Measuring audio system and miniDSP's

[Rockyhill]

It will be interesting to check the response of the mains only and subs to truly gauge what’s going on. I’ll try all that’s being suggested after I do some basic initial tests.

As an aside I’m thinking that I will grab the miniDSP and try using it in a couple of different ways to see how it works out. Wondering if those in the know can confirm if what I’m thinking would work. Both would use the miniDSP 2/4 HD between the preamp and main amp jumpers.

1. Use the miniDSP to EQ correct the main amp based on the specific listening position. Continue to run all speakers full range and subs off speaker level inputs. Crude EQ’ing for all drivers essentially and use the crossover on the subs.
2. Use the miniDSP as an active crossover, 2 of the outputs running back in to the amp to run the mains (LS50’s with the option of a high pass filter) and the other 2 outputs to run an RCA cable directly to the individual subs (crossover on subs deactivated and low pass filtered).

I assume I have the use of the miniDSP right? I wasn’t sure why they offered a subwoofer version of the miniDSP and if I actually needed that instead. But seems like 4 outputs are more than enough for any needs.

The miniDSP can be used with a sub. See:

https://www.minidsp.com/applications/digital-crossovers/subwoofer-integration-with-minidsp

I am in the planning stags of a very similar implementation of the miniDSP 2x4 HD with a system in a non-optimal room setting (bedroom/office).

As you observed, the remote is highly desirable (and practically free).

 

[slobmw]

I’ll do that first for sure and test EQ on the current setup (mains and subs off same amp with speaker level connections). However I’d like to try the subs on RCA’s directly off the miniDSP and I don’t have any cables on hand.

I’ll grab some cheap ones to see if it’s worthwhile.

I have a minidsp nanoAVR HDA and a uMIK, and have been having a huge amount of fun playing with bass management. I'm still on the learning curve and hope to keep improving the sound. But have heard big improvements and it's great to be able to verify what has changed, and experiment in real time with REW and the DSP interface both up on my laptop.

 

[neevo]

Yeah I have to say it’s great fun and I love the idea of “knowing” what the changes are doing vs trying to guess with my ears.

 

[neevo]

Time alignment will most likely not cure frequency null problems caused by the room. It's the room, not the speakers.

Is that true Joe? I guessed that time alignment “might” have an impact on where the null point might be, and therefore could impact nulls placement in a room (and in particular at a listening location), although I don’t suspect it would go away, just move the problem (hopefully away from the ideal spot)?

 

[JoeESP9]

Nulls are usually in the lower frequencies. Although time alignment will make the sound more coherent it's the room and speaker positions that have the most effect on nulls. If you're using a sub(s) try the subwoofer crawl. Quite often the best place for a sub is not near the mains. A variable phase control, not just a switch can be quite efficacious when the sub is a distance from the mains. REW, a DSP and an appropriate mic make setting things up a science instead of a guessing game. A couple of bass traps and some absorption at the first reflection points also IMO significantly improve in room sound.

 

[Drugolf]

I have a 2x4 and Umik sitting here to employ some day too so this is interesting.
Would there be any value in starting the measuring with more of a baseline measurement? Remove the furniture from the room except maybe one listening chair, set the system up in a more logical straight forward triangle configuration and take measurements. Would this give you a better idea of the "possibilities" and "capabilities" of the gear alone w/o the impact of the furniture and room configuration? This could be what you know is possible. I would probably even do it a couple times with the gear on different sides of the room to see if there is one that is just flat out bad or superior for whatever reason.

As an architect I would say your challenge with that particular room is the 3 chairs and yet still achieving some sort of balanced configuration for a music listening position. I really like the look of the room (although i would professionally suggest a different 3rd chair ) These things are always a challenge with the wife if she is fixated as so many are strictly on "presentation" of a room that is rarely ever even used. If this room is going to be used more frequently for music listening and enjoyment she needs to be convinced that it's arrangement should be set-up accordingly. They don't always believe it, but it can be done very nicely. And remember, things can be moved around a bit for different purposes. The LS50's are on stands that could be shifted a bit if you know you are going to have company and the lounge will be used for sitting in. Those chairs could move a bit too and one could even find it's way between 2 different rooms easily enough depending on what is going on.

 

[slobmw]

I have a 2x4 and Umik sitting here to employ some day too so this is interesting.
Would there be any value in starting the measuring with more of a baseline measurement? Remove the furniture from the room except maybe one listening chair, set the system up in a more logical straight forward triangle configuration and take measurements. Would this give you a better idea of the "possibilities" and "capabilities" of the gear alone w/o the impact of the furniture and room configuration?

Now we are getting our heads firmly wedged in the rabbit hole! (Hmmm,the metaphor is somehow a bit off...) Joking aside, this is a great idea if feasible. Though sometimes ignorance is bliss. I have only measured my family room in that way, and introduced other furniture piece by piece, and not my living room. I like the way my living room is arranged for general purposes, and don't want that nagging thought of "if we just got rid of the couch..."

 

[neevo]

Pulled chair 3 out in the critical listening position today and managed to get some more test done. I also managed to adapt my camera tripod to take the umik mic stand so that I can have a very accurate and stable platform to direct the mic around:

Bought a bolt and trimmed it down:



Took my camera quick connect apart to attach the mic holder:



All done. Nice and sturdy:



My camera tripod is the perfect height to replicate ear height:



Mic attached:



To the tests. I'm actually surprised how little the test varied from the position 1 tests, its really not night and day and the hole I "thought: was around sub 80Hz really isn't there... the 120-150 node is still though and is obviously not really position dependent but room dependent by the looks of it.

Mains only:


My observations:
1. Another amazingly flat response from the LS50's above 200Hz and won't need much to get it dead flat I reckon
2. 150Hz room node is going to have to have some treatment (bass traps or maybe playing around with the sub placement)
3. LS50's continue to dig really deep

Subs only:


My observations:
1. Big peak at 40Hz is probably making the subs pretty ineffective as that's controlling the gain position and therefore making the balance of the range much less impactful, however its probably helping to create a pseudo "house curve"
2. Looks like the sub drops off a fair bit from 100Hz but then flat, then again just over 200Hz

Full System:


My observations:
1. Once all integrated it doesn't look too bad. Nothing I can do to EQ out the room node at 150Hz but I think plenty can be done elsewhere to improve things.

I also had a go at EQing this full range result to see what I "might" be able to do once my miniDSP gets here:



This looks pretty respectable to me and I'm keen to see what results I can get once the DSP gets here. Obviously at the moment I'm measuring the system in its entirety which will not be the process once the DSP gets here. Then my plan will be to test each speaker individually (and maybe a few tests to average out the responses) and then EQ each speaker individually.

I've also found references to Linkwitz Transforms for subs. Both of mine are sealed... is this something worth looking in t

 

[slobmw]

... I also managed to adapt my camera tripod to take the umik mic stand so that I can have a very accurate and stable platform to direct the mic around:

Bought a bolt and trimmed it down:

All done. Nice and sturdy:

Very nice!

A few thoughts on your measurements, that I would welcome any corrections to by others. Looking at your graphs, it looks to me like you have a series of modes as multiples of roughly 40 (which shows up at 80, 120 and 160) This doesn't seem to line up exactly with your room dimensions, but the irregular openings and what not effect that.

Your experience of these nodes (you note the 80 Hz in particular) is position dependent. The dips and peaks _will_ exist in an untreated room, but they move around as mic and/or source move. For a fun exercise, put REW into RealTime Analysis (RTA) and walk around the room with the mic while playing a sine wave at, say 80Hz. You will likely hear, and see on the graph some big dips and peaks. If you want to go really nuts, you can try to map the room's response either relative to your preferred listening position, or your preferred speaker/sub placements.

In terms of what you can smooth in eq, versus what you'll need to do with positioning and room treatment, you'll want to take a look at phase and your group delay graph. I am in no way qualified to say much about that, except that it should help you identify what you is possible/relatively straightforward to eq and what is not. See the link:

https://www.roomeqwizard.com/help/help_en-GB/html/minimumphase.html

Thanks for documenting your progress.

 

[Drugolf]

Nice.

 

[neevo]

@slobmw good pickup on the consistent dips, I hadn’t noticed that at all. I tried to read that article but it’s going to need a few reads before I even have half a chance of knowing what it’s saying.

I will do the real time analysis and see what that does for SPL around the 60, 80, 120, 160 regions.

Does REW have a sine wave generator in it or will I need a stand alone option for that?

I haven’t really dipped any toes in to time delay yet, I was going to have a crack at simple EQ first and see how that worked. Given I’ll have a DSP though, if I can really unlock the systems potential by delving deeper, I’m game for that.

At the moment my understanding level is pretty limited though (EQ only really), so a lot more reading to even start to get my head around it.

 

[neevo]

Well I have to say I am really impressed with my miniDSP stuff. I haven't had the chance to test it in anger as its late, so tomorrow will let me open up the amp volume a bit and really see what the seat of the pants experience is like.

Took a bit of playing around as the DSP offers so much flexibility in EQ maps, crossovers etc but I used 3 of the 4 EQ setup slots:

1. Master map for left and right channels individually. However EQ was done across the mains and subs together
2. EQ for all 4 speakers individually (mains L&R and subs L&R), however I don't have the cables to connect the subs yet so that one will have to wait
3. No EQ to see the difference

I think ultimately I will run the mains off the amplifier and the subs off the DSP as not only will this allow me to EQ map each speaker individually, it will also allow me to low cut the mains and high cut the subs. I like the bass from the LS50's and like the option of more bass from speakers, however cutting the LS50's below 50-70 hz will mean I get more amp headroom and will also ensure they are protected from clipping with loud volume a little more. Will be interested to see how that sounds!

So how did I finish up tonight?

Setup 3 (no EQ):


Setup 1 (EQ for L&R individually):


That's all well and good with pretty lines... but how does it sound?

There's a noticeable difference; the bass is richer, warmer and it feels like the subs are doing a lot more than they used to (obviously dialling back those big peaks means I'm getting much more useable linear bass). I'm getting that oomph that I wasn't there before and its tighter and cleaner too. I EQ'd my L&R channels to be flat but then also put a 8dB boost (shelving filter) on the bass as a "house curve". Originally I had the treble dialled down a bit too, but prefer the sound of crisp treble so I took that off in the end.

This is all looking very promising and I think I will find some RCA cables for the subs before I go any further as I see lots of benefits in EQ all 4 speakers individually. Once I've got that done I will proceed with EQing position 2... then sit back and enjoy I suppose?

 

[mkane]

I to also have a 2x4HD and have many days invested. How do you decipher the measurements?

 

[neevo]

I to also have a 2x4HD and have many days invested. How do you decipher the measurements?

I’m probably not the person to ask, as I only really understand the dB and frequency sweep tab and that’s all I’ve focussed on so far.

Essentially with the sine wave sweep and in a perfect (not real world) scenario the line should be flat. So any peaks and troughs are issues that need addressing. Peaks you can EQ down but troughs are much harder as you cannot really EQ them up, so I left them alone and only really applied mild EQ.

There are a few great videos on YouTube that I recommend. Explaining the components:

How to use it all.

The kid is very knowledgeable!!!

I haven’t even gotten in to decays of the frequencies yet and to be honest don’t understand how to use that function yet. So I’m learning a lot myself still.

 

[slobmw]

...This is all looking very promising and I think I will find some RCA cables for the subs before I go any further as I see lots of benefits in EQ all 4 speakers individually. Once I've got that done I will proceed with EQing position 2... then sit back and enjoy I suppose?

Yep! Mostly. My experience has been that as I learn more I tweak based on what I think I'm learning about measurement and dsp. But so far, the obvious improvements from dsp came from the work you've outlined: taming as much of the bass response of the sub (your's looked as peaky as mine did) as I could, and then dialing in the crossover region.

 

[slobmw]

I to also have a 2x4HD and have many days invested. How do you decipher the measurements?

The videos that neevo linked to are good. I found the REW tutorial on this page to be really helpful:

https://www.minidsp.com/support/community-powered-tutorials/295-minidsp-tutorials-by-austinjerry

The help files that come with REW are also useful, though they sometimes left me with more questions than when I started.

 

[neevo]

Well today I got in to the world of seeing what a difference running the subs off the DSP would make. Its laborious, time consuming and when I had it all done, I hate to say it but I wasn't convinced that it was that much better than doing the mains and subs together on each channel. Even the chart didn't look that much better on the screen.

I measured all the speakers individually, mains with the mic pointed straight at them and the subs with the mic pointing inbetween the mains. Once I had all the readings (still only did one reading vs a grid of readings), I did one of the total sound to see what the difference was.

Before:


After (I needed another 5dB of boost on the subs to make it sound nice):


So a lot of work for not much improvement... or so I thought!

Have spun some Pink Floyd's Division Bell and also now on some Supertramp. The sound is another notch up from the initial treatment. The chart doesn't reveal how tight the overall sound is, especially from the subs. There's genuine smooth punching bass that wasn't there before. I compared it to the original setups and its lifting the muddiness off the sound (and to be fair it was already bloody good) and the punch from kick drums really hits you.

I haven't put any filters on the outputs because I couldn't really get them to work properly, so I've resorted to using the filters on the subs at 60Hz and have the mains running full range (I may put a low filter on them around 50hz). Very impressed so far.

I even had a crack at delay on the 4 outputs to really lock in the timing, however couldn't for the life of me work out how to do it. Heres the distance of the speakers from the listening position:

Main L 362 cm
Main R 327 cm
Sub L 430 cm
Sub R 132 cm

Whats next?

1. I want to see if I can put the DSP on a switched socket on the back on the amp so it turns on and off with the amp.
2. I want to get back in a do some more measurements and really have a grid of measurements for each speaker that I'll average out.
3. I want some nicer looking cables for the subs. So far I have some very cheap black RCA cable that I split in two just to test the concept.
4. I will need to repeat the process for position 2

 

[neevo]

I have a couple of hours up my sleeve with an empty house. So I’m going to have a crack at doing both setups properly.

1. Measuring for time aligning
2. Baseline test for each position
3. 9 point measurement for each speaker in each position
4. Average out response for each speaker
5. Create EQ
6. Load EQ’s
7. Test results

Hoping to have the 2 configs setup and ready to go.

 

[mkane]

get with it man

 

[neevo]

I’m all done for the moment! What a learning experience it’s been and I have to say the end results blow the untreated sound out of the water. I have a configuration that’s unused and allows me to check what an untouched audio response sounds like and after listening to an adjusted one, it’s hard to think how I was happy with the sound before

What have I learnt?

Doing an array of readings is significantly better than trying to do it in 1. I took 9 readings for each speaker in position 2 (critical listening) and averaging out the responses yielded a significantly nicer sounding result even though it didn’t look dramatically different on the graph. Goes to show how subtle great improvements can be in the real world.

Position 1 is really sounding good, I’ve time aligned all the speakers and managed to refine the gain in the left channel a little so I get a better placement of the middle of the mix and not have to play with the balance on the amp. It’s always been a great sounding spot and it’s no different now, just the bass is tighter, the mid range and treble is refined and the mix is refined for an off centre listening position. Still need the laptop out to refine the sub gain a little on a few different tracks.

Position 2 is the one that’s changed the most. Not only do I now have a really even bass but doing the time alignment, EQing each speakers and adding some house EQ has made this easily the best seat in the house. It’s also had a minor amount of gain adjustment so I can put the chair in the middle of the room (align it with the fireplace) and the mix is perfectly centred. The LS50’s really shine here and the width and depth of instruments is a little eerie as you have no chance of picking where the speakers are at all. Dire Straits You and your Friend sounds incredible and the guitars seem like they’re playing right in the room behind the speakers.

Is all this crap worth it? My wife would say definitely not. The amount of sine waves that have been going through the house at 90+ dB has been crazy, but the end result is night and day difference and I’ve really enjoyed the process.

I can’t imagine any other audio addition that would be able to transform a system as much as this has for approx $300 too.

If you have an amp with pre outs/main ins I cannot recommend the UMIK-1 and 2/4 HD miniDSP enough

Lastly I’ve found to really extract that last bit of goodness and subtlety out of the setup, there is no substitute for sitting down and listening to music and tweaking on the fly. The great thing about the 2/4 HD plugin is that you can tweak and hear the changes immediately. Even changing configurations it lowers and raises the volume in between the switch to stop any issues with the change. Very slick.

Using your ears to refine the setup is a great final step though, to really dial it in to your specific preferences.