Matching Volume: 32, 24 or 16 bit processing?

vbimport

#1

Hi guys… I do apologize for the (silly and boring) question I am going to ask you.
There are a couple of things I do not understand… maybe it is becasue the language barrier or because often some things are so obvious but confuse me a lot…… so, once again, please excuse my ignorance :frowning:

What I am trying to get is a good “protocol” to burn good quality AUDIO CD.

This is my own method: making an Audio CD compilation I usually extract the audio tracks form the original CD by using EAC. Then if I have to add some MP3s I first convert them into Wave PCM signed 16 bit 44100 Hz.

At this point I have several songs from this different sources, all with the same attribute (all Wave files/16 bit) BUT with different volume levels I want to write to a CD; so I use Goldwave ----> Match volume effect (Root Means Square) to adjust volume levels of each song so they sound similar: I open each file and display the Match Volume effect to see what average value it has, then I apply an overall average value to all the files. To avoid clipping distortion, I use the minimum average across all files. For example, if one file has an average of -18dB and all the other files have a higher average, then use -18dB for all files.

My doubts are:

Am I doing the right thing matching the volume file working on Wave PCM signed 16 bit 44100 Hz files?

If it is wrong, should I convert them FIRST into Wave PCM signed 32 bit 44100 Hz (or IEEE FLOAT 32 BIT or any other attribute you will advice me), matching the volume and then, BEFORE writing, re-converted them into on Wave PCM signed 16 bit 44100 Hz fies?

If this right, it seems that I am going to loose something and that I should apply dithering.

Of course, some people say that dithering is useless and some people say that is absolutely necessary.

So, at this point I am lost!!! … and need a huge help!

Thank you in advance to anyone who can resolve my confusion and HAPPY NEW YEAR! :slight_smile:


#2

This converting thing is absolutely contra productive…

Stay with 16bit RIFF wave stereo, either intel or motorola signed.


#3

[QUOTE=chef;2567337]This converting thing is absolutely contra productive…

Stay with 16bit RIFF wave stereo, either intel or motorola signed.[/QUOTE]

Well, first of all, thank you for answering :wink:

Why are you saying it is contra productive?

I mean… it’s fine, less work to do and I’ll keep going working with my Wave PCM signed 16 bit 44100 Hz, that it should be the attribute EAC extracts wav… but I was just wondering why it is contra productive.


#4

FlyingLuka, I think our friend chef is stating that if your source is 16bit/wav, and your target is standard Audio CD which you know is 16/44.1khz/PCM wav, why on earth waste time converting from 16>24>16, just to apply volume adjustments…
I have to agree it’s “contra productive”…
There are some good burning apps/tools that will burn a mixed CD compilation from different format sources and apply volume changes to the temporary wav files that are created on-the-fly…Burrrn, and CDBurnerXP for example…This way the data from the originals is kept un-motified/untouched…This process is called ReplayGain…If you rip your original CDs to wav or lossless fine, then just keep them as they are, and store them for future use…Good luck!
http://wiki.hydrogenaudio.org/index.php?title=Replaygain


#5

Thanks t0nee1.

I am going to make some test with CDBurnerXP in ReplayGain mode.

At the moment I can tell you for sure the way you adviced me seems to be fine… I mean CDBurnerXP in ReplayGain mode does in few seconds what I manually do in several minutes by GoldWave. Cool!!! :clap:

I want now to compare the wav file made by EAC to to the one made by CDBurnerXP in ReplayGain mode and see if I have different or missing sample.

If I understood what you wrote: “…just keep them as they are” I guess any tracks burned by CDBurnerXP in ReplayGain mode should be exactly the same to the one from the original CD ripped by EAC.

Is it correct? :confused:


#6

No, they won’t be the same as the original rips, since you’ve now altered the volume to the decoded tracks…The burning app will use the [I]temporary[/I] altered/modified wavs and burn them…The Original tracks that you imported in the app will not be altered…This is what I mean by “just keep them as they are”…meaning the original ripped wav tracks…
Hope I’m not confusing you too much…
Most use Replay Gain for SW playback since it is stored in tags…The capable SW player(e.g foobar2k, will then play according to that info…When burning an Audio CD all tag info is lost…So using the burning apps mentioned with RG(wavgain) will only make the changes to the temp WAV files (hint, it’s done in the background) on-the-fly, if you will!..
http://www.hydrogenaudio.org/forums/lofiversion/index.php/t57744.html
http://www.hydrogenaudio.org/forums/index.php?showtopic=85614


#7

[QUOTE=t0nee1;2567616]No, they won’t be the same as the original rips, since you’ve now altered the volume to the decoded tracks…The burning app will use the [I]temporary[/I] altered/modified wavs and burn them…The Original tracks that you imported in the app will not be altered…This is what I mean by “just keep them as they are”…meaning the original ripped wav tracks…
Hope I’m not confusing you too much…
[/QUOTE]

It’s absolutely perfectly clear!!! I mean it. :iagree:

I run the “compare wavs” by EAC:

File A = original from the CD
File B= same of “A” but processed by GoldWave "matching volume"
File C=same of “A” but extracted form the CD burned by ReplayGain

A compared to B = 30 missing samples in A
A compared to C= some different samples in A and C

At this point I do not understand how to “read” EAC result:

Is the track B (by GoldWave) or C (by replaygain) the one less corrupted/more close to the original A ?

:sad:

One more question: the difference between Album gain and track gain: is album if I burn songs all form the same album and track if I burn songs taken from differrnt sources/Cds?


#8

Are you able to audibly distinguish B and C from A(original)…If not, you have your answer…
I don’t think I would use the word corrupted…They may be altered/modified, but I don’t agree that they are necessarily corrupted…
Explain corrupted…If you are editing in GW and saving a new edited file then yes, it is no longer identical to the original, hence your sample differences…But can you hear a difference other than just plain volume difference(s)?
FWIW, except for maybe SW playback, I rarely use Replaygain, normalize, mp3gain etc…I just use the good ole volume knob on my equip…


#9

[QUOTE=t0nee1;2567638]Are you able to audibly distinguish B and C from A(original)…If not, you have your answer…
I don’t think you should be throwing around words like corrupted…They may be altered, but I don’t agree that they are corrupted…
Explain corrupted…If you are editing in GW and saving a new file then yes, it is no longer identical to the original, hence your sample differences…[/QUOTE]

Of course I am not… and I do not trust my ears anyway! :smiley:

I wrote “corrupted” but I was wrong (that is the language barrier, I’m sorry). What I meant to write is basically which one is best because I do not understand the difference between “missing samples” (B) and “some different samples” ©.


#10

No worries…And about trusting your ears, that is precisely what you need to do, trust your own ears…None of us share them and none of us can give you a what is [I]best[/I] answer…


#11

I undesrtand! Thank you.

One more question, the last one: the difference between Album Gain and Track Gain:
“Album” when I burn songs all form the same album and “Track” when I burn songs taken from different sources/album/some kind of various-artists mix ?


#12

You’ve answered your own question(s)…Try album gain when listening to tracks from same Album, and track gain when listening to mixed tracks from several sources…If you choose Album gain it will make adjustments to the album and reach the target gain as a whole…Soft songs will remain soft and strong will remain strong…


#13

I know I answered my own question… but I did not know ReplayGain, so I was supposing (but not sure) about the difference between Track/Album gain. Thank you again t0nee1! :wink:


#14

You’re welcome…;):cool: