XLR to Unbalanced RCA to Phoenix

[rplace]

1 hour ago, babadono said:

-20 to +20, what's 40 db amongst friends

 

Much louder overall and cleaner/clearer at moderate to high volumes. LEDs not moving much on Xilica now. Guessing this is more like what others see. I'll get some more ear time, watch them and report back with screen captures and pictures if I feel it is needed.

[rplace]

New day new questions. Chris has been super helpful giving me Xilica files to flatten out my speakers.  I've been taking the various readings from different days/files and laying them over each other seeing where the peaks and valleys move. What smoothing or other settings do you all use to make the curves look "reasonable" whatever that is? I've been reading some REW tutorials on the MiniDSP site and one of them suggests 1/6 smoothing. Any other suggestions?

[rplace]

22 hours ago, rplace said:

New day new questions. Chris has been super helpful giving me Xilica files to flatten out my speakers.  I've been taking the various readings from different days/files and laying them over each other seeing where the peaks and valleys move. What smoothing or other settings do you all use to make the curves look "reasonable" whatever that is? I've been reading some REW tutorials on the MiniDSP site and one of them suggests 1/6 smoothing. Any other suggestions?

 

I don't think I've ever quoted myself before. @babadono @Chris A @mark1101 @Coytee Any suggestions?

[Chris A]

I use psychoacoustic smoothing because it emulates how we hear more closely than the other smoothing techniques or degrees of smoothing.

 

It's important to note that REW does all of its PEQ optimizations within its EQ facility based on whatever smoothing that you choose.  Psychoacoustic smoothing usually makes the most difference at the lowest frequencies over the other smoothing choices.

 

Chris

[mark1101]

Honestly, I never tried to flatten the response like Chris.  He likes it totally flat.  I knock down any obvious peaks and then adjust further to a preferred sound.  Nothing overly complicated.

 

You may find it easier to use the graphical EQ to get the sound you want.  Initially, I set up my Jubs and Xilica with Roy's settings and never messed with it.

 

I have since completely revamped my Jubs and they are now 100% analog.  I now use ALK ESNs and an Ashly graphical EQ which lacks the precision of a processor.  However, I am really enjoying the presentation this way.  Started as an experiment but I don't think I will go back to digital.  My MCM is all digital so I have that setup for the times I want that type of sound.

 

Right now my Xilica is only used to delay and band pass my subs.

[babadono]

4 hours ago, rplace said:

 

I don't think I've ever quoted myself before. @babadono @Chris A @mark1101 @Coytee Any suggestions?

I'm sorry @rplace I'm just not there yet. Truthfully it will have to wait until fall or winter for me. There are just too many chores during summer here on the rancho.

[rplace]

8 minutes ago, babadono said:

I'm sorry @rplace I'm just not there yet. Truthfully it will have to wait until fall or winter for me. There are just too many chores during summer here on the rancho.

 

So you just let @Chris A do all the heavy lifting, load up your files and listen away   Sounds like Chris is the DSP 2013-2018 version of DeanG who was making passive networks for so many of us in ~2003-2009. Can't thank Chris enough.

 

@mark1101 Nice to have both options. So am I envisioning this correctly? Your EQ is similar or analogous to the old 80s box with all the vertical sliders all over the place for one bank each for L/R? You just move the EQ as the mood strikes you or you find a general set up you like and nudge it a bit one way or the other for types of music or individual recordings?

 

I'm starting to get the feeling there is no "right" answer for setting an active system up. True? Are we back to the age old saying "If it sounds good it is good"???

 

In my spare time I'm trying to look at the room measurement and the resultant PEQs I get from them and reverse engineer in my mind how to get from A to B with what Chris has given me. Do any/all of you think that is reasonable?  Every time I use this tutorial https://community.klipsch.com/index.php?/topic/167089-using-rew-to-find-parametric-equalizer-peq-settings/ to make some PEQs I get a message similar to below. Seems like I'm setting something up wrong. Can't quite figure out what yet.

 

[Chris A]

Sean Olive at JBL/Harman has patented a loudspeaker preference model method, and has described the results of extensive listening trials performed at their Northridge plant on different loudspeaker types (including JBL and other brands).  The results of using that model identified the following factors (taken from Toole's book, 1st ed.):

 

 

These factors correlate between loudspeakers that do well in listener tests and those that don't...to the importance levels shown in percentages in the pie chart.  

 

One of the advantages of the flat on-axis measure and bass response smoothness (and to some degree bass extension) is that they can largely be corrected using DSP crossovers, whereas the other measure can't be corrected: power response is what it is.  The flatter that you can make the on-axis response, the better they sound.  This is something that I've subjectively found to be accurate. 

 

There are of course other performance factors that are important (including but not limited to low compression and modulation distortion), but these factors are described in relative level of importance.

 

Chris

[Chris A]

2 hours ago, rplace said:

Seems like I'm setting something up wrong. Can't quite figure out what yet.

What that's saying is that you're trying to EQ the response lower than most of the response presently there.  Usually what is done is to EQ each channel/driver flat (i.e., closer to the existing median response level) then readjust their relative channel gains after the HF and LF channels are themselves flat.

 

One way to find a better median level to EQ toward is to select "Set Current Level" from the "Target Settings" tab under the EQ facility within REW.

[rplace]

3 minutes ago, Chris A said:

What that's saying is that you're trying to EQ the response lower than most of the response presently there.  Usually what is done is to EQ each channel/driver flat (i.e., closer to the existing median response level) then readjust their relative channel gains after the HF and LF channels are themselves flat.

 

Still trying to get my head around it, but I was visually trying to put the blue, target line between what I though was the average or middle of the highs peaks and valleys. Is there a smarter way to find where that target line should be 85.0 in my example above? Also am I reading your tutorial correctly that you do it separately for the highs and lows? At first I was trying to do 20-20K but in the example above I did 40-300Hz after reading your example a few times. Would that be appropriate given my HF and LF cross points are 200 and 300 I think?

[Chris A]

I added the last sentence, above, while you were typing.  I generally EQ the HF section separately from the LF section, then choose a new gain for either the HF or LF channel to achieve a flat response across the entire listening band.

 

One of the issues with starting with a chosen crossover frequency (like 250 Hz in your case) is that it isn't clear that is the right crossover point.  In your case, I'd store the settings that you have now, then move the crossover frequency up to 400 or 500 Hz, and remove all the PEQs, then start again with the LF and HF channels EQed separately.  That way, you'll generally wind up with significantly fewer PEQs. 

 

Note that all of your present output PEQs are used in the HF channel, and all but one PEQ is used in the input channels.  To start over again, you'll need to zero out the input PEQs, too. 

[babadono]

2 hours ago, rplace said:

So you just let @Chris A do all the heavy lifting, load up your files and listen away  

@rplace yes but I don't put a gun to his head He could tell me or any/all of us to go fly a kite.

[Chris A]

No sweat.  I don't generally bite...at least at first. 🍖

[mark1101]

6 hours ago, rplace said:

 

@mark1101 Nice to have both options. So am I envisioning this correctly? Your EQ is similar or analogous to the old 80s box with all the vertical sliders all over the place for one bank each for L/R? You just move the EQ as the mood strikes you or you find a general set up you like and nudge it a bit one way or the other for types of music or individual recordings?

 

 

Yes, sort of.  I have the EQ on the tape monitor loop of my preamp and basically deal with the 402/driver and room all at once.  Much simpler.  Less amps.  More LIFE in the music.  Noticeably NOT time aligned is the biggest difference.  IMHO the graphic EQ comes close enough that you would probably not be unsatisfied with tehe FR.   Its the time alignment that is most noticeable.

 

Bear in mind Roy tested this solution in front of many of us a few years ago vs. the digital solution in the Klipsch lab.  It works marvelously.  It would be good if Klipsch offered a universal passive for this solution.  Roy's passives have the EQ in the network.  Klipsch changes drivers like I change socks and the settings get improved from time to time.  Guess what?  Toss out those Klipsch networks.  They can't be upgraded along with the speakers as Klipsch makes their improvements.  ALK engineering solves that wonderfully.  Not cheap, but I would spend that money again every day of the week.

 

I don't mean to discount Chris's very critical adjust everything perfectly according to science.  I have that too.  It's good to have both.  

[rplace]

So I was messing around today when I was taking some measurements for @Chris A and decided to measure my other amps. I've been trying to dial in my 3.5 watt Wright SET amps, but today after making the 3.5 watt measurement (Green Below) I ran the same Xilica settings for my 60 watt VRD mono blocks (Red Below). Note that I turned the volume on the preamp down because the VRDs at 60Watts are so much louder than the Wrights no matter where the volume knob is. The LF for both sweeps were taking with the Crown driving the bass bins. No changes to the Crown's gain settings. No change to anything about the Crown.

 

To my untrained eye it looks to me like the peaks and valleys are very similar. I'm also guessing the Red line being below for some portion and above for the other is because of the volume turned down on the pre and no adjustment to the Crown. So I've got lower SPLs out of the crown for the LF portion of the red sweep and more SPLs out of the higher powered amps even thought the volume knob is turned down. Correct?

 

My questions are:

1. Do you active guys switch amps from time to time?
2. If you do, can you just adjust the gain by SPL/ear or is it necessary to re-EQ everything.

 

 

[mark1101]

I always set my amp gains with a measurement mic and PC analyzer software.  This includes the passive setup as I am using monoblocks with gain controls.