Measuring outside is maybe still an option, the next reflecting surfaces are much further away, you can counteract noise with higher spl but that's ofcourse often no real solution, be it neighbours or the measurements need to be at a specific level or power. 50dB noise isn't great but you still got around 35-40dB s/n @1W except if it's not a semi-constant noise but often higher peaks. To measure with 1W up to ~5W should be fine though.
I didn't know you're using an USB mic. They got the preamp already in it and no error source, added distortion or noise in the chain. That means they don't need a loopback (and can't do it either) measurement to check the soundcard and chain. A spl calibration of the chain is also not needed, which is a plus too. However, you're always bound to your USB and can't use it for other uses. Usually not an issue anyway, and since reduces the error sources by a lot of factors, it's often a very good solution, you just should know what it can and what it doesn't.
There is caveat with USB mics and not being able to make dual or semidual measurements. the computer operating system can vary delay on the USB bus or something which renders the USB mics unusable for repeated measurements such as crossover work. Room correction duty is fine where time info is not so critical.
Semidual or dual channel measurements allow to zero out this kind of issues.
Kimmosto has talked a lot about the issue for years
ARTA
Semidual or dual channel measurements allow to zero out this kind of issues.
Kimmosto has talked a lot about the issue for years
ARTA
The USB mic is not well suited to crossover work and is cumbersome at best. The workaround methods on the website have led members to overthink the process.
A USB mic eliminates a lot of error sources and is easier to handle. No sound card, no preamplifier needed, a lot less cables and (possible) calibrated spl with excellent measurement comparison.
No loopback/distortion/2 ch measurements on the down side.
It's not a mistake to buy it, especially if you are just getting started and don't know on which parts of the hardware you can rely on. It is what it is, you can't have it all.
No loopback/distortion/2 ch measurements on the down side.
It's not a mistake to buy it, especially if you are just getting started and don't know on which parts of the hardware you can rely on. It is what it is, you can't have it all.
What are your thoughts on the variations in procedure needed to do crossover work with this mic?
That's a good question. The 2nd channel measurement allows to check the measurement chain and build a compensation for systematic errors like the soundcard, mic preamp or the amp having non linear behaviour, noise etc. Examples for systematic errors are drop off of the chain because of preamp or amp capacitor coupling or phase problems because of the coupling. Of course some of it can still be measured with a different input but not everything can. That are often things that only have to be tested once and it does not pose a problem for every user.
The USB mic got several disadvantages that are limitations or error sources. The USB mic got its own preamp and AD converter. That means the mic recording and the playback are always on two different devices which means it always got a different clock signal/timing. For phase measurements you can get unknown offsets.
Worse: The USB timing isn't very precise and depends a lot on the used hardware (btw: always give the soundcard and mic (if it's USB) its own USB channel). For a frequency response that's rarely a problem but at phase measurements you can get quite big jumps on the measurement. The problem ofcourse increases with the frequency. Not nice if you got a tricky crossover and can't be sure if there's a problem with the crossover, drivers or the hardware. If you use the same soundcard for playback and record, it always got the same clock and latency on both.
There are also practical disadvantages. When measuring, you are tied to the length of the USB cable. Outside that might be a problem. Yes, you can extend it by a certain length but USB 2.0 is specified to just 5m. Using an active hub to get further worsens the latency and clock problems. Heat and coldness could lead to thermal drift (clock!). That can't happen with an analog solution, if the mic cable would be too long you can place the preamp practically wherever you want and extend it after it. A soundcard or laptop chipset reacts much less to that and the clock is the same for record and playback, so the drift hardly matters anyway.
Like I already said, it is what it is. It's not a mistake to buy an USB mic, it definitely got its advantages, you should know the limitations though. I probably forgot the half anyway 🙄 sorry for that.
The USB mic got several disadvantages that are limitations or error sources. The USB mic got its own preamp and AD converter. That means the mic recording and the playback are always on two different devices which means it always got a different clock signal/timing. For phase measurements you can get unknown offsets.
Worse: The USB timing isn't very precise and depends a lot on the used hardware (btw: always give the soundcard and mic (if it's USB) its own USB channel). For a frequency response that's rarely a problem but at phase measurements you can get quite big jumps on the measurement. The problem ofcourse increases with the frequency. Not nice if you got a tricky crossover and can't be sure if there's a problem with the crossover, drivers or the hardware. If you use the same soundcard for playback and record, it always got the same clock and latency on both.
There are also practical disadvantages. When measuring, you are tied to the length of the USB cable. Outside that might be a problem. Yes, you can extend it by a certain length but USB 2.0 is specified to just 5m. Using an active hub to get further worsens the latency and clock problems. Heat and coldness could lead to thermal drift (clock!). That can't happen with an analog solution, if the mic cable would be too long you can place the preamp practically wherever you want and extend it after it. A soundcard or laptop chipset reacts much less to that and the clock is the same for record and playback, so the drift hardly matters anyway.
Like I already said, it is what it is. It's not a mistake to buy an USB mic, it definitely got its advantages, you should know the limitations though. I probably forgot the half anyway 🙄 sorry for that.
Not necessarily. That can happen but it will not happen in all cases. Not a problem for sat/sub either because of the long period of the low frequencies.
And there's still the trick with the reverse of the polarity of one driver, measure the response and you get the best addition by the lowest dip between the drivers. 'Perfect' (or best possible) cancellation means it's swapped back the best phase. You have ofcourse to measure the response at the correct polarity afterwards again to make sure you didn't just created a huge gap. Not perfect but a usable workaround for many 2-way speakers. For more ways it's too time consuming and definitely not a joy.
Last edited:
^^ ICG Yep measuring whole loudspeaker or loudspeaker system at one location USB mic works just fine.
To actually get better results than with pen and paper crossover work needs measurements of all drivers individually on multiple axis. Varying delay on these measurements makes random errors. Maybe the varying delay could be handled with manually tweaking the delays of the measurements? But that would introduce unnecessary errors and hustle into the equation. It would be better to use dual channel system (accurate) or pen and paper (a lot cheaper).
Maybe random delays can be ignored for on axis measurements of individual drivers when speaker and mic stays at the same location? Might be good enough info for some situations. But to get good loudspeaker from this limited info requires prior knowledge of similar systems and what the errors and their magnitude are.
Well anyway, every tool has its purpose. Important thing, in my opinion, is to bring the information out so readers can do their own research and decide what to do in a given situation. Peace 🙂
To actually get better results than with pen and paper crossover work needs measurements of all drivers individually on multiple axis. Varying delay on these measurements makes random errors. Maybe the varying delay could be handled with manually tweaking the delays of the measurements? But that would introduce unnecessary errors and hustle into the equation. It would be better to use dual channel system (accurate) or pen and paper (a lot cheaper).
Maybe random delays can be ignored for on axis measurements of individual drivers when speaker and mic stays at the same location? Might be good enough info for some situations. But to get good loudspeaker from this limited info requires prior knowledge of similar systems and what the errors and their magnitude are.
Well anyway, every tool has its purpose. Important thing, in my opinion, is to bring the information out so readers can do their own research and decide what to do in a given situation. Peace 🙂
Last edited:
I didn't say it does not work. I said you can encounter several issues at certain measurements while others are mostly care-free.
To fix the timing of the impulse is annoying but easy. It's not the varying delay, it's a delay happening while measuring. You can't tell if or how much (or if) the phase changed because of your crossover modification. And that's where you waste your time and nerves.
On some PCs you don't encounter these issues or maybe while measuring which are not critical. You can encounter timing problems because of other things too, very 'popular'🙄 are update checks and antivirus background file checks and the like. Instant messengers can sabotage measurements too, these things are just not measuring system inherited and can be avoided.