filtergtaph~ gain = gain in dB .. how?
Hi All,
I am puzzled on how I make it, so if I input "6" in the gain parameter to filtergraph, I get 6 dB of gain at the cutoff, - 6 = - 6 dB of gain etc.
I use the peak notch filter
Sorry, I get kind of cold water on my cheeks when I try to understand filters / the math behind filters.
all the best, lasse
Hello !
it expects linear input not logarithmic . i know its confusing and it would be great to have control over it internally .
look here at this conversion , its more friendly .
#EDIT
this saved a lot of time. Thank you so much!
can you paste the conversation again? didn't see the link, and would like to understand it, if the conversation explains how to :)
cant u see anything ? maybe its a forum bug . so again ...
and in attachment
[atodb]
[dbtoa] u mean
#EDIT
with [dbtoa] u will notice a bit of decimal offset , that wonders me if its true value or there is precision problem compared to literal conversion
thank you to both. Yes i noticed that too. I do seem to remember however, that a doubling of amplitude gives a 6.02 dB of increase, and not 6.00 dB as commonly stated. Maybe I do remember wrong or misunderstood something?
This would argue that the dbtoa is more precise than your expression - but again, the math side of these things are really not my strong side :)
oh , i didnt know about the result of doubling , good to know , it means im not familiar with how linear formula should work for DB . its offset puzzled me for a while , but i think for such cases it doesnt make a big difference . using [dbtoa] would be the most obvious choice anyway .
no not at all .. i don't think it is possible to hear the difference! i didn't know about the dbtoa - thank you for this!
I sent a mail to cycling support to ask them to include it as an example for filtergraph~ .. i'm most probably not the first or last who would like to be able to control this object by add'ing in dB :)
Good day to all!
Decibels are based on powers of 10 not powers of 2. The fact that 6dB and -6dB are very close to 2.0 and 0.5 is a happy coincidence and a useful rule of thumb, but is not a precise measurement.
thats great , im happy to learn it Christopher . thanks !