Does mono save space?

[Eddie the Editor]

Hi. I'm encoding files for authoring a dvd using TMPGEnc Video Mastering Works 5.
I realise that in most compression situations, mono sound saves space. However, TMPGEnc seems to think I've given it stereo, regardless, and I wonder if I will benefit.
i.e.
Before encoding to mpeg, the program asks for a video source and an audio source. My video is a quicktime reference file exported from the avid timeline. The audio is a wav file exported from the same region of timeline. Up until now, I've been exporting my wav in stereo, but for my latest project, pictures are more critical that sound. Since I'm trying to fit quite a long video onto the dvd, I'm wondering if I would save space better used for improving picture quality, by making the audio mono. I've done some short tests. Obviously the .wavs I export from the timeline are now half the size, but the mpegs TMPGEnc produces are identical in size, and before TMPGenc begins the process of encoding, it displays that the audio will be in stereo - even though the .wavs I've provided it with were mono.

Overall, the mono .wavs for this particular project are 800Mb smaller than the stereo ones, so if I can use that space, it would significantly improve the picture quality I'm able to encode at. I'm just wondering if TMPGEnc simply doubles up the mono provided in order to give me two tracks - and therefore whether I will really benefit.
Many thanks in anticipation!
Eddie

[jagabo]

If you are using AC3 or MP2 audio encoding mono will not save space. The audio bitrate determines the audio size, not the content. Ie, 256 kbps mono is the same size as 256 kbps stereo is the same size as 256 kbps 5.1. Reducing the audio bitrate from 256 kbps to 128 kbps will reduce the audio size and quality but the extra 128 kbps you can then give to the video bitrate will probably make no visible difference to the video.

If you are using uncompressed PCM on the DVD then using AC3 or MP2 will free up over 1000 kbps for the video. That's more likely to make a visible difference.

[jman98]

I'm not sure I agree with Jagabo's post. I would have to test MP2 encoding but MP3 encoding definitely is smaller at mono than strereo. And if you used AC3 with mono you could go to 1 channel which would result in a smaller file, although technically I am deviating a bit from what he is talking about.

Looking to save space in audio is a bit like worrying about saving pennies on the cost of a brand new HDTV. Any savings on the audio side aren't likely to amount to much unless you are doing something like converting 5.1 sound to 1 or 2 channels AND lowering the bit rate.

[Cornucopia]

jagabo, I think I know what you meant (and probably agree with you), but it looks like you may have mis-stated things a bit.

Taking a 2 channel PCM (usually 1500kbps) and converting it to 256kbps 2ch AC3 (or MP2, etc) saves space. Taking the same source and converting it to 256kbps 1ch AC3 (or MP2, etc) WILL NOT SAVE ANY MORE SPACE. However, Taking the source and converting it to 128kbps 1ch AC3 (or MP2, etc) WILL.

As long as you are willing to overlook the loss of stereophonic perspective & ambience, converting to that lower bitrate at lesser channels will still give you an EQUIVALENT quality (per channel), which I believe is what the OP was asking about. This would be similar to doing the common 5.1 -> 2.0 downmix (with resulting lower necessary bitrate).

Note: If you look at the big picture, it may not be enough of a bitrate savings to warrant all the effort. This would only really come into play if you were talking about multiple streams each being lowered, and running times were long.

Scott

edit: jman98, you beat me to it!

[jagabo]

Maybe I wasn't explicit enough.

Taking a 2 channel PCM (usually 1500kbps) and converting it to 256kbps 2ch AC3 (or MP2, etc) saves space.

That's what I said in the second part of my post. Going from 2 channel PCM to AC3 (whether mono or stereo) will reduce the size of the audio by over 1000 kbps.

Obviously, since he doesn't care about audio quality, he won't be using mono PCM (which would save about 750 kbps over stereo PCM). I'm not even sure DVD supports mono PCM.

Taking the same source and converting it to 256kbps 1ch AC3 (or MP2, etc) WILL NOT SAVE ANY MORE SPACE. However, Taking the source and converting it to 128kbps 1ch AC3 (or MP2, etc) WILL.

That's what I said in the first part of my post. The bitrate determines the size:

size = bitrate * running time (true of all codecs, video and audio)

The only way to further reduce the size is to use a lower bitrate. Mono audio may retain more quality at lower bitrates.

[AlanHK]

the .wavs I export from the timeline are now half the size, but the mpegs TMPGEnc produces are identical in size

What you want to change is not the input audio but what TMPGenc produces.
I'd guess it was AC3 stereo, at some fixed rate, regardless of what you give it as input.

Look around its settings and you may be able to change the audio bitrate. The default may be say 320 kB/s. You can use 128 or even lower for spoken word and save 1-2 MB per minute.

[Cornucopia]

Maybe I wasn't being explicit enough last time...

AC3 (or MP2, MP3, etc) at 256kbps for a 2ch signal normally uses 128kbps PER CHANNEL (barring "joint stereo" channel bitrate interactions).
Using a 1 channel signal at that same 128kbps (PER CHANNEL, again) DOES give the equivalent quality (not "may").
(jagabo, I'm sure you know this already; I'm really just clarifying it to the OP - it was that "However" section I wrote about)

The trick is getting your encoder to output that 1 channel (and at that halved bitrate). For TMPGEnc, I know it can be done because I've done it a number of times. It's just unlocking the preset and creating a new preset that specifies 1 ch with lower bitrate. Don't use default presets, as they're ALL geared towards 2 channel material.

Scott

BTW, DVD-Video does support 1ch LPCM.

[Eddie the Editor]

Hi all - thanks for all the information.
TMPGEnc gives a variety of encode formats but I can't see any reference to AC3 in any of the encode formats offered. When selecting the DVD template described as "DVD VIDEO STANDARD MPEG FILE", the three audio options offered within that are "MPEG-1 AUDIO LAYER II", (that's MP2 I think, yes?) LINEAR PCM, and DOLBY DIGITAL. MPEG-1 AUDIO LAYER II is the default. Within this, I'm offered a variety of audio bitrates ranging from 64 to 384. (224 is the default, and what I've used until now). DOLBY DIGITAL offers an audio bitrate range of 128 to 448 (again, 224 is the defaut).
So if I'm already encoding at 224, then going down to 128 won't make much difference? The duration of the video is 2 hours 37 minutes by the way.
So... have I got this right? .... if I export two stereo tracks from the avid to a .wav and select 'mono', all it's doing is producing one (mono) channel instead of two, hence the file is half the size. Then when I use that .wav in a program that produces mpeg for video, the encoding program duplicates the one mono channel I have given it and produces a second mono channel - therefore no savings in data, just a pointless loss of stereo?

[jagabo]

When selecting the DVD template described as "DVD VIDEO STANDARD MPEG FILE", the three audio options offered within that are "MPEG-1 AUDIO LAYER II", (that's MP2 I think, yes?) LINEAR PCM, and DOLBY DIGITAL. MPEG-1 AUDIO LAYER II is the default.

Dolby Digital = AC3

Within this, I'm offered a variety of audio bitrates ranging from 64 to 384. (224 is the default, and what I've used until now). DOLBY DIGITAL offers an audio bitrate range of 128 to 448 (again, 224 is the defaut).
So if I'm already encoding at 224, then going down to 128 won't make much difference? The duration of the video is 2 hours 37 minutes by the way.

Going from 224 down to 128 will let you add 96 kbps to the video bitrate. For 157 minutes on a single layer DVD an additional 96 kbps won't make a visible difference.

So... have I got this right? .... if I export two stereo tracks from the avid to a .wav and select 'mono', all it's doing is producing one (mono) channel instead of two, hence the file is half the size.

Yes, the uncompressed WAV file will be half the size.

Then when I use that .wav in a program that produces mpeg for video, the encoding program duplicates the one mono channel I have given it and produces a second mono channel - therefore no savings in data, just a pointless loss of stereo?

No. If you tell it to create a stereo channel from the single mono channel it will duplicate the mono channel. Or you may have the option to have the audio on one channel and silence on the other. It's also possible to leave it as a single mono channel. But the size of the audio data after compression is determined by the bitrate you select, not the number of channels. At any particular bitrate the audio will be the same size regardless of whether it is mono, stereo, or more. But a mono channel at 128 kbps may sound better than a stereo pair at 128 kbps.

I disagree with Cornucopia that a mono track requires only half the bitrate of a stereo track. MP3 for example doesn't have to blindly allocate half the bitrate to one channel and half to the other. It can encode as a mono track and a right/left difference track (see "joint stereo"). If there's not much difference between the left and right channels a mono track and a stereo track will require about the same bitrate. AC3 uses a scheme where bitrate is allocated to tracks dynamically. So if you had music on one channel and silence on the other, the full bitrate could be used on the music channel. Etc. So a mono track can require anywhere from half the bitrate to the same bitrate as a stereo track.

[Cornucopia]

jagabo,
I understand and kind of agree with your examples, but in practice, stereo usually has enough difference from mono, and the 2 channels have usually roughly equivalent distribution of energy and complexity needs, so I'm going to have to agree to disagree...

Eddie the Editor,
TMPGEnc doesn't do AC3. You'll need a different encoder to do that (see the Tools list on the left). Also, it looks like you're still trying to use the default presets...

Scott

[Eddie the Editor]

Guys - thanks so much for all the detailed help. I just about have my head around it now. You've saved me many hours of needless encoding - so your advice is much appreciated.
Eddie

[jagabo]

TMPGEnc doesn't do AC3.

I think it will if you pay them extra for the AC3 codec.

[Cornucopia]

Also depends on which TMPGEnc we're talking about. VMW does, Plus (what I was thinking of) doesn't.

My bad for not looking more clearly at the OP.

Scott

[Eddie the Editor]

Also depends on which TMPGEnc we're talking about. VMW does, Plus (what I was thinking of) doesn't.

My bad for not looking more clearly at the OP.

Scott

Mine is TMPGEng Video Mastering Works 5, which offers 'Dolby Digital' - which someone said is AC3. Should I be using that then? If so, for better quality or for file size saving?

[jagabo]

Uncompressed PCM will give the best audio quality. MP2 and AC3 are roughly equivalent at equivalent bitrates (and number of channels). At 224 to 384 kbps (stereo) audio quality is very good (most people could not tell the difference between that and uncompressed PCM) and the bitrate savings can go to your video.

Note that the USA DVD spec includes uncompressed PCM and AC3, not MP2. But most players will play MP2.

[Cornucopia]

You know, looking back at the original post, you mention how saving 800MB by going to mono WAV (aka LPCM) would be a big deal to you. This tells me:

Your movie length is 146 minutes. Also, you were comparing savings based on the LPCM sizes. So,

If you used LPCM and it was stereo LPCM it would be 1.6GB (standard 16bit, 2ch, 48kHz=1500kbps), but at mono it's 750kbps.
BUT, retaining stereo and using AC3 or MP2 at say 224kbps is even better. You get to keep it in stereo (and at nearly as high quality), and you get an even better savings that you can pass on to the video requirements. That's 1.33GB of savings as opposed to just 800MB of savings going monoLCPM.

Ultimately, it's not that you're TRYING to make it mono, it's that your trying to lower the audio bitrate to gain savings for your video, right?

Scott