Hello, I'm pretty much a newbie to Vorbis. I am looking for an application that will normalize the volume with .ogg files the same way MP3Gain does with mp3s. I'm creating playlists from different CDs for my flash player and the volume on all of them, vary. I was wondering if I can do something simple like just equalizing the volume of the all the songs in my playlist without any tweaking of the actual content or echoing of each track.

I tried WinVorbis but it changes my .ogg files to .wav files (unless I check the "delete original files" option, then it will delete the original .ogg files) and I don't want that to happen unless I'm missing something in the process.

There really isn't an FAQ or users manual to go along with it, so I was wondering if there are any other options out there. I read the thread at...

http://www.hydrogenaudio.org/forums/index....showtopic=15049

....which helps somewhat, but it doesn't go into any detail about how to use WinVorbis (or OggDropXPd) while this link at....

http://winvorbis.stationplaylist.com

...is basically a download page that describes upgrades done for the last couple of years without much instruction on how to use it.

I generally encode at q5 or q6. I'm using PlexTools v2.17 which encodes .ogg files using Xiph.Org libVorbis I 20020717 (1.0)

Can anybody help? Sorry for sounding clueless but I guess everybody has to start somewhere. Thank you.

Why would you need to normalize when you can just use VorbisGain or something like replaygain in FB2K to apply an album or track gain? . FB2K has a diskwriter plugin for Vorbis that allows you to use it's DSP for post-processing of files. You could try adjusting the equalizer on FB2K to desired levels then using diskwriter plugin.



Check out some of the newer tuned versions of the encoder on recommended settings page. I would recommended Vorbis 1.1 or AoTuv Beta 3, etc. ;-D

Thanks for responding. Well I'm willing to try anything that will have me avoid having to fiddle with the volume control on my flash player everytime a louder .ogg song on my playlist plays.

How does Vorbisgain or replaygain work?

Sorry for sounding ignorant but I'm new to this sort of thing.

Thanks.

Check out the replaygain spec. at http://replaygain.org/. You could also browse through the wiki. Using replaygain in FB2K is pretty self explanatory. FB2k can automatically scan the file for you and apply the appropriate track or album gain. . if you are looking for VorbisGain standalone the documentation is included. http://www.sjeng.org/vorbisgain.html or you can download newer version form Rarewares in the post below.

You can download a Windows command-line version of VorbisGain here. It's very easy to use. FB2k works great too.

Hey thanks. Sorry for the dumb question but where do install these zip files? Do I need to create a new folder off my C: and install them there?

I want to make sure I'm doing this properly. Any help would be most appreciated.

There is no application that is comparable to mp3gain. So far as I'm aware, the volume cannot be adjusted in the same way as for mp3s. Vorbisgain will write vorbis tags that contain the appropriate adjustments but, unless the player is aware of these tags and able to use them, it won't do you much good.

Does that have something to do with the scale factors?



Just too add to that a lot of portables don't support replaygain tags in general ;-D. You might need to use WaveGain first and then encode your files using the appropriate Vorbis encoder. (sorry missed the sentence about flash player).

Hmm...that's interesting.

Currently I use Winamp v5.08e for my computer and a Samsung YP-MT6 1G flash player, both of which support .ogg playback.

I'd like to totally switch over from mp3 to .ogg but this "normalization" problem (for lack of a better term) seems to be my main concern at the moment.

[quote=HotshotGG,Apr 9 2005, 02:25 PM][quote]Just too add to that a lot of portables don't support replaygain tags in general ;-D. You might need to use WaveGain first and then encode your files using the appropriate Vorbis encoder. (sorry missed the sentence about flash player).[/quote]

In other words, rip CDs down to .wav format, and THEN encode to .ogg, instead of ripping directly from CD to .ogg?

yes, If they are from your CD collection try ripping them then adjust using WaveGain and then encode using one the latest Vorbis encoders ;-D. Normalization is generally not recommended due to the fact that it's lossy that's why replaygain is the pusher, but like John33 said MP3 files can only be adjusted due to the their scale factor nature I am assuming. ;-D. I hope that helps to answer your question.

Ok so let me see if I understand this. I would rip the CD down to a PCM (.wav file @ 44100 Hz 16 bit stereo), then using WaveGain, add the .wav files to the WaveGain box, set the Album gain button, hit the 'just calculate' button and it "normalizes" it (for lack of a better term) all those .wav files to roughly the same volume. Then use PlexTools to encode those .wav files to .oog?

Sorry to sound so elementary about it, but if that's the case, then I've been approching this whole thing the wrong way.

I most appreciate any help on this, thank you.

That's what I would do although once again I don't know if WaveGain is tag oriented (try though and see if it works). You will have to consult John about that (I think it might be . If it is then I can't help you, although if I were you I would download the latest Vorbis binaries or some updated DLL's so you can encode with the newer Vorbis builds. That's my take though. . It's unfortunate a lot of portables don't support ReplayGain tags (replaygain is in fact technically superior to normalization) in general or due to the nature of a lot of codecs with the except of MP3 and maybe MPC? you can't scale the volume. especially with Vorbis (if not so I would like a technical explenation why?). MP3 can adjust the scale factor bands in some way I guess in 1.5 dB increments? technically it's normalization, but the way MP3gain does it somehow lossless? I think maybe, because it only uses integer calculations?.

Hey HotShot, thanks for your responses. Who's John, btw?

Too bad about the ReplayGain tags. I guess if my Samsung doesn't recognize them (even though the player itself is .ogg compatible) then I guess it's just a waste of time trying to solve this. I just bought the player a couple of weeks ago so if the volume problem gets too intrusive, I'll just have to go back to using mp3s again. Which is a real shame because Vorbis has a lot of potential and sounds better.

However, If WaveGain ISN'T tag oriented, then I may be in luck and it may be the way to go as far as my portable is concerned. Anybody know?

One more question. If I download the latest Vorbis encoder .dlls, do all I need to do is unzip them to my PlexTools folder? And they'll automatically update the 1.0 version that's already in there? Or is there something else I need to do?

Yeah, I know that's three questions but hey...

-

Developer of WaveGain and OggDropXP he is the dude who mentioned the tidbit about MP3gain



You could always just adjust the volume control for now. I mean you can still encode with Vorbis like if you were sending files across the net or just listening to them on your HD ;-D or you could even use Lossless codec like FLAC to archive them on your HD.




I would assume so more than likely. I don't use plextools, but if you download the lastest Dynamic Link Libraries from rarewares and install them to the apporpriate dir. they should work.

I take it you're using Speek's frontend? You'll need to use the '--apply', or '-y', option. WaveGain directly alters the content of the wave file so that any subsequent encoding will have been replaygained, or 'volume normalised', if you prefer it. No tagging is done since there is no suitable wave chunk defined.

Hi, John.

Yes when I downloaded it, it came with a manual on those options. The only problem is, when I studied the gui interface for it, it doesn't show where those options are supposed to be entered. It presents itself as just a straightforward box with a few options for dither, output and extra gain, but without a place to enter a command line, unlike say: EAC.

For instance, when I direct it to analyze a .wav file, it goes into the sys32 cmd. command prompt that has a little chart showing the gain, peak, scale, new peak and track, along with the filename path to the file. Then it has "press any key to continue...." without leaving me a command line to enter those options

I'm probably missing something from somewhere. What am I missing?

Once you've loaded your files into the 'box', just hit the 'Calculate and Apply' button. Sorry I didn't make that clear before but I haven't looked at the frontend in a while.

Not true! Volume can be scaled in 2.2 dB steps.

Since there's no application yet, could you code one for us, john33?

Hey thanks for the clarification, John.

I don't know why they would include those switches in the WaveGain manual it came with, without any explanation as to how or where they should be applied. It throws a newbie like me a curve.

In fact, learning EAC, PlexTools, MP3Gain and Monkey's Audio was easy compared to this.

It's too bad more detailed instructions on VorbisGain or WaveGain aren't included in the FAQ. But I'm still learning.

Is this Vorbis you are speaking of kjonlee?




If you have any more questions feel free to ask

Well hey, I most appreciate that, HotShot. If I can understand the WaveGain settings and how they work, then you've got a new .ogg convert right here.

Because once I understand that, then it'll be no problem to convert those .wav files to .ogg

For instance, the Radio Gain option would be used for a single individual track while Album Gain would be used for the whole folder or a number of individual selected to appear in the WaveGain selection box. Correct?

Also what does the "6 dB hard limiter" do?

And what does "dither" do along with "noise-shaping"?

And as I click on the extra gain to + or - above zero, I assume this is the actual volume adjustment. Any particular settings you would recommend as a benchmark to start with? I realize I would have to play around with them myself in order to see what's best, but I'd like to know how high I can go before audio fatigue occurs or when clipping begins.

Once again, I really appreciate any information you can give me. I've learned a lot in the last couple of days and I'm really glad I've discovered this forum.

-

Yes, you are correct. They are in reference to 89 dB SPL which is the standard now.




A 6.02 dB hard limiter is used to prevent clipping. If you were to anaylze the waveform, that is the peaks of the waveform would look like a steady-state signal (pushing the signal amplitude to the max limit and holding it steady so it doesn't clip). If you don't know what clipping is it's basically the digital form of analog distortion. Clipping is a major issue for a lot of big label releases these day's as they try to squash the dynamic range of audio signal or take into the "hot levels" just to try and sell more records a.k.a (Loudness Race) this is especially a problem with remasters.



Dithering is the process of adding low level noise into a signal in efforts to minimize the quantization error. From a psychoacoustics standpoint the human air is less sensitive randomn white noise than it is to distortion so white noise coloring is more tolerable. You would use dithering if you were using bit-reduction or downsampling such as 24-bit -> 16-bit. Noise shaping is a technique very similiar to dithering it's not the same process though. Noise shaping is more frequency domain oriented. One example of where it's used is in 1-bit DSD converters for SACD. The ones that claim to be "distortion" free even though it's been proven mathmatically by those researchers at the University Of Waterloo in Canada it's it's not

This is false. If the signal is steady-state at the max limit, it is effectively clipped, regardless of whether it's just below or just at the maximum sample value.

I am pretty sure the 6dB hard limiter actually applies increasingly hard compression exactly to prevent that. Compression is still preferable over actual clipping.



Noise shaping is a very well understood and powerfull technique that shifts the errors introduced by quantization/dithering to an area where they are less audible, for example higher frequencies. I don't understand why you bring up SACD's problems here, it has nothing to do with noise shaping in itself, and only confuses the issue instead of explaining it.

Hmm.. ohh ok that makes more sense to me now.



That must be what I was thinking of when I anaylzed waveform from a CD that was hard-limited once and I think what I was observing was a compressed signal as you mentioned or one that was sent through a compressor.




I was using it as an example (not a good one), but I was right to some extent from what I understand it is frequency domain oriented (SACD does do just this even though it does confuse the issue).

Yep, Vorbis I.

Hey HotShot (or anybody else)

Unfortunately WaveGain doesn't give that 89 dB benchmark that you can look at the same way as MP3Gain.

For instance, what dB is the track recorded as a whole?

With only the 6db hard limiter tiked and the extra gain button untiked, I would get the following for a certain directory:

Gain | Peak | Scale | New Peak | Track
--------------------------------------------------------------------------
-5.66dB | 32025 | 0.52 | 16691 | Track 01

-6.59dB | 32025 | 0.47 | 14996 | Track 02

-2.93dB | 32024 | 0.71 | 22855 | Track 03

-5.38dB | 32025 | 0.54 | 17238 | Track 04

-5.24dB | 32025 | 0.55 | 17518 | Track 05

-6.36dB | 32025 | 0.48 | 15399 | Track 06

-5.66dB | 32025 | 0.52 | 16691 | Track 07

Recommended Album Gain: -5.70dB Scale: 0.5188

~

So what does this all tell me? In layman's terms?

If I tick the gain to -5.70, will that make the album louder?

And should I leave the 6dB hard limiter tiked or untiked if I apply that -5.70dB?

The individual gain per track indicates the amount by which each track would need to be adjusted in order to match the desired 89dB level, and the Album Gain indicates the amount by which each track should be adjusted such that the Album averages the 89dB while preserving the same 'relative loudness' between the tracks.

If you are not applying any additional gain, it is extremely unlikely that you will need to use the Hard Limiter. This is really intended for use when the calculated gain is being boosted by an additional gain and provides for a graceful compression across the 6dB band.

So 89.0dB is the default threshold here on WaveGain as well?

What I'd like to do is make the track louder before clipping occurs. I've noticed with MP3Gain that I can sometimes go up to 92.0dB to 93.0dB with some albums before this happens.

Based on the data I posted for the album I posted up above, how far do you think I can take the gain with the 6dB Hard Limiter unticked (turned off) and the "no clipping prevention" box unchecked?

Yep. I think I'm right in saying that all implementations settled on that.

There is no reason to switch the clipping prevention off, it won't do you any harm. I've found it perfectly safe to increase the gain by 3dB, ie., to 92db, with no ill effects, and without switching on the hard limiter. Leaving the clipping prevention on will automatically adjust the gain down to a safe level, if the extra gain would cause a problem.

So theoretically I could go up to the 94.0db - 97.0db range and leave the "No clipping prevention" box untiked (on) and still avoid clipping because WaveGain would automatically adjust the particular tracks where I wouldn't be able to get away with this. Yes?

Sorry to sound so elementary about all this. I really appreciate all the help you've given me.

Yes.

Ok well that makes sense. I'll just have to do a little math when I create playlists from wavefiles before I encode to Vorbis. As it stands, with 89dB as the benchmark, those album tracks I listed up above are already in the 94dB - 96dB range as it is. Without any gain added (or taken away) since I only scanned them without making any changes.

I'd like to get my .wav files as loud as I can before audio fatigue and clipping kick in.

Thanks for your help, John.