Improvising a turntable for measurements using a 40cm lazy Susan and leftover wood.
Can you adjust the pivot point still? Looks like you need to move the speaker towards the weights slightly, although that may be an optical illusion.
One option for getting the woofers up high is to turn the speaker upside down.
Depending on what furniture acrobatics we manage on the day, we may well need this option.
Yes, but I was not planning to do that. Will do 5deg step to 90 or 120deg. 1.5m mic distance.
I understand priority one is off axis set for each driver.
If we have time left I was then planning to do full speaker with a certain crossover set. 2m distance, 10deg step.
To be honest not sure why.
Third priority would be nearfields of each driver, but as I thought I could always do these at home, priority is low.
Or do I need to do something else while in this large indoor space? We would have about 4 hours (incl. setting up/down).
Last edited:
What I personally would do is much different, so I'll not say too much. These are my thoughts.
Nothing really changes in 5 degree increments.
The software predicts results, but if there's a problem with some of the measurements, you're screwed.
I would work with on axis only. (Initially )
I would rather make a x-over flat on axis, and then measure it two more times at 45 and 60 degrees off axis, to see if it does as intended. If it does, store that x-over information.
If you then want to take another hundred measurements, go for it.
When you get home, and sim some x-overs, you can compare the sim estimate, with the reality of the one real / stored x-over.
The big thing you want, is to have 1k down to 200hz dialed in before you leave. (ie, the woofer to mid x-over. ) Anything else can be done at home if needed. .
Nothing really changes in 5 degree increments.
The software predicts results, but if there's a problem with some of the measurements, you're screwed.
I would work with on axis only. (Initially )
I would rather make a x-over flat on axis, and then measure it two more times at 45 and 60 degrees off axis, to see if it does as intended. If it does, store that x-over information.
If you then want to take another hundred measurements, go for it.
When you get home, and sim some x-overs, you can compare the sim estimate, with the reality of the one real / stored x-over.
The big thing you want, is to have 1k down to 200hz dialed in before you leave. (ie, the woofer to mid x-over. ) Anything else can be done at home if needed. .
Temp25, your method is so old school 😎
The info that AlmaAtaKZ will get shall be fed to VituixCAD simulator, then one can spend rest of the life testing different eq/xos and see summed response, dispersion etc. instantly on the screen. VCAD sims are extremely reliable.
But yes, 5deg steps are not necessary. Using 10deg steps will save 40% of time. When baffle is symmetric one needs to rotate only 0-180deg horizontally, on one side.
Vertical polars can be done with the speaker sideways on the stand, but measurement distance must be as long as possible. Frontal hemisphere should be enough, and up to +/-45deg. Wider angles are prone to artefacs when measuring floorstanders.
Reflex port or radiator output will be included in woofer section measurements, but is relevant only regarging resonances occurring above 3-400Hz (typically 600-1000Hz).
---
I have not been able to measure in optimal conditions, but one can learn to ignore systematic/method errors in the data. Changing right window gating and smoothing will help.
The info that AlmaAtaKZ will get shall be fed to VituixCAD simulator, then one can spend rest of the life testing different eq/xos and see summed response, dispersion etc. instantly on the screen. VCAD sims are extremely reliable.
But yes, 5deg steps are not necessary. Using 10deg steps will save 40% of time. When baffle is symmetric one needs to rotate only 0-180deg horizontally, on one side.
Vertical polars can be done with the speaker sideways on the stand, but measurement distance must be as long as possible. Frontal hemisphere should be enough, and up to +/-45deg. Wider angles are prone to artefacs when measuring floorstanders.
Reflex port or radiator output will be included in woofer section measurements, but is relevant only regarging resonances occurring above 3-400Hz (typically 600-1000Hz).
---
I have not been able to measure in optimal conditions, but one can learn to ignore systematic/method errors in the data. Changing right window gating and smoothing will help.
Maybe too late to change, but that column looks rather close. Also the thing the speaker is sitting on is close to the woofers. I know it would be awkward, but I'd flip the speaker. Hope everything works out! Be really careful. a speaker falling from that height would cause a lot of damage.
Midrange:
The massive 4kHz dip is a complete surprise and I have no idea what has caused it. No such dip was showing on two previous attempts (I think I used the other speaker). Hmmmmm.
Edit: could it be that I forgot to mute the tweeter? Would the effect be like this?
6.5ms
12ms
24ms
The massive 4kHz dip is a complete surprise and I have no idea what has caused it. No such dip was showing on two previous attempts (I think I used the other speaker). Hmmmmm.
Edit: could it be that I forgot to mute the tweeter? Would the effect be like this?
6.5ms
12ms
24ms
Last edited:
Overall experience of measuring in a much larger room:
- noise and reflections were easier to deal with (both in the setup and to see where the reflections are on the impulse response)
- completely puzzled by the 4kHz dip on the mid (which is unlikely related to the room)