setting up first REW charts so they're readable/relevant

[raticusfortuna]

greets,

y'know how a little kid hands you a picture she's drawn and says 'it's a cow!' and it looks NOTHING like a cow? so you smile and nod and say 'what a wonderful moo cow!'

well, here's my cow.
my first REW charts!

i followed the advice given me in another post about what charts to focus in on. (thanks!) and these are my first attempts.

this has nothing to do with the actual room testing. it's about learning how to use REW. i'm not looking for analysis, i'm just trying to get the charts readable so i can then move on to take multiple measurements, and then have something relevant to analyze. (...if i got the gist of how it's done correctly.)

i'll put my equipment and room run down at the bottom of the post.

so here's the first FQ response. both speakers are black, right is red and left green. i set the graph up to focus in on 40hz - 1khz and smoothed it to 1/12 of an octave.


when doing separate speaker readings is it ok to just reach over and turn one of the powered speakers off? that would leave the remaining speaker and the sub for the measurement. ........ or does it call for a more nuanced approach?


here's the etc. again, both speakers are black, right is red and left green. i get that this chart is for early reflections but don't know what information is relevant, so i don't know what to focus in on in terms of showing the most useful information. should i trim the time line and zoom in?


the waterfall..... ....i've narrowed the FQ range to focus in on the lower FQ's, but, i've got this bad boy cranked out to 1500ms and the sound won't die away! is my room that bad? is that a measurement issue? is that a noise floor issue? should i just edit that axis upward till it gives me shapes i can read? and i can't find where in REW to set it up as a muti-color chart like i've seen posted.



the spectrum. once again edited to focus on the low end. i can't quite relate the decay times to the times on the waterfall. i left the time axis long. moving some panels around caused some huge leaps in the low end that looked like solar flares.



those are my first attempts. i'm still a bit fuzzy on how hot to do the readings and how to make those level adjustments with out messing them up. i'm following the level calibrations but i've read post here about skipping that and approaching the levels it as if you're doing a regular digital recording. (dandan?).

so, if i've got REW up and running to the point where it's useful info, the next step is to take apart the studio, drag everything out and start from scratch....doing readings as i rebuild.

udderly yours,
raticus.



ps.
tech and room stuff -
the room is 9 1/2' x 7 1/2' x 8' high.
wood floor, 2 walls wood panel, 2 walls plasterboard, wood pulp tiled ceiling.
i'm using the bluesky 2.1 prodesk monitors

i confess i am using the radiohut spl. meter as a mic. i loaded the specific calibration for that mic that the REW site provided. as i understand it, its weakness is in the higher FQ's. fortunately, i'm focusing on the lows at the moment.

[ritelec]

You have measurements.
I would think you could start with the next process you're thinking about doing.
Not sure how your set up. But you may want to keep whichever computer your using set up somewhere some how , and also to be able to take the computer and monitors, interface into the room when its empty to take some measurements.

[raticusfortuna]

Originally Posted by ritelec

You have measurements.
I would think you could start with the next process you're thinking about doing.
Not sure how your set up. But you may want to keep whichever computer your using set up somewhere some how , and also to be able to take the computer and monitors, interface into the room when its empty to take some measurements.

yes, it will be empty except for the 'box'. i'm operating under the room set up equation - 70% speaker placement, 25% room treatment, 5% eq. except without the eq part.

i'm going to focus on placement by rolling my workstation around, fully operational, listening and testing.

as tedious as this clip is, bob hodus lays out a good approach to setting up a space.

YouTube - ‪- Pensado's Place - #21 - The secrets behind tuning your room!‬‏

raticus

[takman]

man, those graphs are all displayed wrong.

For the frequency response, set the frequency range 20hz---20,000hz. And use less smoothing, so you can identify the problems in more detail.

For the ETC. The initial sound at 0 time should start at 0db. Yours doesnt, not sure why. Set the time to from 0-100ms. We need to look closer at the time.......so the graph shows time in 20db steps. and dont use smoothing on it.

for the waterfall, set the time to 300ms. Bring up the floor, so that it shows a 50db spread between the top of the graph, and the floor...and then tweak it from there. and display 20h-400hz range. Is the room untreated? yes, an untreated room can seem like it goes forever on the waterfall, but dont worry, we are going to tame that with bass trapping.

dont even worry about the spectrogram right now.

post them up as i suggested and lets go from there.

[raticusfortuna]

takman!

exactly what i was looking for! i've copied the post and got it set up as a 'read me' in my REW folder.

to bad the spectrogram isn't needed..... it's the coolest looking chart.

raticus

[DanDan]

So many misunderstandings and I hope typos in this thread it is impossible to know where to start.
Perhaps this http://www.gearslutz.com/board/studi...er-v2-1-a.html

OK. It is not necessary to attempt such sophisticated measurements. A good eye, ear, brain, some reading and some sine waves can treat a room acoustically.

The measurements are taken within simple limits. Plenty of level to get above the ambient noise, without overload anywhere in the chain.
The graphs are just different views of the same information. The graph parameters do not affect or change the measurement.

View LF Freq Resp say 30- 300Hz. No need for smoothing, better to see the detail.
Full range on the other hand looks crazy and presents little or no information as a view, without smoothing. With smoothing on we can see the overall slope, the tone of the room.

Waterfalls have noise at the bottom. This is ambient room noise. Set the minimum level to get above the room noise. Try around -40 to -50dB

The ETC is for viewing acoustic reflections. The most dangerous ones would be within the first 10mS or so. At HF these reflections travel in pretty straight lines, like light. Thus they are usually quite visible in the real world. Walls, Desk, Floor, Ceiling, around the speakers. A mirror can be used to see if they are likely dangerous. Some string cut to length will help tie particular surfaces to the spikes on the graph. Moving the mic in the direction of a suspected reflection should shorten the travel time to the corresponding spike.

But, as I said, this is not mandatory homework. More importantly I reckon the real strength of measurement is in comparing A with B. Move a speaker towards the wall, is it better or worse?
A much better question than how is A or how is B?

DD