Hello all. Ok my question is, aside from using multiband Compression are there any other processors/effects that can be used to get the green & yellow lines to run parallel to each other? (slopes, peaks, etc etc).
For example, on my mixes thers someitmes a peak in the yellow & at the same place a slope in the green, how can i get them both to peak & slope evenly?
Ok another silly question, how much range does an octave cover? Lets say i want to use a multiband comp to compress the 100-150z range, would that be like half an octave in the compressor? If thers a certain slope I want 2 compress in the yellow line whats the easiest way to read the length the slope covers? Hope this makes sense i know they are noob questions. Thanks 2 any1 who answers!
btw i think this program is great blah blah blah what else can I say that no one else hasnt
RED & YELLOW LINES
I wouldn't fret about not having the yellow and Green lines parallel to one another. Just because they are not, is not necessarily an indication of a problem. In fact, if your music is dynamic and multi-thematic then it is highly likely that they won't be, and for good reason.
Take a simple example of a 3 minute track that has a short but intense guitar solo in the middle. Because the yellow trace corresponds to peak spectrum it will clearly show the effect of the guitar solo, but the green trace is the average over the whole track so the short guitar solo won't have much effect on the average spectrum. Put simply, if the peak and average spectrums are different it is an indication that your music changes character in time. An example of something that does't is pink or white noise, but that doesn't make for interesting music.
To control the peak in the peak trace requires you to isolating the offending part of the track responsible for it. HarBal in its current form isn't helpful at doing this though we have plans to fix it. Some people employ multiband compression to control this sort of thing, but if you know which part is responsible, simply applying a volume envelope to it prior to mixdown might fix the problem.
On the subject of an octave and how big it is, an interval of one octave is two frequencies that have a 2:1 ratio between them. For instance the distance between 440Hz and 880Hz is an octave because one is twice the other. Half an octave is the ratio that is the square root of 2 (approx. 1.414) and one twealth an octave is the ratio that is the 12th root of 2 (2 raised to the power of 1/12) which is approximately 1.059. So in your example the interval between 100-150Hz is approximately half an octave and it is centred on the frequency that is the square root of 100 multiplied by 150 (ie. geometric mean) which is 122.5Hz.
Regards,
Paavo.
Take a simple example of a 3 minute track that has a short but intense guitar solo in the middle. Because the yellow trace corresponds to peak spectrum it will clearly show the effect of the guitar solo, but the green trace is the average over the whole track so the short guitar solo won't have much effect on the average spectrum. Put simply, if the peak and average spectrums are different it is an indication that your music changes character in time. An example of something that does't is pink or white noise, but that doesn't make for interesting music.
To control the peak in the peak trace requires you to isolating the offending part of the track responsible for it. HarBal in its current form isn't helpful at doing this though we have plans to fix it. Some people employ multiband compression to control this sort of thing, but if you know which part is responsible, simply applying a volume envelope to it prior to mixdown might fix the problem.
On the subject of an octave and how big it is, an interval of one octave is two frequencies that have a 2:1 ratio between them. For instance the distance between 440Hz and 880Hz is an octave because one is twice the other. Half an octave is the ratio that is the square root of 2 (approx. 1.414) and one twealth an octave is the ratio that is the 12th root of 2 (2 raised to the power of 1/12) which is approximately 1.059. So in your example the interval between 100-150Hz is approximately half an octave and it is centred on the frequency that is the square root of 100 multiplied by 150 (ie. geometric mean) which is 122.5Hz.
Regards,
Paavo.
-
mosthated_99
- Posts: 2
- Joined: Tue Apr 26, 2005 8:18 pm
Thanks very much that was really helpful. I guess i kinda worried about the green & yellow running parallel cuz i was opening commercial tracks with har-bal just to see how the lines would look on a "properly mixed/mastered track".
My song I was monitoring actually does have variations so I guess that does explain y the lines go off sometimes. Now that I check other of my songs I do see they kinda do run parallel.
My song I was monitoring actually does have variations so I guess that does explain y the lines go off sometimes. Now that I check other of my songs I do see they kinda do run parallel.
Mosthated_99mosthated_99 wrote:Thanks very much that was really helpful. I guess i kinda worried about the green & yellow running parallel cuz i was opening commercial tracks with har-bal just to see how the lines would look on a "properly mixed/mastered track".
My song I was monitoring actually does have variations so I guess that does explain y the lines go off sometimes. Now that I check other of my songs I do see they kinda do run parallel.
One thing you will see in well mastered tracks in Har-Bal are the absence of spikes (yellow line) and the absence of holes (green line) where your song's depth resides.
So when you load when of your tracks in Har-Bal, thIs should be your target or goal.
It doesn't take long to get the hang of it. Har-Bal Is a teaching tool and after a while you will be able to recognize bad areas in recordings almost immediately and even call out their frequency areas.
Mastering engineers have been using meters for years to propery align their songs. The difference here is that the same tool allows you to go in and immediately correct those areas.
Earle