DIYMobileAudio.com Car Stereo Forum banner
1 - 20 of 56 Posts

· Registered
Joined
·
6,546 Posts
Discussion Starter · #1 · (Edited)
So as topic says
I want to chat about measurement gear, mics, REW, measurement methods, impulse responses and impulse centering, fir,iir,APF….. 😙

so last few months I’ve just lost confidence in the usb mics from Dayton and minidsp.
they produce good sounding results but it’s not great sounding. It’s just not.

back in the 90s and early 00s I used my audio control 3050A RTA with stereo un-correlated pink noise on CD from the 95 or 99 IASCA CD. Period. I would reject any other noise source because I could not trust it…..

I don’t know what audio control did back then with there noise and IASCA. But they had something very very special. And it’s sad to see that level of accuracy be replaced by these chintzy little gadgets while we wave them around like a magic wand and they’re completely in accurate.

back in the 90s when I first tuned my Jetta with a pair of Alesis MEQ-230s. One on subs and one on highs. I could make a tune that had a harmonic balance that was simply amazing. Vocals and synth had a tone (spectral balance) that was in Harmony and it gave you a sense of realism that would draw emotion from the artist… listening to Mary Chapin carpenter stones in the road on CD on 16/44 and it kicked the crap out of everything today…

here we are 20 years later with a laptop and these devices, Dirac, room correction, fir whizz filtering….. and all you need is a simple 1/3rd octave eq and a good analyzer and a proper install and you didn’t need all this stuff. But we have it. And it’s fun….. but we are not making accurate measurements as a community and promotion of almost gimmickish items.

Mics: usb vs xlr….. I’ve said before , I have the umik1, UMM6, a CM-10, a ECM-8000 and a Dayton IMM10 (phone mic) and now an array of 8 total cm-10s. The cm-10 paired with the audio control 3050A is the only setup I trust…. Hum weird. It’s 30 years old and better then everything else I have.

are the usb mics close. Yes, usually very close. But it’s not precise and I would suggest there’s some anomalies that change the responce…..Hot cold tge compute it’s on, etc etc…. Assuming everyone’s software is configured properly as well.

im a dirac live user. I also make my own corrections in rephase fir. I need a accurate mic and rig. And I believe xlr is choice and you get what you pay for…… I’m happy with the performance of the cm-10 it’s a 150$ mic. I want a earthworks mic , sure….or some 3000$ super high end mic. But I think we can make it work with less expensive stuff….

i think everyone needs to have a reference setup that never changes. Something that is known good and made for eachother. Weather it be a 30yr old rta or one of those handheld analyzer’s with built in mic. Something calibrated and not a plethora of laptops and cables and brands mix matched…. And if it be that, put it up against something that is known good and make an offset.

I’ve personally experienced deviations with usb mics. One day it’s working and next it’s not… same laptop, same REW install. Nothing changed except the weather.

all I know is when I work a set of Eqs with live rta, I shoot for my target to be within 2db max. Any stray peaks needs to be tamed. I don’t just knock the peaks down and leave a jagged response. Although sometimes I tune a car and it comes back months later for a upgrade and retune and it’s changed…. Speakers once broke in should retain the same thing for at least a year.

Why does the umm6 have better sounding HF then the umik1… like the 0deg file, on the rta there super close, but different sounding. There both the same standard, why do they produce a different sound. So so strange. Both 48k mics with Java drivers. Strange

why does the cm-10 have such a different shape and produce a tune that sounds 10X better then minidsp and Dayton… the only problem is getting XLR into your PC faithfully. Like my CM-10 on REW matches the audio control meter to 10khz. Then there’s a cut! I assume it’s the tascam interface, looks like all the analog inputs go directly to an ADC when I open it up. So it’s not the soundcard rolloff from the analog outputs…. What gives…. That’s just one thing I caught. Amongst the many small deviations I run into non stop.

i can’t wait to get my Roland sound card, and fire up some smaart. I hope so much it’s accurate and consistent.

pointing mic up seems the preferred method for reflective environment. The 90deg cal is choice.

And measurements with Dirac. Tonal issues completely solved using xlr mic. Very very different looking responses between xlr and usb. And the xlr mic makes Dirac kick butt!!!

What I’ve been doing lately is using my 3050a and cm-10 and in Dirac make a target and make anchor points at every 1/3 octave ISO. And run Dirac and measure with pink noise from iasca cd and use my Dirac points as a 1/3 octave EQ.

here’s a pic with a Dirac tune using same mic as the 3050A… RTA is set to slow with 2db per step. Dirac is deadly accurate with same mic except 10-20k I had to bring up….. see these comparison pics….they match surprisingly well otherwise, BUT STILL ERRORS that I wouldn’t have known otherwise

bringing that up make the whole harmonic balance fixed and the sound is absolutely breath taking real sounding….

anything anyone can add as far as tips or things they do for a accurate tuning rig please share

Automotive lighting Hood Automotive design Automotive tail & brake light Vehicle

World Gadget Line Electricity Rectangle
 

· Registered
Joined
·
1,157 Posts
Sounds like you are referring to the calibration of each mic. The solution is to use a reference mic to calibrate to every time. Then your gains and trims will be identical and repeatable every time.

I remember in the early 2000’s I had an NTI acoustilyzer and mini-spl mic and my system sounded like Jesus tuned it. Those were the days.
 
  • Like
Reactions: oabeieo

· Registered
Joined
·
6,546 Posts
Discussion Starter · #3 ·
Sounds like you are referring to the calibration of each mic. The solution is to use a reference mic to calibrate to every time. Then your gains and trims will be identical and repeatable every time.

I remember in the early 2000’s I had an NTI acoustilyzer and mini-spl mic and my system sounded like Jesus tuned it. Those were the days.
jesus , he’s a baller tuner … haha 😂

yeah there inconsistencies drive me crazy….
I have cal files, that **** don’t matter…..
some days I get discrepancies for no reason besides temp. Idk what the variable is

it’s like, old faithful is my reference and all this other stuff only gets me in the ballpark….. it’s rather cumbersome
 

· Registered
Joined
·
1,157 Posts
Does the cmt-10 require phantom power or something - why even bother with the others?
 

· Registered
Joined
·
6,546 Posts
Discussion Starter · #5 ·
Does the cmt-10 require phantom power or something - why even bother with the others?
yeah 12-48v

I’m getting to that point. The sheer curiosity of how tight are these things has me wondering.

I mean, unless inconsistent to 2db in spots is average, then a good cal file should make it set.
But even the xlr mics have issues, having to get the signal with no errors to the software, and I noticed a higher noise floor with xlr also.

Idk. It’s very nit picky absolutely. Im Just really paying close attention to detail on some of these measurements and as long as I force myself to be aware I can spot the problems.

it’s so easy to just assume the data is right. I mean, it’s right there in front of you see……(as someone looks at there measurements)
 

· Registered
Joined
·
1,157 Posts
yeah 12-48v

I’m getting to that point. The sheer curiosity of how tight are these things has me wondering.

I mean, unless inconsistent to 2db in spots is average, then a good cal file should make it set.
But even the xlr mics have issues, having to get the signal with no errors to the software, and I noticed a higher noise floor with xlr also.

Idk. It’s very nit picky absolutely. Im Just really paying close attention to detail on some of these measurements and as long as I force myself to be aware I can spot the problems.

it’s so easy to just assume the data is right. I mean, it’s right there in front of you see……(as someone looks at there measurements)
Can you manually calibrate them all so they’re the same? It doesn’t seem too hard to do - there’s apps out there for this.

Here’s a site that plotted FR for 1000 mics:


Rectangle Font Parallel Slope Screenshot


Their conclusion?
If you want to measure more accurately than +/- 2 dB, an individual calibration of the measuring microphone (possibly including preamplifier) is essential!
 
  • Like
Reactions: oabeieo

· Banned
2016 Lexus Ct200h Fsport
Joined
·
2,490 Posts
  • Like
Reactions: oabeieo

· Registered
Joined
·
6,546 Posts
Discussion Starter · #8 · (Edited)
Just get high end and know you got a good mic lol

Baller!!!!


Can you manually calibrate them all so they’re the same? It doesn’t seem too hard to do - there’s apps out there for this.

Here’s a site that plotted FR for 1000 mics:


View attachment 332251

Their conclusion?
If you want to measure more accurately than +/- 2 dB, an individual calibration of the measuring microphone (possibly including preamplifier) is essential!
zum glück verstehe ich deutsch

I get what your saying about a cal file. What if the issue is inconsistent hardware. Chintzy capsules and an in-line xlr to usb ADC that couldn’t be of high quality….

i think he said it right, you get what you pay for

these usb mics that are under 150$ are good, they work, but not dead on balls accuracy. And I don’t think it’s a cal file issue
 

· Registered
Joined
·
6,546 Posts
Discussion Starter · #10 ·
Nice rant, but I suspect if you could go back in time and hear your 1980s/1990s car(s) again, you would find that your memories of them are better than the actual sound.
Ha! Yeah I’ve thought that. But I also remember pretty well when I have a home system that sounds like it should, and listening to the same CDs as I did back then….

but like back then, I have 4” polypropylene cone midrange and aluminum tweets and heavy midbass… I had a hand made passive crossovers all 1st order (except tweets were 2nd order)

i think there was something about resonant type AB amps with very high damping, and a polypropylene cone with insertion damping from passives…. It was a recipe for really good sound in a car….

i had no signal delay. Staging was good but diffuse as hell…. But my goodness tonaly on point…

so maybe ….. maybe not ……
 

· Registered
Joined
·
6,546 Posts
Discussion Starter · #11 ·
This is a quote from another forum. Wow!!!
okay I think I found my culprit maybe!!!
i am guilty as sin for leaving mics in a hot car. I do it all the time…. I’m here it talks about the mics becoming depolarized and loosing sensitivity and other problems….

Thank you. this is my opinion too.
The alternate position is that it is a waste of time to worry about this, just more audiophile obsessiveness like silver speaker cables or precious metal plated mains connectors.
Admittedly there are other, worse sources of problems - if room modes are ± 15 dB then perhaps a few dB error in a measurement doesn't matter?
My view is check what is best practice from professionals in the area, use the benefit of their collective experience.
After we check best practice we can reach a conclusion about what we need.
Before we discuss that there's a bit of introduction.
[Start of the data dump]
Practically all measurement microphones are condenser mics.
This is because condenser mics are basically quite simple - there is a membrane that forms one plate of an air capacitor (or condenser as it used to be called).
Sound pressure moves the membrane and the capacitance varies, easy!
Or not so easy, if you calculate the size of the deflection it's sub-microscopic.
But the simplicity means that basic ones can be made very cheaply and carefully made ones can have their performance calculated quite precisely from first principles.
There aren't too many complexities to mess up the calculation, or the response - a flat line is mathematically simple, and exactly what we want.
The membrane was initially, and in precision mics still is, made of thin metal.
The first ones used aluminium alloy but switched to nickel (for reasons to be discussed).
Some modern mics use stainless steel and titanium also works well.
Cheaper mics use plastic, almost always polyester (Mylar) that has been metalised, gold is oxidation resistant and works well, plus adds sales sparkle for medium priced mics.

There needs to be circuitry to "measure" the capacitance and the only down side of the condenser mic is that the circuit needs to be very close to the membrane so as not to be swamped by cable stray capacitance.
So we need to supply power to the microphone, and from this comes one of our problems because there are many different, mostly incompatible ways to do this.
One way to measure the capacitance is obvious, apply a sine wave and measure just like a capacitor tester.
This will work down to DC and is actually used in speciality, very low frequency mics.
But all common condenser mics use a simpler system - apply a charge to the membrane and then as it moves the potential will vary.
The charge can be supplied in either of two ways.
The first measurement mics just used a 200 V line from the power supply.
Permanently charged plastics, called electrets, were later developed and replaced the 200V, similar to the way field coil speakers were replaced by Alnico, just a little latter.
To measure the potential we need hi impedance, JFETs (a special type of transistor) for all modern measurement mic, tubes in the old days and for retro studio mics.
The circuitry is built into the body of the mic directly behind the membrane.
The first measurement mics had the membrane as part of a "capsule" that included all the mechanically precise and delicate bits.
It screwed on to a body that incorporated all the electronics.
The dimensions were standardised by Western Electric back in the 1930s so are all inch units and Unified Special thread form.
The default was then a 1" capsule, modern precision mics tend to be 1/2" or 1/4", the thread is UNS 60 TPI.
Here's a cutaway that's typical of a metal membrane electret capsule.
M215.PNG



Cheap microphones are built with a small pill shaped capsule that often has the JFET already inside.

So what does this all mean for accurate measurements?
The simplicity of the condenser mic means there are only one or two main sources of error.
The electronics can be made very flat and linear compared to the mechanics, so should not be a major contributor except in very cheap mics.
1. For an electret mic the stability of the electret itself may be a limitation, it can lose its polarisation and the sensitivity of the mic will decrease.
The best electret materials are quite stable at room temperature and low humidity.
ACO Pacific claim >250 years / dB and Bruel and Kjaer similar.
This is far too small to worry about for speaker measurements.
The only problem is that irreversible depolarisation occurs much faster as the temperature and humidity increase.
For reasonable levels of temperature and humidity, say, tolerable to a human, this shouldn't be a major problem.
Don't leave it in a hot car in the sun and that kind of obvious stuff, maybe try to avoid outside measurements where the mic is in the sun on hot days.
Luckily this won't have much effect on the frequency response which is our main interest as audio buffs.
Any depolarisation will show up in a sensitivity test.
That's a problem for noise level measurements that are the central concern of many users but not so serious for us.
At least it makes the problem easy to detect with a simple test.

2. For a plastic membrane microphone the membrane will be sensitive to humidity and temperature.
I mentioned earlier that the first measurement mics had aluminium alloy membranes but they were discontinued.
The problem was that as temperature varied the membrane expansion didn't match the rest of the capsule, tension would vary and thus impact frequency response and sensitivity.
Plastic membranes have this problem but more so.
Whereas a stainless membrane on a stainless steel body, as shown above, will vary very little.
I am not sure exactly how serious the problem is because I can't find much data for plastic capsules.
But the lack of data is itself a clue, I expect the manufacturers would publish it if it looked nice.
There don't seem to be any Class 1 Sound Level microphones with plastic capsules either.
So my assumption is that they can't meet the stability standards.
While the metal capsule manufacturers who meet Class 1 are very happy to announce it.

So, now, back to industry best practice and what it tells us.
The main source is IEC standard 61094 on Measurement Microphones.
This is in 8 parts, isn't cheap, and much of it is of limited use for practical speaker measurements.
So it isn't cited much in the speaker industry and I haven't read it all.
But here's a quick summary for context.

Part 1. Specification of Lab Standard mics.
Part 2. Calibration of LS mics Pressure response. (Reciprocity method)
Part 3. Calibration of LS mics Free Field response.

These parts are about primary standards in labs, as references for the mics that are used for real work.
The calibration is done with a different technique to that used to calibrate a production mic.
It is tedious, slow and difficult, therefore expensive, so only used to establish the reference.
But just in case you ever see the term or wonder how a reference standard is created in the first place.
It relies on "reciprocity", that is, just as a voice coil style speaker can be used as a dynamic microphone, so a condenser microphone can be used as a tiny electrostatic speaker.
The physics ensures that the efficiency is equal in both directions, so, with several mics used both as speakers and microphones it is possible to measure and then calculate the response from fundamentals.

Part 4. Specification of Work Standard mics.
Part 5. Calibration of WS Pressure response. (Comparison method)
Part 8. Calibration of WS Free field response. (Comparison method)

Finally, the bit more relevant to speaker measurement.
Part 4 discusses the accuracy and stability required to be meet the specification.
This is often called a Class 1 or Class 2 microphone.
Technically this is not quite correct, those terms come from the standard for Sound Level Meters - IEC 61672.
But sound level measurements (for noise annoyance and OH&S) are far more common than speaker measurements, so this standard is better known and people tend to use the Class nomenclature.
Also, many measurement mics are built with the same capsules used in SLMs, so the terms tend to transfer too.
Measurement mics should use different terms, a Lab Standard mic is LS, a Work Standard is WS, and the diameters are coded as 1 for 1", 2 for 1/2", 3 for 1/4", finally there is a suffix to denote Free field or Pressure mic.
(If you don't know the difference between the two then there is an excellent tutorial on the Bruel and Kjaer website).
A WS2F is probably the best mic for speaker measurements, for instance it's what Klippel recommend for the NFS.

Part 6. Electrostatic Actuator.
Part 7. Lab Standard Free field to Pressure response.

Electrostatic Actuator is yet another way to calibrate a mic. (Did you think it would be so complicated? It's why I decided to write this!)
And part 7. is yet more Lab Standard stuff we don't really need.

That will do for part 1.
More soon! if people like the story so far, that is.

David”
 

· Registered
Joined
·
395 Posts
Nice rant, but I suspect if you could go back in time and hear your 1980s/1990s car(s) again, you would find that your memories of them are better than the actual sound.
Why do people say this? I've had members say the same thing to me. What I get from this is back in the day when car audio was at it's peak we didn't have good systems until now.

To me, just because things are new and up to date, doesn't mean it's better. It's like an incandescent light bulb, the main concept hasn't changed, maybe minor changes but not necessarily a better product. My point being, there haven't been any major breakthroughs since speakers and equipment were created and refined...the wheel hasn't been reinvented. Everything made today is still based off the same basic principles

As to the OP, I feel you and what you have said. The one thing that comes to mind is you stated the only difference that has changed is the weather...that is the biggest and most important aspect of differences. The atmosphere changes make the biggest impact on everything, especially in audio and the way a car performs. This would be my thinking on why you see the differences that you do. it could change day to day, hour to hour. Maybe I missed something that you stated but I would think these factors would be the biggest culprit.

However you may be onto something with what you discovered about the mics and changes with heat/ environment as well, IDK
I feel many older things were better than what's available today. However, to get equal quality today takes a lot of money it seems. Times have changed too quickly and the economy is too messed up IMO. Today, people want everything cheap...so manufactures have to have something that fits the bill and we know what that gets us. Too many variables at play
 

· Registered
Joined
·
6,546 Posts
Discussion Starter · #14 ·
[/QUOT[/QUOTE]
Before talking about mics that change with temperature, did you ever look at the speaker response how much those change with temperature? As they can change a lot even a few degrees can make a lot off difference!
Yeah that is definitely a very good point….. 2.5k and 200 seem to be changing a lot, with ambiant temp , and speaker temp definitely thins things out , I don’t know what frequencies tho…. It would be a study, and would be system dependent. So no one size fit all. I usually measure with a min of normal playing…. Not nuts out volumes tho.
 

· Registered
Joined
·
6,546 Posts
Discussion Starter · #16 · (Edited)
Why do people say this? I've had members say the same thing to me. What I get from this is back in the day when car audio was at it's peak we didn't have good systems until now.

To me, just because things are new and up to date, doesn't mean it's better. It's like an incandescent light bulb, the main concept hasn't changed, maybe minor changes but not necessarily a better product. My point being, there haven't been any major breakthroughs since speakers and equipment were created and refined...the wheel hasn't been reinvented. Everything made today is still based off the same basic principles

As to the OP, I feel you and what you have said. The one thing that comes to mind is you stated the only difference that has changed is the weather...that is the biggest and most important aspect of differences. The atmosphere changes make the biggest impact on everything, especially in audio and the way a car performs. This would be my thinking on why you see the differences that you do. it could change day to day, hour to hour. Maybe I missed something that you stated but I would think these factors would be the biggest culprit.

However you may be onto something with what you discovered about the mics and changes with heat/ environment as well, IDK
I feel many older things were better than what's available today. However, to get equal quality today takes a lot of money it seems. Times have changed too quickly and the economy is too messed up IMO. Today, people want everything cheap...so manufactures have to have something that fits the bill and we know what that gets us. Too many variables at play
thanks, yeah, I think speaker temp ambient temps are the single largest variable….

i believe (no scientific reason) 1 min of moderate dynamic music before measuring, and control the mic temp as much as possible, take care of the mic, don’t drop them etc etc and make conditions as good as possible….

take 10min cool downs after playing noise for more then 3-5min …. And always put a fixed 13.5V supply on the main batt when tuning (not a charger, although better then nothing)
 

· Registered
Joined
·
395 Posts
thanks, yeah, I think speaker temp ambient temps are the single largest variable….

i believe (no scientific reason) 1 min of moderate dynamic music before measuring, and control the mic temp as much as possible, take care of the mic, don’t drop them etc etc and make conditions as good as possible….

take 10min cool downs after playing noise for more then 3-5min ….
I will be curious if you can find a meaningful answer but I think you are on the right path. best of luck in trying to figure it out.
 

· Registered
Joined
·
6,546 Posts
Discussion Starter · #18 ·
Sent my umik2 to cross spectrum, got it back today

we will see how that goes

the cm-10 I have decided has a slow rolloff past 2.5k (and I almost think that is normal and these other mics are too hot) but now I realize the flat response should sound hot in the HF and the cm10 has a small rolloff

so this brings up “voicing a mic”

it’s crazy how listening to a flat Dirac tune, I am really listening to the mic….. it’s like auditioning a mic!

cm-10 just sounds better imo
Umm6 next best thing

ANU will let me borrow his M30…. This will be a true test….. if his mic sounds that much better
I will absolutely buy one…..

i get the feeling his mic is going to whomp ass on everything I own …… I’m excited to try it actually
 

· Premium Member
2006 Hyundai Sonata V6
Joined
·
784 Posts
I'm curious about how you connect XLR mics to a laptop. I've seen XLR to USB adapter cables, but it seems some sort of interface box connected to laptop analog inputs is more usual?
 

· Registered
Joined
·
6,546 Posts
Discussion Starter · #20 · (Edited)
I'm curious about how you connect XLR mics to a laptop. I've seen XLR to USB adapter cables, but it seems some sort of interface box connected to laptop analog inputs is more usual?
so I have actually been doing a lot of study on this

I have a very nice soundcard that is 96k and a tascam soundcard that is 44.1k

The tascam which I thought was a pile of crap , low and behold I discovered it I turn its latency settings down all the way to lowest latency it performs rock solid, better then the 96k “high performance adc”

The solo or duo are good from what I hear and work okay but nothing to write home about… specs similar to tascam , good, will work and do the job, but not amazing

The only one that I have heard rave reviews from is this studio six iusbinterface2 (at600$)

so you figure 600 for a good mic and 600 for a interface and you should be golden right

i suggest there’s more to it

i think low latency is key tho for measurements

I got the idea to try low latency here from this article

it’s a boring read, but just read it , it’s a wealth of info

EDIT

ITS THIS

Dayton EMM-6 measurement microphone, calibrated by Cross-Spectrum Labs
 
1 - 20 of 56 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top