[07azhhr]

I believe the DBX mic's are corrected in the hardware of DBX like the driverack or such.. So the mic should be calibrated for correct FR measuring

If the played freq. is visual on your screen it means is "hears" the correct freq., the calibration is to adjust the sensitivity of the mic so it shows the "actual relative" level

Thank you Woosey. I kinda figured the cal was good since it matches on the screen. I just figured I would ask since Erin made sure to emphasize the need for a cal file if going for a target curve.

I am not sure what you mean by the mic's hardware or driverack. It is just a black version of that Dayton in Erin's video (well atleast it has the same shape but in black lol) ran thru a micmate then on to the laptop.

 

[req]

awesome stuff erin. i about about to go to tintbox's this weekend to play around and ill snag REW and play around haha.

 

[ErinH]

Hi Erin (I hope I am spelling your name correctly :worried: )

First of all THANKS for a lovely walk through with the REW soft.

I've just downloaded beta version 5 build 13 from their forum and I am about to use it with the new MIC from MiniDSP the UMK-1

UMIK-1 | miniDSP

Up so far I was using True-RTA in conjunction with a RadioShack SPL meter which is actually a standard Phillips MIC (just use the standard calibration file)

I really wanted a better MIC so this is why I am ordering the UMK-1 and I liked the way you used the average calculation, I also tend to balance left and right and take separate measurements of each side and as well each element on its own tone by tone and then at the end run a pink noise test.

Of course - I have a bit 1 so I do not go beyond 1/3 octave but as well, for a car, I do not see a need to go beyond that, right? :surprised:

Any way - I would like to know, be tipped by you as for how to use RTA measurements and understand from it if I have any phase or time alinement issues -> I would tend to believe that such issues would be easier to debug/troubleshoot using a scope no?

Thank you very much!

Eddie (please forgive my englaize - I am not a native speaker of this language)

Use whatever mic you want. ecm8000 works well and you can get them calibrated by Herb @ Cross Spectrum Labs. The Dayton mic I linked will work as will the MiniDSP mic you linked. There are likely numerous others I don't know of or haven't seen. As long as you have the means to use it (ie: you don't need special equipment) and a calibration file is provided, get what you can afford and call it a day.


As for phase/ta, I'm not really sure how you would use a scope to determine the acoustic alignment at your seated position, relative to each speaker.

 

[ErinH]

One question, though: I noticed in the video that all your measurement snapshots are instantaneous. I don't have it in front of me, so this is going by memory, but there's also an option to do multiple samples (4, 8, 16, etc.) for each measurement. Is there any benefit (or detriment) to doing so?

Yep. I actually thought of that after I posted the video (on my way to work). I'll have to check this out again. When I used TrueRTA, I use averaging to essentially slow down the response (keep it from jumping so wildly). That number was usually '50'. But, I'll need to go back and see how REW is set up for this. I really don't use it's RTA feature. I only use the impulse function because, frankly, it's easier as odd as it may sound. I don't have to worry about pausing the CD when I am not measuring. I like using the built in generator from the impulse method. (trueRTA has a built in generator for all sorts of samples)

 

[ErinH]

Thank you Erin for posting this. This should be a big help. I am looking forward to part 2.

A few questions. I noticed your mic was not at what I would think would be ear level so is it just a matter of getting close? Also does the direction matter for example yours is horizontal. Will there be a difference if the mic is angled up say at 30-45 degrees?

I am also wondering about the mic cal. I have a DBX mic but I do not have a cal file for it. But when I play tones via REW into my hu the response is perfectly matched. For example if I play a 3000hz tone it shows the response at 3000hz. I have tried this at various frequencies and it is always correct but I have not tried all frequencies. If a cal is needed are ther places that I can send it to to get a cal file made?

I believe the DBX mic's are corrected in the hardware of DBX like the driverack or such.. So the mic should be calibrated for correct FR measuring

If the played freq. is visual on your screen it means is "hears" the correct freq., the calibration is to adjust the sensitivity of the mic so it shows the "actual relative" level

Thank you Woosey. I kinda figured the cal was good since it matches on the screen. I just figured I would ask since Erin made sure to emphasize the need for a cal file if going for a target curve.

I am not sure what you mean by the mic's hardware or driverack. It is just a black version of that Dayton in Erin's video (well atleast it has the same shape but in black lol) ran thru a micmate then on to the laptop.

Woosey is right.

Think of a calibration file as not one that calibrates the frequency but the SPL measured at a given frequency.

You need this ability if you want to target a curve of any sort. Otherwise, you won't be guaranteed any accuracy.

For example:
Let's say your mic has a natural rise in response above 5khz (most do). If you don't have a calibration file to compensate for this and flatten this out, then this rise will show up in every measurement. In a fictional world, let's say your car measured flat. If your mic has a rise above 5khz, you'd measure an apparent rise above 5khz and you'd start cutting it to flatten it (or whatever you'd do to target X curve). You'd be cutting something that's caused by the mic; not by the car.

Make sense?

 

[07azhhr]

Woosey is right.

Think of a calibration file as not one that calibrates the frequency but the SPL measured at a given frequency.

You need this ability if you want to target a curve of any sort. Otherwise, you won't be guaranteed any accuracy.

For example:
Let's say your mic has a natural rise in response above 5khz (most do). If you don't have a calibration file to compensate for this and flatten this out, then this rise will show up in every measurement. In a fictional world, let's say your car measured flat. If your mic has a rise above 5khz, you'd measure an apparent rise above 5khz and you'd start cutting it to flatten it (or whatever you'd do to target X curve). You'd be cutting something that's caused by the mic; not by the car.

Make sense?

So where do I get it cal'd at? I don't use any driverack or other device with this mic. Also what about the mic placement/direction questions.

 

[ErinH]

So where do I get it cal'd at? I don't use any driverack or other device with this mic. Also what about the mic placement/direction questions.

it's late (to me) so forgive me for slacking...


you'd have to send it somewhere. Herb @ Cross Spectrum could probably do it. Personally, for the price of a cal + shipping, I'd just pick up the Dayton USB mic and call it a day.

Placement should be near the head area. The video I posted was an 'on the whim' video. I didn't necessarily pay attention to putting the mic exactly where I would if I were measuring and tuning. I plan to cover the nitty gritty with later posts. For now I just wanted to focus on getting people up and running with REW's RTA feature.

Direction will alter the measurement some but consider the environment. It has a more detrimental affect and really overrides the issue of pointing. I've read about pointing the mic up in a diffuse environment but I've not found it to be extremely helpful either way. However, microphones are like speakers, in that they are directional. So, there is some plausibility in placing and pointing the mic for higher frequency content. That said, at the end of the day, it's kind of a wash. A good head space average will yield good results.

When tuning via RTA you have to realize that the target curve may or may not result in the best sound you've ever heard. It really depends on a lot of things. But, if you go with this standard method of measurement, you at least get close... and up through about 5khz, you get a really good response. Above 5khz is typically where I have to do the most final shaping by ear to tailor things to my seated position. RTA'ing it gets you at least 70% there (obviously a made up number but you understand what I mean). And once you get experienced and better understand the correlation between measurement and subjective likes/dislikes, you'll get better at closing the gap from 70% to 90%. I still believe a well trained in will always be able to fix flaws a mic cannot, simply due to the environment you're measuring in.

 

[ErinH]

To anyone who has watched the video, be honest...

Did it help?
Would you like me to do anything different?


I ask because I can shoot those videos on my iPhone in minutes and upload them from there. It's incredibly easy on me. I do have a Canon pro-sumer GF10 that is flat out stunning in resolution but if I use that, it requires more effort on the back end and takes longer to get the videos published on youtube. In short, it's just a lot more work. If there isn't an advantage from the viewers' point, then I won't bother doing that. If, however, you guys would like better PQ and a more legitimate video set, then I'll go that route.




Secondly, as we go, let's really try to make this a group effort. Post up your own response curves and methods. Videos, too. I'd like to make this a means to help others learn about the use of measurements and possibly walk people through the basics of tuning. Let's make this an interactive tuning thread. Again, I'm not saying you should lean on measurements but I can guarantee that once you get the basics down and we get a couple more posts on here, you'll be using the RTA to garner some really impressive and quick results. And who doesn't like efficiency?

 

[charliekwin]

To anyone who has watched the video, be honest...

Did it help?
Would you like me to do anything different?

<snip>

Secondly, as we go, let's really try to make this a group effort. Post up your own response curves and methods. Videos, too. I'd like to make this a means to help others learn about the use of measurements and possibly walk people through the basics of tuning. Let's make this an interactive tuning thread.

Being honest, it didn't really help *me,* but that probably should come with a caveat that I'm a web developer and am in front of a computer for 10+ hours a day, so I was up and running with REW in just a couple of minutes after some clicking around. For others, though, I can definitely see the benefit and I'm sure it was quite helpful.

That said, I am by no means particularly knowledgeable about tuning. I've been fumbling through it as I go and feel like while I'm doing *okay,* I'm probably leaving 25% of the "essqueues" on the table, so to speak. A tuning thread is something I'd definitely subscribe to. Seeing how other people are going about getting their results would be interesting -- at least I'd get a better idea if I'm doing it the "right" way myself.

 

[soul786]

Looking forward to following along with this. I'm hoping my install will begin next week so in due time this will be quite useful

 

[Oliver]

Erin has multiple videos on youtube, 21 of em, I think at this time.

I've enjoyed several.

I.E. - Mosconi DSP 6to8 Testing
Mosconi DSP 6to8 Testing Part 2

Nice job Erin !

 

[Woosey]

To anyone who has watched the video, be honest...

Did it help?
Would you like me to do anything different?


I ask because I can shoot those videos on my iPhone in minutes and upload them from there. It's incredibly easy on me. I do have a Canon pro-sumer GF10 that is flat out stunning in resolution but if I use that, it requires more effort on the back end and takes longer to get the videos published on youtube. In short, it's just a lot more work. If there isn't an advantage from the viewers' point, then I won't bother doing that. If, however, you guys would like better PQ and a more legitimate video set, then I'll go that route.




Secondly, as we go, let's really try to make this a group effort. Post up your own response curves and methods. Videos, too. I'd like to make this a means to help others learn about the use of measurements and possibly walk people through the basics of tuning. Let's make this an interactive tuning thread. Again, I'm not saying you should lean on measurements but I can guarantee that once you get the basics down and we get a couple more posts on here, you'll be using the RTA to garner some really impressive and quick results. And who doesn't like efficiency?

Still working with HOLM, but i hope this weekend I have my measuring "cart" ready ( basically a tool cart where i'm building a pc in an a lcd tv on top with mic and soundcard in the drawers haha... ) and then I will try using this software..

 

[voyagerx]

To anyone who has watched the video, be honest...

Did it help?
Would you like me to do anything different?


I ask because I can shoot those videos on my iPhone in minutes and upload them from there. It's incredibly easy on me. I do have a Canon pro-sumer GF10 that is flat out stunning in resolution but if I use that, it requires more effort on the back end and takes longer to get the videos published on youtube. In short, it's just a lot more work. If there isn't an advantage from the viewers' point, then I won't bother doing that. If, however, you guys would like better PQ and a more legitimate video set, then I'll go that route.




Secondly, as we go, let's really try to make this a group effort. Post up your own response curves and methods. Videos, too. I'd like to make this a means to help others learn about the use of measurements and possibly walk people through the basics of tuning. Let's make this an interactive tuning thread. Again, I'm not saying you should lean on measurements but I can guarantee that once you get the basics down and we get a couple more posts on here, you'll be using the RTA to garner some really impressive and quick results. And who doesn't like efficiency?

It helped me a little, as I have never used REW, and have played a little with TrueRTA. What would help me the most is a discussion on what to look for and what to change, as well as things that can't be changed.

I've used TrueRTA in the past to get a relative flat curve and it sounded horrible. I also found areas that I could not EQ, even with different speakers, installation techniques, etc. I'd imagine this was vehicle specific, etc. I think a video/discussion of you thinking out loud and tuning as you use REW, and come across things like this would be extremely helpful.

What you have posted is great and will help many people get started with the software and the use of an RTA to tune their car. What would be incredibly helpful to me would be now that I have it running, what do I look for? What do I do? I realize this is mainly subjective, but your opinion (based obviously on your expertise), would be greatly appreciated.

And finally....thanks for taking the time to do this!

 

[DJTrevLuv]

The video was a big help for me. I bought the “$100 RTA” a couple years ago but didn’t really use it as much as I thought I would because I didn’t really understand how the program worked. I look forward to trying REW.

Keep the videos coming and thanks for sharing all of your audio knowledge!

Trevor

 

[acidbass303]

To anyone who has watched the video, be honest...

Did it help?
Would you like me to do anything different?

Definitely helped me and I am sure many many others like me would benefit a lot from it. MOAR please

 

[Wesayso]

Can you do an impulse response in the car? just curious.

 

[goodstuff]

This helped. Thanks. I've had the omni mic system for a while now and not done anything with it/ This should motivate me.

What are your thoughts on sitting in the seat while you measure?
How would it change the results vs not being in the car?

 

[ErinH]

Can you do an impulse response in the car? just curious.

yes. but your question has multiple meanings. So, can you be more precise as to what you're asking?

This helped. Thanks. I've had the omni mic system for a while now and not done anything with it/ This should motivate me.

What are your thoughts on sitting in the seat while you measure?
How would it change the results vs not being in the car?

I've measured in the car and out of the car. This is the kind of thing that really needs to be experimented with by the user to see how it works. See my response above regarding mic placement/direction.

 

[ErinH]

It helped me a little, as I have never used REW, and have played a little with TrueRTA. What would help me the most is a discussion on what to look for and what to change, as well as things that can't be changed.

I've used TrueRTA in the past to get a relative flat curve and it sounded horrible. I also found areas that I could not EQ, even with different speakers, installation techniques, etc. I'd imagine this was vehicle specific, etc. I think a video/discussion of you thinking out loud and tuning as you use REW, and come across things like this would be extremely helpful.

What you have posted is great and will help many people get started with the software and the use of an RTA to tune their car. What would be incredibly helpful to me would be now that I have it running, what do I look for? What do I do? I realize this is mainly subjective, but your opinion (based obviously on your expertise), would be greatly appreciated.

And finally....thanks for taking the time to do this!

Yes. At some point this thread will start to get in to more of the explanation of what you're seeing and ways to attack it. I fully intend to film my process at some point with the caveat that it's not a one size fits all. But I need to try to get everyone who might be interested in that up to speed with knowing how to measure first.

I'd like to have a friend film this for me sometime but that may not be feasible if time is of the essence.

 

[Justin Zazzi]

I've used REW a bit before seeing this thread and did not know about the averaging method you used with constant pink noise playing. I have instead used the sweep method by pressing the "measure" button, and the program (I believe) will average how ever many sweeps I choose to do.

Do you know if there is a benefit to your method vs using the built-in "measure" method? I do not think your method can view the more detailed time-domain information like impulse response and waterfall graphs, right? Can your method make use of the auto EQ functions or not? From what I remember reading, it relies heavily on time-domain information for making the best suggestions.

-J