2.1 Focal system, questions about measuring L+R channels together

Citysoundman · Jul 8, 2024

Hi there,
I just finished converting a bedroom into a mix room for my audio post work - TV and commercial mixing. I framed out the walls and ceiling and used fluffy insulation throughout. It's a 5.1 setup but most of my work is 2.1 so I'm focusing on calibrating the 2.1 setup now. I have Focal Solo6 BE's and Focal Sub6. The room dimensions are 14' 9" x 11" 5" x 7' 7" height. I'm making measurements with the UMIK-1 USB mic with calibrated file.

The Focal sub has a hi pass filter for the mains, and a low pass filter and phase adjustment for the sub. It's ported on the back. I've found that small changes in the filter frequency and/or phase setting, along with position in the room, really make major changes in the resulting curve. So it's been hours of trial and error, trying different settings and positions. I've landed on a result that seems pretty good.

I'm a newbie with REW but learning a lot as I go. I'd love to get some feedback to see if I'm doing things correctly. I'm not 100% sure about measuring with L+R versus L or R.

1. For checking optimal sub placement and filter/phase settings, I've been making measurements using L+R, and comparing SPL and Waterfall results. I' pretty sure this is correct, to use L+R for this - can anyone confirm?

2. While doing this I noticed that the results in the high frequencies were looking pretty off. The highs have some major dips when looking at the L+R results. (see attached pic, Red trace is L+R). And these dips seem to be a bit random. At first I thought there was some phase problem with my speaker or wiring, but now I'm not sure. When I check the L and R speakers individually the results look normal. So I'd like to know if phasing is normal in the high end when looking at L+R results.

thanks!

moedra · Jul 9, 2024

Measuring with both speakers together (L+R) generally skews the top end of the response. The anomaly has to do with filtering of the higher direct frequencies, often resulting in cancellation around a center frequency. You can see it happening there at 8kHz in your screenshot. It's valuable information for you, but if your goal is to measure for correction purposes, stick to the independent speaker responses.

sm52 · Jul 9, 2024

Citysoundman said:
When I check the L and R speakers individually the results look normal.

Do this if you want to properly align all your speakers. That is, you need to get separate measurements of the left, right speaker, and subwoofer. To do this, acoustic timing must be used to the same speaker having a tweeter. or just a separate tweeter. Then you will have a picture of the timing relationship of your speakers. After this, they can be well coordinated.

Citysoundman · Jul 9, 2024

moedra said:
Measuring with both speakers together (L+R) generally skews the top end of the response. The anomaly has to do with filtering of the higher direct frequencies, often resulting in cancellation around a center frequency. You can see it happening there at 8kHz in your screenshot. It's valuable information for you, but if your goal is to measure for correction purposes, stick to the independent speaker responses.

thanks for the reply, that's what I was hoping to hear! So this confirms I don't have a wiring or speaker issue.

Citysoundman · Jul 9, 2024

sm52 said:
Do this if you want to properly align all your speakers. That is, you need to get separate measurements of the left, right speaker, and subwoofer. To do this, acoustic timing must be used to the same speaker having a tweeter. or just a separate tweeter. Then you will have a picture of the timing relationship of your speakers. After this, they can be well coordinated.

Thank you, I did just start using the acoustic timing feature, and I have it set to output from my right speaker. Can you elaborate on 'properly aligning' the speakers? I'm not sure what you mean.
Also, what do you look for when measuring the sub independently?

What I'm doing now is measuring L+R on the full 2.1 system in order to see the results of different crossover and phase settings on the full spectrum....is this the correct way to check it?

sm52 · Jul 9, 2024

The sub interacts not with two speakers as if they were a single whole, but with each of them. This means that setting up L+R with a sub, as you do, may seem acceptable. But at the same time, L can conflict with the sub and give disadvantages, and R can be friends with the sub and compensate for the failures of L+sub. Or vice versa. If you adjust each pair separately, you get the total addition of all three. The truth is that this is not always possible. Therefore, compromises have to be sought. But the approach to the solution itself is more correct than yours.

DanDan · Jul 9, 2024

Unless the distance from each tweeter to the mic is identical there will be a small time delay between the channels causing comb filtering. So measure L and R separately and combine them using trace arithmetic. Speaker Controllers, DRC, often provide a facility to insert a delay on the early channel in order to virtually move the speakers to equidistant. I have found that sounds quite bad in a highly treated room. Zone Without Early Reflections. Friends and Clients have also noticed this. The Designer of Dirac Live Mathias agrees with us. So turn off or do not enable speaker offsets and move your head or chair if you need to make the centre image shift.

sm52 · Jul 9, 2024

DanDan said:
I have found that sounds quite bad in a highly treated room. Zone Without Early Reflections.

I agree with you. But most still try to drown them out. My opinion is that early reflections should remain, but they should not be louder than the direct signal. The question remains how much early reflections should be attenuated.

DanDan · Jul 9, 2024

Sorry sm52, that was not my point. I was referring to the comb filtering and widening caused by inserting a 'speaker position offset' delay. This detail, like many other small ones, is only audible in a close to anechoic critical listening space, e.g. Studio Control Room. The ability to deliver such tiny details is the very reason why ER is removed, suppressed. In that world the simple rule of thumb is to have no reflections higher than -20dB relative to Direct for at least the first 20mS. 20/20 if you will.

Citysoundman · Jul 9, 2024

DanDan said:
Unless the distance from each tweeter to the mic is identical there will be a small time delay between the channels causing comb filtering. So measure L and R separately and combine them using trace arithmetic. Speaker Controllers, DRC, often provide a facility to insert a delay on the early channel in order to virtually move the speakers to equidistant. I have found that sounds quite bad in a highly treated room. Zone Without Early Reflections. Friends and Clients have also noticed this. The Designer of Dirac Live Mathias agrees with us. So turn off or do not enable speaker offsets and move your head or chair if you need to make the centre image shift.

Thanks DanDan. I rechecked my speaker distances from the mic and got them as matched as humanly possible. I did new measurements of individual L and R as well as L+R, and I tried combining individual L and R results using trace arithmetic like you suggested. But the arithmetic result looks the same as when I do a L+R measurement. I'm still seeing the dip in the high end. And now this high end dip is bothering me.

When I look at the individual L and R results I can see there are phase differences in the high end (from 2K on up). I'm posting some pics of my new results. One of the pics shows individual L and R SPL graphs that are trace separated, to view them easier on one screen. The shape of the waves is fairly different above 7K. What could be the cause - and do I need to worry about it??

I'm also posting some poorly shot studio photos, to show what my room treatment is looking like
.

DanDan · Jul 9, 2024

My main DRC is Dirac Live. It doesn't tell you directly if your mic is equidistant, it just generates an offset delay which you can bypass, but I think it is on by default. Over years, I have NEVER seen a zero. Also I guess no space is truly anechoic so the paths may contain differences. But overall I would trust the individual measures and not take much notice of the single point summing. We have a big lump of meat and bone blocking any chance of HF summing at any ear. There has been a lot of research of the response of real listeners rooms. With the usual amount of HF cancelling reflections, speaker directivity narrowing at HF, Room Modes etc. the typical listener hears +3dB at LF tilted to -3dB at HF. I impose such a Target Curve in my Control Room, and Bingo, the Mix and Tone sound the same in my three domestic rooms, in the car, and on HD650s. Below shows the HF cancelling of two tweeters not equidistant to the single point receiving mic.

Citysoundman · Jul 9, 2024

Gotcha, thanks for your feedback. I will not worry about it!

The other thing that stands out to me on the SPL curve is the big dip between 400 & 550Hz. Do you think this is a room mode thing? Any suggestions on how to figure out if I can do something about it?

DanDan · Jul 9, 2024

Thats a Sherlock Holmes job....... Move the mic. Modes do not change frequency but Reflection issues do. Typically if you get nearer to a reflecting surface which is causing a cancellation the frequency should rise. Speaking of reflecting surfaces....... Use a Mic Stand when Measuring and the delay before sweep, to get a good distance away from the mic.

Citysoundman · Jul 9, 2024

DanDan said:
Thats a Sherlock Holmes job....... Move the mic. Modes do not change frequency but Reflection issues do. Typically if you get nearer to a reflecting surface which is causing a cancellation the frequency should rise. Speaking of reflecting surfaces....... Use a Mic Stand when Measuring and the delay before sweep, to get a good distance away from the mic.

Sherlock Holmes...do you mean a mystery?? or the shape of his pipe

I'll try out some other mic positions and see if it changes.
Funny graphic in your post - do all engineers look dorky like him??

For sure I have my SPL meter and mic on booms stands. I use a 3 sec delay, then scoot to the back corner of the room for each measurement...quite silly the procedure, I'd be embarrassed to be caught on video :laugh:

DanDan · Jul 9, 2024

LOL, I look at my dorky self in the mirror when shaving every morning and I ask myself...... Am I An Artist?
Sometimes when trying to find the main or only reflecting surface causing a particular null of interest I move over to the RTA. Zoom in to around the null and move the mic towards the suspect surfaces. If the frequency rises so should suspicion...... Elementary eh, puff puff......

deercreekaudio · Jul 14, 2024

Citysoundman said:
Hi there,
I just finished converting a bedroom into a mix room for my audio post work - TV and commercial mixing. I framed out the walls and ceiling and used fluffy insulation throughout. It's a 5.1 setup but most of my work is 2.1 so I'm focusing on calibrating the 2.1 setup now. I have Focal Solo6 BE's and Focal Sub6. The room dimensions are 14' 9" x 11" 5" x 7' 7" height. I'm making measurements with the UMIK-1 USB mic with calibrated file.

The Focal sub has a hi pass filter for the mains, and a low pass filter and phase adjustment for the sub. It's ported on the back. I've found that small changes in the filter frequency and/or phase setting, along with position in the room, really make major changes in the resulting curve. So it's been hours of trial and error, trying different settings and positions. I've landed on a result that seems pretty good.

I'm a newbie with REW but learning a lot as I go. I'd love to get some feedback to see if I'm doing things correctly. I'm not 100% sure about measuring with L+R versus L or R.

1. For checking optimal sub placement and filter/phase settings, I've been making measurements using L+R, and comparing SPL and Waterfall results. I' pretty sure this is correct, to use L+R for this - can anyone confirm?

2. While doing this I noticed that the results in the high frequencies were looking pretty off. The highs have some major dips when looking at the L+R results. (see attached pic, Red trace is L+R). And these dips seem to be a bit random. At first I thought there was some phase problem with my speaker or wiring, but now I'm not sure. When I check the L and R speakers individually the results look normal. So I'd like to know if phasing is normal in the high end when looking at L+R results.

thanks!

Have you tried moving the microphone around and averaging?. Here is what typical measurement looks like, mostly additive by 6dB. This is an excerpt from our tech blog: https://deercreekaudio.com/tech-blog/f/using-key-measurements-to-verify-basic-system-setup

Citysoundman · Jul 14, 2024

deercreekaudio said:
Have you tried moving the microphone around and averaging?. Here is what typical measurement looks like, mostly additive by 6dB. This is an excerpt from our tech blog: https://deercreekaudio.com/tech-blog/f/using-key-measurements-to-verify-basic-system-setup

View attachment 72275

Hi Deer Creek! FYI I bought my UMIK mic from you guys

When I move the mic a little bit ( a few inches from the MLP) there's not much change in the 400-500Hz dip - it shifts up and down the spectrum a little. When I move the mic a major amount, like 2 feet up and 1 foot to the side, then the dip seems to go away. But then it is way out of position.
DanDan suggested that it could be due to nearby reflective surfaces, so I covered my desk with some OC703 but this didn't change the dip.

deercreekaudio · Jul 14, 2024

Citysoundman said:
Hi Deer Creek! FYI I bought my UMIK mic from you guys
When I move the mic a little bit ( a few inches from the MLP) there's not much change in the 400-500Hz dip - it shifts up and down the spectrum a little. When I move the mic a major amount, like 2 feet up and 1 foot to the side, then the dip seems to go away. But then it is way out of position.
DanDan suggested that it could be due to nearby reflective surfaces, so I covered my desk with some OC703 but this didn't change the dip.

We have seen in a couple of instances where there has been inverted wiring on one of the drivers, in this case possibly the tweeter. It's pure speculation on my part but worth looking into.

JStewart · Jul 14, 2024

Citysoundman said:
When I look at the individual L and R results I can see there are phase differences in the high end (from 2K on up).

Looks like 2kHz is the crossover for the tweeter on this model.

Chart from

Focal Solo6 Be Review (Studio Monitor)

This is a review and detailed measurements of the Focal Solo6 Be studio monitor (powered active speaker). It is on kind loan from a member and costs US $1,499 each. Unlike most studio monitors that have industrial looks, the Solo6 Be comes across as a high-end home speaker with its nice wood...

www.audiosciencereview.com

deercreekaudio said:
We have seen in a couple of instances where there has been inverted wiring on one of the drivers, in this case possibly the tweeter. It's pure speculation on my part but worth looking into.

I think you may be on to something here.
Measurements with a timing reference would allow removal of small mic distance variations just to be sure and/or @Citysoundman is there a local dealer who could help you check them out?

Citysoundman · Jul 14, 2024

deercreekaudio said:
We have seen in a couple of instances where there has been inverted wiring on one of the drivers, in this case possibly the tweeter. It's pure speculation on my part but worth looking into.

Wow very interesting....
Just to clarify, do you mean you've seen inverted wiring on these Focal monitors? or just on monitors in general?

If one tweeter was inverted then the high end of my stereo playback would be playing out of phase

I just did some testing with a sine wave sent to L+R, then I inverted phase on one side while listening summed in mono. The woofers are definitely in phase. The tweeters may be out of phase, but somehow it's much harder to be certain, especially at higher frequencies. If I shift my head just a little bit the summed signal changes pretty drastically. I will test with a different set of speakers tomorrow.

JStewart said:
Measurements with a timing reference would allow removal of small mic distance variations just to be sure and/or @Citysoundman is there a local dealer who could help you check them out?

Good point - I do use a timing reference, so I checked my Impulses in REW. They all appear consistent - all going in an upwards direction at the initial point. But that's measuring the full range. I'm thinking to measure again with only high frequencies to try and isolate the tweeters. Is it easy to see if one driver is inverted from viewing the impulse curve?

I can also remove the tweeter to check the wiring but they're beryllium so I'd prefer not to mess with it unless I need to.

Edit -
I just found this on another thread - is this the way to check for driver polarity?
Easiest is to look at the impulse response in % FS mode to see if the peak is positive or negative.

JStewart · Jul 15, 2024

Citysoundman said:
Edit -
I just found this on another thread - is this the way to check for driver polarity?
Easiest is to look at the impulse response in % FS mode to see if the peak is positive or negative.

That will work for the entire speaker. Not so sure it will reliably tell you if a driver is opposite phase or put differently, it would work for a single driver but not a speaker with multiple drivers.

Edit: oh wait, John M says that in your link.
”Easiest is to look at the impulse response in % FS mode to see if the peak is positive or negative. That assumes you are measuring a single driver, or if measuring a multiple drive unit speaker you restrict the sweep range to the crossover range of the driver you are interested in.”

Citysoundman · Jul 15, 2024

I measured again with high frequencies only, 10K-20K sweep. The impulses all look consistent with each other.
Then I tested a different pair of Focals - I moved my rears to the fronts, and these measurements look basically the same as the original speakers. So I'm concluding there's no tweeter wiring inversion (phew!)
It would have been awful to think I couldn't tell that the highs were out of phase :gulp:

2.1 Focal system, questions about measuring L+R channels together

New Member

Attachments

Member

Member

New Member

New Member

Member

Moderator

Member

Moderator

New Member

Moderator

Attachments

New Member

Moderator

Attachments

New Member

Moderator

Member

New Member

Member

Senior AV Addict

New Member

Senior AV Addict

New Member

Popular tags