REW+REPHASE=RECONFUSED

[camplo]

Hopefully the title says it all. I am trying to experiment with room correction of the eq and impulse...I know I have the tools but I can't use them right yet. When I load in measurents from REW the spl data is not correct and shows up as a skinny line at 20khz instead of a curve reflecting my actual measurements.

 

[jtalden]

Possibly it's the vertical line we see in rePhase as the indicator for the upper frequency limit? That would be at 22,050 Hz for a 44.1 kHz sample rate measurement.

Your imported .txt file may fall off the default scaling of the rePhase chart. We can go to 'Ranges' and reset the SPL and phase ranges wider to determine that. We can also offset the gain under the 'Measurements' tab to move the SPL trace down to 0 dB where we normally work on it for the purpose of creating filters.

 

[camplo]

wow I feel so silly.....I was limited by the max amplitude view-able, thanks for taking the time =D

 

[camplo]

Ok I;ve created a wav that I load into my convolution reverb vst...and its adding room, reverb time, density, etc etc to the equation and I just want the augmented eq curve....any advice on this?

 

[jtalden]

I don't have any experience with that software, so hopefully someone else can help with that.

 

[camplo]

WEll I think I have that part figured out as well. One last thing on my mind. Step response. I've seen it claimed that rePhase corrects step response. I was just wondering is that something that happens automatically, are there settings for this part of the program? I'm creating linear phase filtering so when I look at the parametric phase eq there's nothing to correct, it just shows a straight line. So whats the deal, with that, does rephase correct step response and do I have to do anything extra to take advantage?

 

[camplo]

I might as well throw another Q out there, where in the world do I find a description of all theses settings in rephase and what they are for. I can figure out most of it but I cannot tell exactly all of them.

 

[jtalden]

...I've seen it claimed that rePhase corrects step response. I was just wondering is that something that happens automatically, are there settings for this part of the program?

It's not automatic. The settings are manually set.

I'm creating linear phase filtering so when I look at the parametric phase eq there's nothing to correct, it just shows a straight line. So whats the deal, with that,...

Normally, it's recommended to use minimum phase EQ filters for amplitude correction as that also helps to improve the phase response. Linear phase EQ filters will leave the phase unchanged.

...does rephase correct step response and do I have to do anything extra to take advantage?

To improve the appearance of the step response you can do that by removing the speakers phase rotation. Setting the overall speaker phase response to match its minimum phase response is the common recommendation.

The overall approach to optimizing the system depends on your equipment capabilities. The rePhase thread on DiyAudio.com is very detailed particularly for those with independent FIR XO and EQ for each driver. The MiniDSP site has a good example for setups using their devices. There is another excellent 'how to' linked within the main DiyAudio thread, but I could not quickly spot it.

 

[jtalden]

Another good thread.

I don't know of any sites that explain the various rePhase controls. It's best to invest in reading the main DiyAudio thread.

 

[camplo]

Ty for your time and help. In the regards of phase alignment. Could you be a little more specific about the phase rotation?Or maybe point me to the guide?

 

[jtalden]

Please confirm your situation.
What is the objective?
What are the speakers?
Are you using a DSP crossover (XO) to EQ and phase align the individual drivers, or are you applying IIR XO to the drivers and then intend to remove the phase rotation at the end of the process, or??

A better understanding of your situation and objective is necessary if we are to provide any specific help.

By overall phase rotation I am referring to the changing phase with frequency that is a result of 2 or more drivers within a speaker that is crossed over with a passive XO or active IIR XO. Below is a phase overlay chart showing the initial phase rotation (red) and the resulting phase rotation (green) after an FIR phase correction filter was applied. The application is the speaker shown in my avatar along with passive SWs. The XOs are active IIR at 100 Hz and 2 kHz.

If the XOs are done well then removal of the remaining phase rotation is as shown here has no significant impact to sound quality in my opinion. It does make the step response look more ideal however. I can't hear a difference, but my hearing is substandard and others say they can. I think a well timed XO is more likely to impact sound quality

than is the overall phase rotation and improved step response. I use this FIR filter for music, but it is not active for AV use.

​