Those "video VBR" are really ABR modes. (Of course, ABR is a kind of VBR...)



It should probably just give an error when you try that, because I really have no idea what happens in this case, and it's certainly untested, so I'd strongly recommend not to do it




66kbps

-q 0.12

If you try lower quality levels, you'll get an errror. Hmm, I think I'll change this so the lowest is really -q 0.0 too.

Hmm, what awards?

Can't download. Any mirrors?

And please, pretty please: could we have a "recommended settings"?

And maybe some more info on command line usage for foobar and EAC?

The server is alive. Problem must be at your end.



I would have no idea what would be supposed to go in such a thread.



What more info do you need?

I was experimenting with 2-pass vs single pass, and I noticed similar results. I had quality set to 0.6, and encoded Thursday's "A City by the Light Divided." Single pass came out around 74.4MB, with a bitrate ~224, whereas 2-pass came to 70.7MB with a bitrate of ~215. Interestingly enough, that's exactly the bitrate I was looking for. Almost every other album I've done w/2-pass averages ~215kbps.

i've have waited a long time for the aac encoder to be separated from Nero and now it finally happened, and best of all for FREE

thx to all that made this possible!

Nope. Never had any problems downloading from Nero servers (I usually get Nero updates on the first day). Could it be overcrowded?
That is why I asked for mirrors.


Rough translation of q values to bitrates. Just having naming the presets won't do.

You know, like the lame V settings table.

And I feel-- and I'm sure I'm not the only one-- that


is not clear enough.

I'm assuming that the -q setting is to be used exactly like in oggenc, except the number is divided by 10? That seems to be about the equivalent.

Though overall I am quite impressed (and happy to finally have access to this encoder), I do have one complaint. The encoder gives no indication of progress. Once you hit enter it just sits there for a seemingly random period of time while it's encoding. Showing the percentage of completion would keep me from thinking that it isn't working.

Maybe not really the answer you wanted, but this can be handled by the frontend (foobar2000 will show progress, for example).

No, as I already said, the server is alive.



http://www.hydrogenaudio.org/forums/index....showtopic=44283

Thank you so very much.

Still no luck downloading, though.

Great stuff lads.

Been needing a good quality gapless AAC CLI encoder for a while now.
Tested out a couple of files and they all seem to be encoding at roughly 8.5x (1.4g Celeron laptop).

Liking the quality aswell of ~100kbits aswell. I'll post some ABX logs in a minute but I'm finding it very hard to hear a difference...

Here's the ABX report:

Works wonderfully on my P4. Many thanks to the Nero MP4 devs for this and taking the time to be a part of this community.

perhaps bittorrent sources will resolve the problem, also here, ftp6 doesn't respond

Thanks to Ivan and the whole Nero team for this release and their support of the community.

My question is related to tags: right now it doesn't seem like any tag info is being transferred and I'm not sure if that's a problem with my setup or I'm using the wrong command line parameters.

I'm using foobar2k 0.8.3, with the Nero encoder just released today and the command line I've settled on for testing at this point is one posted earlier in this thread:

q 0.425 -if - -of %d

Saw the suggestion about the fractional amounts and that 0.425 being approximately 128Kbps or so, which is just fine with me for my iPod. The FLAC files that I'm using as source material are fully tagged and the files are named Artist - Album - Track - Title.flac, so again I'm not sure where the issue lies.

My guess it I'm just missing something in the parameters, or perhaps the tags aren't going to transfer. If anyone can offer a solution or advice on this I'd greatly appreciate it. I've got 100GB of FLAC files that would fit nicely onto my 30GB Video iPod as m4a files.

Have fun, always...
bb

AMD problem has been solved, but it will take until tomorrow for the new version to be uploaded.

Apologise for the inconvinience caused to our AMD users.

Edit: Illegal link removed.

I'm using foobar 0.9.1's converter with FLACs and everything's tagging okay for me. My parameters differ slightly from yours, only due to the quality I've chosen (0.525)

I generally use Mp3tag to tag my FLACs. I doubt whatever you use (if any different) should cause the problem.

Thanks Nero Audio team. This is a realy nice surprise indeed. You guys are amazing.

Can you try with foobar2000 0.9? I wouldn't know why 0.8.3 doesn't work, but I know for sure 0.9 will because it was tested

Actually, the bug was not AMD-related, but related to CPUs without SSE2 - so latest Intel and AMD chips were not affected by this bug.

Sorry. I didn't know it was illegal. I'll take the torrent down.

Is there a change log for Nero's AAC encoder? Just to see the different versions and what work was done/bugs fixed and that sort of thing? I tried searching HA, but couldn't find anything.

I couldn't download with my browser too (guess however I have a general ftp problem with my browser).
I had no problem downloading with a specific ftp file transfer tool (FTP-Commander).

Starting from the command-line version, we will maintain a changelog for every revision.

And, yes - expect regular revision and updates

We don't have any changelog (yet)[1], however, you can find some discussion in the Nero 7.2 thread. Several bugs which are discussed there are fixed in this encoder.

[1] It would generally look like:
1) Quality improved
2) Bugs fixed
3) Made it faster

going into more detail is tricky because it might reveal our secret sauce But we'll see what we can do.

Too bad there wasn't one up 'til now. I was just curious to see how much was changed. But, the fact that there will be one now is awesome. Thanks Ivan!



Well, I definitely don't want the magician to reveal the secret to his magic. It might just spoil the whole show.

No idea why this is happening (just for me it seems), but foobar2k 0.9.1 transfers the FLAC tag info properly, but 0.8.3 doesn't. Weird, very weird. Anyone else having this issue, or can do some testing with 0.8.3 installed into a different directory?

bb

If I encode my speak audio files with this, is it legal for me to distribute these audio files?

Also, is it legal for me to distribute these exe's with my program?

AAC does *not* require you to pay patent royalities for distributing music, so I believe the answer to the first question would be "yes".

The answer to the second question can be found in the bundled "license.txt":



So the answer is: "no".

PS. This is a personal opinion. I am not the Nero legal department!

yeah w/2pass the bitrate comes out more of what i wanted. But Garf said that we shoudlnt do 2pass vbr so i guess i gotta go lower w/the q values.

Hi,

I asked Peter and he says MP4 tagging in foobar 0.9 is entirely rewritten. So quite likely, 0.8.3 has some bugs or issues that cause it to fail to work that are solved in the latest version.

It also takes way too much time to encode w/2-pass. If there's really no point in using it, then I figure I'm not going to waste my time. I did the same and lowered my q values.

Although, I believe the reason it takes so long is that I have all my music archived on DVD+R's, and foobar2k seems to like to convert 2 files at a time, which slows the whole process down quite a bit when it comes to transcoding directly from a DVD+R.

It's good to have floating point bitrate. Now I can have exact bitrate 50.0 ..... 130.0 .... . and It's free and less than 1 mb. Thank you

Bitrate is specified in bits/second, not kilobits/second.

Any mirror to download it? The ftp6 url don't work

I'm using -q VBR mode.

To Garf or Ivan:

I guess I'm just looking to clarification as to why 2-pass isn't a good idea with VBR. I know you mentioned video codecs doing 2-pass and its association with ABR. I also know that VBR should automatically figure the best way to allocate bits... I guess maybe I'm just looking for re-inforcement as to why I should opt out of using 2-pass (aside from longer encode times).

Cheers

i dont think its hte media source. it takes quite a bit w/hdd sources using 2pass. but then again it is going @ it twice ...

Please! Have tried everything!

I know your server is alive and well, but it is not working here!

ftp3 works well, for example.

What about Rarewares?

Well, unless I'm mistaken, data is burned onto a disc in linear fashion, and is read in linear fashion. Having two separate files being read causes the laser to jump around repeatedly, trying to multitask, which slows the spin-rate of the drive down significantly. (Off topic, sorry gang)

It is a question of terminology - 2-pass is of course VBR, but it does not make sense to use 2-pass with the -q option.

-q triggers "Quality VBR" mode - where bits are allocated per each frame to reach certain quality, so you don't really need two passes

-2pass triggers "2 Pass" mode - where encoder triggers to reach certain size - specified in average bits/second - file is still 'VBR' but bit rate allocation obeys target file size and the quality is distributed equally by using analysis in the 1-pass

Thank you very much for making this available for free!

Now onto being pesky: included readme.txt files says "Package build date: Apr 27 2006" while console output states "Package build date: May 1 2006".

Thanks Ivan. I'm terribly obssessive-compulsive when it comes to audio, and sometimes it just helps hearing it from a pro (it's all in the mind, I know).

I'd say 2 pass audio encoding is a bit ridiculous, both from a time perspective (taking twice as long) and from a quality perspective (how good do you think your ears really are, anyway?). Just seems like the same old situation I commented about here at HA a few years ago: "When is it really good enough?" If you want that kind of quality, lossless is the way to go for most people.

As for the time it takes to encode from some format to another, it's not two steps, it's three formats unless you're encoding directly from a WAV file:

1) The original source file (stuff like mp3, flac, alac, ogg, etc).
2) The in-between data file (temporary) created when the source file (#1) is converted to WAV format since almost all encoders require WAV files to work with to encode.
3) The output format of your choice.

So when reading the data from media like a DVD or wherever, the data is read, decoded to WAV (directly into RAM would be the best possible method for speed but I think the temporary files are still hitting the hard drive), then the encoding phase starts. It would be nice if it were possible to just transcode directly from one format to another, but unless the source is a real WAV file, it's always 3 parts.

bb

I think 2-pass makes a lot of sense with low-bitrate encoding of specific long content, such as long-duration tracks with human dialogs, silence and music - in this case, 2-pass encoding would make very good use of bits vs. quality as it would know that there is a lot of speech and/or silent content, and reuse this to improve more important parts.

2-pass is useful when you want to do an optimal bitrate allocation and you have to meet a specific average bitrate (or fixed size, which is the same thing). The first pass analyses the song with the psychoacoustics engine, and finds out where the hard and easy parts lie. The second pass uses this information to decide how to allocate the fixed number of bits it has.

The whole scheme does not make any sense when the encoder can spend as much bits as it needs (VBR mode).

(This whole thing made my think about ways to use the fact that with 2 pass mode we can see ahead further than in normal encoding, and there might actually be ways to exploit this, but they are most certainly not implemented let alone tested in the current encoder. For all I know, using 2 pass VBR with the current binary might make your Britney Spears tracks come out as Madonna, or something).




We prepared it last week, but when it wasn't up on the website this weekend I sent a new encoder with further improvements (and apparently, not a clear enough indication that, yes, the SSE2 build is faster, but no, it's not gonna work on all machines... sorry people). The commandline options didn't change in those few days.

Thanks for that bb. Although my ears are quite sensitive when it comes to audio. I'm sure thousands have said it before and, of course, it's completely relative. I do see the point in staying away from 2-pass now though, so thanks again.

As for my ears, take a listen to "Going Under" by Evanescence. At around 0:15 into the song there's a pop as Amy's saying "bleeding." I noticed that when I first heard the CD and it drove me so mad that I actually found the sample in the WAV of the song (there was quite the spike in the sine wave) and smoothed it out, which goes quite against a purist's best interests, I know. Most listeners don't pay attention, but I actually listen to my music.

It's maddening actually. You have no idea how hard it was for me to break down and use lossy codecs. Alas, for convenience certain sacrifices need to be made.

I'm crazy, I know.

Thanks for the gift. I would include the configuration into MAREO.

A little comment, why not use ISO date format for the filename?

NeroDigitalAudio_050106.zip is 2006-05-01, or 2006-01-05?

NeroDigitalAudio_yyyy-mm-dd.zip is much better

What would be the command line equivalent of what was used on sebastian mares' listening test?

EDIT: there is no artist, album, etc tag options right?

No worries, didn't want my post to come across as "Oh the hell with that, stop being stupid and just use it." That's not the case.

A few years ago I made a pretty intense posting here explaining how I'd been using mp3 encoders since *before* they became anything Shawn Fanning ever heard of and created Napster. Audio quality is and always has been a big BIG thing to me, but after almost 8 years of dealing with it, encoding, testing, re-encoding, re-testing, redoing entire collections of CDs just because a "better" encoder came out, the whole cycle repeats, ad nauseum.

There comes a time when you have to make a decision, sorta:

Either listen and enjoy the tunes, or continue searching for perfection and that next step, which rarely if ever actually becomes a reality. Such is the nature of psychoacoustic audio compression, basically. I just got tired of preaching EAC + LAME for years and worrying so much about perfect rips, perfect sound, perfect etc... I finally gave it all up and said to myself, "Screw it, it sounds good enough to me.

So yeah, I understand the quest better than most, actually. I was there where you're at, could slip right back into it easily, but it works like this:

I was sitting down early this morning to begin using dBpowerAMP Music Converter and the iTunesEncoder plugin to change all my FLAC files (on DVD) into plain old VBR AAC files (128Kbps nominally) then add the album artwork for all of 'em and slap 'em on my new 30GB Video iPod I just bought a few days ago. Someone here in Vegas won two of 'em when his wife hit a bonus on a slot machine at a casino, so they sold the second one to me for $200.

Lo and behold, the tagging was becoming an issue of sorts, so I popped into HA this morning and was downright blown away by Ivan's announcement about the free MP4/AAC encoder and snagged it and started working with it. While iTunes is ok for the encoding, it's damned slow, so since I didn't get too far into the stack of DVDs, I just deleted the work I had already done and will be redoing all 100GB with this Nero encoder - and now that I've worked out the tagging issue with 0.8.3 (it's still there, I just mean that I'm aware of it now so I'm using 0.9.1 for all the encoding), I'm all set and ready to spend the evening encoding from DVD to M4A.

</hijack_off>

Thanks again to Ivan, Garf (forgot to mention him earlier) and the whole Ahead/Nero gang for their wonderful products and continued support.

bb

Garf: And yep, I understand the usefulness of 2-pass encoding for very specific situations, but for general purpose audio encoding (meaning portable playback on portable devices which is the most common reason for using audio compression in the first place, I think), 2-pass just seems like a waste of that much more time, but to each his own. My opinions on the matter aren't meant to sway anyone, heaven forbid.

Is there any shot of this being available for us OS X users?