Full Version: iTunes 9 offers HE-AAC encoding on Windows too
Just wanted to let you know that iTunes 9.0.0.70 for Windows also allows you to encode to HE-AAC up to 80 kbps I guess. At least, if I have HE-AAC checked and select a bitrate higher than 80 kbps, I get an error that the selected combination of sampling rate (which is set to Auto) and bitrate is not allowed and the bitrate jumps to 64 kbps. Selecting 80 kbps works, though.
Any other changes of interest, such as true VBR?
At a whopping 88.8mb, it better include he-aac.
At 88.8mb it better include a whole album in he-aac. 
Just did a quick test at 32 kb HE-AAC VBR vs. Nero (Feb 2007 build) @ q 0.15 (33kb SBR+PS) with about a dozen older LAME test clips (i.e. not necessarily "killer samples" anymore), ranging 5-30 secs. in length - not ABX, just which one sounded closer to the source.
When played back in fb2k, Nero won easily on almost every track - far fewer artifacts and much clearer sound. iTunes defaults to 32 kHz sampling at 32kb, and while forcing it to 44 kHz definitely helped, Nero still had fewer artifacts and was noticeably clearer on most tracks. I'm assuming this is largely due to Nero's use of PS (HE-AAC v2).
Interestingly enough, while iTunes would play back the Nero-encoded tracks with (apparently) full SBR high-frequency content, it would only play them in mono...evidently some bug with PS playback (even though it would correctly identify them as "High-Efficiency v2" in the Get Info box). iTunes did, however, play Nero's 48 kb SBR-only tracks correctly, and iTunes' own encodings at 48 kb sounded MUCH better, sometimes even with fewer artifacts than the Nero tracks. However, the iTunes tracks were usually - though not always - 5-10kb/sec higher than Nero (as reported by fb2k).
One odd thing is that the iTunes tracks were only half as long as the originals in fb2k, and only played the first half of the track, as well - tagging issue?
Also confirmed that HE-AAC radio stations could play back just fine, although they were all 48kb and higher - I'm assuming any that use PS would have the same mono bug as the Nero files.
When played back in fb2k, Nero won easily on almost every track - far fewer artifacts and much clearer sound. iTunes defaults to 32 kHz sampling at 32kb, and while forcing it to 44 kHz definitely helped, Nero still had fewer artifacts and was noticeably clearer on most tracks. I'm assuming this is largely due to Nero's use of PS (HE-AAC v2).
Interestingly enough, while iTunes would play back the Nero-encoded tracks with (apparently) full SBR high-frequency content, it would only play them in mono...evidently some bug with PS playback (even though it would correctly identify them as "High-Efficiency v2" in the Get Info box). iTunes did, however, play Nero's 48 kb SBR-only tracks correctly, and iTunes' own encodings at 48 kb sounded MUCH better, sometimes even with fewer artifacts than the Nero tracks. However, the iTunes tracks were usually - though not always - 5-10kb/sec higher than Nero (as reported by fb2k).
One odd thing is that the iTunes tracks were only half as long as the originals in fb2k, and only played the first half of the track, as well - tagging issue?
Also confirmed that HE-AAC radio stations could play back just fine, although they were all 48kb and higher - I'm assuming any that use PS would have the same mono bug as the Nero files.
Interesting news... any word yet on which iPod/iPhone model(s) support HE-AAC playback, or will that presumably require additional firmware updates?
- M.
- M.
According to the tech specs on Apple's website, only the new shuffles and nanos (i.e. the low(er) capacity models) are shipping with HE-AAC support .
Haven't seen any news regarding firmware upgrades to other models, old or new...
Haven't seen any news regarding firmware upgrades to other models, old or new...
3.1 is coming out for iPhone and probably iPod Touch and I hope it adds support to HE-AAC if it isn't there already (cannot test because my damn iPhone is at T-Mobile or wherever for three weeks because of some problems - crappy customer service as you can see).
QUOTE (M @ Sep 9 2009, 16:46) 
Interesting news... any word yet on which iPod/iPhone model(s) support HE-AAC playback, or will that presumably require additional firmware updates?
I'm playing a 64K HE-AAC file converted in iTunes 9.0/Mac on my 2008 8Gb iPhone 3g, 3.1 OS while I'm posting this. I don't have a 3.0 device to test with anymore.
SomaFM's Groove Salad 48K AACplus stream (http://somafm.com/aacdemo.pls) works in iTunes 9, too, btw.
-d.w.
Yes but are you getting proper playback or is the iPhone just decoding the LC portion of the HE-AAC file? Nothing on Apple's websites (in the tech specs sections) say anything about HE-AAC compatibility for firmware 3.1 and current hardware. ANY iPod and iPhone can technically playback HE-AAC files but they will only decode the LC portion of the files. As previously said, Apple's website only mentions HE-AAC compatibility for the 5G iPod nano and the 2G iPod shuffle.
Anyone done a 80kbps playback test using iTunes 9? I've just encoded it and listened to it via my laptop speakers and although there is a very-very-very slight difference it sounds good enough to use for ordinary iPod playback. What does everyone else think?
PS: They've just released 7.6.4 of Quicktime - it apparently isn't for Snow Leopard.
PS: They've just released 7.6.4 of Quicktime - it apparently isn't for Snow Leopard.
QUOTE (kornchild2002 @ Sep 9 2009, 21:18)
Yes but are you getting proper playback or is the iPhone just decoding the LC portion of the HE-AAC file? Nothing on Apple's websites (in the tech specs sections) say anything about HE-AAC compatibility for firmware 3.1 and current hardware. ANY iPod and iPhone can technically playback HE-AAC files but they will only decode the LC portion of the files. As previously said, Apple's website only mentions HE-AAC compatibility for the 5G iPod nano and the 2G iPod shuffle.
I can't be certain, but the fact that this link: http://www.apple.com/ipod/compare-ipod-models/#moreinfo indicates HE-AAC support on the 8GB touch with OS 3.1 (a device that, AFAIK, saw no hardware changes at all today) I think chances are that OS 3.1 supports true HE-AAC on everything (with the possible exception of the original iPhone?)
-d.w.
QUOTE (marmoset @ Sep 10 2009, 13:44)
QUOTE (kornchild2002 @ Sep 9 2009, 21:18)
Yes but are you getting proper playback or is the iPhone just decoding the LC portion of the HE-AAC file? Nothing on Apple's websites (in the tech specs sections) say anything about HE-AAC compatibility for firmware 3.1 and current hardware. ANY iPod and iPhone can technically playback HE-AAC files but they will only decode the LC portion of the files. As previously said, Apple's website only mentions HE-AAC compatibility for the 5G iPod nano and the 2G iPod shuffle.
I can't be certain, but the fact that this link: http://www.apple.com/ipod/compare-ipod-models/#moreinfo indicates HE-AAC support on the 8GB touch with OS 3.1 (a device that, AFAIK, saw no hardware changes at all today) I think chances are that OS 3.1 supports true HE-AAC on everything (with the possible exception of the original iPhone?)
-d.w.
OS 3.1 is an across the board upgrade. I wouldn't be surprised if even the very first iPhone optimised hardware had support for AAC-HE but it wasn't taken advantage of.
It would be interesting to see whether there is a battery sacrifice because of higher cpu utilisation via the using of a more complex audio compression algorithm.
QUOTE (marmoset @ Sep 9 2009, 19:44)
I can't be certain, but the fact that this link: http://www.apple.com/ipod/compare-ipod-models/#moreinfo indicates HE-AAC support on the 8GB touch with OS 3.1 (a device that, AFAIK, saw no hardware changes at all today) I think chances are that OS 3.1 supports true HE-AAC on everything (with the possible exception of the original iPhone?)
-d.w.
-d.w.
That chart is interesting as it shows that the two iPod touch models are the only ones to get HE-AAC support. However, you can click on a tech specs page for the iPod shuffle and it shows HE-AAC compatibility, the tech specs for the 5G iPod nano show HE-AAC compatibility, and the specs for either the 8GB iPod touch or the 32/64GB iPod touch don't mention anything about HE-AAC compatibility. Definitely conflicting information from Apple. The tech specs for the iPhone 3GS don't mention HE-AAC compatibility either.
Although I can ABX both, in my cursory testing (i.e. using tracks from 311 - 311 and Bad Religion - Stranger Than Fiction) of iTunes 9 HE-AAC 80 kbps VBR v. Nero 1.3.3.0 -q 0.30, Nero was more difficult to ABX. The iTunes encoder seemed to have a more difficult time with percussion instruments and those happens to be the artifacts to which my ears are the most sensitive. Unfortunately, the Nero HE-AACs skip periodically during playback on both my computer and iPhone, so I'll be sticking to good old Nero. 1.3.3.0 -q 0.53 for now.
I've just tested several songs encoded in iTunes HE-AAC in my iPod nano 4G. Based on what I am hearing, I would say that the 4G nano is decoding iTunes' HE-AAC files. The tracks certainly don't sound dull (ie. like when playing Nero HE-AAC files).
Apparently the new encoder creates buggy format info that causes problems with some programs. Reported at the foobar forum: http://www.hydrogenaudio.org/forums/index....showtopic=74722
QUOTE (ambient_88 @ Sep 10 2009, 13:09)
I've just tested several songs encoded in iTunes HE-AAC in my iPod nano 4G. Based on what I am hearing, I would say that the 4G nano is decoding iTunes' HE-AAC files. The tracks certainly don't sound dull (ie. like when playing Nero HE-AAC files).
A firmware update was not released for the 4G iPod nano and their current firmware cannot properly decode HE-AAC files. Things may have changed overnight but I haven't seen anything regarding firmware updated for older iPods over the last few days.
I have tested the files from this thread (downloaded two years ago, no longer available), on my iPhone 3GS running the new 3.1 firmware.
And I can hear the 3kHz sine, but not the pan in the White Noise, when the file is encoded HE-AAC v2.
I also tested to encode the source wav from that thread with Nero 1.3.3.0. With the -lc, -he and the -hev2 switch.
The files are recognised in iTunes 9 as "Low Complexity", "High-Efficiency", and "High-Efficiency v2".
The iPhone can definitely decode HE-AAC, but does not play Parametric Stereo.
Anders
And I can hear the 3kHz sine, but not the pan in the White Noise, when the file is encoded HE-AAC v2.
I also tested to encode the source wav from that thread with Nero 1.3.3.0. With the -lc, -he and the -hev2 switch.
The files are recognised in iTunes 9 as "Low Complexity", "High-Efficiency", and "High-Efficiency v2".
The iPhone can definitely decode HE-AAC, but does not play Parametric Stereo.
Anders
I have encountered a problem today. I have one track that neither iTunes or QuickTime will encode at 128kbps VBR constrained. The name of the song is "Strong And Smart" by In Flames off of their Clayman album. At first I thought that it could be an issues with dBpowerAMP's ALAC encoder so I converted the file to WAV using foobar2000, iTunes, and dBpowerAMP. Both iTunes and QuickTime Pro always hand at the same spots when encoding the track at 128kbps VBR constrained. I was able to use both iTunes and QuickTime to convert either the ALAC or WAV files to AAC using every other setting. They both just hand at 128kbps VBR constrained. This is with iTunes 9 + QuickTIme 7.6.4 or when using just QuickTime 7.6.4
QUOTE (kornchild2002 @ Sep 12 2009, 16:24)
Both iTunes and QuickTime Pro always hand at the same spots when encoding the track at 128kbps VBR constrained. I was able to use both iTunes and QuickTime to convert either the ALAC or WAV files to AAC using every other setting. They both just hand at 128kbps VBR constrained.
*Hang. Proofread.
Well, I can't edit my post. I read through my post once, that is more than enough for forum posting. It is not like I am writing a research paper or anything along those lines. People will still understand what I said just fine, I am not the first person to make spelling/grammar errors here and I won't be the last.
i noticed a new change in iTunes 9,selecting 320 kbps with VBR enabled now let encode with VBR without giving any warning. In previous version I couldn't select 320 with VBR,the maximum was 256. It is VBR constrained though. It looks like Apple little by little is introducing all the quicktime encoding modes to iTunes. We miss true vbr and we're done lol
Regarding the "mono bug:"
THIS IS NOT A BUG.
HE-AACv2 (PS) will play in mono using an HE-AACv1 Decoder. This is by design and the nature of the technology.
iTunes does not support HE-AACv2. It only supports HE-AACv1. Therfore any HE-AACv2 files or streams will play in mono using iTunes, or Apple HE-AAC in iPhone.
It took Apple about 7 years to get HE-AACv1, since we, Orban, were the first to use it commercially on the Internet for streams. I just wonder when or if they will add HE-AACv2?
The Tuner2 HiFi iPhone App uses the FhG AAC/HE-AACv1/v2 Decoder. It supports HTTP/ICY SHOUTcast/Icecast2, RTSP/RTP MPEG-4/3GPP, and Adobe Flash RTMP streams. It is the only known iPhone App to play Flash Audio streams.
http://www.indexcom.com/iphone
There is lots of HE-AAC technical information and graphs on the indexcom.com/iphone website, with more to come in the near future.
-greg.
THIS IS NOT A BUG.
HE-AACv2 (PS) will play in mono using an HE-AACv1 Decoder. This is by design and the nature of the technology.
iTunes does not support HE-AACv2. It only supports HE-AACv1. Therfore any HE-AACv2 files or streams will play in mono using iTunes, or Apple HE-AAC in iPhone.
It took Apple about 7 years to get HE-AACv1, since we, Orban, were the first to use it commercially on the Internet for streams. I just wonder when or if they will add HE-AACv2?
The Tuner2 HiFi iPhone App uses the FhG AAC/HE-AACv1/v2 Decoder. It supports HTTP/ICY SHOUTcast/Icecast2, RTSP/RTP MPEG-4/3GPP, and Adobe Flash RTMP streams. It is the only known iPhone App to play Flash Audio streams.
http://www.indexcom.com/iphone
There is lots of HE-AAC technical information and graphs on the indexcom.com/iphone website, with more to come in the near future.
-greg.
QUOTE (mixminus1 @ Sep 9 2009, 13:42)
Just did a quick test at 32 kb HE-AAC VBR vs. Nero (Feb 2007 build) @ q 0.15 (33kb SBR+PS) with about a dozen older LAME test clips (i.e. not necessarily "killer samples" anymore), ranging 5-30 secs. in length - not ABX, just which one sounded closer to the source.
When played back in fb2k, Nero won easily on almost every track - far fewer artifacts and much clearer sound. iTunes defaults to 32 kHz sampling at 32kb, and while forcing it to 44 kHz definitely helped, Nero still had fewer artifacts and was noticeably clearer on most tracks. I'm assuming this is largely due to Nero's use of PS (HE-AAC v2).
Interestingly enough, while iTunes would play back the Nero-encoded tracks with (apparently) full SBR high-frequency content, it would only play them in mono...evidently some bug with PS playback (even though it would correctly identify them as "High-Efficiency v2" in the Get Info box). iTunes did, however, play Nero's 48 kb SBR-only tracks correctly, and iTunes' own encodings at 48 kb sounded MUCH better, sometimes even with fewer artifacts than the Nero tracks. However, the iTunes tracks were usually - though not always - 5-10kb/sec higher than Nero (as reported by fb2k).
One odd thing is that the iTunes tracks were only half as long as the originals in fb2k, and only played the first half of the track, as well - tagging issue?
Also confirmed that HE-AAC radio stations could play back just fine, although they were all 48kb and higher - I'm assuming any that use PS would have the same mono bug as the Nero files.
When played back in fb2k, Nero won easily on almost every track - far fewer artifacts and much clearer sound. iTunes defaults to 32 kHz sampling at 32kb, and while forcing it to 44 kHz definitely helped, Nero still had fewer artifacts and was noticeably clearer on most tracks. I'm assuming this is largely due to Nero's use of PS (HE-AAC v2).
Interestingly enough, while iTunes would play back the Nero-encoded tracks with (apparently) full SBR high-frequency content, it would only play them in mono...evidently some bug with PS playback (even though it would correctly identify them as "High-Efficiency v2" in the Get Info box). iTunes did, however, play Nero's 48 kb SBR-only tracks correctly, and iTunes' own encodings at 48 kb sounded MUCH better, sometimes even with fewer artifacts than the Nero tracks. However, the iTunes tracks were usually - though not always - 5-10kb/sec higher than Nero (as reported by fb2k).
One odd thing is that the iTunes tracks were only half as long as the originals in fb2k, and only played the first half of the track, as well - tagging issue?
Also confirmed that HE-AAC radio stations could play back just fine, although they were all 48kb and higher - I'm assuming any that use PS would have the same mono bug as the Nero files.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.