Hello, I love your videos! I'd like to know what are the settings in the preferences prior to taking measurements? I have been trying your approach and failing at some stages so I wonder if it's because of the premeasurement settings so would it be possible to make another video with premeasurement adjustments in REW or maybe you can write it here I don't know which will make more sense. I'm using Umik-1.
The settings are modified from the defaults, though recently I have been using Focus Fidelity's Impala to take measurements. I had one instance in which I was not able to get Impala to work, however, and in cases like that it does help to know how to configure REW for measuring. I'll take a few screenshots and put links to them in the description.
Why did you specifically use psychoacoustic smoothing before creating your averages? Also, would that be different than averaging with no smoothing, then changing the averaged trace to psychoacoustic?
I don't think it would be different, no. I smoothed the response to better match the way we hear and also to make it easier for REW to match it to the target in the EQ section. But recently the updated versions of REW are able to handle matching better at higher resolutions. I can also use 1/6 or 1/12 smoothing now for the EQ filters. I still use psychoacoustic for optimizing group delay.
The way I approached the process there was in two stages. The first stage is the magnitude linearization, which calls for RMS or dB average. The second stage is for the timing aspect of the filter, which calls for a vector average to generate the necessary data.
@@davidbrancato is convolution the only tool to adjust the timing aspect? I'm very familiar with EQ'ing things, I use MiniDSP's and Audyssey Multi-EQ-X (which turns PEQ to their FIRs). But I've never been able to figure out how to manipulate phase.
@@joek6207 With regard to processing like this, yes. As far as I'm aware you cannot correct phase or timing with equalization alone. I could be wrong about that in some sense but I've never seen an equalizer correct phase issues by itself.
@@joek6207 In this video we are only dealing with the frequency response. You'd need to use the vector data from the measurements in addition to rePhase to correct the phase.
Try it both ways and see which you prefer. Generating these magnitude correction filters is a relatively quick process and you can A/B the results in your own acoustic space to audition them. Using a single-point measurement you will get more accurate results at your listening position than you'll get if you use a multi-point array. However, you can give the main listening position more influence in an average by measuring it several times. In the video here it had been done twice, but you can do it as many times as you want. The more you do, the more influence it will have. Let me know how your results turn out.
Let me ask you, why have the overall max boost value set at 2 but the individual value set at 9? Shouldn’t it be vice-versa? How can you ever exceed 2db if the max is 2db overall? I have never seen anyone (albeit I have seen people set the two at the same value but more often than not I have seen people set the max/overall at 1db higher than the individual) set the max lower than the individual………hence you definitely have piqued my curiosity!
I actually set them both to zero most of the time. In this case it didn't matter. As long as it gives you the result (predicted response) you are happy with it's alright. Check the default values after you first install REW. I think they are similar to what I have in this video... at least they were before. I don't know about now because I am always adjusting them, which is what I had done here prior to making the video until the predicted response looked the way I wanted it.
A great step-by-step explanation of a relatively complex topic. Keep up the good work!
Thank you!
does the time alignment is in the same filteR ?
Hello, I love your videos! I'd like to know what are the settings in the preferences prior to taking measurements? I have been trying your approach and failing at some stages so I wonder if it's because of the premeasurement settings so would it be possible to make another video with premeasurement adjustments in REW or maybe you can write it here I don't know which will make more sense. I'm using Umik-1.
The settings are modified from the defaults, though recently I have been using Focus Fidelity's Impala to take measurements. I had one instance in which I was not able to get Impala to work, however, and in cases like that it does help to know how to configure REW for measuring. I'll take a few screenshots and put links to them in the description.
Can Eclipse Audio Speaker Measure be used instead of REW?
Why did you specifically use psychoacoustic smoothing before creating your averages?
Also, would that be different than averaging with no smoothing, then changing the averaged trace to psychoacoustic?
I don't think it would be different, no. I smoothed the response to better match the way we hear and also to make it easier for REW to match it to the target in the EQ section. But recently the updated versions of REW are able to handle matching better at higher resolutions. I can also use 1/6 or 1/12 smoothing now for the EQ filters. I still use psychoacoustic for optimizing group delay.
Hello David, great Tutorial. The only one I can follow as beginner. Could I ask some questions? In which forum are you available?
avnirvana, under the REW section. My avatar/name on the forum is Moedra.
Why do you use RMS average vs Vector average?
The way I approached the process there was in two stages. The first stage is the magnitude linearization, which calls for RMS or dB average. The second stage is for the timing aspect of the filter, which calls for a vector average to generate the necessary data.
@@davidbrancato is convolution the only tool to adjust the timing aspect? I'm very familiar with EQ'ing things, I use MiniDSP's and Audyssey Multi-EQ-X (which turns PEQ to their FIRs). But I've never been able to figure out how to manipulate phase.
@@joek6207 With regard to processing like this, yes. As far as I'm aware you cannot correct phase or timing with equalization alone. I could be wrong about that in some sense but I've never seen an equalizer correct phase issues by itself.
@@davidbrancato how is phase corrected?
@@joek6207 In this video we are only dealing with the frequency response. You'd need to use the vector data from the measurements in addition to rePhase to correct the phase.
hi ，would me tell me which method is the best for stereo . single point or average? Thank you.
Average if you want some room to move your head at the listening position and remove some of the room influence in the measurements.
Try it both ways and see which you prefer. Generating these magnitude correction filters is a relatively quick process and you can A/B the results in your own acoustic space to audition them.
Using a single-point measurement you will get more accurate results at your listening position than you'll get if you use a multi-point array. However, you can give the main listening position more influence in an average by measuring it several times. In the video here it had been done twice, but you can do it as many times as you want. The more you do, the more influence it will have. Let me know how your results turn out.
Let me ask you, why have the overall max boost value set at 2 but the individual value set at 9? Shouldn’t it be vice-versa?
How can you ever exceed 2db if the max is 2db overall? I have never seen anyone (albeit I have seen people set the two at the same value but more often than not I have seen people set the max/overall at 1db higher than the individual) set the max lower than the individual………hence you definitely have piqued my curiosity!
I actually set them both to zero most of the time. In this case it didn't matter. As long as it gives you the result (predicted response) you are happy with it's alright.
Check the default values after you first install REW. I think they are similar to what I have in this video... at least they were before. I don't know about now because I am always adjusting them, which is what I had done here prior to making the video until the predicted response looked the way I wanted it.