Topic: How much can I boost the volume?

[paxhell]

Hi, did a record a few days ago and It come out with a very low volume level, I had to boost +15 dB but I dont know if I can go higher?? Look at the screenshot from soundforge.... I had -2.5dB peaks from the fireworks and I lowered them first then I boosted the volume.




Will a record with a low level affect the sound quality?

Here is a sample
http://www.box.net/shared/t73vg6imf3

Thanks for any help

[Gordon]

looks like you don't have too much more room to boost and it doesn't look like you need to.

Will a record with a low level affect the sound quality?

not really.  it will just be very quite and you will have to turn it way up to hear it.  edit: I'm wrong and should have known    I am a fan of bumping the gain/volume a bit when needed but you got to be careful.  you are also raising the volume of all the room/self noise etc.

[Brian Skalinder]

I had to boost +15 dB but I dont know if I can go higher?? Look at the screenshot from soundforge.... I had -2.5dB peaks from the fireworks and I lowered them first then I boosted the volume.

Judging from the waveform snapshot you posted, the only way you could increase gain would be to apply compression.  Compression reduces the dynamic range, i.e. the difference between the loudest portion and the softest portion.  By reducing the dynamic range - think of it as squishing the loud parts so they're not quite as loud - you would be able to add more gain and the whole recording would sound louder.  Some people like to use compression, some don't.  It's easy to over do, in which case it often sounds terrible.  Personally, I generally apply compression only if I have HUGE dynamic range on the recording - like some acoustic amplified sets, or unamplified performances.  Live field recordings will never sound as loud as studio releases b/c most studio releases include gobs of compression and limiting, resulting in very little dynamic range.  Recordists don't have as much control in a live field recording situation as in the studio, so it's more difficult to effectively achieve the same levels of compression and limiting as in a studio release (and frankly, I don't think it's desirable).

Lots of posts here on compression, and your Help file may have some useful information, too. 

Will a record with a low level affect the sound quality?

Yes.  But will you notice it?  Maybe, maybe not.

The lower your levels, the worse your signal to noise ratio.  Say the noise floor is at -90 dBFS (just an example).  If you record with your levels set perfectly so they peak at 0 dBFS without clipping, then you have 90 dBFS of range between your loudest signal and the noise floor.  No need to add gain after the fact.

Now let's say you set your levels so they peak at -18 dBFS.  The range between your peak levels and the noise floor (still -90 dBFS) is now 72 dBFS (90-18).  When you boost gain after the fact, in this example +18 dB to get your peak levels up to around 0 dBFS, you'll raise the levels of both the audio (from -18 dBFS to close to 0 dBFS), but you'll also raise the noise (from -90 dBFS to -72 dBFS)

For 16-bit, I think peaking in the -6 dBFS to 0 dBFS range puts you in great shape.  For 24-bit, I think peaking ~-12 to -6 dBFS is just fine due to the lower noise floor.  Over time, you may start to learn how your gear's levels respond in specific venues / locations / amounts of gain, and it's easier to be more aggressive with levels while still not risking a nasty clip.  You're far better off running your levels a little conservative and boosting in post than running aggressively and clipping.

[paxhell]

Thanks for the explanation, this is a great site 

[Liquid Drum]

Not a good idea to guess when boosting (in my opinion).

I normally just Normalize to -0.5db (anything over 0db will cause distortion). That will boost the recording to that level automatically without having to manually input the amount of db you want to boost (which could end up being too much or too less). It saves time and is more accurate imho.

[Gordon]

Not a good idea to guess when boosting (in my opinion).

I normally just Normalize to -0.5db (anything over 0db will cause distortion). That will boost the recording to that level automatically without having to manually input the amount of db you want to boost (which could end up being too much or too less). It saves time and is more accurate imho.

in wavelab you can highlight the whole file and hit "get peak level".  so then you can know just how much you can raise the gain.

[Liquid Drum]

Not a good idea to guess when boosting (in my opinion).

I normally just Normalize to -0.5db (anything over 0db will cause distortion). That will boost the recording to that level automatically without having to manually input the amount of db you want to boost (which could end up being too much or too less). It saves time and is more accurate imho.

in wavelab you can highlight the whole file and hit "get peak level".  so then you can know just how much you can raise the gain.

Ah, didn't know that. Thanks for the info (I normally use Adobe Audition). 

[RobertNC]

in wavelab you can highlight the whole file and hit "get peak level".  so then you can know just how much you can raise the gain.

Ah, didn't know that. Thanks for the info (I normally use Adobe Audition). 

You can do the same in Audition (I think, assuming it has not changed too much from CEP)

In CEP 2 you select Amplify and in the dialog there is an option to "Calculate Normalization Values".  Enter the final dB you want to boost the gain to and it will calculate how many dB gain you need to add.

This is useful especially when you are working with very large files such as 24/96 files.Although people have indicated on this board that CEP (Audition) will handle large files, I personally have had problem s with files over 3 GB  - maybe because I am not using NTFS file system? or maybe because I did not have enough swap space or whatever...

I use this feature to estimate how much gain I want to add  to keep it consistent across multiple wave files without having to join them into a single huge file.  Makes the workflow a lot easier.