[robabeatle]

I was pointed to a post on FB that went through the process of using a feedback llop to measure phase issues with REW. I cannot find the tab for the following step:

2b. after measurement in "SPL&Phase" tab click "Estimate IR delay".

Any help?

 

[cathul]

I guess that was my post on FB.

You can find that under "Controls", i.e. the gear icon right next to "Limits".

 

[robabeatle]

I guess that was my post on FB.

You can find that under "Controls", i.e. the gear icon right next to "Limits".


View attachment 359104

Your post?! Thank you very much. I have no idea what I am doing, but trying to follow your step by step.

 

[cathul]

If in doubt just drop me a note on FB, or add me there. I‘m more active there these days.

 

[robabeatle]

Okay, I think I ran the first phse check. I used a sine sweep 1500 to 20 Khz for the tweets. The "length" is set at 512K, no idea if that is right. I measured Rtweet and got an estimate IR delay of -1.7976 ms and then the Ltweet at -.0406 ms. Here is the plot? What do I do with this? This plot is a jumbled mess. What am I doing wrong?

 

[cathul]

Nothing. Phase at high frequencies is hard to read. That’s why i use the impulse responses for aligning left and right tweeter. You can do that in the overlay window of REW.

 

[robabeatle]

These plots?

 

[robabeatle]

 

[cathul]

Yeah, but switch from dbfs to %fs. Makes it easier to read. And normalize everytime you zoomed in.

 

[Speedhunter]

@cathul Hermanns is that you?

 

[dumdum]

Nothing. Phase at high frequencies is hard to read. That’s why i use the impulse responses for aligning left and right tweeter. You can do that in the overlay window of REW.

Erm… you need to change that up to make the phase readable… in smaart you add delay to the measurement signal to achieve this, I can’t recall how rew does it to correct the view

now I’ve double checked… it’s estimate ir delay (impulse response delay) this explains it well

 

[robabeatle]

Thanks everyone. I basically gave up as I was overly confused but the few tips here I will try to employ tomorrow.

 

[cathul]

Erm… you need to change that up to make the phase readable… in smaart you add delay to the measurement signal to achieve this, I can’t recall how rew does it to correct the view

now I’ve double checked… it’s estimate ir delay (impulse response delay) this explains it well

I know, and if you read what I wrote this „Estimate IR delay“ is exactly what I recommend. Still the phase graph in REW is hard to read compared to Smaart.

 

[cathul]

@cathul Hermanns is that you?

Yes

 

[Speedhunter]

Yes

Good tutorial on TA with aucustic reference; had good results.

 

[robabeatle]

I just was reading this post: The Essentials of Sound Quality IMHO

And wonder if I should go through that method "by ear" around the crossover, before measuring time alignment?

 

[Kalmangar]

Erm… you need to change that up to make the phase readable… in smaart you add delay to the measurement signal to achieve this, I can’t recall how rew does it to correct the view

now I’ve double checked… it’s estimate ir delay (impulse response delay) this explains it well

If I remember correctly you need loopback for this application.
Correct me if I am wrong, but great video love him.

 

[dumdum]

I know, and if you read what I wrote this „Estimate IR delay“ is exactly what I recommend. Still the phase graph in REW is hard to read compared to Smaart.

The phase graph in rew can be made to look exactly like smaart, and no you never mentioned estimate ir delay, you said you used impulse response to time tweeters…. It’s written above and you definitely never mentioned the ‘estimate ir delay’ button in any way… perhaps you meant to write it but didn’t

use estimate ir delay on a midrange as that is where the phase will be flattest and so will give the best comparison of data 👍🏼

 

[dumdum]

If I remember correctly you need loopback for this application.
Correct me if I am wrong, but great video love him.

No you can use a timing chirp from one tweeter to measure phase of the system

 

[cathul]

I even put a screenshot in here where I circled in red.

and here is a screenshot from my posting in the RTA group on Facebook. See point 2b?