Reply to thread

Message

<blockquote data-quote="Ivan Beaver" data-source="post: 82749" data-attributes="member: 30">Re: Listening Get TogetherYou will get &quot;data&quot;-it just doesn&#039;t mean much of anything if the window is to small.&nbsp; The shown data can vary over a huge range-and in some cases make the &quot;measured&quot; response look worse than it actually is.What I like to do is to use a resolution that is 1/10th of the lowest expected freq of the cabinet.So it the expected low freq is 50hz-I like to have 5 hz resolution.&nbsp; It is one thing to &quot;measure&quot; a cabinet-and quite another to do it properly and display it properly.When a graph that is displayed and is obviously full of &quot;errors&quot; (such as going out to 100KHz for a loudspeaker), to me that says that the person &quot;measuring&quot; doesn&#039;t know what they are doing-so I cannot trust any of the &quot;data&quot;.And showing low freq with a short time window-is the &quot;data&quot; actually showing the response or just the noise of something nearby-did the person measuring have enough signal to noise ratio-or did they just &quot;accept&quot; what was shown on the screen as being accurate?You should ALWAYS QUESTION your measurement.&nbsp; But that of course requires a knowledge of what you are doing and being able to tell if the data is valid or not.If you don&#039;t know what to show-then how valid is it.&nbsp; Things like 100dB vertical scales are another indicator-what do you expect people to get from that?Measurements are one thing-USEFUL measurements are quite another.</blockquote>

[QUOTE="Ivan Beaver, post: 82749, member: 30"] Re: Listening Get Together You will get "data"-it just doesn't mean much of anything if the window is to small. The shown data can vary over a huge range-and in some cases make the "measured" response look worse than it actually is. What I like to do is to use a resolution that is 1/10th of the lowest expected freq of the cabinet. So it the expected low freq is 50hz-I like to have 5 hz resolution. It is one thing to "measure" a cabinet-and quite another to do it properly and display it properly. When a graph that is displayed and is obviously full of "errors" (such as going out to 100KHz for a loudspeaker), to me that says that the person "measuring" doesn't know what they are doing-so I cannot trust any of the "data". And showing low freq with a short time window-is the "data" actually showing the response or just the noise of something nearby-did the person measuring have enough signal to noise ratio-or did they just "accept" what was shown on the screen as being accurate? You should ALWAYS QUESTION your measurement. But that of course requires a knowledge of what you are doing and being able to tell if the data is valid or not. If you don't know what to show-then how valid is it. Things like 100dB vertical scales are another indicator-what do you expect people to get from that? Measurements are one thing-USEFUL measurements are quite another. [/QUOTE]

Verification