Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How... - Page 7

Thanks again, Feri. I will give it a try.

This article from what has become one of my favourite audio blogs is interesting and germane to this thread:

 

 

Should you get your measurement mic calibrated

 

 

This article is amusing and OT though (but contains some good stuff):

 

How to read Stereophile without wanting to throw it across the room

Hey Guys
Long time lurker here! The short and sweet is, after starting to read this thread,decided to download REW and give it a shot. So after just playing with some of the functions yesterday without any mic or HDMI connections I got a better understanding of what I was reading about. It looked overwhelming and complicated.
So today after reading about the Audessey mic working decided to throw my ARC mic in the mix. Guess what it works.[[So far]] I even believe I have it calibrated to the program [[Yessss]] I`m not very technical with computers and they frustrate the heck out of me sometimes LOL.
Anyways question 1 is I dont have a laptop nor a comp with HDMI. So will any HDMI soundcard for my computer work [money and quality wise], second question my HDMI run will be about 25 ft will this be ok.
Thanks guys and keep it going......

@Heybrook,

Any current generation sound card with HDMI should work. A 25' HDMI cable should not be an issue either. If you are going to use a desktop computer, make sure that the noise (cooling fans, etc.) are not loud enough to affect the REW measurements. Also, note that the mic you are using is OK for learning and testing, but when you want to take accurate REW measurements, you will need a mic that has a calibration file for REW.

Man I hope this thread doesn't get too long by the time I get my mic. Guess I'm not helping matters. Can't wait

 

The wait will seem for ever... I'm betting at least the beginning of Feb before I get mine - depending how CPL dispatch it, it could take 10 days to get here - so you guys will have a headstart on me - which may not be a bad thing ;)  Meantime, I am going to have a practice run as per Feri's suggestion, using my Audyssey mic.

so any cheapo spl meter would do?

 

One would think so.  However, you will find many of us use the Radio Shack SPL, so that may be a better choice.  For REW use, make sure the SPL has an output connection.  The RS SPL, for example, has an RCA connection that sends a signal to a soundcard or your laptop.

On that subject--analog or digital?

If all you are using is the mic, the metering format doesn't matter.

Thanks - I've read some comments where the analog format is still preferred by the DIYers, so good to know.

What are your graph axis settings? I just did a quick measurement it only went as far as 200, even if I set the right limit to 300 or 400.

Also how do I post a graph in this tread?

When you say 200 do you mean 200 Hz on the x-axis? My graphs show full range, i.e. 20 Hz to 20 kHz. If you go back to my original post and click on a graph it will expand on your screen for easy reading.

For posting a graph (jpg), in the reply window you will find an icon on top for embedding images. Click on it, them look up the file in your computer and hit Submit. That's all. You can check how it will look like prior to sumbitting by clicking on Review.

Mostly because it is easier to read the analogue dial I think. You can just 'look and see' rather than having to 'read'. Like the difference between an analogue and digital watch. How many times do you see someone with an analogue watch glance at it and then, if you ask them the time, they have to look again?  First time, the position of the hands was all they needed to see to know the time. To convert it to a specific time (it is 27 minutes past six) they need to look again. On my 911. I have reorientated the rev counter so that the red line starts when the needle is pointing straight up on the dial - no need to read anything - the swiftest glance tells me I need to change up... :)

It is good to see you in this thread, Arny. Some of us need all the help and expertise we can get with REW... and acoustics in general once we have figured out how to make REW work ;)

 

Some reviews I have read stated that the analog SPL is/was more accurate than the digital one.  However, when I checked the RS web site, only the digital one seems to be available anymore, BICBW.  The SKU for the new digital one is 3300099, and for the old digital one is 33-2055 (can't find this one either).

Yes, except we do not have a calibration file for it... And it is hard to trust the linearity of the mic. input on notebook... at least I don't trust mine.

The only purpose of using the Audyssey mic with REW is to get acquianted with measuring techniques and REW itself until more people can get their UMIK-1 usb PNP mic. Worry not, nobody is going to make serious measurements with the Aud mic on this thread.

 

Here is a trick question, Feri:

 

Which of the following two response curves looks better?

 

 

 

The answer:  they are the same curves.  The difference, of course, is the smoothing and the vertical axis resolution.  How you set your graph parameters has significance in the value of what you are showing. 

 

I think we should agree in this thread on certain standards for the graphs we publish.  My recommendations:

 

- Use 1/24 smooting for full-range graphs, and no smooting for graphs that focus on bass frequencies, say below 300Hz.

- Adjust vertical axis resolution so that the scale shows in 5 dB increments (I believe Jason made the same recommendation).

- Always include a full-range graph (20-20,000 Hz), as well as a graph that focuses on a problem range (e.g. bass frequencies 20-300 Hz).

 

Having said that, would you care to re-publish your averaged graph with 1/24 smoothing, rather than 1/3 smoothing?

 

Thanks.

would it be a problem if my soundcard's input doesn't work. but the computer onboard sound input works and i'm sending out audio signal to receiver via video card's hdmi output?

Got it, thanks.

Hi Jerry,

Well noted you recommendations, so here goes again with 1/24 smoothing for full-range.



Thanks.

testing

 

That doesn't look right.....   There shouldn't be a 30 dB increase in your lower bass.  Either REW is improperly measuring, or you have a serious issue in the low end.

 

That looks better, Feri.  If we all follow the guidelines, we will be comparing apples to apples. 

 

AFAIK you are right - the analogue RS meter is discontinued...

 

I agree. This ought to be in one of the first posts in the thread if we are going to remember to do it and what it stipulates. HST, I always do 1/24th anyway, but I also often make a 1/6th as I understand that it can give us a better idea of what our ears actually hear. 1/24th tells us better what is going on, 1/6th tells us better what it might sound like. Please correct me if I am wrong.

 

EDIT: looking at your two graphs Jerry - the 10db peak to peak difference between 970Hz and just over 1kHz all but disappears on the 1/3 smoothed graph - as it is quite deep but also very narrow, does this not correspond more to what we hear?  IOW, we could obsess over that small frequency band as shown in the 1/24th graph and spend a lot of time trying to smooth it, but a quick glance at the 1/3 (or pref 1/6th) smoothed graph would show us that is probably inaudible anyway? Have I got this right or am I way off?

Edited by kbarnes701 - 1/10/13 at 1:42pm

 

This, to me, is one of the BIG benefits of REW as opposed to OM - it is easy to go back and re-present the graph in a different format as you just did there. With OM you'd have to go back and do the measurement again and make one graph at 1/3 and another at 1/24th. 

+2