Time for treatment

#31

You could try flipping the LF switch to "deep" (which is how I have the ones set that I use for live sound), and see how that goes. The problem then is that it won't give flat response, but to be honest it's more useful to get good bass out of them, rather than accuracy, for a live room.

Darn. I will do a few tests with the LF on deep and see what we get. I am hoping the other tests give you enough information because it took a while to test for all those instruments and it would be a bummer to have to do it all over again.

That's just a gain control on the back of the K12: it cuts or boosts all frequencies equally. Did you re-calibrate REW after you adjusted that?

Yes I did.

For some of the tests, do set up the K12 where you intend to have your PA speakers, and in that case set up the mic where you expect the keyboardist and the guitarist to be when they are playing. But also set the speaker up in various other locations, where you expect to have instruments while tracking.

I did exactly that!!!!

I'm downloading those, and I'll take a look tomorrow.... That's a lot of data!

Thanks Stuart. I know it's a lot.

#32

So, I took a quick look at the data:

The first thing that jumps out is that there are two sets of tests, done at very different levels. The ones done on August 1 are all at the correct level: around 85 dBC, but the ones done the next day are about 30 dB lower! Around 55 dB. You most definitely would have noticed such a large difference in level if that were a speaker setting, so I'm assuming that this is just a soundcard calibration issue: you didn't re-calibrate the levels on August 2nd. If that's the case, then I can just boost all of those measurements by 30 dB within REW... but that is only valid if this is just a REW calibration issue. I can't do that if the levels in the room were actually lower.

Next up: there's a lot f useful data in there! It's showing a general set of problems in the region 80 to 300 Hz, roughly (depends on locations in the room, to a certain extent), so that's where some of the treatment should target. Which is good news, since it's not that hard to treat in that range.

There's also a noticeable dip in the response at around 2 kHz (covers 1k to 4k, roughly), which I'm suspecting might be the speaker itself: it might naturally have a dip in that region, which doesn't matter too much for FOH use. Maybe you could do one test with your speaker and mic outdoors, as far as you can get it from any buildings, walls, cars, or other large surface, so we can see what the actual free-field response of that speaker really is. I suspect it isn't very flat! And that's fine.. as long as we know about it, and in what part of the spectrum it isn't flat.

The corner test also cam through:

Even though the speaker doesn't go down low enough to provide good detail, at least we can see the 70 Hz issue. I suspect that it's a modal thing, but not properly triggered because of the lack of speaker power down there... however, 70 Hz is firmly in the territory of kicks, bass, and floor toms, so it does need some treatment. But it would be good to get more data, going even lower down the spectrum. There's the pressure zone down there, possible structural resonances, and doubtless more modal stuff. There's a hint of something at around 45 Hz, but very unclear, and almost certainly stuff going on even lower that would need dealing with. I think there's a slight trace of something around 36 Hz, but that's already off the bottom of the spectrum that the speaker is producing... but still within the spectrum of bass guitars, keyboards, and a couple of other things.

Do you have a good, powerful sub you could use to do that, along with the K12? Something that gets down below 30 Hz? Maybe something you could lborrow form a friend, if you don't have one yourself? You don't need to re-do all of the tests: just a couple of key points, such as the corner-to-corner one, the drum overhead one, and "guitarist PA-R".

But anyway, there's enough data there now to get a reasonable picture of what the room is doing, and start putting together a treatment plan....(except for the deep end).

- Stuart -

#33

Hi Stuart,

Much thanks for checking these out.

I'm assuming that this is just a soundcard calibration issue: you didn't re-calibrate the levels on August 2nd. If that's the case, then I can just boost all of those measurements by 30 dB withing REW...

I did not recalibrate levels on Aug. 2nd. and the speaker settings were left the same as Aug. 1st. So I think adding 30dB would work. I mistakenly thought the calibration would remain at same level. My bad.

It's showing a general set of problems in the region 80 to 300 Hz, roughly (depends on locations in the room, to a certain extent), so that's where some of the treatment should target. Which is good news, since it's not that hard to treat in that range.

Ok. Great. Thanks.

Maybe you could do one test with your speaker and mic outdoors, as far as you can get it from any buildings, walls, cars, or other large surface, so we can see what the actual free-field response of that speaker really is. I suspect it isn't very flat! And that's fine.. as long as we know about it, and in what part of the spectrum it isn't flat.

Ok. I can try that tomorrow.

Do you have a good, powerful sub you could use to do that, along with the K12? Something that gets down below 30 Hz? Maybe something you could lborrow form a friend, if you don't have one yourself?

I don't have a powerful sub myself. The K-12's are usually my monitor speakers. My bandmate has Klipsch passive speakers and an amp, which we use for our rehearsals. He is gone until this Sunday. I don't know how low those speakers go, but I will get the model number later today and search for specs. I also own a pair of passive Fostex Reference Monitors RM900 (late 80's) coaxial bass reflex, 8" woofer, 2" tweeter speakers and a 300 watt amp. I haven't used them in a while but they have been stored well and probably work fine. I couldn't find any specs on those online.

You don't need to re-do all of the tests: just a couple of key points, such as the corner-to-corner one, the drum overhead one, and "guitarist PA-R".

Great. That is very doable. I will wait to see if I can get some better speakers. If not I will use the k-12 on the deep setting.

#34

If not I will use the k-12 on the deep setting.

Great! When you do the outdoor testing, do one test with the "Flat" response, and one with the "Deep". Also do one test with the "Vocal Boost" switch up, to see if that might do something for the 1 kHz dip.

For those outdoor tests, set the mic about 3 or 4 feet away from the speaker, and roughly on-axis. Try to get them both as high above the ground as you can, to avoid ground-plane issues. Best of all is to have the speaker laying on its back, facing the sky, with the mic suspended about 3 to 4 feet above it... but don't forget that the cooling fan is on the back, along with the cooling vents! So don't have it resting directly in the ground like that: maybe support it on bricks, or some such, to allow for air circulation.

(I said "best of all", but this is really "close second best". True "best of all" is to have the speaker set into the ground, with the front face flush with the ground... but that's pretty hard to do, and not necessary!)

- Stuart -

#35

Hi Stuart,

Here are four tests of my K-12 speaker outside lying down on some 2x4's facing the sky.

Test 1 is the settings I used previously: LF- Norm; HF- Flat;
Test 2 is LF- Norm; HF- Vocal Boost;
Test 3 is LF- Deep; HF- Flat;
Test 4 is LF- Deep; HF- Vocal Boost;

Howie

Outside Speaker Tests.mdat: (6.36 MiB) Downloaded 1215 times

Outside Speaker Tests.mdat: (6.36 MiB) Downloaded 1215 times

#36

Cool! Very interesting, actually. Here's what you are getting with your K12s....

First, I normalized the frequency response curves as close as I could get, so they are all at the same relative levels, then I looked at them in pairs, to see what each control is doing. First, the "Vocal Boost" control:

Red is "flat" response, green is "Vocal boost". As you can see, the "flat" setting is really more like a vocal CUT! And with it set to "boost", that's actually flatter. :shock:

Go figure...

As I suspected, that returns some of the missing upper mids. So I would keep that "vocal boost" switch on for now, for all future measurements. It gives you flatter response, with a roughly 3 dB increase, centered around 2.5 kHz, and covering a range of about 4 octaves. That solves one mystery: the missing mids are not a room problem, but rather an EQ problem.

The other interesting thing you can see there is that, even with the LF set to "flat", there is actually a pretty big LF boost! It's roughly 4 dB boost at 71 Hz, with fairly tight EQ. That's probably the bass reflex ports, and there is some low-level ringing going on there, so that makes sense. You can see that here too, in the waterfall plot, below...

Notice all the low-level "muck" between about 50 Hz and 100 Hz: that's typical of bass reflex ports: in order to do their job, they have to resonate... but sometimes they resonate too much:

That's WITHOUT the bass boost on. When you apply the bass boost, this is what you get:

A massive boost around 68 Hz, with an even more massive increase in ringing, in the reflex ports. This amount of air "chuffing" in the ports also introduces some distortion. Here's the "bass flat" and "bass boost" THD graphs:

So, what does all that say?

1) Only use these with the vocal boost ON, to get the mid range flatter.

2) These speakers don't go down low enough to properly trigger the room. Here's the final comparison,

Green is "bass flat", orange is "bass boost". The cursor marks the point where the FR crosses over the average SPL level, at 44 Hz... with the bass boost turned ON! And below that, it rolls off very steeply. At 30 Hz, it is 24 dB down. By 25 Hz, it is down into the noise floor... nothing at all down there. So there's no "bottom end" to these guys.

3) It looks like you are going to need to borrow a sub to excite the low end of the room. Hopefully something that is reasonably flat down to 30 Hz.

That low end, below about 80 Hz, is the only remaining mystery in your room: that's what we need to see, to complete the picture.

- Stuart -

#37

Thanks Stuart. Almost there.

It looks like you are going to need to borrow a sub to excite the low end of the room. Hopefully something that is reasonably flat down to 30 Hz.

These subs might be available at my Costco store. Would they suffice? Freq. response says 29hz-120hz.

#38

That should be fine, Howie. That would do the job.

- Stuart -

#39

Hi Stuart,

I got the R-12SW sub. It is passive. I am going to use my Yamaha sound bar sub out from my home tv room, to connect the R-12SW. So I won't be using the K-12 for these tests. I hope that is ok. Otherwise I would have to purchase some adapters to connect the R-12SW to my K-12.

Stay tuned!

Howie

#40

Hi Stuart,

Here are some tests I did this morning. The first and last had no clipping but the other tests did. The last test I added the calibration file for my Dayton mic as seen here:

: Screen Shot 2020-08-07 at 11.14.02 AM.png (33.98 KiB) Viewed 34809 times

: Screen Shot 2020-08-07 at 11.14.02 AM.png (33.98 KiB) Viewed 34809 times

I didn't do that for the other tests.

Here are the mdat's for the corner trick only. I was only able to get to 75dBc where the mic was in the top corner. The sound bar was at full volume. I did a few with different settings on the sub woofer. I hope this gives us what we need.

Howie

Sub woofer tests.mdat: (9.55 MiB) Downloaded 1210 times

Sub woofer tests.mdat: (9.55 MiB) Downloaded 1210 times

#41

It says 200/400 amps but it is passive.

I think you mean Watts, not Amps! I didn't check the manual, but I'm guess that means 200 watts continuous, 400 watts peak? But in any case, 200 watts should be plenty for what you need. You don't need to blow the roof off! Just put out enough power in the low end to trigger all the bad stuff.

Also, I'm confused about the "passive" comment: that's an active sub: it has a built-in amp.

I plan on putting the LF-EXT SUB and connecting the line out post gain XLR to on the K-12 to the RCA in on the sub. Am I thinking about this right? Missing anything?

Right! That will do the job. There's no need to worry about careful cross-over settings here, or power balancing: that's only needed for control rooms and home theater applications. All we want here is decent low-end output to make that room roar at its worst, the same way it will roar when you have a bass guitar, drum kit, keyboards, and growling electric guitar going in there.

- Stuart -

#42

The quotes from your last post were from a reply that I had edited so you must have seen it before the edit. Anyhow, I didn't have the right adapters to go from the K-12 into the sub so I used my sound bar from my home tv, which has a sub out, to do the tests. You can download my mdats from my post from earlier today.

Thank you!

#43

howiedrum wrote:Source of the post You can download my mdats from my post from earlier today.

Oh yeah! Now we be TALKIN'!

Now we have some serious low end.... in fact, now it's too MUCH :shock:

Try turning your sub down by about 30 dB, to get it at more or less the same level as the rest of the spectrum. That's why you were getting clipping... and you still are! Here's where your distortion is right now:

That's pretty high (the lower curve). Just the distortion on your sub all by itself is actually louder than the mains...

But at least the setup is working, and we have low end now!

- Stuart -

#44

Hi Stuart,

I believe I got it this time. I did one corner trick test, two tests where bass amp would go, and two tests where bass drum would be.

Thanks again,

Howie

Sub Woofer Tests 2.mdat: (7.97 MiB) Downloaded 1202 times

Sub Woofer Tests 2.mdat: (7.97 MiB) Downloaded 1202 times

#45

Hey Stuart,

Were you able to check my mdat subwoofer tests on my last post? Even a quick "they look good" would allow me to return the subwoofer.

Thanks a bunch!

Soundman2020 - Studio Design Forum