RMS/Limiter question...

[sambrn]

In my experience (rather short) I can say that different conditions bring to a good Rms average level before working with HB mastering process (it's great!). Eg: using limiters and compressors in each track during mixing or making well balanced mixing.

However I have experimented that mixes with more mid-high freq and less kick drum bass around 40-90Hz (mostly in rock style) allow to reach lower thresold in the limiter without squashing the sound and so higher Rms avg values.

On the contrary, when handling mix with kick drums very showy (hip hop style) it's impossible for me reaching low thresolds with the limiter: the sound squash on kick beats!

My question is:

Do higher freq sounds (like snare) can be better limited while kick drum sounds squash easly?
How can I manage a song like White America with a big kick and reach a RMS avg of - 10.5 db without squashing it?

I hope my english will be intelligible and if my question is not clear i will make it more detailed.

Thanks

[har-bal]

In my experience (rather short) I can say that different conditions bring to a good Rms average level before working with HB mastering process (it's great!). Eg: using limiters and compressors in each track during mixing or making well balanced mixing.

However I have experimented that mixes with more mid-high freq and less kick drum bass around 40-90Hz (mostly in rock style) allow to reach lower thresold in the limiter without squashing the sound and so higher Rms avg values.

On the contrary, when handling mix with kick drums very showy (hip hop style) it's impossible for me reaching low thresolds with the limiter: the sound squash on kick beats!

My question is:

Do higher freq sounds (like snare) can be better limited while kick drum sounds squash easly?
How can I manage a song like White America with a big kick and reach a RMS avg of - 10.5 db without squashing it?

I hope my english will be intelligible and if my question is not clear i will make it more detailed.

Thanks

Hello

I want you to research a few terms.

1. Peak to average ratio
2. Dynamic Range
3. Difference between Limiting and Compression.

If you familiarize yourself with these terms the mystery will be removed and your question will be answered.

Cheers

Earle

[sambrn]

Hi,

I've read some articles about "peak to average ratio" and "dynamic range" as you suggested me. I've heard many times the difference between limiter and compressor, but the readings on the first two terms make it sound more clear now.

Thanx

Sam

[har-bal]

Hi,

I've read some articles about "peak to average ratio" and "dynamic range" as you suggested me. I've heard many times the difference between limiter and compressor, but the readings on the first two terms make it sound more clear now.

Thanx

Sam

Sam

Perfect so now you understand that the loudness of a track normally comes from the reduction of the dynamic range. Lets talk about something in laymans terms for a minute.

If you are working on a track that has a constant high hat hitting you will notice when looking at the wave that there are quite a bit of spikes. Believe it or not these spikes can actually give you an average RMS reading that makes you believe your loudness is in line with commercial music (but it isn't).

Try this.......load a commercial song into your editor and notice the distance between the loudest and softest parts in your level meter. You will notice that the loudest songs (which is not usually good) remain at the top of the meter most of the time. This is what determines perceived loudness. So we can conclude thIs. The less dynamic range the louder the perceived volume. The more dynamIc range the less the overall loudness and the more the song can breathe.
The compressor is the tool of choice that narrows the dynamic range and increases overall loudness by reducing the loudest parts and increasing the softest parts thus creating a fuller sound.
The limiter keeps the loudest parts the same and brings up the rear. Of course there is a reduction in your overall quality.

Does this clear up a few things?

BTW: I am currently writing a book that will clear up all the mysteries of audio. It will be 14 chapters.

Cheers

Earle

[tcatzere]

Earle,

Do you usually feel it's necessary to use limiting on most all recordings -- even those employing primarily "acoustic" instruments? And, when it comes to compression, what are the guidelines that determine whether it should be single or multi-band?

Lastly, do you usually find yourself utilizing both of these processes on a given project or do you tend to pick and choose on an "as needed" basis?

Tom

[har-bal]

Earle,

Do you usually feel it's necessary to use limiting on most all recordings -- even those employing primarily "acoustic" instruments? And, when it comes to compression, what are the guidelines that determine whether it should be single or multi-band?

Lastly, do you usually find yourself utilizing both of these processes on a given project or do you tend to pick and choose on an "as needed" basis?

Tom

Tom

It depends on the music genre. I will place a Limiter at the end of the mastering chain sometimes just so the song won't clip at all. I just set the outceiling to -0.1 and the threshold to 0.0
This insures no clipping at all, especially when mastering acoustic instruments. This would be a good technique for you since you usually master the music from orchestras which does not have to be loud and requires real dynamic range.

Your question regarding the decision on when to use a wideband compressor or a multiband compressor is an easy one.

Remember this: If the music sounds as though it could do with a little more remixing...use a mutlband compressor. I usually recommend using a multiband because I run into a lot of projects where the mixing is questionable.

If the music sounds pretty balanced across the spectrum, use a wideband compressor.

Cheers

Earle

[tcatzere]

Thanks, Earle, for your input -- much appreciated.

Tom

[doogle]

Hi all, Earle, when will your book be available. Im sure all of us are hanging out to have a read.

Doogle

[sambrn]

Just another question while waiting for your new book...

Let's assume that you're mastering an LP of 10 tracks. Nine songs have a similar spectrum and using the har-bal mastering process you obtain about the same loudness (eg: Rms avg -12 db). The 10th track is 5 minutes long and has got for the first 90 seconds only piano, lead vocal, hit hat and for the rest a full sound with drum, bass, elec. guitars, strings, ect... Even if you've worked well at the mixing stage trying to make the difference in loudness be minimal between the two parts, probably there will always be a gap between them. Now, after mastering you obtain a Rms avg lesser than the ones of the other 9 songs. You suppose that the first 90 seconds which are less loud than the rest of the track take down the average level.

What to do?

1) Must you refer only to the part of the song which sounds louder and if its Rms avg is about the same of the other Lp's tracks it's OK, sounds good?

2) Do you have to make sound the entire song (1st 90 seconds embodied) at the desired average loudness? (In this case probably the second part with fuller sound will be louder than the other songs of lp which have got a similar arrangement)

3) Do you have to come back in the mixing and fix some stuff?

Bye

Sam

[HarBal]

What I'd do is to match the loudness by ear by switching between the soft track and one of your other loud tracks as a reference in HarBal. Alternatively, if you want to use HarBal match loudness to come up with the figure, create a cut down version of the soft track with the piano intro removed. Load this cut down version and your saved EQ filter for the full track into Harbal and use match loudness to come up with the more appropriated match loudness level.

Regards,


Paavo.

[har-bal]

Just another question while waiting for your new book...

Let's assume that you're mastering an LP of 10 tracks. Nine songs have a similar spectrum and using the har-bal mastering process you obtain about the same loudness (eg: Rms avg -12 db). The 10th track is 5 minutes long and has got for the first 90 seconds only piano, lead vocal, hit hat and for the rest a full sound with drum, bass, elec. guitars, strings, ect... Even if you've worked well at the mixing stage trying to make the difference in loudness be minimal between the two parts, probably there will always be a gap between them. Now, after mastering you obtain a Rms avg lesser than the ones of the other 9 songs. You suppose that the first 90 seconds which are less loud than the rest of the track take down the average level.

What to do?

1) Must you refer only to the part of the song which sounds louder and if its Rms avg is about the same of the other Lp's tracks it's OK, sounds good?

2) Do you have to make sound the entire song (1st 90 seconds embodied) at the desired average loudness? (In this case probably the second part with fuller sound will be louder than the other songs of lp which have got a similar arrangement)

3) Do you have to come back in the mixing and fix some stuff?

Bye

Sam

Sam

I absolutely agree with Paavo.

Let me just add one more item. Generally if you have a song that starts out with acoustic instruments and then changes to a full range and it follows a full range song on an album try this:

Add a longer space between these two songs. This changes the volume expectation of the listener for the next song and it becomes a fresh new song. It is only when a song follows immediately that the listener expects the volume level to be the same. Make sure you place a longer fade out on the previous track.

Not sure how old you are but this trick was used quite a bit in the 70's during the disco era. The music would be pumped up.....one after another and then finally they would allow a song to finally end instead of mixing in a new one. There was a brief moment of silence and then they would bring in a slow song. The transistion seemed to be natural because it allowed us to wind down as the volume level decreased in the last song

Cheers

Earle

[sambrn]

Many thanks! Now it's all more clear and I can work better than I've done till now.

Actually I've got another question (the last!), this time more related to har-bal than the other ones which were general, but probably still linked to compression and/or limiting.

By the har-bal demo I've imported different commercial songs trying to study their spectrums. I've seen that the most of them have got a graph very sweet with valleys and peaks very light. On the contrary someone of them has got (mostly in the low freq area) peaks and valley more apparent.

What is the cause of this behaviour? Is only the compression to attenuate so much the peaks and fill the valleys? Do deep valleys and high peaks mean resonance?

I know that har-bal can be used for solving just these problems, but the engineers who mastered commercial songs produced 5 years ago I supposed they didn't have har-bal to make their spectral graphs so sweet. How can they do it without HB?

Sam

[har-bal]

Many thanks! Now it's all more clear and I can work better than I've done till now.

Actually I've got another question (the last!), this time more related to har-bal than the other ones which were general, but probably still linked to compression and/or limiting.

By the har-bal demo I've imported different commercial songs trying to study their spectrums. I've seen that the most of them have got a graph very sweet with valleys and peaks very light. On the contrary someone of them has got (mostly in the low freq area) peaks and valley more apparent.

What is the cause of this behaviour? Is only the compression to attenuate so much the peaks and fill the valleys? Do deep valleys and high peaks mean resonance?

I know that har-bal can be used for solving just these problems, but the engineers who mastered commercial songs produced 5 years ago I supposed they didn't have har-bal to make their spectral graphs so sweet. How can they do it without HB?

Sam

Sam

I can tell you this. They sure mixed the songs a lot better. Now a lot of folks just use a compressor to crush the song into shape instead of delicately sculpting it. It really is an art.

Years ago it was more the application of corrective eq'ing, just a minimal touch of compression to control the peaks and safe limiting. The whole idea was to preserve the dynamIcs and not compromise the sound by making it as loud as possible.

Cheers

Earle

[sambrn]

Thanks Earle!