Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > General > Audio encoding

Reply
 
Thread Tools Search this Thread Display Modes
Old 16th January 2006, 12:42   #61  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 6,890
@desta
Maybe...
When you decode in Beligth, for reencode purpose, don't use Dynamic compression (Azid settings). Each time you use Dynamic compression the peaks are attenuated. With Dynamic compression unchecked you have the full original Dynamic range.
tebasuna51 is offline   Reply With Quote
Old 16th January 2006, 13:19   #62  |  Link
desta
Registered User
 
Join Date: Jan 2006
Posts: 220
Hi tebasuna51...

Dynamic compression was turned off at all times in BeLight. Cheers for the suggestion though, and replying.




edit: actually, while I think of it... two things:

- should I have ripped the wavs from the original ac3 as 16bit, rather than 32bit?

- when re-encoding in softencode, with no actual changes to the wavs themselves, should all the pre-processing filters and compression be turned off or left checked?


Thanks again.

Last edited by desta; 17th January 2006 at 23:11.
desta is offline   Reply With Quote
Old 22nd February 2006, 23:11   #63  |  Link
Hans Ohlo
Registered User
 
Hans Ohlo's Avatar
 
Join Date: Apr 2003
Posts: 123
Quote:
Originally Posted by desta
Hi tebasuna51...

Dynamic compression was turned off at all times in BeLight. Cheers for the suggestion though, and replying.




edit: actually, while I think of it... two things:

- should I have ripped the wavs from the original ac3 as 16bit, rather than 32bit?

- when re-encoding in softencode, with no actual changes to the wavs themselves, should all the pre-processing filters and compression be turned off or left checked?


Thanks again.
hi,
i have the same problem. i extract 6 32bit wav files (also tried 16bit) with belight (nothing turned on) and use acid to encode them to ac3.

maybe i did not handle acid right but the result sound dimmer. i have serveral questions:
1. how do i map the wave files to the distinct channels (i use the sourround panner from acid, switched all channels off except the one of the file, i also upped the volume to 0db of each file/track. in the sourround panner there stays a -6db at the used channel...)
2. i can't switch off the center and sourround mix level in the bitstream options

how does encoding 6 discrete files exactly and with no lowering or attenuating to the channels and how do i get the same dynamic and volume as the original source (i only had to cut some minor stuff)?
Hans Ohlo is offline   Reply With Quote
Old 26th February 2006, 04:54   #64  |  Link
leonid_makarovsky
Registered User
 
Join Date: Oct 2004
Posts: 129
Dialog Normalization (WAV to AC3)

My expirience with Dalog Mormalization is following. I recorded a VHS to my hard drive and wanted to make a DVD out of it. After doing research I concluded to have Dialog Normalization set to -31dB. I unchecked all other options in all tabs. My DVD player was Philips DVP-642 which was connected to my Onkyo stereo receiver using RCA (analog connection). During playback I noticed some sort of clipping on right channel during loud parts. When played back in my computer using WinDVD and PowerDVD, the clipping didn't take place. (my computer's soundcard M-audio was also conected to my stereo receiver using analog connectors). I thought for a while and re-encoded with -27dB. Clipping diappeared, but the overall volume was lower than if I had this DVD with Uncompressed PCM from the same WAV file.

Recently I bought the 7.1 receiver. I connected my DVD player to the receiver digitally. When I played the DVD that I enceded with -31dB, clipping didn't take place. And the overall volume was louder than the reference DVD with Uncompressed PCM soundtrack. Playing -27dB DVD one against the reference (PCM) DVD showed the identical volume of soundtracks.

Basically I use AC3 2.0 to have the soundtrack that is maximum close to the original WAV file. I usually use Uncompressed PCM soundtracks, but when the video is longer than 75 minutes in order to fit it on DVD I prefer to use AC3 than to sacrifice video bitrate to make it lower than 8mbs.

--Leonid
leonid_makarovsky is offline   Reply With Quote
Old 3rd May 2006, 16:03   #65  |  Link
craftech
Registered User
 
Join Date: Apr 2006
Posts: 4
It is hard to believe this post hasn't drawn much criticism yet. There is absolutely no way that the method of determining dialnorm at the start of this thread will produce anything but a muffled, muted, and terrible sounding audio movie track.

I am sure that any of the subsequent posters who asked the original poster about this method and never reposted their results came to the same conclusion.

The reason is simple. The standard for dialnorm was not based upon the overall soundtrack, it is based upon the dialog. The simplistic method of calculating it described at the start of the post cannot work because the original posted doesn't understand the basis for dialnorm in the first place.

Considering the level of equipment necessary to determine it properly you are better off setting the dialnorm to -31 dB to start with and working your way down by trial and error leaving the other two parameters set to "None".

Last edited by craftech; 3rd May 2006 at 16:15.
craftech is offline   Reply With Quote
Old 3rd May 2006, 16:26   #66  |  Link
leonid_makarovsky
Registered User
 
Join Date: Oct 2004
Posts: 129
Quote:
Originally Posted by craftech
Considering the level of equipment necessary to determine it properly you are better off setting the dialnorm to -31 dB to start with and working your way down by trial and error leaving the other two parameters set to "None".
And what would the error be? Would there be clipping?

--Leonid
leonid_makarovsky is offline   Reply With Quote
Old 3rd May 2006, 16:34   #67  |  Link
craftech
Registered User
 
Join Date: Apr 2006
Posts: 4
And what would the error be? Would there be clipping?

========

Distortion, but moreover the "levels" would be audibly too high. Ear-piercing dialog or singing voices if they are present.

Craftech
craftech is offline   Reply With Quote
Old 3rd May 2006, 17:44   #68  |  Link
leonid_makarovsky
Registered User
 
Join Date: Oct 2004
Posts: 129
Quote:
Originally Posted by craftech
And what would the error be? Would there be clipping?

========

Distortion, but moreover the "levels" would be audibly too high. Ear-piercing dialog or singing voices if they are present.

Craftech
Ok, as I previously said, when I use analog RCA output for audio going from my DVD player to my receiver, I do notice clipping (distortion) if I set it to -31db level. When I use SPDIF out, I don't notice any clipping or distortion at -31db. Now my in my WAV file (which I use as a source) the peak level is no more than -0.1db to 0db. And the average loudness is most likely ranging from -3db to like -7db. So you're saying there's a chance to get the distortion? Thanks.

--Leonid
leonid_makarovsky is offline   Reply With Quote
Old 9th May 2006, 16:30   #69  |  Link
drob
Registered User
 
Join Date: May 2004
Posts: 255
Very helpful guide, however one point is still unclear, is it preferable to use the "use equal loudness contour" when calculating the dialnorm (referring to sound forge)?
drob is offline   Reply With Quote
Old 13th June 2006, 00:03   #70  |  Link
3ngel
Registered User
 
Join Date: Mar 2005
Posts: 457
I was compressing a 2 chan 48khz to ac3 using Soft Encode.
The wav had a freq peak around 23khz, but the resulting (decoded) ac3 has a hard upper limit of 20khz (a kind of mp3 freq view).
Is that normal? I'm doing something wrong?
Thanks
3ngel is offline   Reply With Quote
Old 13th June 2006, 08:12   #71  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 6,890
Is normal.

You can see in Soft Encode -> Encode Settings -> Audio bandwidth, the max value is 20.3 KHz for a 48 KHz wav with max Data Rate.

A wav 48 KHz is really 48000 samples/sec, then a 24 KHz tone have only 2 samples for period (a triangular wave instead a sine curve). The hard upper limit 20.3 KHz is a reasonable value for this input signal.
tebasuna51 is offline   Reply With Quote
Old 13th June 2006, 09:41   #72  |  Link
3ngel
Registered User
 
Join Date: Mar 2005
Posts: 457
I see, thank you very much.
Dts has this same limitation?
3ngel is offline   Reply With Quote
Old 13th June 2006, 10:07   #73  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 6,890
To preserve 23 KHz info you need use at least 96000 samples/sec.
And in DTS specs:

"DTS X96k Stream: DTS extended audio stream that enables encoding of original LPCM audio at up to 24 bits per
sample with the sampling frequency of up to 96 kHz"

But I can't help you about the soft needed. Maybe another user..
tebasuna51 is offline   Reply With Quote
Old 13th June 2006, 17:42   #74  |  Link
3ngel
Registered User
 
Join Date: Mar 2005
Posts: 457
But, to obtain a 23khz would not to be enough to have 48khz (Nyquist theorem) instead of 96khz?
3ngel is offline   Reply With Quote
Old 15th June 2006, 00:37   #75  |  Link
raquete
Registered User
 
Join Date: Nov 2004
Location: Brazil
Posts: 745
first i want to thank SomeJoe for this magnific guide.
congratulations!

edit: obsolent.....perfect answer here: http://forum.doom9.org/showpost.php?...1&postcount=90

as english is not my "mother language", i have some doubts and if someone could clarify me i will be thankfull:
1-
Quote:
The default Dialog Normalization setting is -27 dB, and the default Dynamic Range Compression is set to "Film Standard".
where the -27 value for Dialog Normalization came from? i can't find this in "anywhere" inside Dolby.Inc or in any .pdf file(i download lots).

2-
Quote:
The decoder will perform an attenuation of (31 + dialnorm) dB to the program material when played back.
what i found is that the dialnorm is "centralized" in -31 db(as the the pictures posted) and not (31 + dialnorm).please correct me if i'm wrong and explain me.

2-
Quote:
I loaded it into Sound Forge and measured the RMS level of the entire file as -20 dBFS.
a long time i don't use sound forge that have lots of options in the normalization dialogue(i don't remember all),i'm using audition and for each db choosed to normalize give different result.
-16dB was used as "reference" in the dialogue normalization tab like show the picture in the link.why not use -31,-27 or other value because each value choosed give different volume in the result.

thanks for answers.

Last edited by raquete; 18th June 2006 at 08:55.
raquete is offline   Reply With Quote
Old 15th June 2006, 10:04   #76  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 6,890
Quote:
Originally Posted by raquete
i can't find this in "anywhere" inside Dolby.Inc or in any .pdf file(i download lots).
Try with this another .pdf:
http://www.dolby.com/assets/pdf/tech...data.Guide.pdf
tebasuna51 is offline   Reply With Quote
Old 15th June 2006, 11:30   #77  |  Link
raquete
Registered User
 
Join Date: Nov 2004
Location: Brazil
Posts: 745
Quote:
Originally Posted by tebasuna51
perfect tebasuna51.
this .pdf answer my 2 first questions,thank you so much

about the last question:
measured the RMS level of the entire file as -20 dBFS in audition "Group waveform normalize" :

source with 0dB( 100% volume)
http://img228.imageshack.us/my.php?i...urce0db6bs.png

same source with -3dB ( 70.79% volume)
http://img149.imageshack.us/my.php?i...urce3db9fi.png

from audition help:
Group waveform normalize
Eq-Loud
Is the final loudness value with an equal-loudness equalization curve that takes into account frequencies to which the human ear is most sensitive. If you select the Use Equal Loudness Contour option in the Normalize tab, this value determines how much to amplify the audio to normalize it.
Loud
Is the final loudness value without equal-loudness equalization. If you don't select the Use Equal Loudness Contour option in the Normalize tab, this value determines how much to amplify the audio to normalize it.
Max
Is the maximum RMS (Root-Mean-Square) amplitude present. This value is based on a full-scale sine wave being 0 dB, and it conforms to the width specified in the Advanced section of the Normalize tab.
Avg
Is the average RMS of the entire waveform. This value isn't used for normalization.
% Clip
Is the percentage of the waveform that would be clipped as a result of normalization. Clipping won't occur if limiting (in which loud passages are decreased in volume) is used; instead, the louder portions of audio are limited to prevent clipping. In general, avoid values higher than 5% to prevent audible artifacts from occurring in the louder portions of audio.

Quote:
Quote (from guide):
I loaded it into Sound Forge and measured the RMS level of the entire file as -20 dBFS. .
...as each source have different volume, we have different volume in result.
how much dbs we have to use before to load the source and find the RMS level ?

thanks!


edit: changing the too big screenshots (that are ugly) for the imageshack urls.

Last edited by raquete; 16th June 2006 at 13:16.
raquete is offline   Reply With Quote
Old 16th June 2006, 09:20   #78  |  Link
3dsnar
Registered User
 
3dsnar's Avatar
 
Join Date: Dec 2005
Location: Proxima Centauri
Posts: 315
Quote:
Originally Posted by 3ngel
But, to obtain a 23khz would not to be enough to have 48khz (Nyquist theorem) instead of 96khz?
Exactly, 48 kHz is enough. It is not important how many samples are representing each period, such sine can be (during resampling for example) restored nearly perfectly (assuming it is below the Nyquist frequency).
More here.
http://en.wikipedia.org/wiki/Nyquist_theorem
__________________
Aud-X MP3 5.1 Format
3dsnar is offline   Reply With Quote
Old 16th June 2006, 10:16   #79  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 6,890
@3dsnar
@3ngel
I agree with you for uncompressed wav, or using flac and similar.
But I doubt the info at 23 KHz is well preserved with encoders (ac3, mp3, aac, ...) at normal bitrates. I don't know dts.
tebasuna51 is offline   Reply With Quote
Old 16th June 2006, 10:36   #80  |  Link
3dsnar
Registered User
 
3dsnar's Avatar
 
Join Date: Dec 2005
Location: Proxima Centauri
Posts: 315
Yeah, exactly. But this is more related to the compression technology, than to the sampling theorem (i.e. people normally do not perceive sounds above 20 kHz, so there is not sens to encode such high frequencies - so it is probably cuted out to save some bitrate).
Cheers, 3d
__________________
Aud-X MP3 5.1 Format
3dsnar is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 08:36.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.