-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Can anyone help with Alignment?
- Thread starter Robbie010
- Start date
For Fq-SPL FFT analysis, instead of REW's rapid sine-sweep, have you ever tested "FFT averaging of DSP-processed flat white-noise rich accumulated (recorded) sound data"?
References:
- Frequency response measurements by "cumulative white noise averaging": #392, #404, especially the end portion of #297(remote thread) by Dr. Floyd Toole, #315(remote thread) by Dr. Floyd Toole, #125(remote thread)
- The latest Fq-SPL (re-confirmation) of multiple amplifiers SP high-level output signals and that of room air sound at listening position: all measured by “FFT averaging of recorded cumulative DSP-processed flat white noise” (as of June 8, 2025): #1,009
References:
- Frequency response measurements by "cumulative white noise averaging": #392, #404, especially the end portion of #297(remote thread) by Dr. Floyd Toole, #315(remote thread) by Dr. Floyd Toole, #125(remote thread)
- The latest Fq-SPL (re-confirmation) of multiple amplifiers SP high-level output signals and that of room air sound at listening position: all measured by “FFT averaging of recorded cumulative DSP-processed flat white noise” (as of June 8, 2025): #1,009
Keith_W
Master Contributor
Yeah, I agree. @Robbie010 Ignore the tweeters for now until you can verify the measurement. The next step is to generate crossovers, then convolve them with the measured response, and then sum them. You don't need to take more measurements for that, it's purely a simulation. Would you like to try that?
I would recommend to at least have some off-axis measurements of the horn and woofer around the presumed crossover point to match directivity.
You also cannot avoid comb filtering of the horn and the super tweeter. The CTC distances are too large for the crossover point. You could minimize this with a steep filter. Or just run without the Visaton and see how well that works.
You also cannot avoid comb filtering of the horn and the super tweeter. The CTC distances are too large for the crossover point. You could minimize this with a steep filter. Or just run without the Visaton and see how well that works.
Last edited:
Last edited:
- Thread Starter
- #67
Yes, absolutely!
Please allow me a short while to gather my info.
Today I will be completing the re-lining and stuffing of the bass cabinets and the installation of the new front baffle.
I will post further details of the finished speakers, along with details such as the C to C of the drivers and and new "fresh" frequency response graphs for each driver at multiple distances, volumes and positions on axis and off axis. Hopefully that will give a clear idea of my starting point.
- Thread Starter
- #70
Obviously, the tweeter is free standing so can be moved backwards/forwards, left/right and tilted as necessary.
The DIY bracket I made to hold comp driver & horn allows me to tilt forward/backwards, left/right and swivel left/right if needed.
The DIY bracket I made to hold comp driver & horn allows me to tilt forward/backwards, left/right and swivel left/right if needed.
- Thread Starter
- #74
One final note, none of these measurements have any time reference or gating applied. However, the drivers were placed in such a way that they were as far away as possible from reflective surfaces, at least 1m(ish).
- Thread Starter
- #75
I've just put in some very basic crossover settings of 1,200hz & 4,500hz to have a listen. In order to manage background hiss I have the input gain on the Crossover / DSP set at -8db and can still hear it..... that's another issue.
To get anything like a normal sound I'm having to set the woofer output gain at +12db, the Midrange at +6db and the Tweeter left at 0db.
If I plot this in REW using to 50cm sweeps I get this:
Not sure if thats relevant but it was just an observation.
To get anything like a normal sound I'm having to set the woofer output gain at +12db, the Midrange at +6db and the Tweeter left at 0db.
If I plot this in REW using to 50cm sweeps I get this:
Not sure if thats relevant but it was just an observation.
- Thread Starter
- #77
To be fair, the midrange horn seems to deliver what I’m wanting, which is a clear vocal midrange. I listen to a lot of acoustic / stripped-back male and female vocals and wanted a good central image and slightly forward presentation. It feels unbalanced on some tracks, weak in the bass maybe.
The issue with the tweeters will annoy me. I’m sure I can’t hear much beyond 13/14khz but if I paid for 20khz I want my missing 6,000hz…..
The issue with the tweeters will annoy me. I’m sure I can’t hear much beyond 13/14khz but if I paid for 20khz I want my missing 6,000hz…..
I know I'm tedious.
To start, assemble two bands of Sovereign and 1464. Equalize them flat on-axis and apply some
crossover (you can later try different options (including some rather non-trivial ones)) around 800 Hz.
You can then add a Visatone in the upper octave if you want. Or not. I consider this a matter of personal preference. But neither I nor a dozen of my friends use UHF horns or waveguides. Although one friend sometimes uses super tweeters.
I won't mention the bass in the room for now. But the link is in the signature.
To start, assemble two bands of Sovereign and 1464. Equalize them flat on-axis and apply some
crossover (you can later try different options (including some rather non-trivial ones)) around 800 Hz.You can then add a Visatone in the upper octave if you want. Or not. I consider this a matter of personal preference. But neither I nor a dozen of my friends use UHF horns or waveguides. Although one friend sometimes uses super tweeters.
I won't mention the bass in the room for now. But the link is in the signature.
Keith_W
Master Contributor
FYI, it is conventional to specify lateral sweeps in terms of degrees, e.g. 15deg, 30deg, 45deg, etc. rather than "10cm left", "20cm left". There are two ways for home hobbyists to measure speaker directivity:
1. Construct a turntable with degree markings. Place your speaker on the turntable and the mic 1m away. Rotate your speaker to the marking and take the sweep.
2. Tape a protractor on top of the speaker. Cut a piece of string to 1m and stick it to the centre of the protractor. Move your mic around the speaker using the piece of string to gauge distance and angle.
Assuming you have taken precautions to keep reflections late and attenuated, both methods will give you the same information. But in practice, the second method changes the pattern of reflections, so the results aren't the same.
Measuring "20cm to the left" at 1m away (blue) gives you an angle of 11.3deg compared to on-axis (red). You can see the curves are almost the same, they only diverge above 13kHz. Measuring at 11.3deg is pretty useless information, if you want to see the directivity you will need these measurements: 15, 30, 45, 60, 90, and 180 (i.e. behind the speaker), in the horizontal plane. Some crazy people check at 5 deg increments and feed all that data into VituixCAD to generate nice looking graphs. You don't need to be that crazy - even a blurry photograph gives you an idea of the general shape. And you only need the general shape. If your driver is mounted asymmetrically in the cabinet, you will need to repeat the measurement to the left and right of the speaker.
I personally wouldn't worry about the directivity of your speaker for the time being. It is a physical design thing, and it's not something you can easily change at this point. Not without constructing a new cabinet or buying new drivers / waveguides / horns anyway. At some time in the future you should check the directivity, but for now your priority should be to get the loudspeaker working.
1. Construct a turntable with degree markings. Place your speaker on the turntable and the mic 1m away. Rotate your speaker to the marking and take the sweep.
2. Tape a protractor on top of the speaker. Cut a piece of string to 1m and stick it to the centre of the protractor. Move your mic around the speaker using the piece of string to gauge distance and angle.
Assuming you have taken precautions to keep reflections late and attenuated, both methods will give you the same information. But in practice, the second method changes the pattern of reflections, so the results aren't the same.
Measuring "20cm to the left" at 1m away (blue) gives you an angle of 11.3deg compared to on-axis (red). You can see the curves are almost the same, they only diverge above 13kHz. Measuring at 11.3deg is pretty useless information, if you want to see the directivity you will need these measurements: 15, 30, 45, 60, 90, and 180 (i.e. behind the speaker), in the horizontal plane. Some crazy people check at 5 deg increments and feed all that data into VituixCAD to generate nice looking graphs. You don't need to be that crazy - even a blurry photograph gives you an idea of the general shape. And you only need the general shape. If your driver is mounted asymmetrically in the cabinet, you will need to repeat the measurement to the left and right of the speaker.
I personally wouldn't worry about the directivity of your speaker for the time being. It is a physical design thing, and it's not something you can easily change at this point. Not without constructing a new cabinet or buying new drivers / waveguides / horns anyway. At some time in the future you should check the directivity, but for now your priority should be to get the loudspeaker working.
Last edited:
Similar threads
