Home Theater Forum and Systems banner

1 - 5 of 5 Posts

·
Registered
Joined
·
12 Posts
Discussion Starter #1
I am trying to use rephase so I need an usable phase plot.

when I make measurment with rew the phase plot is really steep (because of a delay I guess)
so when I export it it is unusable.

Usually I usew the minumum phase function but it cannot be exported as a phase plot txt for rephase.

How can I have a phase plot that is usable with rephase?
I mean it cant be so steep because I cannot make adjustments.
I am going crazy with that. the plot shows a delay that depends on what filters I use (apo or jriver live input), it goes from 300ms to over a second.

I should take care of that but I don't know how :(
 

·
Premium Member
Joined
·
1,562 Posts
I am trying to use rephase so I need an usable phase plot.

when I make measurment with rew the phase plot is really steep (because of a delay I guess)
so when I export it it is unusable.
Do not unwrap the phase for this purpose. Just ignore the wrapped sections when using RePhase.

Usually I usew the minumum phase function but it cannot be exported as a phase plot txt for rephase.
We should use the actual phase plot for the RePhase correction, not a min phase plot.

How can I have a phase plot that is usable with rephase?
I mean it cant be so steep because I cannot make adjustments.
I am going crazy with that. the plot shows a delay that depends on what filters I use (apo or jriver live input), it goes from 300ms to over a second.
For phase correction using a RePhase filter we are looking to remove the phase rotation of the direct signal from the speaker.

The LP phase plot in REW is often very difficult to read when we are looking for the direct signal phase as the numerous reflections and room modes interfere. In REW no smoothing should be applied as smoothing seems to make matters much worse for this purpose. Applying a proper window can often make the chart much easier to read, but takes some experience to adjust the window properly as it can also distort the results. Ideal window settings change with the range of frequencies we are looking at.

If we are using full range speakers or only interested in the main speaker phase rotation (not the SW) then a solution is relatively easy. We can move the mic closer to the main speaker to dimish the impact of the room and let us see the direct signal phase rotation more clearly. How close we move, is dependent on the type and size of the speaker in question. The mic should be kept on the listening axis, often identified as going from LP to the TW. I [prefer to move to a point midway between the TW and MW (or MR) drivers. The difference is no doubt trivial unless we are getting very close. We will still get some phase (and SPL) variation at the lower end of the range, but it will be much easier to recognize the account for.

If we are intending to include the SW phase rotation into the correction then the LP mic position is necessary. We could mentally paste the low freq result at the LP position to the near field result from the close mic position or use REW trace arithmetic to "merge" the 2 traces at the XO point. We then can work from there using an iterative process between REW measurements and Rephase adjustments to the Filter settings.

HolmImpulse does a much better job of exposing the direct signal phase when smoothing is applied. That can be a very useful tool for exposing the direct phase.

There are probably other options to help with this problem depending on the specifics of the situation. These general comments should be helpful, but if more specific advice is needed I would need a measure files and more specifics on your situation and questions.
 

·
Registered
Joined
·
2 Posts
Hi jtalden,

I also have a few questions about exporting to rephrase.

I have been using rephrase for amplitude and phase linearization as well as crossover creation, the resulting .wav files being convolved in Jriver. Seeing this post and a few others I wanted to check with you as to what is the correct phase to export.
Above you say not to use the minimum phase but to use the direct phase, can you explain why?
If using the direct phase should this be measured using a loopback reference to remove any system delays? Should I adjust for any estimated delays reported by REW?
Or should I just use Holmimpulse as suggested?

Any help you can offer will be much appreciated.
 

·
Premium Member
Joined
·
1,562 Posts
Well, I think this is more a technical difference rather than a practical one. I think I was trying to assure the OP that it is not necessary to use excess phase. The Phase trace can be used so the concern that excess phase could not be exported is not critical. If you have your choice, then either should be okay. The excess phase should be less influenced by room reflections and may make the phase easier to read.

Best to follow a guide from someone with lots experience with your type of setup. I don't know best practices in your type of setup.

Generally:
HolmImpulse is very valuable for phase response as it does a very good job when smoothing filters are applied. It helps to show the phase of the direct sound and help us see past the impact of the reflections. If the phase trace at the LP is too cluttered in REW it can be helpful to take a measurement with the mic close to the driver. That way a relatively clean direct sound phase response is captured. The filter should address that response. I would not try to smooth out the phase irregularities measured and the LP due to reflections.
 

·
Registered
Joined
·
2 Posts
Thanks for your prompt reply.
I will try a few measurements in REW and Holmimpulse to see which method gives the most reliable and repeatable results.
 
1 - 5 of 5 Posts
Top