System wide dynamic range compression

It might be easier for you to just buy a hardware compressor and put it in the signal chain between your computer and your speakers. Plus, then you don't have the compressor tying up resources. If you're really set on using a software compressor, there is an application included with the Developer Tools that basically functions as a mixer for your system audio. You can load effects in it. So, you could just open it and set up your compressor (theres a fairly good multiband compressor AU included with OS X) and then just hide it and let it run in the background.

Out of curiosity, why do you want a system-wide compressor?

Wolfgang

kaz219 said:

The AU Lab seems compress sound from an external input source. Not able to apply this a the sound coming from a particular sofware (like itunes or vlc)...

Still, thank you for the answer!

Click to expand...

Ah yes, I forgot to include a very crucial piece of software. You will need soundflower (http://www.cycling74.com/products/soundflower)for this to work. Basically, soundflower allows you to route audio to and from one application to another. Once you have soundflower installed, set your default audio out to "Soundflower (2ch)". Then, set the input device in AU Lab to "Soundflower (2ch)" as well. This will cause every app in OS X to send its audio into AU Lab. Then, in AU Lab set the output to either internal audio or whatever audio interface you desire to use. You can then use effects in AU Lab on audio in OS X. As a side note, Soundflower is probably one of the most useful free little audio add ons. If you do any audio production with more than one piece of software, it is extremely useful.

kaz219 said:

It looks like Hear is doing the job. But it's 50 bucks, and a bit overkill (lots and lots of features).

I must admint I still don't really understand how the dynamic range compression feature (called "limiter") works. Have to make some tests, I have a 30 days trial.

Click to expand...

Generally, a limiter prevents the audio from passing a said threshold. I think the OP really would find a compressor much better as it not only lowers peaks in the audio, but raises quiet parts as well. Never thought about using it for movies before, but I wouldn't recommend using it while listening to most music. Alot music is already heavily compressed when it is mastered before being released.

Wolfgang

get a set of closed cup headphones or even noise cancelling ones. Putting a Multi-band compressor with some kind of limiter into your audio chain is going to result in significant hearing damage. Your going to end up boosting the audio levels immensly and because you've removed the transients you will have virtually no perception of the damage that's say being done at the high end of your hearing because the entire sound will be saturating your ears. Transients are a good thing

kaz219 said:

I don't understand that. Dynamic range compression only reduce the difference between the loudest part of the movie/music and the quietest ones, isn't it?

So I can then adjust the volume of the movie/music as I want to. It's just that quiet parts will be louder, and loud parts will be quieter.

Click to expand...

In theory, yes, in practice you simply end up massively increasing the overall sound pressure levels of the entire mix. It works on amps at home as it usually just compresses the bottom end or maybe applies a frequency cut at 100hz to remove boom and thump. In a headphone your going to boost the highs and compress the lows but your ears will tire very fast and you will push the volume to re-instate the bass while your perception of the treble will get muted in just a few minutes. Over weeks and months you will end up burning out you high frequency hearing in your ears. This is why music producers (myself inc) NEVER mix with headphones and mix at low levels on monitors. I'm not saying something can't be accomplished but the odds are in favor that you'll damage your ears.

surferfromuk said:

In theory, yes, in practice you simply end up massively increasing the overall sound pressure levels of the entire mix. It works on amps at home as it usually just compresses the bottom end or maybe applies a frequency cut at 100hz to remove boom and thump. In a headphone your going to boost the highs and compress the lows but your ears will tire very fast and you will push the volume to re-instate the bass while your perception of the treble will get muted in just a few minutes. Over weeks and months you will end up burning out you high frequency hearing in your ears. This is why music producers (myself inc) NEVER mix with headphones and mix at low levels on monitors. I'm not saying something can't be accomplished but the odds are in favor that you'll damage your ears.

Click to expand...

i was always wondering why those compressors are not standard in all consumer devices. your excellent post finally explains that. thanks.

surferfromuk said:

In theory, yes, in practice you simply end up massively increasing the overall sound pressure levels of the entire mix. It works on amps at home as it usually just compresses the bottom end or maybe applies a frequency cut at 100hz to remove boom and thump. In a headphone your going to boost the highs and compress the lows but your ears will tire very fast and you will push the volume to re-instate the bass while your perception of the treble will get muted in just a few minutes. Over weeks and months you will end up burning out you high frequency hearing in your ears. This is why music producers (myself inc) NEVER mix with headphones and mix at low levels on monitors. I'm not saying something can't be accomplished but the odds are in favor that you'll damage your ears.

Click to expand...

Does this only apply to multiband compression that's set up to cut boomy bass and boost highs? Because, at least for regular non-multiband compression, all frequencies that pass the threshold in either direction are boosted/cut by the predetermined ratio, yes? If not, then I have had an entirely incorrect understand of how compression works for the past 4 years

And as far as not mixing with headphones, I don't quite understand that. I know that headphones don't give you an accurate representation of the depth or power of the mix, but I had never heard that it causes ear damage any more than monitors (both at high volumes). Or is that specifically in regards to multi-band compression resulting in the highs becoming too loud?


As far as closed/noise canceling headphones are concerned, I certainly would recommend them for noisy environments so that you don't have to push the volume up to compete with the ambient noise. However, I don't think that this would really solve the OP's problem of some sections being too quiet, unless they are only too quiet due to the louder ambient volume, i.e. when listening to them at home or in a quieter environment, the sections that are too quiet on-the-go are fine. I also somewhat retract my previous recommendation of a hardware compressor. They're usually not very portable and I didn't realize that the OP needed the compression while out and about.

Wolfgang

kaz219 said:

I'm not at all an expert, but I think I want a downward AND upward compression. (see wiki page on DRC)

Let say the gain can go from value GainMin to value GainMax.

I choose a value A and a value B, A being smalle than B. Then I do an upward compression for all audio with a gain lower than A, and a downward compression for all audio with a gain higher than B.

I don't understand why I would want to differenciate between frequencies. I just apply the same compression to all frequencies.

I probably don't understand something, because it seems too simple. Somebody would have already done it if it was that simple.

Click to expand...

Assuming I understand the concept of compression correctly (I am either misinterpreting surferfromuk's post, or have had an entirely incorrect understanding of compression for the 4 years that I've been recording music!), this is what a standard (not multi-band) compressor does. You first set a threshold and a compression ratio. Anything above the threshold is lowered by the set ration and anything under the threshold is raised by the set ratio. Most compressors also have a setting for attack and release times to specify how quickly you want the compressor to raise/lower the volume and how long you want it to keep the gain increase/decrease at that level (how I understand it, feel free to correct me surferfromuk)

Wolfgang

kaz219 said:

@Wolfgang:

Hello,

sorry to bother you again, but I can't manage it with Soundflower and AU Labs.

1 – installed Soundflower 1.3.1
2 – AU Lab 2.0.0
3 – set my sound pref panel to use "Soundflower (2ch)" as the *output*.
4 – When I launch Au Lab I got a "Document configuration assistant" window asking to "Specify the output configuration of the document". I can add mono or stereo output channels. I take 1 stereo channel (actually taking 2 channels. This is where I'm probably wrong I think)
5 – I click then next, and it's the same for input. (I take 1 stereo channel again)
6 – Then I have two tabs in the window: ouput and input channels. Whenever I change the audio device for input channels, it also changes the device to output channels, whereas my understanding is that Soundflower (2ch) should be my audio device for input channels and "built-in audio" should be my audio device for ouput channels.
7 – I choose Soundflower as audio device for both.
8 – The AU Lab window opens, and I can see that sounds comes in (itunes running in background), but I don't hear anything coming out of my built in speakers...


What am I doing wrong?

Click to expand...

Sorry to put this thread back on top, but is there an answer to the question from kaz219.
I'm new in this forum. I have done the same setting and have the same problem. What am I doing wrong?

Thanks in advance.
J