Random inaccurate stupid measurements of CLR and K10

so if it's poor information, why post it? how does it inform anybody? not trying to be a jerk, but why not post something rigorous, accurate and useful? the internet dearly needs more stuff like that.
 
This is the equivalent of people posting "this amp has girth and juicy tone", sure on its own it doesn't mean much but if a hundred people start to say/find the same thing then there may be something to look into
 
Well, I did the best I can with the measurements, they're all average of 4 15 second sweeps, but my measuring equipement costs less than $500 and this was not done in a scientifically accurate reverbless chamber that costs millions, but outside, where I painstakingly carried all these heavy things and cables by hand, to get away from walls and ceilings, so you'd get the best representation of the speaker, not the room. So scientficially speaking the measurements are totally worthless, and I don't want to say they are accurate. If I was an internet jerk, I'd claim they are almost science or at least usable, and Genelec is unsurprisingly the flattest in the crucial mid frequencies and K10 has a problem in the crossover area, and CLR is pretty good. But I'm not. Also Jay Mitchell would kill me for making less than perfect measurements with anything less than a 99,9% reverbless space, even if it shows his product is flatter than another speaker, and a Genelec is flatter than either of them, and therefore it looks legit. Instead I'm just presenting the data as is and let you do the interpreting instead. Ok ok, it's useless shit, I'll delete it.
 
Last edited:
I hope you meant 4-15 millisecond gated sweeps.

Edit: Only reason I ask is you took the time to write out "seconds"
 
"15 milliseconds is too short a time to do any good." That's what she said. Also, I have no idea what you are talking about. One 66hz wave takes 15 milliseconds, so I can't see how one would measure anything in such a short time. Happily however, I don't have to know anything about such things in order to use Room EQ Wizard's default settings.
 
This is the equivalent of people posting "this amp has girth and juicy tone", sure on its own it doesn't mean much but if a hundred people start to say/find the same thing then there may be something to look into

Posting knowingly inaccurate measurements is misleading and is a type of objective information. Posting an opinion is subjective by it's very nature. The two types of information are very different. The hypothetical situation of having lots of correlating opinions is empirical and this type of data is never left to one subject, opinion, or finding, whereas a scientific approach involving measurements must be rigorous to not be potentially misleading.

Not all information has the same value.
 
Yea, it's better to post "this has hyped mids" than a measurement that didn't cost a million. What the fuck, this is worse than photography forums
 
Last edited:
Yea, it's better to post "this has hyped mids" than a measurement that didn't cost a million. What the fuck, fuck this forum, this is worse than photography forums

It wasn't a comment on your hard work. It was a comment on the idea that all information has value. It doesn't.
 
"15 milliseconds is too short a time to do any good." That's what she said. Also, I have no idea what you are talking about. One 66hz wave takes 15 milliseconds, so I can't see how one would measure anything in such a short time. Happily however, I don't have to know anything about such things in order to use Room EQ Wizard's default settings.

Such long sweeps, in the range of seconds, start to become pretty meaningless when trying to make precise audio measurements. Listen, none of us have access to million dollar anechoic chambers and $100k+ test equipment, so we need to cheat a little to make our measurements as precise as possible. Using gated measurements is one of those methods (There seem to be three main methods for those on a budget: LMS (gated toneburst), TEF (swept sinewave with time-delayed swept bandpass filter), and MLS (Maximum Length Sequence) testing. Read more at Measuring Loudspeakers, Part Three | Stereophile.com) When measuring for 15 seconds outdoors, just think of all the extraneous noise that is being added to your averaged measurement! All that added stuff skews your freq. response chart to being pretty much meaningless.

The main reason for a gated measurement is to remove reflections from your measurement. Even when measuring outdoors, you will get a ground reflection in your measurement. Here is a pretty good link with steps to follow for taking semi-accurate freq. response measurements in room using a UMIK-1 microphone and REW. If you follow those steps, you should be able to make a decently accurate freq response plot without having to haul your gear outdoors. Although, if you can swing a calm day, gated measurements outdoors are the way to go.

Edit: I'm not trying to criticize your efforts. You've at least made an effort at contributing. I just wanted to give you some advice on how to obtain much more useful measurements, both for your own personal joy and maybe for the benefit of the community, assuming you haven't been soured by the responses so far and choose to continue sharing!

Double Edit: You are correct about the gated measurements. They are more accurate, but you give up frequency resolution due to the shorter sweep time. You just need to be conscious of the compromises you are making.
 
Last edited:
Posting knowingly inaccurate measurements is misleading and is a type of objective information. Posting an opinion is subjective by it's very nature. The two types of information are very different. The hypothetical situation of having lots of correlating opinions is empirical and this type of data is never left to one subject, opinion, or finding, whereas a scientific approach involving measurements must be rigorous to not be potentially misleading.

Not all information has the same value.
To be fair there is absolutely nothing inaccurate about this information. It is exactly the resulting data captured from the test performed.
 
Hey thanks, I knew the REW uses windowing but the default values are quite long. I had the speaker and mic at about 150cm from ground and mic 1m from speaker. Next time I'll try to hang the speakers from a tree as high as possible. (Thats what Genelec did so it must be good!) :)

I did 12 different lenght sweeps as a first test, just to test the effects of external noises, but they all agreed perfectly with each other, and I thought I read somewhere the rew is smart enough to only listen to the right frequency at the right moment, for blocking out noise. Is that the TEF method?
 
Last edited:
Sorry, I don't use REW, but have visited their forums. Your best bet is to seek advice over there on the peculiarities of REW.

Edit: I think the freq response graph given in the Atomic CLR manual was measured at a distance of 2m. If your frequency response plot does not look like the one provided in the manual when you are measuring from 2m then you are probably doing something wrong. I don't believe it was taken at a certified testing facility, but I'm going to just go on faith that the equipment and methodology they used were of a sufficiently high caliber to trust, certainly better than the equipment and facilities at our disposal. So, use the plot they give you in the manual as a good target to shoot for when refining your measurement technique.
 
Last edited:
To be fair there is absolutely nothing inaccurate about this information. It is exactly the resulting data captured from the test performed.

...and if the test is sloppy then the information is not a true representation of the performance of the item in question. This is the definition of inaccurate.

To the OP: Suspending the speakers is actually a great way to measure. There are always ground reflections that interfere with the measurements and they can skew frequency response to a surprising degree. For example, a point source in a free field radiates it's sound power equally in a spherical manner. When a flat surface is close by some frequencies are reflected back, which in some frequency bands can result in a doubling of sound pressure at the measuring point as the radiation is effectively hemispherical, and this can result in phase interference, giving a skewed response. If doing the measurements outdoors not only is wind noise a factor, but air currents defract sound upwards or downwards depending on the difference in temperature. Just food for thought! :)
 
Last edited:
I would have posted it as a frame of reference info and not hard data given the lack of correct measuring environment. It would still be a good experiment for others to see the results you came up with in the environment you had to work with. No one plays there speakers in a perfect setting anyway so any skewed data should be the same for any speaker used providing you use the same measure method on each one.

I say post your results and forget about what others think ;)
 
I'd be careful with the whole "hanging the speaker" in a tree thing. Accurate audio measurements require that your setup be completely still. Just picture in your head a sine wave coming out of a loudspeaker pointed at a mic. Now take the mic and loudspeaker and move them back and forth along the x-axis. How accurate do you think that measurement would be in measuring the sine wave's frequency? Now think of how you're going to ensure your test rig remains absolutely still while hanging in a tree!?!?

Your best bet is to start in your house where you can control more variables and graduate to outdoors when you have refined your technique.

Edit: I agree with Sixstring. It would be nice to have a log of your data so we can see your test results become more accurate as you refine your technique over time.
 
Last edited:
Back
Top Bottom