Yes, flat they are very bright. But I'd rather have it and be able to cut it than not have it.
Sent from my iPad
Sent from my iPad
New DIY Mid High (90deg) - AKA PM90
- Thread starter Peter Morris
- Start date
The important thing to realize is that if you cut down the VHF too much, you quickly lose some of the clarity this cabinet has. I personally haven't had the time and equipment to tune it to the sweet spot between being bright and losing clarity yet.
Target Frequency response, or Power response?
I tried going the of-axis route while EQ' ing, mainly to see if maybe of-axis peaks and dips would explain the difference in what I heard and what my on-axis measurements showed. This task is not trivial, as you need to measure of-axis and on-axis for every move you make - or is there another way? Does this approach have a name?
I have a couple of d&b Q7 tops with a D12 amp, and they definately doesn't measure flat, and to my ears they sound very ok and hi-fi like( with minimum eq). With processor in the amp, the developers could make it flat out of the box, - but they didn't.
I guess most of the bigger brands in this industry has some kind of target FR?
Flat target can be ok, as long as the speakers are operated by someone who knows how to make it sound nice in the given environment. Problem is, not everybody knows how to operate EQ, and sometimes there is not much time for that. Further in situations where the job is dry-hire DJ, - they expect the box to sound nice, as is, also when it gets really loud?
Please share your preferred DSP settings for this box as it is very helpful to others not so experienced - like me :0)
I tried going the of-axis route while EQ' ing, mainly to see if maybe of-axis peaks and dips would explain the difference in what I heard and what my on-axis measurements showed. This task is not trivial, as you need to measure of-axis and on-axis for every move you make - or is there another way? Does this approach have a name?
I have a couple of d&b Q7 tops with a D12 amp, and they definately doesn't measure flat, and to my ears they sound very ok and hi-fi like( with minimum eq). With processor in the amp, the developers could make it flat out of the box, - but they didn't.
I guess most of the bigger brands in this industry has some kind of target FR?
Flat target can be ok, as long as the speakers are operated by someone who knows how to make it sound nice in the given environment. Problem is, not everybody knows how to operate EQ, and sometimes there is not much time for that. Further in situations where the job is dry-hire DJ, - they expect the box to sound nice, as is, also when it gets really loud?
Please share your preferred DSP settings for this box as it is very helpful to others not so experienced - like me :0)
Of course i forgot when i got home. So ill send a reminder to myself for the actual PEQ settings...[/QUOTE said:
I think there are a couple of reasons why we don’t like the flat response we measure using Smaart or Systune - the environment and the measurement compromises we often make do not give us a “real” anechoic response especially with respect to the lower frequencies; some additional LF energy still tends get into the measurement. When we EQ it out, it sounds thin. In addition the distortion products of the compression driver tend to make it sound a little louder than it is … quantifying these is difficult.
You can use a half space measurement to see what the LF end is doing and then measure the box as far away as possible from any boundaries, and about 4m in the air for the mid and HF response. Then try and splice to two measurement together remembering the half space measurement is +6db at the same distance.
The only problem is loading this horn into half space tends to flatten its response a little (1 or 2 dB) compared to full space.
If you want to do a power measurement you should have look at Acoustic Power Labs stuff – that’s their approach, measure the power response of the speaker then use EQ so the power response matches a correction cure. A suitable correction curve will be something like + 8 dB at the low frequency end relative to about 2K – 3K and flat- ish there after depending of the directivity of the speaker.
At the end of the day I aim for a smooth response and “voice” the speaker with a little LF boost so that it sound natural to me.
You can use a half space measurement to see what the LF end is doing and then measure the box as far away as possible from any boundaries, and about 4m in the air for the mid and HF response. Then try and splice to two measurement together remembering the half space measurement is +6db at the same distance.
The only problem is loading this horn into half space tends to flatten its response a little (1 or 2 dB) compared to full space.
If you want to do a power measurement you should have look at Acoustic Power Labs stuff – that’s their approach, measure the power response of the speaker then use EQ so the power response matches a correction cure. A suitable correction curve will be something like + 8 dB at the low frequency end relative to about 2K – 3K and flat- ish there after depending of the directivity of the speaker.
At the end of the day I aim for a smooth response and “voice” the speaker with a little LF boost so that it sound natural to me.
Last edited:
Did you get to try out the base dialings? For me the BMS outruns the 12's easily so the first step is to dial it back. -2.5 to -3.5 is a good start...
Appreciate your thoughts and input - thanks a lot.
I tried many different approaches of measuring, - not including the free space 4m thing yet though :0) All methods showed the same pattern, except for the off-axis measurements - thus the idea of trying out the power response approach.
Actually my main concern with this speaker has been to achieve a smooth response in the 3-12kHz area. I have a 6 db dip in the response at 4.8kHz requireming a massive narrow band rise on the EQ, - and a strange sudden rise in the 10-18kHz area that needs a lot of EQ to get a flat FR. I am also cutting some 2.9kHz and some other small corrections, but they are not that important. The result after applying the EQ specifically in the 4.8kHz and damping the 13kHz peak is not just bright but unpleasantly sharp, as if something is missing.
If I introduce a small rise in the 7.5kHz Q:4 +2dB things start to fall into place in the treble area, where hi-hats and metal intruments etc. starts to sound smooth and real, - but the measured FR is definitely not flat anymore.
Could it be due to the mic's I am using? (the rest of the signal path is calibrated - I even made an L-PAD and measured the output of my amps - to rule that out also :0))
I have tried two mic's one is a DBX RTA M1 and the other is a Behringer UMC8000 and found that there is quite a bit of variation in the VHF. (see enclosed screen shot of HF and VHF separately on BMS passive filter. Pink is RTA and green is UMC). Notice the 10dB scale.
Could the mic's and their respective calibration files be the cause of all my trouble?
Thanks again for all the inputs, suggestions and explanations.
I tried many different approaches of measuring, - not including the free space 4m thing yet though :0) All methods showed the same pattern, except for the off-axis measurements - thus the idea of trying out the power response approach.
Actually my main concern with this speaker has been to achieve a smooth response in the 3-12kHz area. I have a 6 db dip in the response at 4.8kHz requireming a massive narrow band rise on the EQ, - and a strange sudden rise in the 10-18kHz area that needs a lot of EQ to get a flat FR. I am also cutting some 2.9kHz and some other small corrections, but they are not that important. The result after applying the EQ specifically in the 4.8kHz and damping the 13kHz peak is not just bright but unpleasantly sharp, as if something is missing.
If I introduce a small rise in the 7.5kHz Q:4 +2dB things start to fall into place in the treble area, where hi-hats and metal intruments etc. starts to sound smooth and real, - but the measured FR is definitely not flat anymore.
Could it be due to the mic's I am using? (the rest of the signal path is calibrated - I even made an L-PAD and measured the output of my amps - to rule that out also :0))
I have tried two mic's one is a DBX RTA M1 and the other is a Behringer UMC8000 and found that there is quite a bit of variation in the VHF. (see enclosed screen shot of HF and VHF separately on BMS passive filter. Pink is RTA and green is UMC). Notice the 10dB scale.
Could the mic's and their respective calibration files be the cause of all my trouble?
Thanks again for all the inputs, suggestions and explanations.
Attachments
Jason:
Yes I tried your settings and the lower cross over point seems to be working well. I am still having issues with the voicing of HF and VHF and I was hoping to see some suggestions for PEQ's in that frequency area.
I am using 16/16 Ohm BMS and two RCF 8 Ohms in parallel so I should theoretically have 3 dB more SPL on the LF, but I have only +4dB dialed in on the LF. Different amounts of PEQ cutting peaks in the LF section would also reflect on the SPL difference?
Do you have PEQ in the HF and VHF that you could share?
Andy ... a random thought, by chance you are not using an 8 ohm BMS crossover with a 16 ohm driver? What are the values of the two capacitors?
Both the mics you are using have issues between 5 & 10 kHz ,,, the manufactures show some peaks of around 2 dB, but that could easily be 3 or more in reality.
Both the mics you are using have issues between 5 & 10 kHz ,,, the manufactures show some peaks of around 2 dB, but that could easily be 3 or more in reality.
Peter, The values cannot be seen as they are face down. I will find a way to check.
Could that be the reason for the big dip in the 5 kHz area you think?
So I will definitely buy a new mic. Any suggestions?
I think this is the crossover, however some of mine use a 4.7uF
Attachments
I have now measured the component values, and they are same as yours, and according to this:
Have measured Ohms values with a multimeter and got 6.8 Ohm on the HF and 8.5 Ohm on the MF.
I can't seem to find the RE data on BMS website and I don't know if that is normal for a 16/16 Ohm version?
The meter is an old FLUKE 77.
Ok. so tried to put 8.2 Ohm in series with the HF and VHF, but the changes in frequency response was surprisingly small.
I am still having issues with a sharp dip at 4.8 kHz, which is something going on with the HF section, and the a strange rise/peak at 14 kHz. Both 'findings' are present even when using another mic.
This is a measurement of the HF Near field, laying on the edge of a table without the passive XO. Notice the dip at 4.8 kHz is on the HF, not at the XO point.
Measurement at 60cm on edge of a table with passive XO and the dip and peaks are showing quite well:
I am still having issues with a sharp dip at 4.8 kHz, which is something going on with the HF section, and the a strange rise/peak at 14 kHz. Both 'findings' are present even when using another mic.
This is a measurement of the HF Near field, laying on the edge of a table without the passive XO. Notice the dip at 4.8 kHz is on the HF, not at the XO point.
Measurement at 60cm on edge of a table with passive XO and the dip and peaks are showing quite well:
Attachments
DSP, DRIVERACK VENUE 360
HF: BMS 4594nd HE 16/16 Ohm
Passive BMS crossover
Delay: 1.96ms
Gain: -3.0db at crossover point
MID: 2 x B&C 12NDL76 16 Ohm.
Delay: 0ms
Gain: +2.0db at crossover point
Updated...
HF: BMS 4594nd HE 16/16 Ohm
Passive BMS crossover
Delay: 1.96ms
Gain: -3.0db at crossover point
| Type | FREQ | Q | dB |
| Highpass LR24 | 473 | ||
| Lowpass LR6 | 18k | ||
| 15.7k | 16 | +2.5 | |
| 3.5k | 9 | -3.0 |
MID: 2 x B&C 12NDL76 16 Ohm.
Delay: 0ms
Gain: +2.0db at crossover point
| Type | FREQ | Q | dB |
| Highpass BW18 | 95 | ||
| Lowpass LR24 | 500 | ||
| 143hz | 1.4 | 2.8 | |
500 Hz is getting quite low for that driver and horn combination ... no problem, but if you want max SPL I would cross a little higher
I've taken it into consideration so i have two settings. One above and another recommended 600hz LR24.
Last edited:
Jason.
Thank you for sharing dsp settings. The settings will not work on my setup. I have some peaks in the Mid section and one serious at about 400 Hz that must be corrected. I have a peak in the HF at about 2.8 kHz and I am using a much lower Q for that, comparing to your correction at 3.5 kHz.
Next attempt to make it work will be to try to make a new passive filter for the coax.
I am expecting my new Audiomatica Clio Pocket measuring tool to arrive within a few days, and will start this up with hopefully new more reliable measurements.
Thank you for sharing dsp settings. The settings will not work on my setup. I have some peaks in the Mid section and one serious at about 400 Hz that must be corrected. I have a peak in the HF at about 2.8 kHz and I am using a much lower Q for that, comparing to your correction at 3.5 kHz.
Next attempt to make it work will be to try to make a new passive filter for the coax.
I am expecting my new Audiomatica Clio Pocket measuring tool to arrive within a few days, and will start this up with hopefully new more reliable measurements.
Does anybody have any experience with the RCF ND950 1.4 compression driver?
I've read an article about a German company that build the PM90 with the ND950.
I also thought RCF used the ND950 in the newer ART745A from 650 Hz and up.
I'm aware that the ND950 can be fully used from 800 Hz, but I am curious of it will go a bit lower.
Also would the B&C 12NW76 be suited driver for the low/mid?
Thanks for any info!
I've read an article about a German company that build the PM90 with the ND950.
I also thought RCF used the ND950 in the newer ART745A from 650 Hz and up.
I'm aware that the ND950 can be fully used from 800 Hz, but I am curious of it will go a bit lower.
Also would the B&C 12NW76 be suited driver for the low/mid?
Thanks for any info!
Stef, I can bring you in contact with Christian (from Germany) if you like. Christian built the PM90 with the ND950. Send me a PM.
I think Peter Morris used the 12NW76 (or was it the 12NDL76?) in one of his builds, but he liked the RCF more.
Sent from my iPhone