spent a lot of time listening to acoustic guitar, and can hear no difference between Vorbis at q6 and the original CD.

~Phewl.

Quote:



Well, there is always a couple of things to blame - music (undemanding), electronics/cables (unresolving), hearing and attention to cetain details. In Papa Roach's Infest track 1 there is a guitar passage before vocal introduction, which has got complicated timbral content, and -q6 loses quite much of it, -q7 partly, and -q8 sounds fine in the midrange region for me.
Regarding the mp3 - I mostly use FhG as I find it superior to LAME in 320kb/s bitrate.

Quote:



Yes I believe so, the main tuning in b5.5 was for lower bit rates, so Q6 and above should not be affected.

Quote:



I do, it is not 100% guaranteed every car saleman is a conman. I'd rather ask him than phone my Gran.

This is why working on commission is evil, i.e. the more you sell the more you earn. A car-salesman on commission would always be tempted to lie through their teeth.

Quote:



I personally use 'oggdropXPd V.1.9.0 using aoTuVb5.5' myself
Thing is that yes, it isn't transparent to me at -q6 but for file size/SQ quality, I argue that -q6 is the best setting.

Quote:



How is it harder for MDCT to handle even harmonics than odd harmonics?

I would expect the phase offset for a harmonic relative to the MDCT window to have a detrimental effect (since MDCT has no way to separate out phase and magnitude information the way the FFT does), but I'd expect this to apply to all parts of the signal.

Quote:




majkel - it seems I do owe you an apology. I've been sitting here tonight, and I've convinced myself that my initial testing can't have been detailed enough.

The differences, IMO, are extremely slight, but they are there. I've sat here listening to "Jennifer Warnes - Famous Blue Raincoat", and I'm convinced that Vorbis q6 is not -quite- CD quality.

I can't make the ABX script for Amarok work, I guess it's incompatible with the later releases. So, I can't ABX, but I'm happy enough that I can hear extremely subtle differences, time after time.

The difference is not great enough for me to wish to re-encode all 600-odd albums, but I'll be using FLAC in future. I dare say I'd find the differences vanished at a higher bitrate with Vorbis, but I'd never know if I was missing out on something when I started using different gear.

I'm a FLAC-convertee from this day on.

For anyone using Vorbis, the differences are really, really, extremely small IMO, and not worth losing any sleep over. You have to be paying total attention to the intricate details to find any differences.

~Phewl.

Quote:



I don't think so. I respect people who dare to admit they were wrong. I find full timbral transparency at q8, and full spectral (no treble roll-off) at q10 but it needs very good records/amplification/headphones or speakers to hear it. For headphones like Sennheiser PX100, q8 is the limit.