This post is the direct following of this previous post:
http://www.hydrogenaudio.org/forums/index....howtopic=16000&
To be short in the first part of my considerations I was defending that the use of a linear-scale would be easier than a exponential-scale for newbie end-user despite its being illogical at first look for audio specialists.
A lot a people disagred with me but I am quite bull-headed
& none of the arguments they advanced convince me that I was wrong so I decided to go even further in my "wrong logical" & pushing it further made me realize that the actual pseudo-standard heritated from MP3 Lame weren't accurate for use with advanced codecs like Xiph's Ogg[Vorbis] & Nero's MP4[AAC(+SBR)]Second Part:
Actually in MP3 Lame the generally accepted standard is "alt preset standard" which give 192VBR, this standard has lead most people to consider that 192/2=96, 96 Fraunhofer was the magic number for MP3 use in Movie audio, so for most people the standard in MP3 are:
192 Music Audio
96 Movie Audio
48 "Speech"
I must say that if these standards were good for MP3 Lame/Fraunhofer I disagree much with their use with both Xiph's Ogg[Vorbis] & Nero's MP4[AAC(+SBR)] as these two codecs have lowered the point of audio transparency from 192 to 160Kbps, & as I already said in my previous post habbits die hard ... & with audio I noticed they REALLY die hard ... weither if 160 is really the point of transparency or not doesn't matter, it is the virtal point at which people generally think transparency is reached, & if we apply the same logic that most people once applied to MP3 the result gives:
160 Music Audio
80 Movie Audio
40 "Speech"
what I would like to point is that if we continue to "Think MP3" OGG96Kbps will become the standard for Movie Audio in the same way many friends of mine that jump from Lame to OGG start using OGG at 192 for the very clever reason that "they always ripped in MP3 Lame 192" ... while this is IMHO a mistake ...
there is no good reason that 192/96/48 would become the standard in OGG over 160/80/40 further than that & even this is not at all an argument I would like to make you notice that with a linear scale you notice that 160/80/40 as standard bitrates are multiple of 8kbps ... & by a wonderfull hazard 160/80/40 are perfect multiple
8Kbps X 20 = 160Kbps
8Kbps X 10 = 80Kbps
8Kbps X 5 = 40Kbps
... again don't understand me the wrong way, the real argument is the OGG 160NKbps virtual point of transparency (divided by 2 & by 4), not at all round numbers ... but still I make you notice that round numbers are easier for end-user audio newbies in the same way I thought the linear scale is easier for them.
Third Part:
Now having determined that 160/80/40 were the cleverer standard nominal bitrates for my personnal Ogg Vorbis use, I decided to compared Xiph's Ogg[Vorbis] & Nero's MP4 HE[AAC(+SBR)] at lower bitrate both with very simple listening tests & cool edit frequency analysis graph,until now I had only quickly compared OGG 160NKbps with Nero MP4 LC 110-150, & was not that much convinced of MP4 being really better than OGG, & I must say that I am still not convinced of MP4 LC being better than Ogg Vorbis at bitrate superior to 100Kbps
... but MP4 HE AAC+SBR results almost made me fall from my chair !!!
1: I know very well of audio placebo effect
2: I know very well I have to trust my ears
3: I know very well an image will never tell the quality of audio
this said both frequency analysis & listening tests leads me to the conclusion that:
aac alone = vorbis, aac+SBR > vorbis
at 40/80 average bitrates aac+sbr sounds better & looks better ... all my ears, eyes & brain agreed on that ... I was first planning to post frequency JPGs but I finally forget about it as I was sure it would lead to negative reactions & furthermore what I say is mainly due to my ears not to JPGs at all.
Does that mean that with rumors of winamp 5.1 supporting MP4, Ogg Vorbis time is over, the anwser is definitly NO.
The friendly battle between Ogg & MP3/MPC was win from start by Vorbis due to its bigger potential, & Vorbis rivalize with aac alone, the real challenge for Ogg Vorbis is SBR.
I personnaly think that Ogg Vorbis has never needed our support more than nowadays & will really needs it till it rivalizes with aac+SBR.
I think I already read on this forums that Ogg Vorbis doesn't need SBR due to its work, & that applying SBRlike technology to Ogg Vorbis was not the solution, I don't know what Monthy is planning, but seeing how great is all he already accomplished I trust him for V1.1 being as good as aac+SBR at low bitrates ... 80Nkps is what I will use with Theora to the condition that theora rivalize with Xvid. So on this I put high hopes in Monthy & in Theora developper (which I don't know the name yet ...)
Now that I compared MP4HE to Vorbis, there is one thing that I would like to tell to other OggZealots spreaded among these forums, a lot of them don't understand Monthy's choice to tweak the low bitrates before the high bitrates as they only use Ogg Vorbis for Music Audio, honestly even with the pre-echo issue & such I am quite happy with the quality Vorbis has achieve in high bitrate (I never used Garf versions) ... but after realizing the quality of MP4HE I must admit that Monthy is right & he really knows what he is doing ... he has all the time to tweak the higher bitrates later as most end-user won't hear the difference between vorbis & aac alone ... but reading doom9 forums you realize that many Divx/Xvid won't wait for Ogg Vorbis to achieve aac+sbr quality at the bitrate they use ... Ogg Vorbis MUST be optimized at 80Nkps(or 96 ...) for 2 majors Vorbis use:
1: The Coming of Theora
2: The Use of Vorbis in portable player (I-River) ...
plz stop complaining that Monthy doesn't tweak High Bitrate ...,
IMHO Vorbis should have other priority:
1: Low Bitrates AAC+SBR Killer
2: Bitrate Peeling (Day Dreaming???)
3: High Bitrates MPC Killer
So in conclusion to my posts:
1: Wonder the use a Linear Scale
2: Wonder the use of 40/80/160 as standards for Ogg & (Nero MP4)
3: Support Monthy for focusing on Low Bitrates
4: If you think that I am completly nuts ... well maybe you're right !!!
Now I expect mainly 2 sorts of reactions to my post:
Part2: People contradicting me on the use of 80Nkbps over 96Nkbps for movie audio
Part3: People adding info on Monthy plan for MP4's SBR Killing, how will he do ?
Have good time listening OGGs
