Hi guys,
I'm having trouble again with goldwave.
This time I'm trying to increase the loudness of certain sound files but after increasing the volume the sound gets messy and distorted.
Is there a way to increase it's volume (in the way I turn up the volume on my speakers) without messing up the sound?
Sounds I'm trying to amplify:
http://www.megaupload.com/?d=46MW5HEN
I hope to hear from you soon!
Terribly distorted sound
-
- Posts: 1171
- Joined: Fri Mar 11, 2005 11:15 pm
- Location: Washington DC Metro Area
- Contact:
For starters, your source files are 8-bit. 8-bit does not have a whole lot of resolution and when you attempt to amplify 8-bit audio, the results are pretty horrible.
the second reason is while this file has a low RMS volume, it's got a LOT of spiking up to 0 dB, and when you amplify anything over 0 dB, you get clipping and distortion.
the other thing to remember is these files are just plain garbage quality wise, garbage in, garbage out.
your only option is to make them "sound" louder by applying lots of dynamic compression; reducing peaks, quieting loud parts and boosting quiet parts. however, being 8-bit source, your quality is likely to still suffer.
boosting to 16-bit won't undo these problems, in fact, goldwave does it's processing in 24-bit floating point.
the second reason is while this file has a low RMS volume, it's got a LOT of spiking up to 0 dB, and when you amplify anything over 0 dB, you get clipping and distortion.
the other thing to remember is these files are just plain garbage quality wise, garbage in, garbage out.
your only option is to make them "sound" louder by applying lots of dynamic compression; reducing peaks, quieting loud parts and boosting quiet parts. however, being 8-bit source, your quality is likely to still suffer.
boosting to 16-bit won't undo these problems, in fact, goldwave does it's processing in 24-bit floating point.
Some background-
There is an upper limit to the volume. The digital limit is set by the number of bits. With 8 bits, you can "count" to 256. With 16 bits, you can "count" to 65,535. These maximum levels are calibrated (referenced) to 0dB. (So, a maxed-out 16-bit file doesn't sound louder than a maxed-out 8-bit file.)
If you try to go above the 0dB maximum you will get clipping (distorted, flat-topped waves). You can get analog clipping if you try to push 110 Watts out of a 100 Watt amplifier.
GoldWave (and most other audio editors) use floating-point numbers for temporary data storage. This allows you to go over 0dB during processing. However, if you try to save the file as a standard 8 or 16-bit file, it will be clipped.
With GoldWave, this is a 2 or 3 step process.
1. Effect -> Volume -> Maximize. (Your file is probably already maximized. But just to make sure, this will adjust the volume so your peaks are at the digital maximum of 0dB.)
2. Effect -> Compressor/Expander. Select the Reduce Loud Parts preset. This will reduce and "round-off" the peaks, with minimal distortion.
3. Effect -> Volume -> Maximize. This will boost the overall level, bringing the peaks back-up to 0dB, and increasing the average volume level. (It will also increase any low-level noise.)
You can repeat for more compression, or try different compression settings.
When you digitize, you also get something called quantization noise which is much worse at low resolutions. I think compression will make this worse.
There is an upper limit to the volume. The digital limit is set by the number of bits. With 8 bits, you can "count" to 256. With 16 bits, you can "count" to 65,535. These maximum levels are calibrated (referenced) to 0dB. (So, a maxed-out 16-bit file doesn't sound louder than a maxed-out 8-bit file.)
If you try to go above the 0dB maximum you will get clipping (distorted, flat-topped waves). You can get analog clipping if you try to push 110 Watts out of a 100 Watt amplifier.
GoldWave (and most other audio editors) use floating-point numbers for temporary data storage. This allows you to go over 0dB during processing. However, if you try to save the file as a standard 8 or 16-bit file, it will be clipped.
As DewDude said, dynamic compression.But is there no way at all to make them sound louder...?
With GoldWave, this is a 2 or 3 step process.
1. Effect -> Volume -> Maximize. (Your file is probably already maximized. But just to make sure, this will adjust the volume so your peaks are at the digital maximum of 0dB.)
2. Effect -> Compressor/Expander. Select the Reduce Loud Parts preset. This will reduce and "round-off" the peaks, with minimal distortion.
3. Effect -> Volume -> Maximize. This will boost the overall level, bringing the peaks back-up to 0dB, and increasing the average volume level. (It will also increase any low-level noise.)
You can repeat for more compression, or try different compression settings.
Probably not. When a file is saved as 8-bits, you loose information. (i.e. A resolution of 256 data-points vs. 65,535.) You can't recover the missing information by saving at a higher resolution....and decent?
When you digitize, you also get something called quantization noise which is much worse at low resolutions. I think compression will make this worse.
-
- Posts: 423
- Joined: Wed Jun 16, 2004 8:33 pm
What I'm going to say hopefully won't be construed as an insult to our OP.
There is an old adage, "you can't make a silk purse from a sow's ear".
It particularly applies to music or sound.
With a digital photograph, I can easily manipulate individual pixels (I've done it countless times) to improve an image, but in music that's another story.
Glenn
There is an old adage, "you can't make a silk purse from a sow's ear".
It particularly applies to music or sound.
With a digital photograph, I can easily manipulate individual pixels (I've done it countless times) to improve an image, but in music that's another story.
Glenn
-
- Posts: 1171
- Joined: Fri Mar 11, 2005 11:15 pm
- Location: Washington DC Metro Area
- Contact:
I've said garbage in, garbage out.
Photos are drastically different than audio and there's lots you can do to cover up imperfections. I suppose because it's the whole visual aspect of images..we can fool our eyes moreso than out ears.
There's also lots of stuff you can do to photos you can't for audio. For example, if you have a slightly blurry photo you can run blind convolution to unblur it. You can, in theory, apply blind convolution to audio to remove hallow/echoy sound...but it's entirely complex and has only been done once that I know of. There's no software that allows for this, it was implemented into a audio workstation created by SoundStream and apparently never done or implemted again, it's also supposedly terribly difficult to even get the results right.
Photos are drastically different than audio and there's lots you can do to cover up imperfections. I suppose because it's the whole visual aspect of images..we can fool our eyes moreso than out ears.
There's also lots of stuff you can do to photos you can't for audio. For example, if you have a slightly blurry photo you can run blind convolution to unblur it. You can, in theory, apply blind convolution to audio to remove hallow/echoy sound...but it's entirely complex and has only been done once that I know of. There's no software that allows for this, it was implemented into a audio workstation created by SoundStream and apparently never done or implemted again, it's also supposedly terribly difficult to even get the results right.