Hi Nvphotos, Very nice looking build ! great woodwork
When you get them tuned to that same degree of excellence, methinks you are going to be a happy camper
I very much agree with Wesayso...start with just examining SPL.
Try to get a sense of overall tonal balance, and big valleys and hill tops in response.
Move the mic around, and see if the frequency of sharp dips changes...this will help you determine where reflections are coming from. Either fix those dips with speaker placement, or room treatments etc, or ignore them.
Big valleys and hill tops that show up everywhere you measure... go ahead and try to EQ with PEQs. Stay broad.
Same with tonal balance...go ahead, perhaps using high and low shelving
I wouldn't even look at other measurements, unless they help you understand the SPL curve, until you have the SPL curve looking good enough to say, hmmm, what next? Good luck !
When you get them tuned to that same degree of excellence, methinks you are going to be a happy camper
I very much agree with Wesayso...start with just examining SPL.
Try to get a sense of overall tonal balance, and big valleys and hill tops in response.
Move the mic around, and see if the frequency of sharp dips changes...this will help you determine where reflections are coming from. Either fix those dips with speaker placement, or room treatments etc, or ignore them.
Big valleys and hill tops that show up everywhere you measure... go ahead and try to EQ with PEQs. Stay broad.
Same with tonal balance...go ahead, perhaps using high and low shelving
I wouldn't even look at other measurements, unless they help you understand the SPL curve, until you have the SPL curve looking good enough to say, hmmm, what next? Good luck !
Did a simple test comparing my old Magnavox CDB 582 cd player to the computer sound card with all EQ and effects turned off. The cd player is subjectively much better sounding. My guess is that I just have an awful sound card.
My long term plan is to get a miniDSP with Dirac. Hopefully that will solve my issues. The cd player sounds great but is not flat to the low end.
My other issue is that all my audio equipment is vintage. I am using an MCS 3125 at 125wpc to power these and a Fisher RS-180 at 170wpc on standby but I think the Fisher could use a recapping to sound it's best. I only listen at 65-75db. I was hoping that at that "low" volume I could get away with these receivers. Assuming a miniDSP solves the signal problem, are my vintage receivers holding me back? Both of them were monsters back in the day.
My long term plan is to get a miniDSP with Dirac. Hopefully that will solve my issues. The cd player sounds great but is not flat to the low end.
My other issue is that all my audio equipment is vintage. I am using an MCS 3125 at 125wpc to power these and a Fisher RS-180 at 170wpc on standby but I think the Fisher could use a recapping to sound it's best. I only listen at 65-75db. I was hoping that at that "low" volume I could get away with these receivers. Assuming a miniDSP solves the signal problem, are my vintage receivers holding me back? Both of them were monsters back in the day.
Don't worry... most of my listening was done with an ancient 100 watt Pioneer amp, I've used that one for about 3-4 years with the arrays.
I now have a 350 watt monster amp and while it does make a difference, I was quite satisfied before.
Any plans to get some basic curve correction going until you get your miniDSP? One of the biggest things to get right is having a good gain structure.
I now have a 350 watt monster amp and while it does make a difference, I was quite satisfied before.
Any plans to get some basic curve correction going until you get your miniDSP? One of the biggest things to get right is having a good gain structure.
Wesayso, I did work on speaker positioning today as that is something I'm familiar with. I ended up positioned almost exactly like yours. I worked with just the CD player rather than the computer as if sounds better. The bass is much improved and it is with no EQ or tone adjustments. I really want to see how this new position measures.
I plan to buy the miniDSP next month. Budgeting is no fun but works really well once you get into a groove
This whole project has been a huge learning experience for me being my first speaker build. In every step I have been learning as fast as I can to come to a decision. Point being, this signal processing is a whole new world to me and I am absorbing it as fast as I can. Wesayso, I have no idea what that means but I will get right on it!
Here is the SPL graph as well as the IR Filtered in dbFS and %. This is from a new measurement after moving them with left and right together...
Attachments
Not quite what I was hoping to see. First a question, why do we get such a strange ratio oh graph height to width? This way we still make huge jumps between the values at the left that represents the dB scale.
I'd expect to see a screen filling graph in REW. With about a 5 dB division between the numbers. Any reason for it being this wide compared to its height? Its kinda likecwatching this stuff on an ultra wide screen right now.
It seems we do see a very strong reflection. Usually we see a peak at zero.
After that peak there should be a drop, followed by the slope of arrivals. The bigger that drop, the less reflections we have (good!). Here we have reflections as high as the first peak. It doesn't help that the top end is down in SPL level, but your earlier graphs did show that peak, a healthy drop and that slope I mentioned.
(be sure to set the left boundary to - 0.001 s so we see that peak more clearly)
Edit: looking more closely (I'm on a phone) I do see it is at -0.001. However the main peak doesn't look centered at the 0 (zero) mark.
Now that makes me curious about what happens on the left. Setting it to -0.010 might show us that.
I'd expect to see a screen filling graph in REW. With about a 5 dB division between the numbers. Any reason for it being this wide compared to its height? Its kinda likecwatching this stuff on an ultra wide screen right now
.It seems we do see a very strong reflection. Usually we see a peak at zero.
After that peak there should be a drop, followed by the slope of arrivals. The bigger that drop, the less reflections we have (good!). Here we have reflections as high as the first peak. It doesn't help that the top end is down in SPL level, but your earlier graphs did show that peak, a healthy drop and that slope I mentioned.
(be sure to set the left boundary to - 0.001 s so we see that peak more clearly)
Edit: looking more closely (I'm on a phone) I do see it is at -0.001. However the main peak doesn't look centered at the 0 (zero) mark.
Now that makes me curious about what happens on the left. Setting it to -0.010 might show us that.
Last edited:
I'm sorry, I moved but the speakers and the microphone. The microphone was at the couch which is further back from the original measurements.
That tiny nub between the drivers could probably be left out without any serious consequence. It’s surely not adding anything to the stability of the cabinet. Depending on how tightly I try to squeeze my drivers together, I’ll either have a slightly wider gap, or remove that bit entirely, and maybe put some foam strip or a piece of rubber between each driver in its place.
I’ve got animals that will think of these as a scratching post, or my dear old dog will stumble into the drivers in a moment of excitement, so I plan to put some sort of covering over the speakers. The perforated plastic from the originals looks like it will work, and I think I’ve found a slightly more aesthetically pleasing way to attach it, though I’m still exploring those options.
I hope you figure out getting the EQ right on your speakers! I wouldn’t worry about a bigger amp. For quiekt listening, I’m guessing you would need a lot less power than you would think, even with lots of boost to bring the bass up. At rated sensitivity, you’re using 1w with no eq to get 83.5dB. Supposing you’re listening at 74.5 dB, you’re using half a watt for 80.5db, 1/4 watt for 77.5 dB, and 1/8w for 74.5 dB. If you need ~ 18dB correction to get the bass numbers close to flat down to 30hz, you’ll need a whopping 4W of power from your amp to get the bass up equal with the mid range. All of this is making me realize I’m going to need more than the 20W amp that I currently have to get any kind of volume out of my future towers while reaching low in the bass region. But that’s a problem to deal with after I build them
Thank you!
My end goal is to be able to listen from my couch 3/4th into the room (current measurements). My original measurements are from a listening chair 1/3rd into the room.
It looks like the room is the problem, which I can't address right away. Should I eq for the 1/3 seating position or the 3/4th position?
My end goal is to be able to listen from my couch 3/4th into the room (current measurements). My original measurements are from a listening chair 1/3rd into the room.
It looks like the room is the problem, which I can't address right away. Should I eq for the 1/3 seating position or the 3/4th position?
Nice build.
With localization (imaging), There is a distinct difference between the intensity cues falloff between a point source, line array, and planar wavefront. Planar really doesn't fall off much, a line array wavefront expands as 1/R, and a point sources falls as 1/R squared.
I would expect the imaging to be different between point and line as you discovered. I just have no idea what to do about it, you seemed to have found a solution that works for you.
Congrats
jn
Don't go for the closer listening position. A little distance can be a good thing with these arrays. If you can't eliminate the reflections, we can use some tools/options to go around that. I bet you're going to be thrilled once we add a little adjustment.
Unless your closer listening position is your favorite.
Last edited:
Can we tell from the impulse plot which of the two is the worst offender: first reflection or rear wall behind couch? There is half wall behind the couch forming a hallway to the front door.
It isn't always clear, as the IR does not show us the direction of the sound .
But that puzzle can be solved with a few measurements if needed.
First thing we should try is giving you a better tonal balance to listen to, for that we might need a few measurements around the (desired) listening position.
Measure in the center sweet spot (equal distance to both speakers), and a few spots around it, say each 10 cm apart doing 5 measurements total. Separately measured for both left and right speaker.
If you get 5x left and 5x right that way, spanning just over half a meter with the exact sweet spot being the central measurement, we'll average those for the left and the right and work on a balance.
.But that puzzle can be solved with a few measurements if needed.
First thing we should try is giving you a better tonal balance to listen to, for that we might need a few measurements around the (desired) listening position.
Measure in the center sweet spot (equal distance to both speakers), and a few spots around it, say each 10 cm apart doing 5 measurements total. Separately measured for both left and right speaker.
If you get 5x left and 5x right that way, spanning just over half a meter with the exact sweet spot being the central measurement, we'll average those for the left and the right and work on a balance.
Should these measurements be recorded as stereo l+r or l and r separate for each position.
In addition, should the microphone be pointed it with 90° off axis calibration or aimed straight towards the speakers?
L and R separately. I point it forward, slightly tilted up towards the center of the array. Keeping the mic at average ear height.