2 Attachment(s)
Re: PE Cnote Bookshelf Speakers Build
Finally tuned these last weekend.
Here are the REW results: I tuned everything for about a 0.6db/oct slope downward. I tuned the sub to be that same level as both speakers playing but then added a low shelf filter as a global eq. Sub is a cheapo polk audio 10" psw110. I use the term sub loosely with this thing... This isn't a room issue. I tried it different spots and same issue with this. I guess I don't need 20-50hz....
Attachment 15252
That was pre-smaart phase aligning too.
Here are the smaart results:
Attachment 15253
The crossover is set to around 130hz. My speakers are 3' 9" from the wall which is conveniently around 80hz-100hz for them...So to get rid of that wall interference, I had to up the crossover higher than I would like. I also tried aligning the crossover region up the best which is about 100hz - 150hz when they speakers are within 10db of each other.
Re: PE Cnote Bookshelf Speakers Build
Do you have a trace of all three speakers playing together? How well does that crossover region sum together?
Re: PE Cnote Bookshelf Speakers Build
I thought I did but apparently didn't save it. I did measure it to make sure I was good and it had the best summation I could get.
Re: PE Cnote Bookshelf Speakers Build
Quote:
Originally Posted by
Jdunk54nl
Finally tuned these last weekend.
Here are the REW results: I tuned everything for about a 0.6db/oct slope downward. I tuned the sub to be that same level as both speakers playing but then added a low shelf filter as a global eq. Sub is a cheapo polk audio 10" psw110. I use the term sub loosely with this thing... This isn't a room issue. I tried it different spots and same issue with this. I guess I don't need 20-50hz....
Attachment 15252
That was pre-smaart phase aligning too.
Here are the smaart results:
Attachment 15253
The crossover is set to around 130hz. My speakers are 3' 9" from the wall which is conveniently around 80hz-100hz for them...So to get rid of that wall interference, I had to up the crossover higher than I would like. I also tried aligning the crossover region up the best which is about 100hz - 150hz when they speakers are within 10db of each other.
do you have all pass filters available in your dsp? If so you can make a way better job of that crossover, the sub has too much delay on it also, flip the phase and it should be a whole lot nicer through the crossover
Re: PE Cnote Bookshelf Speakers Build
Ya I do have all pass somewhere in it. It is an adau 1701 chip ran by sigma studio (Dayton Audio DSP-K). It has everything.
The sub has 0 delay set in the dsp, but on smart it was like 38ms to "Track" it. The other speakers were around 3.5ms to "track" them and I added as much delay as I could without running out of RAM on the dsp chip (500 samples which I am not sure on how many ms that is, but Justin showed me how I can do that). I can't add that much delay.
I also need to do some more tests. This is where I am practicing my smaart skills as it is easier to set everything up in an office compared to my truck :).
Edit* just swtiched to ms instead of samples on the program and left is delayed 12.1042ms and right is delayed 11.6667 ms. Man sigma studio is precise.... I can change the time delay by .0001 ms.
Re: PE Cnote Bookshelf Speakers Build
Quote:
Originally Posted by
Jdunk54nl
Ya I do have all pass somewhere in it. It is an adau 1701 chip ran by sigma studio (Dayton Audio DSP-K). It has everything.
The sub has 0 delay set in the dsp, but on smart it was like 38ms to "Track" it. The other speakers were around 3.5ms to "track" them and I added as much delay as I could without running out of RAM on the dsp chip (500 samples which I am not sure on how many ms that is, but Justin showed me how I can do that). I can't add that much delay.
I also need to do some more tests. This is where I am practicing my smaart skills as it is easier to set everything up in an office compared to my truck :).
Edit* just swtiched to ms instead of samples on the program and left is delayed 12.1042ms and right is delayed 11.6667 ms. Man sigma studio is precise.... I can change the time delay by .0001 ms.
how do you mean track it? You didn’t use different delay locator settings for both drivers or try and match them using the delay locator? It will lock on the peak of the high freqs impulse response, so it does work for similar drivers like mids and tweeters, but a sub vs main alignment I would t use it for... I’d swap the polarity of the sub using a fixed delay inserted and then use the same delay inserted to time the main speakers to it, also use a 2nd order all pass on the sub at 200ish hz with a high q to better align the phase through the crossover after that
it sounds like the mic wasn’t perfectly between the speakers of the delay is slightly different, I must admit if I’m doing a pair of speakers I will measure/move/measure/move the mic until the phase overlays with equal delay if I am setting them up for a centred listening position as a best listening position
Re: PE Cnote Bookshelf Speakers Build
I tracked/find delay via smaart with the far right speaker and used that number for the rest. I would add time or subtract time in smaart until I could get the best alignment. Then find the difference in the smaart time and appropriately put that into the dsp and check it to make sure.
I tried flipping polarity on the sub but will try it again. Right now it definitely sounds better but can be even better with some more knowledge, so I appreciate the help. If I can get these nailed, then using it on my vehicles will be easier.
I just can’t believe the sub is like 35ms delayed...I expect a little beyond the other, just didn’t expect that. The Dayton amp that is on it must have a lot of delay for some reason...
Re: PE Cnote Bookshelf Speakers Build
I would start by equing/adjusting the crossovers a little for the two sides to a correct acoustical response for 24db slopes linkwitz Riley with a 150hz crossover point, then see what the phase looks like
if you look at the origin of the subs and tops you can see they are out of phase
you don’t have 35msec of delay on the sub as such, it’s the phase characteristics of the driver/box
i suspect correcting the acoustic responses will make it look smoother, then inverting the phase will allow you to add some delay to the satellites, but take some off the delay finder delay to make it all level again in the display
4 Attachment(s)
Re: PE Cnote Bookshelf Speakers Build
Played more today with making sure EQ and the LR 24db slopes at 100hz were nailed prior to doing any phase stuff.
Attachment 15261
Attachment 15262
Attachment 15263
Attachment 15264
2 Attachment(s)
Re: PE Cnote Bookshelf Speakers Build
So flipped polarity and delayed the right and left by about 15.5 ms, which is about my max before my dsp runs out of RAM. The +/- 10db interaction between the sub and the speakers is about 75hz to about 130hz. That was my area that I was most interested in, but got it pretty dang close up until the sub coherence dropped drastically due to output.
@dumdum. thanks for the suggestion on flipping the polarity, this allowed for both slopes of the phase and the value of phase to be lined up pretty well!
This was also a great resource: https://www.merlijnvanveen.nl/en/stu...bsolute-method
Attachment 15265
This was the end results: NOTE THESE MEASUREMENTS WERE TAKEN AT ONE SINGLE POSITION WHERE AS THE ABOVE REW MEASUREMENTS WERE USING MOVING MIC METHOD (MMM). I also did not confirm my mic calibration file was loaded this time.
The red is no sub, the green is sub without global low shelf filter, and the blue is low shelf filter at 100hz.
Attachment 15267