Voxengo Premium Membership - All Voxengo Plugins For a Fixed One-Time Fee
Forums     Plugins     CurveEQ Equal Loudness Curves - Waves or Presets

This topic was created before release of the latest product version, and it may contain details irrelevant to this version.  Replying is disabled for this topic.

Does anyone happen to have a link to a wave file of either white or pink noise that has had equal loudness compensation applied to it.  Here's a curve, for example:

http://www.sfu.ca/sonic-studio/handbook/Phon.html

I thought I'd pick a shape from the 83-85 dB SPL range and play around a little.  Maybe even use it on Sonic Finalizer and get a equal loudness settings.

I'll try and make one if no one else has one.

I also want to use CurveEQ spectrumatch to listen to an equal loudness compensated noise wave file and match a couple of mixes, then fine tune it to my needs.  Then see how it sounds in different places.

Thanks for any tips, wave links, or keywords to search the internet !

kylen

ps I'd mention why I'm doing this but the post would be real long, OK I'll mention that it starts with "bad listening environment..."


Grey or gray noise - that's white noise that has had an 'A' weighted equal compensation curve applied to it.  More to resemble a balance the way the ear hears - they say.

Anyone have any samples or waves of grey noise or eq settings to apply to white noise to make it grey ?

Thanks,

kyle


I guess you can get gray (pink, brown, etc) noise in CurveEQ itself.  You have two control points at the very start.  Simply set the spectrum slope global control to the desired value and then move those two control points so that white noise's spectrun becomes a horizontal line.  Taking into account the implied spectrum slope you will have the necessary compensation EQ line to transform the white noise.

Hi Aleksey,

I just found an 'A' weighted loudness compensation curve I'm going to apply to white noise and try to make grey noise.

Frequency (Hz) Curve A (dB)

16 ...... -56.7

31.5 .... -39.4

63 ...... -26.2

125 ..... -16.1

250 ..... - 8.6

500 ..... - 3.2

1000 ......  0

2000 ......  1.2

4000 ......  1.0

8000 .... - 1.1

16000 ... - 6.6

(Note: I'm not sure if the db above are in db SPL or dbm so I have to check that one first.)

I've tried the white noise and pink and the CurveEQ slope per octave settings and want to see how the grey noise works if I morph to it.  I am expecting balanced octaves but I could be making a mess too - it's all part of the journey !

Thanks,

kylen


BTW, the easiest way to do it is first put enough points to those frequencies (1000, 2000, 4000) and lock them all (turn them red).  After this you can adjust their gain easier, not paying attention to frequency.

I've just did that EQ curve...  Dunno what is the actual goal of it.


Hi Aleksey,

A good guy over at the audiomasters forum (Cool Edit Pro) just gave me a weighting calculation tool that looks pretty good:

http://www.measure.demon.co.uk/Acoustics_Software/a_weight.html

Now I can try my 1/3 octave experiment.

I think my first experiment here is to try and do a little EQ pre-emphasis on a bad mix (both EQ and dynamics are off) so the multiband compressors don't kick in (detector threshold) for the wrong reason.  In other words I don't want the compressor to kick in when it detects a frequency band I don't want compressed - remember this is a bad mix and those kinds of things are properly adjusted during mixdown.

At any rate I believe if I rebalance the EQ a bit first towards a neutral zone or flatten it - so to speak - before the compressor sees it then the compressor will be better able to adjust incorrect dynamics.  The neutral zone I'm trying to adjust the EQ to is a cross between the fletcher-munson curves (at around 80 db SPL) and the 'A' weighted compensation circuitry both of which are said to resemble the way the ear hears naturally.

Using the up-down arrows on CurveEQ I can add or subtract as much pre-emphasis as I like.  I'll probably do this in conjunction with a good single or multiband compressor like Elephant and Sonic Finalizer.

After balancing the dynamics this way I am imagining that I can readjust the EQ by either adding back in a certain degree of the inverse pre-emphasis - or more likely rebalance the EQ to the type of program material and my liking (remember the EQ is bad on this mix also).

I might even readjust the dynamics once they are actually balanced properly.  Maybe adding a little punch or what ever - maybe not.

So this is just another repair job I've brought into the shop here to have fun with !

This technique would also be somewhat similiar to flatening out the curve according to a 3 db/octave slope that you already have in the CurveEQ.  I'm just trying to add another degree of accuracy (?) by putting some equal loudness (fletcher-munson & 'A' weighting) compensation points on the curve.

I'll post back as to how it goes.  Maybe others have already tried this or can think it thru - let me know if you think it's baloney that's ok too !  I'm just trying out different things here.

kylen

ps I might try to use spectrumatch with grey noise too, maybe later.  I've already tried it with pink and white noise for fun - but the pre-emphasis idea didn't hit me till now.

pps I just figured out that doing an experiment using pre-emphasis may have a similiar effect as sending a control signal thru a sidechain to a detector.  Hmmm


These are the center frequencies and 'A' weighted relative db I'm going to try in a few day.  Right now I'm comin down with the flu so don't get to close to this post, stay far away - Europe should be far enough !  The chart looks a little squeezed toghether due to the forum formatting.  There's a prettier chart over at: http://www.audiomastersforum.org/phpbb/viewtopic.php?t=161

==================================================================

31 ISO Standard, 1/3 octave bands per channel from 20 Hz to 20 kHz

==================================================================

20 25 31.5 40 50 63 80 100 125 160

200 250 315 400 500 630 800 1k 1.25k 1.6k

2k 2.5k 3.15k 4k 5k 6.3k 8k 10k 12.5k 16k

20k

==================================================================

'A' Weighted Equal Loudness Curve in relative db

==================================================================

-50.4 -44.8 -39.5 -34.5 -30.3 -26 -22.4 -19.1 -16.2 -13.2

-10.8 -8.7 -6.6 -4.8 -3.2 -1.9 -0.8 0.0 0.6 1.0

1.2 1.3 1.2 1.0 0.6 -0.1 -1.1 -2.5 -4.2 -6.7

-9.3

==================================================================

kylen

ps Please let me know if you see anything wrong here so I can fix it.


kylen, you are right about the 'pre-emphasis' you are talking about (EQing before and inverse EQing after).  Sometimes earlier I suggested to chain two EQs: before Sonic Finalizer and after it.  It is a 'messy' approach, but it can give a bit of additional control.
This topic was created before release of the latest product version, and it may contain details irrelevant to this version.  Replying is disabled for this topic.