GUIDE: How To Properly Encode Dolby Digital Audio (AC3)

[SomeJoe]

How to Properly Encode Dolby Digital Audio (AC3)


Introduction

Many people on the forum have experienced problems when encoding audio using Dolby Digital. These problems are primarily volume-related, with some dynamic range compression issues as well. This guide aims to educate about how Dolby Digital audio should be encoded, and how to make it sound best.


References

The primary references for the information contained in this guide are two guides on Dolby's web site. The first is Standards and Practices for Authoring Dolby Digital and Dolby E Bitstreams (http://web.archive.org/web/20031206104650/http://dolby.com/pro/digaudio/pa.ma.1102.Standards.S.pdf), which has the best information on Dynamic Range Compression. The other is Dolby Digital Professional Encoding Guidelines (http://web.archive.org/web/20040716131627/http://www.dolby.com/tech/L.mn.0002.DDPEG1.pdf) which gives an excellent explanation of the dialogue Normalization parameter. You will need Adobe Acrobat Reader (http://www.adobe.com/products/acrobat/readstep2.html) to view these .pdf documents.


Philosophy of Dolby Digital

Dolby Labs has been doing high-quality audio with cutting-edge techniques for a long time, using their past experience as a guide. As such, there is often confusion about their methods and philosophy to those of us who are not privy to that information. Of prime example is the current problem: Why is Dolby Digital so much quieter compared to my original sound?

Most audio destined for DVDs is audio originally recorded for use in the movie theater. The movie industry has a huge advantage when producing audio for the theater -- the theater has large speakers and amplifiers, and a quiet, near-ideal listening environment. Huge dynamic ranges are possible, where the slightest whisper of dialogue is audible, yet gunshots and explosions can be earth-shattering. Dolby's dilemma was: "How do we bring this audio, with its huge dynamic range, into the home?" This is a major problem -- most homes don't have the speakers and amplifiers necessary to shake the living room. Further, background noise in the home can easily drown out those subtleties in the soundtrack.

Dolby's answer is to allow the decoder to modify the sound to compensate for these problems. Low-volume sounds are boosted automatically so they can be heard, whereas high-volume sounds are quieted down so that speakers aren't blown and other persons in the home are not disturbed. Further, Dolby Digital allows for different program material to be equalized, so that volume does not have to be adjusted when switching between inherently quiet programs and inherently loud programs.


Decoder Specifics

The methods I'm about to present here for encoding Dolby Digital are generic and do not apply specifically to any one encoder. All Dolby-certified encoders (and some non-certified ones) will have the appropriate parameters available to follow this procedure. I have personally tested the Sonic Foundry 5.1 Plug-In Pack for ACID Pro, as well as Sonic Foundry Soft Encode. These methods should also work for BeSweet, Vegas Video + DVD, Scenarist, and other software-based encoders.


Basic Parameters

Every Dolby Digital encoder has some basic parameters that need to be set.

The first is the channel combination, presented as (Number of front channels)/(Number of rear channels), with an optional ".1" added to represent a low-frequency-effects (LFE) channel if present. i.e. 2/0 represents normal left and right stereo sound. 3/2.1 represents a standard "5.1" setup, of Front Left, Front Right, Front Center, Rear Left, Rear Right, and LFE. This parameter should obviously be set to the number of channels of program material you will be encoding.

The other major basic parameter is the bitrate. Obviously, higher bitrates allow for less compression. Typical bitrates used are 192 kbps for 2/0 program material, and 448 kbps for 5.1 program material.


Referencing Volume to a Known Level - Dialogue Normalization

To meet the Dolby Digital requirement that different programs should have approximately the same listening level (thus the consumer does not have to adjust volume level between programs), Dolby Digital incorporates a parameter called dialogue Normalization. This metadata parameter tells the decoder how far away from the reference level the average sound pressure level of the material's dialogue is.

The movie industry masters their soundtracks in a specific way. The maximum rated sound level (where all amplifiers are putting out their rated power) is 0 dB. Sounds below that level are rated in terms of how many decibels (dB) they are down from that maximum level. As such, these values are negative. The movie industry typically masters the "normal" listening level of dialogue (where people are speaking in a normal voice) at -31 dBFS. In other words, a speaking voice is at an average of -31 dB when referenced to the 0 dB maximum sound level, hence the term decibels of full scale (dBFS).

Since movie content is the largest class of programs to go on DVD, Dolby chose -31 dBFS as the reference level for audio on DVD, where 0 dB represents the maximum encodable digital sound level (full scale).

The dialogue normalization parameter needs to be set to the LAeq level of your program material's dialogue. LAeq stands for the long-term A-weighted sound pressure level. Loosely, this is the average volume level of your source material's dialogue. Us lowly consumers really don't have a tool that can measure this parameter, but we can get close. Sonic Foundry's Sound Forge has a "Normalization" feature that can measure the RMS level of a .wav file (or the portion thereof containing dialogue). CoolEdit may also have a feature like this. To use it in Sound Forge, open your .wav file containing the movie audio. Select a section containing dialogue (no sound effects or music). Go to "Process"/"Normalize". Select the "Average RMS Power (Loudness)" radio button. Then click the "Scan Levels" button. The displayed "RMS" level is very close (within 1-2 dB) to the LAeq level.

That RMS level is the number that the dialogue normalization parameter should be set to. In other words, if the RMS level in Sound Forge shows as -17.6 dB, set the dialogue normalization parameter in your Dolby Digital encoder to -18 dBFS.

The decoder will perform an attenuation of (31 + dialnorm) dB to the program material when played back. So, in this case, the decoder will attenuate by (31 + -18) = 13 dB. This will bring the average sound level of the material to (-17.6 - 13) = -30.6 dBFS. The program is now played back at approximately -31 dBFS, the reference level.

-31 dBFS is a lower average volume level than what is typical from other sources. It will be noticeable that you will have to turn the volume up on your system when playing a DVD versus playing broadcast, tape, or other non-Dolby Digital program material.


Allowing Comfortable Listening - Dynamic Range Compression

Meeting the other end of the requirement, that the consumer should be able to listen to quiet and loud sections of the program without having to adjust volume levels, requires a decrease in the dynamic range of the program. A movie, with whispers at -50 dbFS and explosions at -5 dBFS can't be comfortably listened to in the average home. The whisper is drowned out by extraneous background noise, and if the explosion is played at a tolerable level that doesn't wake up the neighbors, regular dialogue at -31 dBFS requires straining to adequately hear.

Dolby solves this problem by compressing the dynamic range of the program material. Quiet sounds are automatically boosted in volume so that they're audible, and loud sounds are automatically cut down in volume to tolerable levels.

There are several dynamic range compression profiles available that are custom tailored to the particular flavor of program material. However, all of them share the same basic features. All of the compression profiles can be thought of as an input-output "black box", where certain input volume levels are mapped to certain output volume levels. Observe this graph (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/dddrclabeled.gif), which is a graph of one of the Dolby Digital compression profiles (Film Light).

The blue line is the "unity gain" line, also referred to as the "no compression" line. This line represents that the dynamic range compressor feature of the decoder is essentially turned off, and no boost or cut of the program material is done.

The purple line is the compression profile for "Film Light". It is divided into 5 different sections, as are the other Dolby Digital compression profiles:

Unity Gain = Volume neither boosted nor cut
Variable Boost = Beginning of increasing volume of soft sounds
Constant Boost = Increase volume by a fixed amount for very soft sounds
Early Cut = Beginning of attenuating volume for loud sounds
Cut = Very loud sounds almost clamped to a maximum volume level

The application of a compression profile like this allows the soft sounds to be heard while preventing speaker overdrive and disturbances by the loud sounds.

The Dolby Digital encoder offers 5 different compression profiles that can be specified depending on the nature of the program material being encoded. This graph (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/ddcompprof.gif) illustrates all of the available profiles. The profiles range from no compression ("None"), to fairly light compression ("Music Light") all the way to extremely aggressive compression ("Speech").

For the exact dB numbers where each range of dynamic range compression is located on the graph, see the Dolby documents cited in the references at the beginning of this guide.

Many authors, when compressing audio to Dolby Digital, are turned off by the idea of dynamic range compression. You have this well-mixed audio with nice dynamic range and are then going to kill it by compressing that dynamic range. This is a valid concern, but should be answered by looking at what the listening environment is going to be. If you are authoring a DVD only for yourself, and you have a home theater room that can deliver theater-like sound, perhaps a compression profile of "None" is suitable for you. However, this profile may not sound good in a more mundane living room. Some experimentation may be in order to determine what compression profile will sound best for you. Most Hollywood DVDs use "Film Light" or "Film Standard".

The following point, however, cannot be stressed enough: In order for the Dynamic Range Compression to work as designed, the Dialogue Normalization parameter MUST be properly set first!

All of the dynamic range compression profiles assume that the average volume level of the program material's dialogue being fed to the dynamic range compressor is -31 dBFS. If that is not the case, boost or cut will be applied to the material when it shouldn't be!

A prime example is the situation where an average volume .wav file (with an LAeq of -16 dBFS, for example) is fed to Soft Encode using Soft Encode's default dialogue Normalization and Dynamic Range Compression settings. The default dialogue Normalization setting is -27 dB, and the default Dynamic Range Compression is set to "Film Standard". Because of the misadjusted dialnorm parameter, only (31 + -27) = 4 dB of attenuation is applied to the audio, so the average volume level is (-16 - 4) = -20 dBFS instead of the expected -31 dBFS. This places the audio on the DRC graph at the incorrect position, and now most of the audio is being played back in the Early Cut and Cut ranges. This causes the audio to sound flat and dull, with a possible audible "pumping" of the volume up and down as the decoder changes between Early Cut and Cut based on average volume level. Here is a representative graph (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/ddwrongdialnorm.gif). If dialnorm had been properly set to -16 dB, the audio would be centered at -31 dBFS, and would sound like it is supposed to.


Line Mode and RF Mode

The Dolby Digital compressors have the ability to further alter the compression profile to compensate for different transport mediums. Most of the time, audio is transported between devices in "Line Mode", where a line-level is used. There is also "RF Mode", meant for broadcasting of Dolby Digital and devices that send audio via RF cables to a TV set. RF mode sound from the decoder uses a higher average volume level (-20 dBFS vice -31 dBFS) in order to correlate volume level well with other, non-Dolby broadcast audio, and also can use a more aggressive Dynamic Range Compression to prevent overmodulating the signal. There is an option in most Dolby Digital encoders to turn on that overmodulation limiter (in Soft Encode, it is labeled as "RF Overmodulation Protection"). Since we are primarily interested in authoring for DVD which will operate in Line Mode, we do NOT want to insert the additional Dynamic Range Compression that RF Overmodulation Protection will add. Therefore, for DVD authoring, the RF Overmodulation Protection option should be turned off.


Example Compression Settings

For this example, I will use an audio file that was captured from analog material. It is a plain stereo .wav file, 48 kHz, 16-bit. (Note that this audio was from a filmed seminar, and thus the entire audio file was dialogue. Because of this, I did not select a particular range to compute the RMS level, but I rather selected the whole file. In your file, you should select only a portion that is representative dialogue.) I loaded it into Sound Forge and measured the RMS level (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/ddexsfrms.gif) of the entire file as -20 dBFS.

Knowing that, here are some screen shots of the proper settings to encode this file (Sonic Foundry's 5.1 Plug-In Pack for ACID Pro was used as the encoder. Your encoder may not have all options available.)

The first page (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/ddexacid1.gif) in ACID is the Audio Service Configuration, where the coding mode (2/0), the data rate (192 kbps), and dialnorm (-20 dBFS) are set.

The second page (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/ddexacid2.gif) is the Bitstream Information. Set these parameters are appropriate for your source material.

The third page (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/ddexacid3.gif) is the Extended Bitstream Information page. Your encoder may not allow these options to be set. I typically do not set any options here.

The final page (http://web.archive.org/web/20050208121358/http://pages.sbcglobal.net/wilsondr/ddexacid4.gif) is the Preprocessing page. Here is where you set the Dynamic Range profile you want to use for the material. The Line Mode setting is the one that will actually be used by the DVD player or Dolby Digital receiver, since the sound will be coming from the line outs. The RF Mode generally won't be used, but I typically set it to the same profile as the Line Mode. Make sure RF Overmodulation Protection is not checked.


Conclusion

I hope that this guide has given some insight into the proper methods that should be used to get the most out of your Dolby Digital encoding.

[kxy]

Thanks SomeJoe, for your clear and detailed explanations. :)

[JensG.]

Great work, SomeJoe!

(Dammit, now I will always have to think about sub-optimal audio in my divx!)

;)

[Julio]

Hi,

I think I've understood the whole post of SomeJoe, but I've got some more questions with a WAV (of a rock concert) I'm trying to encode to AC3, using Soft Encode :

Before reading this post, I thought it was best to set Dialog Normalization to -31 dB, because when I used lower settings it gave me an AC3 that, when decompressed to WAV, had extremely low dynamics.

So I encoded it with -31 dB dialog normalization, and I got what SomeJoe calls "up and down pumping of the volume" at certain moments in my soundtrack ! Therefore I think my dialog normalization parameter is not right.

Here are the values I got when analyzing my soundtrack with Cool Edit 2000 :

Min RMS Power : -87.2 dB
Max RMS Power : -12.0 dB
Average RMS Power : -22.8 dB
Total RMS Power : -22.4 dB

So, should I set dialog normalization to something like -23 dB ?

[SomeJoe]

Originally posted by Julio
... a WAV (of a rock concert) ...

Min RMS Power : -87.2 dB
Max RMS Power : -12.0 dB
Average RMS Power : -22.8 dB
Total RMS Power : -22.4 dB


Yes, I would set Dialog Normalization to -23 dB. Since your sound is of a rock concert, and the Max RMS power only goes to -12 dB (about 10 dB above the average level), you can probably set Dynamic Range Compression to None. If you find that the loudest parts of the concert are too loud to comfortably listen to, go back and set Dynamic Range Compression to Music Light.

[nuked]

So I've heard there are some gain markers in the AC3 that tell how to do dynamic range compression when decoding. Then you can either do it or not. I asume ACID does not encode these markers, it's just compressing the raw tracks, right? Now if I don't compress, is there a filter that I can compress with on playback that will use the same -31 db standard and that do as good a job as besweet? Why compress if you can leave it optional?

nukeD

[SomeJoe]

Originally posted by nuked
So I've heard there are some gain markers in the AC3 that tell how to do dynamic range compression when decoding. Then you can either do it or not. I asume ACID does not encode these markers, it's just compressing the raw tracks, right? Now if I don't compress, is there a filter that I can compress with on playback that will use the same -31 db standard and that do as good a job as besweet? Why compress if you can leave it optional?

I am not well-versed on the particulars of the AC3 file structure, but as far as I know, both the Dialog Normalization parameter and the Dynamic Range Compression parameters are metadata in the ac3 stream. In other words, the actual audio is not altered to apply these parameters, the parameters are only there as flags to tell the decoder what modifications to apply to the audio.

As such, if the decoder is properly programmed, it could ignore some or all of these metadata parameters, and not apply those modifications to the decoded audio. However, I know of no common hardware or software decoders that will do this. (There are some Dolby Digital decoders in some high-end home receivers that give the user some control over the Dynamic Range Compression, but that's about all I've seen). I know of no software decoder or filter that can optionally implement user-defined parameters on decode.

[nuked]

Ac3filter useses the metadata and leaves the decompression optional.
It's very cool.
If you turn it on you can watch the little DRC gain bar go up
and down while it plays back. If ACID really uses the metadata,
I should be able to see that little bar working. Maybe I'll give it
a try. That option could be reason enough for me to encode in stereo
ac3 instead of mp3.

nukeD

[nuked]

oh... I mean did I say ACID... I meant something cheaper ;P
no disrespect... it's probably worth it, but not for me.
I misunderstood and thought this was available as a plugin
for besweet now... guess it's still just the ac3enc.

UPDATE:
For the record, I can't get ac3enc to use this metadat trick for encoding
dynamic range control, but actually I'd have been very surprised if it had since ac3enc isn't what's actually doing the dynamic range compression in besweet. This would be a very neat feature to have in ac3enc though(as if DSPGuru has nothing else to do :) ).

[KoVaR]

so....
is it possible to make 6.1 and 7.1 files ?
it should be set to 3/3.1 for 6.1 ?
but for 7.1 ?
i think that 3/2/2.1 should be correct

:p

[resonator]

Works well with 2.0 material, however, if I want to mess around with 5.1 encodings (i.e. combining 2 sloppy truncated ac3 files by resyncing them using VDub), how am I supposed to find the RMS level of a 5.1 mix? Should I load every mono file into SoundForge, let it detect the RMS value and then use the avarage of these 6 values? Any suggestions?

[echooff]

Severyone on this thread seems knowledgable. Maybe someone could tell me what I'm doing wrong. Everytime I try to encode to ac3 (both 2.0 and 5.1)with either Besweet or softencode all I get is a File full of squeal. It seems to be the proper size and besweet's log file is normal. I have ac3 filter installed. My stero handles surround but all it plays is squeal also. I tried the different methods on the sitcky's in the audio forum. I have found several guides on the web and tried them. I am still coming up empty. Any suggestions?

[SomeJoe]

Originally posted by resonator
Works well with 2.0 material, however, if I want to mess around with 5.1 encodings (i.e. combining 2 sloppy truncated ac3 files by resyncing them using VDub), how am I supposed to find the RMS level of a 5.1 mix? Should I load every mono file into SoundForge, let it detect the RMS value and then use the avarage of these 6 values? Any suggestions?

I wouldn't use the average of all 6, because some channels of a 5.1 mix are more important than others.

I might start with loading just the left and right channels into sound forge and making a stereo .wav, and finding the RMS of that. Then, if you feel there's a decent amount of content in the center and rear channel mixes, that would raise the overall RMS level. For example, if you make a left/right .wav, and find the RMS is -18 dBFS, and there is a pretty good amount of surround and center, assume the RMS level is 3 dB higher then that. So I'd set dialnorm in the encoder to -15 dBFS.

Another way to do it would be if the mix is a true 5.1 mix with all dialog in the center channel. You might be able to just measure RMS of the center channel by itself and use that for dialnorm.

These are just a couple ideas. You may have to modify this depending on your content.


Originally posted by echooff
Everytime I try to encode to ac3 (both 2.0 and 5.1)with either Besweet or softencode all I get is a File full of squeal. It seems to be the proper size and besweet's log file is normal. I have ac3 filter installed. My stero handles surround but all it plays is squeal also. I tried the different methods on the sitcky's in the audio forum. I have found several guides on the web and tried them. I am still coming up empty. Any suggestions?

If the encoded file you're making is a .wav, that's AC3 frames encapsulated in a .wav header. On the computer, it will try to play this as PCM audio, not AC3, and you'll get noise.

You should use SoftEncode or BeSweet to make a standard .ac3 file instead, and try to play that.

If that doesn't work, there may be something wrong with your filter setup. You should probably start a new thread to get help with that since the problem is really not an encoding parameters issue.

[bitbrain2101]

Hi,

I have come over a strange problem encoding AC3-files with SoftEncode or Digigram Multichannel Encoder. I have encoded AC3 5.1 and 2CH Stereo files from mono wav-files with 48k/16bit.I can play and hear these files with PowerDVD and the Dolby Logo is shown in PowerDVD.But if I try to import these AC3-files in TMPGenc DVD Author it tells me "illegal File".If I import AC3-Files ripped from DVD everything is ok.I examined these files with GSpot,my self-encoded AC3-files are unknown,but the DVD-ripped AC3-files are shown as AC3-files with their bitrate.Is there perhaps some kind of header missing that my self-encoded files aren´t recognized ??

bitbrain2101;)

[nuked]

hmmm... I just found the same issue, where windvd works but matrix mixer doesn't. I re-encoded with intel byte order unchecked and it worked. I think that's the only thing I did diferently but I'm not sure. This could certainly make sense though.

correction: ac3filter, not matrix mixer of course.. anyway.. I guess it's mute.

[bitbrain2101]

Hi,

I have found a solution for the problem.I re-encoded the AC3-file with AC3Machine and now it works fine.

@nuked I didn´t have the Intel byte order checked any time

bitbrain2101;)

[FulciLives]

Hello :)

I don't have Sound Forge but I do have CoolEdit Pro 2.0 and here is what I get when I analyzed a WAV file:

---------------------------------------------------

Left Right
Min Sample Value: -29437 -29493
Max Sample Value: 29848 29591
Peak Amplitude: -.81 dB -.89 dB
Possibly Clipped: 0 0
DC Offset: -.001 -.001
Minimum RMS Power: -52.88 dB -53.46 dB
Maximum RMS Power: -10.38 dB -10.2 dB
Average RMS Power: -18.14 dB -17.95 dB
Total RMS Power: -17.54 dB -17.36 dB
Actual Bit Depth: 16 Bits 16 Bits

Using RMS Window of 50 ms

---------------------------------------------------

So what value do I use here for setting the DIALOG NORMALIZATION?

BTW I'm using the AC-3 encoder that comes with Scenarist in case that makes a difference (I doubt it does as the available settings seem to be the same as in your examples).

So basically I'm just confused by the output above that CoolEdit Pro 2.0 provides in relation to your guide which uses Sound Forge.

Thank you! :)

- John "FulciLives" Coleman

[SomeJoe]

Originally posted by FulciLives
Average RMS Power: -18.14 dB -17.95 dB


Average RMS power is what you're interested in.

-18 dBFS is the appropriate dialog normalization value for this material.

[FulciLives]

Originally posted by FulciLives
Hello :)

I don't have Sound Forge but I do have CoolEdit Pro 2.0 and here is what I get when I analyzed a WAV file:

---------------------------------------------------

Left Right
Min Sample Value: -29437 -29493
Max Sample Value: 29848 29591
Peak Amplitude: -.81 dB -.89 dB
Possibly Clipped: 0 0
DC Offset: -.001 -.001
Minimum RMS Power: -52.88 dB -53.46 dB
Maximum RMS Power: -10.38 dB -10.2 dB
Average RMS Power: -18.14 dB -17.95 dB
Total RMS Power: -17.54 dB -17.36 dB
Actual Bit Depth: 16 Bits 16 Bits

Using RMS Window of 50 ms

---------------------------------------------------

So what value do I use here for setting the DIALOG NORMALIZATION?

BTW I'm using the AC-3 encoder that comes with Scenarist in case that makes a difference (I doubt it does as the available settings seem to be the same as in your examples).

So basically I'm just confused by the output above that CoolEdit Pro 2.0 provides in relation to your guide which uses Sound Forge.

Thank you! :)

- John "FulciLives" Coleman

*** EDIT ***
Just wanted to say thank you for answering my question SomeJoe :)

[Sunix]

Hi,

I just read your guide and found it very usefull.

However I am a little confused about what value to use for normalisation.

In Sound forge, I get these values:

RMS: -18.4 db WITHOUT use equal loudness contour option checked
RMS: -21.1 db WITH use equal loudness contour option checked

And if I run Cool Edit on the same file, I get these values:

In FS Square mode:

Minimum RMS Power: -57.74 dB -54.83 dB
Maximum RMS Power: -5.01 dB -6.13 dB
Average RMS Power: -31.11 dB -31.37 dB
Total RMS Power: -22.34 dB -22.91 dB


In FS Sine mode:

Minimum RMS Power: -54.73 dB -51.82 dB
Maximum RMS Power: -1.99 dB -3.12 dB
Average RMS Power: -28.1 dB -28.36 dB
Total RMS Power: -19.33 dB -19.9 dB


I read in a previous message that Average RMS Power from cool edit should be used when you just have cool edit as program, but these value are far away from those in Sound Forge. Even in Sound forge value change depending of selected options.

So what is the best value to use for normalisation?

Thanks in advance,
Sunix

[SomeJoe]

Originally posted by Sunix
In Sound forge, I get these values:

RMS: -18.4 db WITHOUT use equal loudness contour option checked
RMS: -21.1 db WITH use equal loudness contour option checked

And if I run Cool Edit on the same file, I get these values:

In FS Square mode:

Minimum RMS Power: -57.74 dB -54.83 dB
Maximum RMS Power: -5.01 dB -6.13 dB
Average RMS Power: -31.11 dB -31.37 dB
Total RMS Power: -22.34 dB -22.91 dB


In FS Sine mode:

Minimum RMS Power: -54.73 dB -51.82 dB
Maximum RMS Power: -1.99 dB -3.12 dB
Average RMS Power: -28.1 dB -28.36 dB
Total RMS Power: -19.33 dB -19.9 dB


I read in a previous message that Average RMS Power from cool edit should be used when you just have cool edit as program, but these value are far away from those in Sound Forge. Even in Sound forge value change depending of selected options.


Interesting. :p I don't have Cool Edit, so I haven't been able to read the documentation to see what the difference is between "Average RMS Power" and "Total RMS Power". However, if you look back in this thread, previous posters who have asked me this question about CoolEdit, their Average RMS and Total RMS values were very close to each other. Yours are very different.

Also I haven't seen any documentation on what the difference is between "FS Square" mode and "FS Sine" mode, but if I had to guess I would pick FS Sine, because that seems to say that the RMS value is being computed from a Fourier series based on sine wave computations, which is how it should be done.

As far as Sound Forge goes, I'm reasonably confident that you should use the RMS value you get without the equal loudness contour.

Since the RMS value is a definite, fixed value for the file, the differences we see in values between all methods must come from algorithm differences in the way each is computed. I would say that from your file, it appears as if Cool Edit's FS Sine, Total RMS Power is the closest algorithm to Sound Forge's RMS without equal loudness contour.

Because of that, I would say your appropriate DialNorm setting is -19 dBFS.

If you or anyone else has appropriate documentation from CoolEdit on an explaination of their reported RMS parameters, I'd be able to make a more educated recommendation as to which value to use. Also keep in mind that the true setting of dialnorm is the LAeq level -- using RMS is our low-budget, attainable, close approximation. :D

---------------------------------------------------------------------

Edit:

I went and looked at the help file for Sound Forge and found this regarding the equal loudness contour:

Checking the Use equal loudness contour option causes the scan to take into effect the Fletcher-Munson Equal Loudness Contours. Essentially, very low and high frequency material is less audible than mid-range audio. This option forces the scan to weigh this factor into the RMS calculation.

In view of that, I would recommend that you use the equal loudness contour setting when performing a scan for normalization. This is contrary to my above recommendation. :eek:

Also in view of that, I would then change my recommendation for your DialNorm level. Since Total RMS Power, FS Sine from CoolEdit is -19 dB, while RMS, equal loudness contour enabled level from Sound Forge is -21 dB, you could use either with confidence or use their average of -20 dB. A +/- 1 dB error will not make a very (if any) audible difference in the resulting AC3, and indeed the error imposed by using RMS instead of true LAeq could be more than this anyway.

[macik-pacik]

Hi!! I am just a new here, at this forum and DVD burning as well. I do not have problems with a quality and processing AC3 files, I use Maestro to authoring, but sometimes -e.g. Matrix reloaded, I got large pieces of audio streams - 2x370MB /english and czech/. that means I am to reduce video quality through CCE. Are there possibilities to reduce AC3 to a smaller size??
Thanks a lot. j.:)

[clapper]

quote:
--------------------------------------------------------------------------------
Checking the Use equal loudness contour option causes the scan to take into effect the Fletcher-Munson Equal Loudness Contours. Essentially, very low and high frequency material is less audible than mid-range audio. This option forces the scan to weigh this factor into the RMS calculation.
--------------------------------------------------------------------------------

Are you sure that the curve should be included? It sounds as thou the program would, in essence, apply an EQ curve to the program (non-destructive to the file) before analysis, and thereby raise or lower the result?

[SomeJoe]

Originally posted by clapper
Are you sure that the curve should be included? It sounds as thou the program would, in essence, apply an EQ curve to the program (non-destructive to the file) before analysis, and thereby raise or lower the result?

That is true, and that's essentially what Sound Forge would be doing. I would still recommend it be included because the objective is to make use of a quantifiable scale (dB of RMS power) that is as closely as possible related to the perceived volume level of the material. (Which should be as close as possible to the actual LAeq level of the material). Since the equal loudness contour applies a psychoacoustic model to the sound when computing the perceived loudness, you can view the application of this curve as an updated algorithm for computing the RMS power (if you're trying to associate RMS power with perceived loudness).

At any rate, if you don't believe the equal loudness contour should be used or is not applicable to the material, it can be turned off in Sound Forge's normalization dialog box.

Furthermore, I believe the difference in final measurement with and without equal loudness contour may be well within the tolerance/error that results from using an RMS power measurement rather than LAeq. I would state that if the error associated with using RMS is too high for your application, that you should probably be using a higher end tool to measure LAeq directly instead of RMS power.

I would like to actually test and see how close the various RMS measurements from Sound Forge and CoolEdit/Audition are getting to LAeq, but I unfortunately don't have any tool that can measure LAeq directly. Thus my "poor man's" approach to the entire problem. ;)

[Kilyan]

Hi!

I'm making dolby ac3 soundtracks in my language for dvds like saving private ryan and a lot more, taking it from vhs source. My problem is that if I use your method (For example I get -14 dB rms in Sounforge) and encode in softencode I get a much quieter audio track (2.0 dolby encoded )than the english 5.1 . If I use the -27dB I get better results but the audio is pumping (companding and expanding). So what's the problem here? How could I get a resonably loud audio without pumping?

BTW why all ac3 tracks Ive seen so far on dvds use the -27dB level?

Thanks

[SomeJoe]

Originally posted by Kilyan
How could I get a resonably loud audio without pumping?

If you want it louder than the reference -31 dBFS level but don't want the pumping, you will have to turn the dynamic range compression off.

[tbrunner]

Hi

I have also a lot of problems with bad AC3 sound resulting in to low volume dialogue (very difficult to understand) and to loud sound (explosions, music and so on).

My scenario is the following: I take a divx/xvid avi file and encode it to mpeg2 with AC3 sound to burn it on a dvd and watch it on a stand alone dvd player. I prefer AC3 sound as it is supported both for NTSC and PAL dvd's. As I would like to have the highest possible bitrate for the video, I encode the sound to two channel stero with 192 KB.

The sound in this divx/xvid files is either
a) MPEG Layer 1/MPEG Layer 2/MP3 or
b) AC3 5.1 or
c) AC3 2/0

My way of processing is:

1) convert the sound to WAV: For this I use the MPEG2 Encoder (Mainconcept in my case) to produce elementary video and audio streams. For the audio stream I select wav format. As the Mainconcept encoder uses direct show to read the file, AC3 Filter can be used to decode the AC3 sound and downmix it to 2 channel. I assume this works with every encoder who supports either direct show directly or can read avisynth files.

2) convert the WAV to AC3 using AC3 machine (I do not use AC3 Machine to downmix AC3 5.1 sound directly to two channels as proposed in the DOOM9 guide, because as result, I still get a 5.1 AC3. This surprises me not as the resulting command line is the same regardless if channels mode 5.1 or stero is selected).

Now my question is, where and what do I set to archive the proper 5.1 downmix to 2 channels, Dialog Normalization and Dynamic Range Compression. I would like to use as much as possible freeware tools.

I assume that the solution depends on the sound type in the avi.

a) MPEG Layer 1/MPEG Layer 2/MP3 (I assume that the low volume problem also occurs with this sound types as the people are not properly encoding the input AC3).
How and with what tool do I have to reencode the WAV to a WAV so that if feeding the WAV to AC3 Machine the sound is correct in the resulting AC3?

b) AC3 5.1
As the AC3 Filter provides so many settings, I assume that it is feasible to create a WAV file that can just be feed to AC3 Machine and the resulting AC3 sound is correct. The most obvious setting is to choose 2/0 stereo as output. To what do I set the other settings.

c) AC3 2/0
Some as AC3 5.1 but with different settings in AC3 filter.

Can someone give advice on the correct settings?

[tbrunner]

Hi

I have encoded a wav to ac3 exactly following the instructions in the first post of this thread. The result sounds very satisfying. I had to set the dialog normalization to -24 db. For the dynamic range compression I used "Film: Standard".

After encoding to ac3 I analyzed the resulting ac3 file. Opening it again with Soft Encode it showed -24 db for dialog normalization in the stream settings. I played it back with Radlight 3.03 over AC3 Filter. In the AC3 filter settings (right click on radlight, Advanced/Filters/AC3Filter) the DRC level slider was moving.

This seams to prove that setting the dialoge normalization and the dynamic range compression "only" affects the meta data of the ac3 stream.

I have encode the same wav with AC3 Machine. Afterwards Soft Encoded showed -31 db for dialog normalization and the DRC slider was not moving in AC3Filter.

For me this means that AC3 Machine expects as input a wav file which is "properly" (dynamic range compressed, loudness attenuated) mixed.

Does somone know how to mix the wav file (drc, loudness attenuation)?

[tbrunner]

Hi

If the source is a rip from a cd or the sound of a DV-AVI file from a camcoder, is it then also necessary to apply dynamic range compression? I would expect not, as the input has a normal "dynamic range" and needs not to be changed. Is this assumption correct?

Is there a way to measure the dynamic range of a given file to find out if and what dynamic range compression should be applied?

Thank you for any hints.

[Mug Funky]

@kilyan:

if you're getting your soundtracks from VHS (especially if it's taped from live broadcast), it will already have a compressed dynamic range. compressing further will inevitably result in pumping.

easiest way to check if you need to compress at all is to take a look at the waveform overall. if everything is at a similar volume then it's already been limited and further compression will just reduce quality.

so i'd turn DRC off altogether.

as far as why most DVDs use -27dB dialog normalisation? probably because it's the default. the commercial DVD author i know had no idea what dialog normalisation is all about, and apparently they generally leave it at -27. If the sound is bad (very rarely on the whole) they just turn it off and re-encode (after reading this thread i set him straight on this. hehehe).

@tbrunner:

the situation is similar with DV tapes... most DV camcorders will have an internal limiter that prevents clipping. they do, however have a much wider range than VHS (especially recorded from broadcast through RF), and so could benefit from compression. again, look at the waveform for loud transients and quiet speech, etc. (the DV cams i've used seem to only do significant limiting when you plug an external mic into them. the internal mic i'd say is also limited, but far less).

[Soapm]

Great faq Somejoe! Thanks. I now know how bad I've been destroying the audio.

Can you give us more input on how to fix an AC3 5.1 file?

Ex. After reading your faq, when I demux a file that has 5.1 AC3 audio I use the file verify option in soft encode to check it. If the diagnorm is -27 and it sounds ok I don't touch it. However, I have one that sound encode will not open. Any ideas how to fix it without destroying the quality?

I also have one that the diagnorm says reserved. Any idea what that means? Is this a good thing?

What if I just want to add light compression to a 5.1 AC3 file? Ieas on that?

[SomeJoe]

Originally posted by Soapm
Can you give us more input on how to fix an AC3 5.1 file?

Ex. After reading your faq, when I demux a file that has 5.1 AC3 audio I use the file verify option in soft encode to check it. If the diagnorm is -27 and it sounds ok I don't touch it. However, I have one that sound encode will not open. Any ideas how to fix it without destroying the quality?

I also have one that the diagnorm says reserved. Any idea what that means? Is this a good thing?

What if I just want to add light compression to a 5.1 AC3 file? Ieas on that?

If Soft Encode won't open the AC3 then the file is probably corrupted. You may be able to fix the file with one of the tools that fixes bad AC3 frames (not sure of the names of them ... maybe one is called AC3fix?)

I have no idea what the "reversed" message means. Never have seen that before, but I only use Soft Encode rarely.

Changing the metadata parameters of an existing AC3 file (i.e. changing the dialog normalization parameter or the DRC mode) should be theoretically possible by altering the metadata in the AC3 stream. But I know of no software which can do this. Writing it wouldn't be too tough for some of the guys around here IF we could find some detailed specs on the AC3 file structure -- but I doubt Dolby is going to post that. ;)

[magilvia]

Thanks SomeJoe for this really instructive post.
I've only question: how to set in Soft Encode "Audio production information", which are "Mix level" (default to 105db SPL) and "room type" ?
Is it best to leave "Info exist" checked?
Thanks again

[magilvia]

One other question too :p :D
Has anybody verified if the average RMS of CoolEdit and SoundForge is compatibile with that measured with Goldwave ?
I like Goldwave MUCH more than CoolEdit plus it is SHAREWARE!

[SomeJoe]

To get further information on the audio production information and the room type, I think you'll have to go to those .pdf files that are on Dolby's web site. I'm not positive what those settings do, nor do I know if typical AC3 decoders/receivers will use that metadata to alter the playback sound. I typically leave those settings at the default, with the "Info Exist" unchecked.

I have never used Goldwave, so I'm not positive if it's measured RMS readings are on par with Sound Forge of CoolEdit. But if the reading says that it's an RMS measurement, it's probably very close.

[oldiexyz]

Hi all,

emmm... may I come back to an older reply by bitbrain2101?

On August 6, 2003, he asked about a problem with Digigrams MC en- / decoder. I do have exactly the same problems: AC3s produced by this program cannot be used in other apps like TMPGEng DVD Author and cannot be "reconverted" into 6 waves by BeSweet. On the other hand, the Digigram DEcoder cannot produce WAVs from AC3s made by e.g. the AC3Machine which uses the ac3enc.dll.

Originally posted by bitbrain2101
Hi,

I have come over a strange problem encoding AC3-files with SoftEncode or Digigram Multichannel Encoder. I have encoded AC3 5.1 and 2CH Stereo files from mono wav-files with 48k/16bit.I can play and hear these files with PowerDVD and the Dolby Logo is shown in PowerDVD.But if I try to import these AC3-files in TMPGenc DVD Author it tells me "illegal File".If I import AC3-Files ripped from DVD everything is ok.I examined these files with GSpot,my self-encoded AC3-files are unknown,but the DVD-ripped AC3-files are shown as AC3-files with their bitrate.Is there perhaps some kind of header missing that my self-encoded files aren´t recognized ??

bitbrain2101;)

In the following reply (by "nuked") there was an remark about the "intel byte order" (vs. "motorola"?). I guess, this could be a reason, but...


...sorry, where ist the flag set? What do I have to change to make Digigram's en- / decoder work compatible???

Trying to fix the output file using BeSplit / BeSliced leads to an empty (0 kb) file. Log:

BeSplit v0.9b6 by DSPguru.
--------------------------

Logging start : 02/16/04 , 15:51:45.

d:\tools\besweet\BeSplit.exe -core( -input E:\DVDs\6channel-dg.ac3 -fix -logfile d:\temp\BeSliced.txt -type ac3 -output E:\DVDs\6channel-dg_Fixed.ac3 ) -profile( BeSliced v0.3 )

[00:00:00:000] +------- BeSplit -----
[00:00:00:000] | Input : E:\DVDs\6channel-dg.ac3
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] | Channels Count: 5, Bitrate: 448kbps
[00:00:00:000] | Output : E:\DVDs\6channel-dg_Fixed.ac3
[00:00:00:000] +---------------------
[00:00:00:000] | Writing E:\DVDs\6channel-dg_Fixed.ac3
[00:00:00:000] +---------------------
[00:00:00:000] Operation Completed !
[00:00:01:000] <-- Process Duration
Logging ends : 02/16/04 , 15:51:46.


The "demux" log using BeSweet is:

BeSweet v1.5b25 by DSPguru.
--------------------------
Error 59: Failed to sync to payload's start position : "e:\DVDs\6channel-dg.ac3"Using azid.dll v1.9 (b922) by Midas (midas@egon.gyaloglo.hu).

Logging start : 02/16/04 , 15:56:15.

D:\Tools\beSweet\BeSweet.exe -core( -input e:\DVDs\6channel-dg.ac3 -output e:\DVDs\6channel-dg- -6ch -logfile D:\Tools\beSweet\BeSweet.log ) -azid( -c normal -g 0.95 -L -3db )

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : e:\DVDs\6channel-dg.ac3
[00:00:00:000] | Output: FL, FR, SL, SR, C, LFE
[00:00:00:000] | Floating-Point Process: No
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Total Gain: 99.554dB, Compression: Normal
[00:00:00:000] | LFE levels: To LR -3.0dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: No
[00:00:00:000] | Rear channels filtering: No
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] +---------------------
[00:00:00:000] <-- Transcoding Duration

Logging ends : 02/16/04 , 15:56:15.


What does "error 59" mean, and how can it be fixed?

Thanks in anticipation!!


I.

[pixita]

can someone point me some links to download the necessary software to do this?
Thanks

[ursamtl]

Originally posted by pixita
can someone point me some links to download the necessary software to do this?
Thanks

The BeSweet homepage at http://dspguru.doom9.org/ should be a good starting point.

You can also find some of the freeware at http://www.needfulthings.webhop.org/ in the audio/tools folder.

In addition, you can check Doom9 in the Downloads section. And, if in doubt, you usually find just about anything legal at www.google.com.

Happy encoding!

[skobipe]

Hi
I have 6 mono wav. files to make an .ac3 from.
I got those results in sounf forge:

* Merging the front left/right channels into streo file an
checking the rms gives: -29.8 DB.
* Merging the surround left/right channels into streo file an
checking the rms gives: -34.5 DB.
* Checking the rms of the center channel gives: -34.4 DB.

How could I know the right value to put in the dialog normalization
according to these results?

Thank You

[ursamtl]

Originally posted by skobipe
Hi
I have 6 mono wav. files to make an .ac3 from.
I got those results in sounf forge:

* Merging the front left/right channels into streo file an
checking the rms gives: -29.8 DB.
* Merging the surround left/right channels into streo file an
checking the rms gives: -34.5 DB.
* Checking the rms of the center channel gives: -34.4 DB.

How could I know the right value to put in the dialog normalization
according to these results?

Thank You

First look for your loudest result of these three, which is the fronts. Then take 31 + -29.8, to get an attenuation of 1.2dB..

Here are a couple of good references:
http://www.macprovideo.com/aPack/dialNorm.html
http://www.hometheaterhifi.com/volume_7_2/feature-article-dialog-normalization-6-2000.html

[skobipe]

@ursamtl

Thank you!
The guides were very helpful.
If I may, I have another minor question:
I use soft encode for encoding and in the "preprocessing" tab
I can choose "90 degree phase shift" and/or "3DB attenuation" from the "surround channel processing" menu.
I wanted to know which of the two, if any, will produce better results.

Thank You

[ursamtl]

Originally posted by skobipe
@ursamtl

Thank you!
The guides were very helpful.
If I may, I have another minor question:
I use soft encode for encoding and in the "preprocessing" tab
I can choose "90 degree phase shift" and/or "3DB attenuation" from the "surround channel processing" menu.
I wanted to know which of the two, if any, will produce better results.

Thank You

You're very welcome skobipe. The more we share our collective knowledge, the more progress we all make.:)

The 90° phase shift is not essential for decoding through a Dolby Digital 5.1 system, but it is if your mixed might be played back on a Dolby Pro Logic or surround system. These expect the surrounds to be phase shifted by 90°

The 3dB attenuation is usually necessary for files that will end up being played back on consumer home theater equipment.

If you want details just Google them and you'll find there's tons of stuff around on the net.

Hope this helps. Happy mixing!
Steve.

[ursamtl]

One other good reference for AC3 of course is the www.dolby.com site in their Information menu.

Steve.

[ursamtl]

In another thread, I read that SurCode Dolby Digital v2 and the AC3 Encoder for Acid use newer Dolby libraries than SoftEncode. Has anybody been able to compare the results from encodes done with both libraries?

Steve.

[keithmac]

"90 degree phase shift" and/or "3DB attenuation" from the "surround channel processing" menu.

Are these 2 options actually altering the sound itself or just metadata tags that are read by the decoder?

The 90 degree phase shift option, if it actually alters the sound data itself, and you apply it to waves derived from a dvd ac3 source (that should already have been phase shifted) surely it will be shifted by 180 degrees on the 2nd ac3?.

Same with the 3db attenuation, is it just a metadata value or is the sound itself altered while encoding?

Would be interesting to have a list of what options are just metadata for the decoder to read and what options actually modify the sound during encoding.

[ursamtl]

Originally posted by keithmac
"90 degree phase shift" and/or "3DB attenuation" from the "surround channel processing" menu.

Are these 2 options actually altering the sound itself or just metadata tags that are read by the decoder?

The 90 degree phase shift option, if it actually alters the sound data itself, and you apply it to waves derived from a dvd ac3 source (that should already have been phase shifted) surely it will be shifted by 180 degrees on the 2nd ac3?.

Same with the 3db attenuation, is it just a metadata value or is the sound itself altered while encoding?

Would be interesting to have a list of what options are just metadata for the decoder to read and what options actually modify the sound during encoding.

Hi keith,

There's a good explanation of this at YOU ARE SURROUNDED ([URL=http://www.soundonsound.com/sos/dec01/articles/surround5.asp).

Regards,
Steve.

[keithmac]

Thanks for the link, some good reading there!

[Capturebat]

In The AC3 machine it has dialog normalization reduction, but you can't input a value. Does it automatically scan and adjust for you?

[Paulcat]

It seems i'm in the right thread for this (I hope!)

I have a video file which has AAC 5.1 audio. I can use FAAD to convert that to either a 5.1 WAV file or a stereo WAV file. I would like to take the audio from AAC 5.1 to AC3 5.1 in order to put this video onto a dvd and still have surround. I use TMPGEnc DVD Author for mastering DVD's which I assume will remux my video with AC3 5.1, but for converting video files I use TMPGEnc Plus which downmixes everything to stereo.

What would be the SIMPLEST way (eg the LEAST amount of different software packages!) for me to convert a 5.1 WAV file into a 5.1 AC3 file?

I attempted this with BeSweet 1.4 and AC3 Machine (and BeSweet GUI 0.6) but it stated the AC3Enc.dll was missing (which I have since found) only to read that this DLL is next to useless for creating proper AC3 audio streams (or at least that's what people appear to be saying).

I'm new at this and the terminology is still way over my head, HELP!

[00diabolic]

Hi

I have been encoding my moviez with mp3 audio for as long as I can remember. I know all the tricks when it comes to video. Yet audio has always been mp3.

Recently I got ahold of a movie and was shocked that the audio was ac3 and file size was still normal. I figured ac3 had to be huge cuz thats whats on dvd's. So I started testing and found ac3machine, ac3enc.dll, and besweet (which I have used with gknot before) and discovered I could make an ac3 that was 128kps and was the exact same size as my audio in mp3 format. That blew my mind. 5.1 Ch 128kps audio same size as mp3 2 ch.

In the ac3machine guide it says "Set whatever bitrate you see fit. Going below 224kbit/s for a 5.1ch AC3 doesn't make much sense and going above the input bitrate doesn't make any sense either."

Then I read the above. Yet At 224 that audio was twice as big 220megs, no good. 128 same size as Mp3 @ 128 around 125megs.

So my question is this. Is ac3 5.1 128kps audio better then mp3 2ch 128kps audio that I have been using for years?

I have a 5.1 ch system and from what I can hear the 5.1 sounds good in 128kps except for a lil hiss. So what am I missing. Is it better quality then mp3 or not?

Would that ac3 encoded to 128kps be better with a different encoder?

Ac3machine uses the ac3enc.dll and I know that its not as good as other encoders. Can I use the sonic foundry soft encode dll with ac3machine, and how would I do that? or do I have to use soft encode to do it all? Soft encode is slow, is there a fast way to encode with ac3machine and not use the ac3enc.dll?

Any suggestions would be helpful. I know that both AAC and mp3 both have 5.1 surround options would using one of them result in a file around 125megs.

THANK YOU