Holmimpulse allows convolutions too, but impulses need to be imported in frd format. Praxis also allows many manips. But as i got tired of messing with free soft which partly allowed to get things done, i finally invested some pocket money into Acourate which, way beyond a mere DRC solution , is a perfect fir pocket calculator.
Attachments
Last edited:
Anyway many convolvers allow cascading convolutions, though this has a cost in computing resources. This can be done with brutefir for example.
Hello, is there a way to export only phase data from a measurement with REW5? I tried to export the minimum phase "Measurement as text" , but i get the spl data also. What i need is an impulse response as wav or text which has the minimum phase data of the measurement but flat frequency response. Then i can create a time reversed impulse of the measurement which i will use for corrections (example to compensate high pass behaviour of a subwoofer). Thanks 🙂
You can export measurement as text with REW.
Using excel (or other),
delete the magnitude column,and place a column with "0" data.
re-import the file with REW.
and export as impulse.wav
But minimum phase is implicitly linked to the magnitude .
Using excel (or other),
delete the magnitude column,and place a column with "0" data.
re-import the file with REW.
and export as impulse.wav
But minimum phase is implicitly linked to the magnitude .
Thanks for reply, i tried to open it with excel but it reads it like a txt file, then the numbers are not centered on columns. rephase doesn't let me to do a solo time reversal of the phase of the loaded txt.
Hello,
Seems to work fine (with openoffice) and REW export (.txt).
header is not useful,just as information.
the middle column is often (always) the magnitude.
Seems to work fine (with openoffice) and REW export (.txt).
header is not useful,just as information.
the middle column is often (always) the magnitude.
An externally hosted image should be here but it was not working when we last tested it.
In the tab "file"
import frequency measurement.(.txt files)
perhaps,you need to remove dot's or comma's from the .txt file.(depends how excel or OpenOffice saves the file).
import frequency measurement.(.txt files)
perhaps,you need to remove dot's or comma's from the .txt file.(depends how excel or OpenOffice saves the file).
Last edited:
Yes but if i import it as "frequency response" i can't re-export it as wav impulse needed for the convolution.
Question for all. I need to cascade rephase filters with other rephase or drc filters.
Do the impulse centers have to be in the exact same location? Or can i have a filter with an impulse center at 10 samples cascaded with a filter that jas an impulse centwr at 22050 samples?
Do the impulse centers have to be in the exact same location? Or can i have a filter with an impulse center at 10 samples cascaded with a filter that jas an impulse centwr at 22050 samples?
That will provide a pure delay.
At 48K,
22050-10=22040*1/48k
0.459 second.(only in linear phase configuration).
At 48K,
22050-10=22040*1/48k
0.459 second.(only in linear phase configuration).
thank you, but is it ok to use that filter with another one with no delay or a different delay?
Hello,
i don't understand exactly the purpose.
taking a look at the impulse with any .wav editor,
in linear phase,impulse's are symetrical.
delay is given by the max peak of IR (or energy max).
in minimum phase (or non-min),delay is given by the "beginning" of the impulse.
once coefficient start to rise.
Otherwise,an acoustic measure would provide the delay to apply.
(like "use loopback as time ref" in REW,or cross correlation with ARTA).
i don't understand exactly the purpose.
taking a look at the impulse with any .wav editor,
in linear phase,impulse's are symetrical.
delay is given by the max peak of IR (or energy max).
in minimum phase (or non-min),delay is given by the "beginning" of the impulse.
once coefficient start to rise.
Otherwise,an acoustic measure would provide the delay to apply.
(like "use loopback as time ref" in REW,or cross correlation with ARTA).
hi,
here is the purpose.
I am running active crossover using rephase filters. they are linear phase but not centered.
for example, a crossover im using right now is 8192 taps but the impulse center is at 5%, not 50%(rephase allows one to change this as you know)
in addition to that, using brutefir, i want to also cascade EQ filters for the speakers that will be separate from the crossover filters. I use drc for this and drc does not have an option to set the impulse center of the filter(unless using a true linear filter which the creator does not recommend)
so can i use a drc eq filter and a rephase xo filter at the same time, with different impulse centers?
also, can they be different tap lengths?
i guess i could just go and test this except im travelling right now so no access to my htpc.
thanks
here is the purpose.
I am running active crossover using rephase filters. they are linear phase but not centered.
for example, a crossover im using right now is 8192 taps but the impulse center is at 5%, not 50%(rephase allows one to change this as you know)
in addition to that, using brutefir, i want to also cascade EQ filters for the speakers that will be separate from the crossover filters. I use drc for this and drc does not have an option to set the impulse center of the filter(unless using a true linear filter which the creator does not recommend)
so can i use a drc eq filter and a rephase xo filter at the same time, with different impulse centers?
also, can they be different tap lengths?
i guess i could just go and test this except im travelling right now so no access to my htpc.
thanks
Last edited:
You can merge brutefir EQ with each of rePhase file.
using REW,Holm impulse,
this is a simple multiplication.(frequency domain)
A x B (REW or HOLM).
EQ do not imply pure delay,but reaching 🙂o for a nice measurement) a linear phase.(on the bandwidth).
using REW,Holm impulse,
this is a simple multiplication.(frequency domain)
A x B (REW or HOLM).
EQ do not imply pure delay,but reaching 🙂o for a nice measurement) a linear phase.(on the bandwidth).
1201, there is no problem running two convolutions one after the other, regardless of their individual delays.
As Thierry38 suggests you can also "merge" the two corrections by convoluting one FIR with the other, using SoX or Holm for example.
As Thierry38 suggests you can also "merge" the two corrections by convoluting one FIR with the other, using SoX or Holm for example.
Last edited:
Thank you POS. thats what i was looking for.
I prefer not to preconvolve the multiple filters as if I want to change,say, a drc configuration, then i would have to generate the drc filter and preconvolve it with the xo filters every single time.
cascading the filters, i can just keep my crossover filters in place and place the drc filter on top of them, or even just an eq filter. I think this would be the easiest and fastest way.thanks.
in Europe:
use decimal separator , (comma)
use field separator ;
column headers, which mean name of columns are frequency, amplitude, phase
when you open the export file into excel and if it is .txt type file,
excel will ask step by step how to open the file and convert data into columns.
The key part is what you use as decimal separator and what you use as field separator.
In case you have all data in one column case, you can use in excel menu
DATA Text to columns
and excel will ask the field delimiter used and it can be manually selected
in case US exxcel with US country settings
- you need first batch the source file with text editor that has find and replace function
- use that in case the source file is done with international settings and field delimiter needs to be comma (,) and decimal separator dot
I recommend using international settings and then import and export back is easier..... if you use excel in between as column values editor
Dear Pos, dear all,
I use Rephase since years, but being only an enthusiast some doubts still remain on its proper usage notwithstanding your invaluable help, here as well as elsewhere.
In particular, regarding the compensate mode, you stated that it is only a temporary state and recommended to use it in conjunction with additional filter(s). Could XO filters be sufficient? Might I have a practical example? To this aim I attach the txt file of my B&G Neo 10 and the corresponding Rephase setting (BP2R). Is it right?
Finally, regarding the functionality phase linearization (BOX) under Filters Linearization, could it be used to flatten the phase of any driver (for instance the above-mentioned Neo 10, see for instance the setting BP2-IIR) ?
Many thanks in advance to all for the attention
Kind regards
Antonello
I use Rephase since years, but being only an enthusiast some doubts still remain on its proper usage notwithstanding your invaluable help, here as well as elsewhere.
In particular, regarding the compensate mode, you stated that it is only a temporary state and recommended to use it in conjunction with additional filter(s). Could XO filters be sufficient? Might I have a practical example? To this aim I attach the txt file of my B&G Neo 10 and the corresponding Rephase setting (BP2R). Is it right?
Finally, regarding the functionality phase linearization (BOX) under Filters Linearization, could it be used to flatten the phase of any driver (for instance the above-mentioned Neo 10, see for instance the setting BP2-IIR) ?
Many thanks in advance to all for the attention
Kind regards
Antonello