Question about dual sub, rew and allignment

[Pieter1267]

The darker is A+B. Almost the same.

sub2 with inverted measure

The darker line is with very little peq. The brighter is the autorew.

Both measured individually. They actually look very flat already, with 2 nulls from sub2. I cross at 80(edit: also tried some different configs here, now one preset at 65 and another at 94 in MINIDSP). Decided to change just a little bit and it sounds nice. That said so many tracks have so varied productions, so one track the bass sounds great where the other it sounds too loud.
Tomorrow another day for fiddling.

Edit @sigbergaudio indeed, lots to learn on rew too. I discovered rta, wich helped immensely. I keep the manual peq since the autorew file boosted the nulls too much. I now lowered the 45-62hz -4db at 4q on both subs a bit and sounds much less bloated. And then fiddled some more. I took out some dbs all across 44-94 and now it sounds better. Cant measure yet. Tomorrow. I do love how much space for improvement there is. 2 subs are fantastic! But you deffinitly have to love to dig into REW for sure. Luckily I am a proud nerd, just like all of us on this here forum.

 

[sigbergaudio]

Your measurement results where the two subs together are LESS even and also lower in SPL looks pretty weird. May I ask how this is connected and what you do to measure individually as well as together?

 

[Sokel]

You must also adjust Y axis to 50dB range as this is the way we usually see charts.

 

[Pieter1267]

Your measurement results where the two subs together are LESS even and also lower in SPL looks pretty weird. May I ask how this is connected and what you do to measure individually as well as together?

It really is weird, isnt it? Hope im just making a silly config mistake somewhere.

Both subs are DIY cabinets Dayton 15" ultimax, one mki the other mkii.

They are in opposite corners of rectangular room. 3.4x4 meter. I have a rug on the floor, covering 70%. Back wall, sidewall and ceiling have several acousting damping plates. Ceiling is rather low and together with one wall, made of plaster. The room still has some rather present echo/reverb. Ill ask my wife to put some curtains on the walls, hope to dampen more.

I suspect the problem might be there. Partly.

I Have dragonfly usb stick as pc output, the source. That goes into Quad34. Out that pre amp goes into minidsp ddrc24.

Minidsp is fully reset, no dirac. I put input left and right in minidsp output 3. I do the same in 4. (Should i maybe make a change there, only left on 3, right at 4?
There it splits to quad606mki for main, and inuke6000dsp for subs.

The inuke is set at dual mono, no filter or dec. Limiter but very high. The idea is both subs produce bass from both left and right inputchannels.

Umik1 at 90deg in a good mic stand (as per keith article). I measured at listening pos but since the middle of room is close, as of today right in the middle of the room and even distance on each sub. As the individual graphs show, both subs work well, very similar and since i removed the big chair, almost no nulls.

I measure in rew, first test output and aim for 75db with that test sub noise rumble. I then set rew for 0 to 200 sweep.

Then open minidsp, remove any peq or xover, mute one and measure the other. Then vice versa. Then both. Immeadiatly, both, audible less output. And the graph shows a much curvier line.

@Sokel Ill set the graph at 50db tomorrow. Ill also post measurements from one position as above shows 2 different positions and makes a bit difficult reading. Thanks and apologies.

 

[Flaesh]

Ill ask my wife to put some curtains on the walls

This will not affect the summation of the subwoofers. Can you upload the .mdat file?

 

[jeffaegrim]

What was the issue with just EQ-ing the response of both subs playing together? That‘s what I have been doing for a long time now. Could never get a good result with EQ filters for individual subs. Admittedly, my room is very asymmetrical and so is the positioning of both subs.

 

[Pieter1267]

This will not affect the summation of the subwoofers. Can you upload the .mdat file?

You mean here? Atm the yesterday and previous look messy bcs its not the same measuring position.

This afternoon ill make new measurements from one position and when im done, upload the .mdat.

@jeffaegrim Both subs using the same peq showed even more weird curves the first day i started testing. Yesterday, however, i ran RTA and adjusted one peq doing both subs until the spectrum looked more equal. I saved some configs and havent measured them yet.

I have quite a few more things to try. I think im also going to change the subwoofer amp to even less dsp stuff. Switching to stereo, making sure slope is off, etc. Etc. Ill figure it out, also thanks to you guys. And im learning tons about REW. Did a lot of reading about MSO.

 

[goat76]

The inuke is set at dual mono, no filter or dec. Limiter but very high. The idea is both subs produce bass from both left and right inputchannels.

I would definitely go for a stereo configuration with two subwoofers. It will not solve the overall loss of energy in the bass when both subs are playing so you still need to find another solution far that, but it may solve the problem you experience with too much or too little bass from track to track, as stereo bass content will not double up as it will otherwise do when both subwoofers receive everything equally.

Before I bought my second subwoofer, I constantly felt the need to change the level on the single subwoofer as it didn’t sound correctly balanced from album to album. That is completely gone now with two subwoofers set up in stereo, and I think the reason for that is that I now receive the intended level of bass.

 

[sigbergaudio]

It really is weird, isnt it? Hope im just making a silly config mistake somewhere.

Both subs are DIY cabinets Dayton 15" ultimax, one mki the other mkii.

They are in opposite corners of rectangular room. 3.4x4 meter. I have a rug on the floor, covering 70%. Back wall, sidewall and ceiling have several acousting damping plates. Ceiling is rather low and together with one wall, made of plaster. The room still has some rather present echo/reverb. Ill ask my wife to put some curtains on the walls, hope to dampen more.

I suspect the problem might be there. Partly.

I Have dragonfly usb stick as pc output, the source. That goes into Quad34. Out that pre amp goes into minidsp ddrc24.

Minidsp is fully reset, no dirac. I put input left and right in minidsp output 3. I do the same in 4. (Should i maybe make a change there, only left on 3, right at 4?
There it splits to quad606mki for main, and inuke6000dsp for subs.

The inuke is set at dual mono, no filter or dec. Limiter but very high. The idea is both subs produce bass from both left and right inputchannels.

Umik1 at 90deg in a good mic stand (as per keith article). I measured at listening pos but since the middle of room is close, as of today right in the middle of the room and even distance on each sub. As the individual graphs show, both subs work well, very similar and since i removed the big chair, almost no nulls.

I measure in rew, first test output and aim for 75db with that test sub noise rumble. I then set rew for 0 to 200 sweep.

Then open minidsp, remove any peq or xover, mute one and measure the other. Then vice versa. Then both. Immeadiatly, both, audible less output. And the graph shows a much curvier line.

@Sokel Ill set the graph at 50db tomorrow. Ill also post measurements from one position as above shows 2 different positions and makes a bit difficult reading. Thanks and apologies.

The rugs will do nothing in the frequency range of the subs.

"I put input left and right in minidsp output 3. I do the same in 4. (Should i maybe make a change there, only left on 3, right at 4?"

"There it splits to quad606mki for main, and inuke6000dsp for subs"

Both the above statements are unclear.

Does this mean you are sending a summed mono signal to each sub? In that case, yes, I would perhaps test changing this to just left=3 and right=4, assuming 3 goes to sub#1 and 4 sub#2?

Then do a measurement of L, then R and then L+R in REW and see what happens. Perhaps also disable any EQ before doing this, so we see a completely basic, raw response from each sub + together before going further.

 

[Pieter1267]

You must also adjust Y axis to 50dB range as this is the way we usually see charts.

maybe a stupid question but how and where do I set this? You mean top to bottom a 50db range?

 

[sigbergaudio]

EDIT: This should typically say 45 and 105, not 15 and 105 as in the screenshot. So 60dB range. This is the most common way to present these graphs.

 

[Sokel]

maybe a stupid question but how and where do I set this? You mean top to bottom a 50db range?

You go here:

and you set 100dB for the upper limit and 50dB for the low one.

 

[Pieter1267]

I started the day by fixing the measuring position. See pic.

I then openend REW and made some testings with the generator and the SPL gauge. This worked as expected: both subs measured above 76db, and single output measured about 2 db per sub lower. Callibrated, and set. Figured out the Y axis settings range. Ready. Lets go!

I then did some measurements of the on the fly PEQ adjustements whilst running RTA and playing songs (soft and waysoft, rew65) while running both subs. Wasnt looking that great.

Then measured the subs with those peq settings and again, higher.

So I decided to do more basics. I turned off MINDSP completly. No change. I changed input in MINIDSP from both left and right per sub, to only left input on left sub, and right input on right sub. This gave some lower SPL but still, both went lower than individually.

I then changed the subwoofer amplifier, the Inuke. Put it in stereo mode, removed any xover, any filter, any limiter. Again no change after measuring: both subs lower spl than individually. Put MINIDSP also completly in basisc, no PEQ but xover to 120 on both channels.

Then I decided to test on the main speakers. Interesting as there the same happened: both have lower SPL as individually. See above graph. It kind of rules out a lot of things like subs and amp, wich im glad but what can it be?

Then also turned off xover MINIDSP and any filter i could imagine and let REW measure a last time both subs and individually: no change.

I then went back to the REW generator and the SPL gauge. There too, both subs had less SPL than individually.

I did notice the read out of the SPL gauge was a bit dodgy at times but since I now tested for 3 days and constantly see both speakers produce lower SPL than individually, I dont think there is something wrong with REW or UMIK1. It is strange that when I started the day, the callibration test, it all worked as expected, but aftera few measurements suddenly that too did not go back to how you would expect it to be. The generator and the SPL gauge measured less SPL when both subs were playing than individually. That is odd, right.


I can actually only think of 2 things: either MINIDSP is faulty, or some windows settings or maybe even motherboard is doing some kind of automatic enhancement or fiddling that I didnt ask for. What do you guys think? Going to take a small break now and continue this afternoon. Uploaded the .mdat file. Some pics and above graph is the main speakers, the purple line being both playing.

 

[sigbergaudio]

@Pieter1267 Is it possible to bypass MiniDSP completely? Put cable directly into both subs? Based on your graph, one would expect the green curve to be both playing at once.

And what on earth is happening with the blue curve from 1-1.5khz?

 

[sigbergaudio]

For reference, here is what you'd expect to see. The two green ones are left and right, the red one is a stereo measurement.

 

[Pieter1267]

@Pieter1267 Is it possible to bypass MiniDSP completely? Put cable directly into both subs? Based on your graph, one would expect the green curve to be both playing at once.

And what on earth is happening with the blue curve from 1-1.5khz?

I ran a measurement without the minidsp. No difference. The left woofer especially, runs about 2 db more alone when i test the generator. When i run the measurement, this is the graph. The green line is both.

I started out with the main speakers, though dsp, and then the measurement was as it supposed to be. The red line is both main speakers.

The mdat file. Im kind of lost guys. How can I fix this?

the attach file button also gives an eror message when i want to upload the .mdat file

 

[sigbergaudio]

I ran a measurement without the minidsp. No difference. The left woofer especially, runs about 2 db more alone when i test the generator. When i run the measurement, this is the graph. The green line is both.

Then I think we're back to suspecting that one of your subwoofers is wired wrong. (negative and positive terminal switched).

 

[Flaesh]

eror message when i want to upload the .mdat file

just zip the file

 

[Pieter1267]

just zip the file

with what? winzip demands paid subscription.

just zip the file

 

[Pieter1267]

Then I think we're back to suspecting that one of your subwoofers is wired wrong. (negative and positive terminal switched).

The one that puts out 2db more? Also: On suggestion I tested and inverted polarity. It didnt make any change. I mean, if they are wrongly connected, shouldnt clicking the inverted option in minidsp immediatly show results?