Is Audacity the best program for my application?

[Special K]

I've been researching freeware audio editing tools to try and find the best one for my application. I have a digital voice recorder that will only generate WMA files. I would like to cut sections out of these recordings and either discard them or save them as separate files. These files will either be left as WMA or possibly converted to MP3. The way I see it, there are a couple of ways to approach this:

1. Convert the WMA files to MP3, and then use an MP3 editor on the resulting files
2. Edit the WMA files directly, and then save them as a WMA or MP3 file

I would prefer to edit the files while minimizing the amount of quality lost due to recompression. It seems the only way to ensure zero quality loss due to recompression would be to use a program that can edit WMA files directly without having to perform a recompression to save the changes. Otherwise I would be looking at 1-2 losses due to compression by using option 1 above - once to convert the WMA file to MP3, and another loss I have to recompress the MP3 file to save the changes. Option 2 would only result in a loss due to recompression if the program was incapable of saving changes to a WMA file without recompressing it, or if I wanted to save the WMA file as an MP3 file.

I've considered the fact that the audio I am working with consists only of spoken voices and background noises. I would think that the quality of these recordings wouldn't suffer much due to recompression as a result. Is that correct? I'm also aware of the fact that Audacity is incapable of saving changes to an MP3 or WMA file without performing a recompression. Having said that, I have another couple of questions:

1. What is the bitrate equivalence between MP3 and WMA?
2. If I edit a WMA/MP3 file in Audacity and save it as the same type of file with the same bitrate, will the loss of quality due to recompression be negligible?

Given this information, will Audacity work fine for my application, or are there better freeware programs out there?

Thanks

[kozikowski]

<<<1. What is the bitrate equivalence between MP3 and WMA?>>>

There isn't one. They're based on very different technologies. Windows Media is much the newer and based on MPEG4/h.264 compression algorithms and technologies. MP3's real name isn't "MP3." It's MPEG1, Layer 3 and it's over 20 years old.

Yes, you do need something that can directly edit in WMA in order to eliminate as many compression steps as possible. That might be interesting because h.264 uses temporal compression, not just clever re-arranging of the harmonics. In English that means where you think you're editing and where the edit actually goes may be different.

Koz

[Special K]

<<<1. What is the bitrate equivalence between MP3 and WMA?>>>

There isn't one. They're based on very different technologies. Windows Media is much the newer and based on MPEG4/h.264 compression algorithms and technologies. MP3's real name isn't "MP3." It's MPEG1, Layer 3 and it's over 20 years old.

Yes, you do need something that can directly edit in WMA in order to eliminate as many compression steps as possible. That might be interesting because h.264 uses temporal compression, not just clever re-arranging of the harmonics. In English that means where you think you're editing and where the edit actually goes may be different.

Koz

Thanks for the information. So there's no way to make a general statement like "an MP3 at X kbps will sound identical to a WMA file of Y kbps"? I realize that anything other than a direct WMA editor will result in at least 1 compression step, but is it possible to minimize the damage enough so as to make it imperceptible to a casual listener? I might want to convert these files to MP3 anyway, so one compression step would be unavoidable in that case. What about using Audacity to edit the WMA file directly and export it as a WMA file? Could I go that route and minimize the damage due to compression enough such that it would be imperceptible to a casual listener?

[whomper]

<<<1. What is the bitrate equivalence between MP3 and WMA?>>>

There isn't one. They're based on very different technologies. Windows Media is much the newer and based on MPEG4/h.264 compression algorithms and technologies. MP3's real name isn't "MP3." It's MPEG1, Layer 3 and it's over 20 years old.

Yes, you do need something that can directly edit in WMA in order to eliminate as many compression steps as possible. That might be interesting because h.264 uses temporal compression, not just clever re-arranging of the harmonics. In English that means where you think you're editing and where the edit actually goes may be different.

Koz

Thanks for the information. So there's no way to make a general statement like "an MP3 at X kbps will sound identical to a WMA file of Y kbps"? I realize that anything other than a direct WMA editor will result in at least 1 compression step, but is it possible to minimize the damage enough so as to make it imperceptible to a casual listener? I might want to convert these files to MP3 anyway, so one compression step would be unavoidable in that case. What about using Audacity to edit the WMA file directly and export it as a WMA file? Could I go that route and minimize the damage due to compression enough such that it would be imperceptible to a casual listener?

personally i would consider a new recorder that gives wav files
$99 and up

for voice the $99 model would do the job

is the hassle of the conversion and editing and quality loss risk
worth $99 ?

[kozikowski]

<<<So there's no way to make a general statement like "an MP3 at X kbps will sound identical to a WMA file of Y kbps"? >>>

I see where you're going with this. You can't go between two different compression systems. The quality of the work always go down. Compression damage, no matter where it came from is cumulative and permanent.

If you can figure out a way to directly edit WMA, then it's possible the quality of the original will not change. If you open the work in Audacity and export an uncompressed wav file, the quality won't change -- or very much. There's very minor housekeeping errors that Audacity almost always has.

That's pretty much the whole list. If you pull the work into Audacity, edit it, and export as MP3, then you will gain the damage that encoding MP3 generates, plus the interaction between the WMA damage and the MP3 damage. Each compression generates damage and when you convert, you're forced to convert the damage, too, not just the show.

WMA is pretty good at compressing into a small filesize. It's always a shock when somebody decompresses one of these things and sees how much hard drive they really take up. We had a posting of someone who was trying to edit a multi-hour show and kept running out of hard drive space. Yes, that soda can worth of compressed data turns into three-bedroom, two-bath ranch style when you decompress it, and Audacity always decompresses the work before managing it.

You can minimize the overall damage by Exporting a really, really good MP3. Pick a bitrate north of 300 and nobody can tell what you did. Of course, the file sizes go way up when you do that, but you should only hear the original WMA damage.

Koz

[Special K]

So basically my best bet is to import the WMA file, make my changes, and then export it as a WAV file? This will result in only minor errors due to the way Audacity handles files?

[Special K]

Is there a voice recorder that supports WAV files? Won't the file size be huge? Some of my recordings could be up to a few hours long.

[kozikowski]

<<<So basically my best bet is to import the WMA file, make my changes, and then export it as a WAV file? This will result in only minor errors due to the way Audacity handles files?>>>


Yes. We're forced to say that because Audacity converts its jobs to 32-bit floating for internal use. That's super high quality and lets you apply effects, tools, and edits without damage. The down side to that is when you export to a normal WAV file like 44100, 16-bit, Stereo (Audio CD standard), Audacity has to downconvert the work from totally excellent to only really, really, really good.

Every so often, somebody posts indignantly that they found a teensy data error somewhere where "it shouldn't be." Yes. That's correct. Audacity does not do perfect bit-for bit processing.


Stereo 16-bit WAV files are going to come in at about 700 MB per hour. Fuzzy numbers, but that should be close. When you open them up in Audacity, triple that. As if that weren't painful enough, your computer hard drive must never, ever go over 90% full and the drive space must be perfect, error-free, and thoroughly defragmented...

Defrag
-- Right Click Start > Explore > Right Click Local Drive C: > Properties (Used, Remaining) > Tools > Error Checking & Defragmentation

There are some other restrictions with really long shows like WAV files can't go over 4GB. You start approaching magic with really long shows. We urge strongly that if possible, you chop up your long shows into small fragments and edit those.

Koz

[Special K]

There are some other restrictions with really long shows like WAV files can't go over 4GB. You start approaching magic with really long shows. We urge strongly that if possible, you chop up your long shows into small fragments and edit those.

Koz

Of course, in order to chop the show up into small fragments you must first have enough space to open and decompress the entire file at once, correct?

[kozikowski]

<<<Of course, in order to chop the show up into small fragments you must first have enough space to open and decompress the entire file at once, correct?>>>

No, that's a little more Audacity Magic. Unless you changed it in Preferences, Audacity does not pull sound inside itself for editing. It points to an existing sound file at the right time. So if the sound file already exists anywhere on your machine, Audacity can be used to break it up without all the severe drive space limits.

You're in trouble the first time you need to do special effects or other serious sound management. Boom. Drive space requirements go up instantly. This isn't black magic, but it's seriously dark gray and I may have missed one or two of the finer points, but that's the general idea.

The question comes when the original sound file is compressed. I think Audacity has to manage the exported segment and not the rest of it. Someone will correct me.

Koz