FCC Leaning Toward a 1 t0 5 dB boost for digital FM stations..not 10dB.

[RolfTaylor]

Ibiquity does not talk about this because of NIH (not invented here) syndrome. A number of people "in the know" have told me the core is AAC. This is also completely consistent with the reports from folks that have evaluated the HDC codec - it is better than MP3-pro.

Plus, HDC has a scalability profile that is not part of AAC-plus/MPEG HE-AAC.

As I said earlier, SBR can be a bit of a compromise and from my personal experience seems to have a greater variability with regards to noticeability (e.g. not perfect) among different listeners.

I was not entirely clear earlier that it is comparable to HE-AAC v.1 NOT HE-AAC v.2 which transmits mono plus a low bit rate Cue/hint stream to generate stereo (in this case) or 5.1 (in the case of WZLX).

Rolf

You are misinformed. HDC is based on AAC+ Yes, the SBR portion is not perfect (seems to be much more detectable to some individual vs other) and it does not support the newer methods of AAC-HE for stereo encoding (the same tool used for 5.1 broadcast but extended to stereo).

From the very obscure information from iBiquity, I understand that the HDC is a proprietary codec that uses SBR. Coding Technologies is quite specific that iBiquity uses their SBR, so I don't think iBiquity would use (or that Coding Technologies would allow them to) any inferior version of SBR than the original one from Coding Technologies (which is the same as in HE-AAC v1, aka AAC+).

That leaves the core codec to be different. If the core codec was indeed AAC, then the codec would be HE-AAC. As it is not, it's called HDC. An additional argument is that the codec was called HDC even before they added the SBR. If you remember, that was when they did some listening test and testers found the audio quality to be poor. I understand HDC is based on MPEG-4, but it is not the same nor compatible.

Btw, I would LOVE to get all the specifics on HDC. But no one's talking. Back when I was doing research for an article, I tried to get information from iBiquity but I got no response from them what-so-ever...


Regards,
Goran Tomas

 

[gtomas]

Ibiquity does not talk about this because of NIH (not invented here) syndrome. A number of people "in the know" have told me the core is AAC. This is also completely consistent with the reports from folks that have evaluated the HDC codec - it is better than MP3-pro.

Is it possible to send me those reports? My e-mail is goran dot tomas at gmail dot com.

As I said earlier, SBR can be a bit of a compromise and from my personal experience seems to have a greater variability with regards to noticeability (e.g. not perfect) among different listeners.

Yes, that seems to be the case. I can tell SBR codec at any bitrate, because of the characteristic sound of the high-end. I like to call it artificial, well, because it is artificial

In addition to some people noticing it more, I believe there is the effect of "acquired sensitivity". At first, you might not notice the artifacts at all. But after you listen for a while and once your ear "figures it out", the artificial high-end becomes very obvious, every time.

I said it before and I'll say it again - IMO, SBR is a great tool to extend frequency range and improve the sound quality of coded audio at very low bit-rates. However, there is a point where waveform based codec starts to sound better than the SBR-enabled ones. Simply because SBR-enabled codecs can't entirely get rid of their synthesized high-end. And, as bitrate rises, waveform codec can. Again IMO, with HE-AAC/AAC that turning point is 96 kpbs. At and higher than 96 kbps AAC sounds better than HE-AAC v1/v2. People have started using HE-AAC (aka AAC+) as the greatest codec ever regardless of the bitrate, but that is simply not true. Additionally, no SBR-enabled codec was tested transparent, as in indistinguishable from the original. Waveform codecs have. In particular, in EBU testings AAC was found transparent at 128 kpbs (which doesn't mean broadcasters, due to specifics of broadcast transmissions and additional audio processing, should use source material coded with AAC @ 128 kbps!).

In that regards, I sincerely hope that the HDC scalability you talked about means that at 96 kbps (simulcast only mode) HDC runs without SBR.


Regards,
Goran Tomas

P.S. Due to common misunderstanding, I feel the need to keep pointing out that 'HE-AAC' generally assumes 'HE-AAC v1' which is the same as 'AAC+' and the same as 'aacPlus'.