Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How... - Page 33

Hm, IMHO you only have to spend 500 and not 520 to effectively perform the same task.

 

Yes, of course, no argument.  But the mic needs to be calibrated, one way or another.  That is what JohnM is saying.

Well, for those of us who spent more like 15 bucks for the analog RS meter something like 10 years ago it's all academic anyway. I had to calibrate speaker distances manually back then, after walking to and from work in the snow, uphill both ways.

Now get off my lawn.

 

In fact dStew's method is no good. Playing the AVR test tones bypasses Audyssey's calibration. It is true that Audyssey has calibrated the system to 75dB and it will indeed measure that when Audyssey is turned on. But if Audyssey is turned off, which it will be for many measurements with REW, or if the internal AVR test tones are used (which bypass Audyssey) then the alleged 75dB may or may not be 75dB at all simply because Audyssey may have boosted certain frequencies by as much as 9dB, thus creating an average SPL quite different from the non-Audyssey result.

 

I can't understand what this pointless discussion is for - John M has responded to Jerry's question in unequivocal terms. He says you need a SPL meter with the exception of one specific mic which few appear to be using. As you point out, a SPL meter costs a few bucks. WTF is the discussion about - saving a few dollars when the systems involved probably cost as much as 500 to 1,000 times the cost of the meter??!  Why mess about - if John M says you need a SPL meter, who are you going to believe - an argumentative forum member or the guy who created the program we are all discussing?  Calibrating the mic with a SPL meter takes a few seconds and you do it precisely one time. I can't see any sensible reason for going against John M's advice.

You are misunderstanding his answer. He was asked if the right amplitude would show up in REW if you don't calibrate. He was not asked if you must calibrate to get measurements you can use to optimize your room.

Point us to a link that says you can't use REW to optimize your room without calibration and I will be happy to debate him here or there .

 

And of course, unless the definition of 'calibrate' has been changed, you need another known reference to calibrate against. One cannot calibrate a mic using the mic under test as the known reference. Sheesh. 

 

Oh really. This is the reason to do it in a way contrary to the author of REW's dictum?  If anyone here has spent more than $5,000 on their system and cannot afford a $20 SPL meter, if they PM me, I will send them the 20 bucks. FFS!

Keith, we are not talking about level calibrating a mic, but about level calibrating the whole REW measurement loop, aren't we?

 

You walked to work in the snow?  Bloody luxury. In my day we had to drag ourselves on our bellies, uphill in the snow (both ways), because our feet had been deformed beyond use by working 21 hours a day in the mines.... you young guys don't know you're born.... :)

Well, when I say 'walked' it was really staggering along on the bloody stumps where our legs used to be, but it was walking to us!

I think we all know what we are discussing, Feri.

That's not what he said. John said
Think outside the box a little. I did post a new thread @ HTS asking if the post audyssey AVR tone at 75db is a valid way to calibrate REW's SPL level. If he has a valid reason why it's not a good idea or not accurate enough, fair enough. But don't go around misquoting people as if that is some kind of valid rational. Do you have any technical reason why you don't think it should be done this way?

BTW everyone keith can't see this because he blocked me for telling him he was wrong once. Now I'm doing it again for the sake of accurate helpful information. This is probably going to get ugly, geeze I hope not.

 

You had stumps? I'd have sold my sister into slavery for stumps. We could only dream about the lucky guys with stumps, from the Big House at the end of the lane. We'd had to feed what was left of our legs to our little brothers, to keep them alive for their job sweeping chimneys...

 

No - I unblocked you some time back. Now you can tell me I was wrong, again... But yes, you did tell me I was wrong. That, of course, is entirely different to being wrong ;)

 

The pointless argument has descended into pedantic semantics now. I suggest all users follow Jerry['s excellent guide. If they choose not to, that is their prerogative.

 

Using the internal test tones is no use in an Audyssey calibrated system because the tones bypass Audyssey. So every measurement you make with Audyssey on will stand a very good chance of being done at a different SPL than you think it was done at. If that is OK for you, then by all means do it that way.

You were lucky to have little brothers, we had to sell ours as dog food to earn enough money to bribe the project manager for permission to come to work each evening.

Fair enough . We'll get further communicating IMO. If I hear something back from John I'll certainly let you all know

 

Cool!  Really, and seriously, it's a non-topic IMO. A 20 dollar accessory... a few seconds to do it the 'approved' way. No downsides at all. What's to discuss?  I'll be glad when we all have our mics and we can start actually measuring and getting on with learning how to improve our rooms!

 

Yes, we were lucky to have little brothers. Especially at Christmas when we'd boil one of them up for a special dinner. It gave meaning to the phrase "having your family over for dinner" and it was a great honour to be picked. One year, our smallest brother, having failed to be chosen for the pot, cut off his own arms and legs and roasted them in the fire for us. It was such a sad sight to see him, holding the fire irons in his mouth as he raked the coals. He'd have given anything - anything at all - for the chance to own a RadioShack meter and calibrate his software the way its author advises. 

fixing quotes

If it can be done without an extra tool then I think there is value in that. Like I said, I have the RS SPL meter so it's a non-issue for me but for those without one on hand this is easier. Can I have 20 dollars? Please?

Not sure who decides what the 'approved' way is, when is the board meeting?


Here's what I said over at HTS
And John came back with the very detailed reply

It seems obvious to me this is a valid way to do it and since you all are keeping a guide this may be helpful info for some folks. The cool thing about the USB mics is less stuff IMO.

Well, this is a small group, so there's a certain degree of 'founders' effect'. So the 'board' has already met virtually LOL . This is starting to remind me of the different liberation groups in The Life of Brian...splitters

I've stayed outside of this, but it seems to me that the whole reason we're here is because we've moved beyond the 'trust Audyssey' stage, and the goal is to build knowledge in external measurement accessment to tweak our HT above and beyond correction that Audyssey (Pro) or electronic room correction provides. Otherwise we'd be sitting back and enjoying our systems, which would make AVS a dull place .

Having said that, I don't see the harm, given that the whole USB mic usage is a work in progress, in using an standalone SPL outside of REW as a first step to define parameters, whether or not electronic means from an AVR may 'do the job'. Philosophically, starting with an absolute standard outside of the measurement software, so that we're sure the yardstick is always in relation to a known standard is attractive. And I'd rather not wonder if there's subtle effects being applied in processing those test tones. It's classic A/B/X thinking. End of story. At least in the REW world....

We are trying to optimize our systems to as close to the best of their ability we can get them. Frequency response, RT60 measurements, smooth reverb decline, verifying impact of upgrades, measuring impacts of changes, etc... We can do better than Audyssey by itself with time and patience, however this topic is on the SPL reading. I have a master volume control on my system, it's OK if the SPL meter was off 1db, that is not hurting my systems performance. It is likely the Audyssey mic's SPL reading is better than a $20 meter. My radio shack SPL meter and Audyssey agree what 75db is though.

All though you make several valid points about "trust but verify" mentality I don't think it applies to the SPL reference point IMO.

Do what you will with the info. My concern was a valid approach was being rejected and extra unnecessary hardware purchases were being recommended. I feel I've done what I can to get the right info out there.

Woohoo! My microphone came in today, complete with calibration chart and USB stick.

I shall use it in tribute to all my lost little brothers, tasty or not.

 

 

What I mean by 'approved' is the way the guy who wrote the software says to do it.

 

 

 

It seems John perhaps is not up to speed on the way Audyssey works wrt to the test tones in the AVR. As I've said three times now, the AVR test tones bypass Audyssey. So the tones will be right when Audyssey is on (for 75dB) but not for when Audyssey is off. Half the time the tones will be incorrect. Audyssey can boost by up to 9dB and cut by up to 6dB. This means that there is every chance that running the test tones with Audyssey off will NOT deliver 75dB at the MLP. This means that using the AVR test tones is not a valid way to ensure that the AVR is delivering any known value of SPL after calibration. Just use a nice cheap Ratshack meter, like John suggests, and all will be guaranteed to work as designed.

 

I've said more than enough on this pointless topic now, so this is going to be my last word on it.

Exactly. Why deviate from John M's suggested way of using REW when there is no need to?

Time to give up Keith my friend! If your peculiar logic penetrates we will all end up sooner or later thinking JohnM is another unreliable soure, next to Chris K. eh?!

Because the whole exercise is absurd!

Here is an easy way to prove that.. Let's say we get to the end of this process, and I perform my measurements using your calibrated settings. I then apply my optimizations (EQ, acoustics, etc.) based on those measurements. Here is the main question: Will the optimization be valid at other listening volume levels?" The answer better be yes. Or else I have to have many sets of acoustic treatment plans and EQ that I switch in and out! So as you see, if the system performance is invariant of volume, then the measurements better be too. Yes, you don't want to measure down in the noise or clip. REW nicely warns you if you are too high or too level for that.

Lack of experience is providing the anxiety you feel with not calibrating the inputs Keith. Once you use REW a few times, learn how to read its measurements and convince yourself that what you see is what is going on, that will subside. Don't insist on your gut feelings to be right when you don't have any business having any yet!

Of course this whole thread is ridiculous, but let's reduce it a more basic level: REW v5.xx is beta software in the first place, and using a mic other than the UMIK-1 USB mic is 'beta' as well. As such, regardless of what any authority says, the calibration factor needs to be established outside of REW to know that any measurements we're conducting are achieving what we think they are, before they're used as a standard for REW, period. I don't trust John on this anymore than I trust Jason absolutely. Get the Ratshack SPL and be done with it. At least if we're being 'scientists'.

The other reason I agree with Keith is that coming from OmniMic like him, I found it too easy to import data from multiple runs, and wind up with incompatible graphs (and having to use 'compatible' offset factors to make a best attempt at comparability), without knowing how much of the effects I measured were due to non-trivial deviations in mic calibration factors. Better to have one source, with an absolute standard.

it's no different than building new software and seeing if repeated results in that software can be replicated elsewhere: internal vs. external validity. It's paranoia, but it's 'best practice' paranoia.

He said using your AVR's test tone is a valid way to do it, you're welcome to ask him again, but I provided the information from the author which disagrees with what you are saying. I feel you should stop using him to support your position as you are not paraphrasing him accurately.

Audyssey EQ is bypassed when playing the test tones, Not the trim levels. The test tone is calibrated to 75db. That's the whole point of the test tones, for levels. Really unsure where you are going with that????????????

The ratshack meter confirms this approach is valid. If you want to waste the time and money acquiring one to verify what is true (and multiple people are telling you which have tested it), go for it. But if you are going to make recommendations to other people, it should not be to go buy an SPL meter for this purpose.

I'm still not hearing a technical reason why the simple, no extra hardware/money approach does not work.


I hope I'm not coming off argumentative as some of the posts on this topic are, just trying to get accurate info out there.

Graph compatibility over multiple runs I can get on board with. However now your making my point for me. The AVR test tone REW SPL calibration approach is better in this respect. It's up to the user how careful the USB mic placement is. But AVR test tone is better approach as you don't have to get two device in identical placements between runs, nor do you have to stand there reading the SPL meter bouncing the sound around. Hold an SPL meter in one place and move around it. Where you are changes the reading. Your body can be far away for the mic with the AVR test tone. Or are you now recommending two boom stands and some binoculars too?