What's the most neutral sounding mic below 1000$?

[rob aylestone]

I've got a Behringer and another similar one somewhere that I used to use for PA setting up, but I have long held the belief that mics do have something weird going on with them, over and above a simple frequency/amplitude plot. Has anyone ever bought a microphone that has a specific plot that was very similar to another mics and heard two totally different beasts? Science tells us our ears cannot detect a less the 2dB change in level, so seeing perhaps a little bump at 5K, you imagine this and then discover you have a rather thin and weak sound, yet another mic with virtually the same trace on the graph behaves much better on the voice you have? We treat these graphs as solid science, yet clearly, mics sound different - and often the little bumps and dips are 2dB or less up or down, so we should not be able to hear those? Here are three graphs - they are currently on sale for £30, £500 and £4500. I can't tell which is which?

 

[R D Smith]

Way back in 2010, Harvey Gerst did a cheap mic shootout and the Behringer ECM-8000 was one of the mics he tested. That was one of the first threads I read when I joined this site several years ago. Long story short, I bought the Behringer and MXL V67g. I thought his analysis and tests were very insightful and well done.

Harvey Gerst Mic Test

 

[TalismanRich]

One problem with looking just at the published mic response graphs is that they are smoothed out to a large degree. I took a short section of a song, and looked at the response curve at 3 different level of smoothing. If this was going in a microphone sales brochure, which do you think they would choose? You never see a company say 1/3 octave smoothing or 1 octave smoothing, and if it moving averaged or Gaussian. There are multiple ways to analyze data!

 

[TalismanRich]

It always amazed me that most audio gear has specs shown like 20-20kHz +/-.01dB, but just about all microphones say 20-20kHz... with no deviations spec. You look at the graph (already smoothed) and there's a +5dB peak at 8 or 10kHz and the bottom end is rolled off 3dB at 40Hz and 10dB by 20Hz.

This is the chart of a $1300 mic by a major manufacturer. 20-20kHz spec. It really tells you a lot, doesn't it.

 

[LazerBeakShiek]

If you are singing, you will find a dynamic hyper to be much more useful. You got that 6" of distance to work with and a microphone technique in that range.

I don't see intimate recordings from a LDC, just sound reinforcements. Like you cannot stand back and sing rock n roll.

 

[rob aylestone]

It always amazed me that most audio gear has specs shown like 20-20kHz +/-.01dB, but just about all microphones say 20-20kHz... with no deviations spec. You look at the graph (already smoothed) and there's a +5dB peak at 8 or 10kHz and the bottom end is rolled off 3dB at 40Hz and 10dB by 20Hz.

This is the chart of a $1300 mic by a major manufacturer. 20-20kHz spec. It really tells you a lot, doesn't it.

View attachment 152914

This is exactly the snag - flat response through the primary range then all above a bit brighter. Some of the others seem to have bad piano responses - as in where some strings seem to ring out differently to the neighbours - making chords sound a bit odd. That trace shows an audible increase - 4dB or so Above 5K doesn't do much for male voice, far more for female and for instruments make a big difference. Enough to make people go wow, a nice mic, but would they have said the same if you had just done the same thing with the EQ?

 

[plautus]

manufacturors lie abotu their frequency response. On the first graph u can see the fake (hgihly modified) frequency reponse curve of the Rode nt1 4th gen, On the second graphc u see the more closer to real frequency respones of the rode nt1 4th gen.

I never trusted frequency response curves publihsed by these companies not even these. People are realy gullible....

 

[rob aylestone]

No Sir, that is quite incorrect. The vertical axis of the two plots is so different and you have ignored this totally! One has a spread of 50dB, the other 25, and because this is a logarithmic scale the information presented above 1K in both these plots is perfectly compatible. Clearly, the marketing department prefer the flat looking curve - the one the extra wide amplitude scale reveals, but it's a science plot, that relies on a level of understanding in the viewer, and we all know many musicians/recordists are technically very advanced, while others think digital watches are a really neat idea. Mic plots have NO standard, so you cannot compare them without some thought. They are useful tools, but need analysis to read. You cannot use the word fake or even highly modified. I'm not defending Rode for using test methodology that appears to suggest something the mic isn't, but the plot makes perfect sense and I suspect is perfectly accurate. They have 2dB increments in the horizontal scale in one and bizarrely use 1.6dB per increment in the other. If you take the results of that plot and rescale to the other, they are very similar - just a very odd way to display the data. I suspect that legislation exists to make false data claims rather silly, but the method of display is not illegal, or dodgy - just a bit 'sharp'.

 

[plautus]

i give up on this forum, the mod is an ignorant consumer who cant read graphs and falls for marketing lies.

 

[rob aylestone]

Sadly, the mod is a 50 year professional, who is also a qualified teacher in Music Technology and certainly does understand graphs. He is also immune to the non-science based flowery language promoted by some members. He also posts as an individual, expects no special treatment for the sin of being a mod and has on many occasions, been wrong. This, however, is not one of them. I simply do not agree with your statement that manufacturers lie. Suicide for a tech company. They do of course, present material in favourable ways. Not my fault you don't appreciate the difference.