Sir Rab of Everest Posted January 22, 2012 Posted January 22, 2012 (edited) My calibrated Behringer mic finally arrived this week, so today i started playing around with REW for the first time, and i have just made my first ever room measurements. My aim in this project is just to find (i) a good listening position, then (ii) good positions for my Lenehan ML-3 speakers, and then (iii) possibly some room treatments, if i can get them past SWMBO. Unfortunately, my mic stand has not yet arrived, so these measurements were taken with the mic supported between two cushions on my couch (!) -- yes, far from ideal, i know, but the kids were out of the house so i had to make the most of the opportunity. This is with both speakers running -- i just read the post in another thread stating that you should do one at a time. OK, so here's my first attempt at frequency response - a bit wobbly, but it doesn't seem to look a lot worse than other plots I have seen posted elsewhere. Some very sharp suckouts at ~35 and 65 Hz though... [ATTACH=CONFIG]40371[/ATTACH] On the other hand, the waterfall plot seems to show nasty room ringing below ~100 Hz - note this is to 1500 ms!! It seems so bad that i am suspicious... 1.5 seconds seems wayyy long, and i don't hear anything so dramatic... Looks ok above 100 Hz. [ATTACH=CONFIG]40370[/ATTACH] Here's RT60: [ATTACH=CONFIG]40369[/ATTACH] So that's it for today. Hopefully my mic stand will arrive this week and i can get more serious next weekend... Any comments, suggestions, etc are of course welcome! Edited January 22, 2012 by rab
88Pro Posted January 22, 2012 Posted January 22, 2012 Hey, All the attachments return "Invalid Attachment specified." error.
Sir Rab of Everest Posted January 22, 2012 Author Posted January 22, 2012 Fixed - Sorry about that... my token timed out while i was writing the post and then somehow the attachment numbers messed up... I just re-attached them.
Drizt Posted January 22, 2012 Posted January 22, 2012 Rab can you please change the y axis to 45 to 95db, its hard to get a good feel for what is happening with that scale.
Drizt Posted January 22, 2012 Posted January 22, 2012 That waterfall plot is extremely wierd. I wonder if that is a measurement anomoly ?
Sir Rab of Everest Posted January 22, 2012 Author Posted January 22, 2012 (edited) Rab can you please change the y axis to 45 to 95db, its hard to get a good feel for what is happening with that scale. Sorry Drizt, i assume you mean the waterfall. That does look cleaner. Maybe four modes, mainly the one at ~ 70Hz. Edited January 22, 2012 by rab
Paul Spencer Posted January 22, 2012 Posted January 22, 2012 Rab, try these settings: http://redspade-audio.blogspot.com/2011/09/recommended-rew-settings.html
Drizt Posted January 22, 2012 Posted January 22, 2012 (edited) rab, I would wait until you get a mic stand and retake the measurements. Put the mike pretty much where your head would be and see what you get. Only connect one speaker up at a time in mono. If your listening chair is leather it would be better to raise the mic at least 30cm above the back of the chair to remove the reflections that play havok with the measurements. Also use Pauls advice regarding REW settings as it makes it easier to make sense of the readings. Edited January 22, 2012 by Drizt
Sir Rab of Everest Posted January 22, 2012 Author Posted January 22, 2012 Rab, try these settings:http://redspade-audio.blogspot.com/2011/09/recommended-rew-settings.html Thanks, Paul. Here it is: [ATTACH=CONFIG]40385[/ATTACH] ... and the same but to 1500ms as per my originals: [ATTACH=CONFIG]40386[/ATTACH] rab, I would wait until you get a mic stand and retake the measurements. i will certainly do that... i was just a bit excited and wanted to get these out. Also, it's not often that i get the house to myself and my boys are extremely noisy or else they want to be playing on the computer or xbox, both of which exclude me from making my measurements!
Drizt Posted January 22, 2012 Posted January 22, 2012 Understandable rab. Good luck with it all, ill be following this with great interest.
joz Posted January 22, 2012 Posted January 22, 2012 I can lend you my mike stand if yours is going to take a somectime to arrive.
Drizt Posted January 22, 2012 Posted January 22, 2012 (edited) Rab, this thread might also give you some info that you might find helpful -> http://www.stereo.net.au/forums/showthread.php/42121-Interpreting-listening-position-measurements This part of svenr's post maybe helpful for your investigations - Take an inverse measurement approach - place the main speaker at the LP (midspan between tweeter and midrange driver at ear height) and measure the transfer function at the possible speaker positions with microphone at speaker midspan height. Take measurements at a grid of say 30cm over an area as wide as 1m and as deep as 2m for each speaker position (mark mic positions with tape on the floor). - Then shift the LP forth and back in say 30cm increments and repeat the full set of measurements. - Import all in Excel, discard all frequency bins above say 400Hz (that's all reflection dependent), calculate the magnitude, and normalise so that the mean is zero dB. Calculate the variance from zero mean for each response, and rank the responses based on the variance. Plot the 10..15 best responses. You have now identified a range of positions resulting in the flattest frequency response in the modal frequency range. Edited January 22, 2012 by Drizt
Paul Spencer Posted January 22, 2012 Posted January 22, 2012 including graph limits as well! Try 20 - 300
Sir Rab of Everest Posted January 23, 2012 Author Posted January 23, 2012 including graph limits as well! Try 20 - 300 Apologies, Paul: i was in a rush this morning (as usual) and missed the graph limits: [ATTACH=CONFIG]40398[/ATTACH] I can lend you my mike stand if yours is going to take a somectime to arrive. Thanks, Joz, but i bought one today. Should arrive by the end of the week.
Paul Spencer Posted January 23, 2012 Posted January 23, 2012 You still have some weird settings ie the windowing settings.
Sir Rab of Everest Posted January 23, 2012 Author Posted January 23, 2012 You still have some weird settings ie the windowing settings. D'oh! I generated 2 graphs: one with your exact settings and another the same but to 1500ms - and then forgot to upload the first one! *sigh*... here it is: [ATTACH=CONFIG]40407[/ATTACH]
Paul Spencer Posted January 24, 2012 Posted January 24, 2012 I was curious to see this because when I was at the GTG I noticed quite a bit of bass boom, then noticed brick walls! I could tell it was a room issue. But at the same time you have a fairly big open space that helps. What is interesting is that it's confined to 35 - 100 Hz then above it looks very good. Quite curious, I don't recall seeing similar. What this means is that it's possible to improve things with pressure traps, which are less obtrusive. They are also more difficult to design and build, but they can be made as flat panels.
Sir Rab of Everest Posted January 24, 2012 Author Posted January 24, 2012 (edited) Thanks Paul. Yes, i didn't know what to expect due to the complex shape of the room. pressure traps, which are ... more difficult to design and build, but they can be made as flat panels. Are you talking about something like this? : photo, description Edited January 24, 2012 by rab
Paul Spencer Posted January 24, 2012 Posted January 24, 2012 If sealed up with a thin front ply panel then yes
mrrock Posted January 25, 2012 Posted January 25, 2012 Hi rab...you should actually measure at 90db if possible. I'm interested to see how they measure at your seating position, because a vast majority of speakers will give you up to 20k 30 cm from the speaker. Alot of peaople measure from 1metre away which is pointless as you usually don't sit that close. I don't know how you measured, but keep a distance, extend the rta responce to 3 secs and then you can get a half decent reading. If your too close to the speaker, you will be picking up the speakers own inacuracies. Also move the mike up, you will notice an increase in highs...So it's a game of averages.
Paul Spencer Posted January 25, 2012 Posted January 25, 2012 Hi rab...you should actually measure at 90db if possible. When taking room measurements, the main three issues are: 1. avoiding clipping, over driving anything, or causing room rattles 2. (often) being able to measure various positions without clipping because you get a different room transfer function 3. measuring far enough above the noise floor The actual number isn't all that important. Alot of peaople measure from 1metre away which is pointless as you usually don't sit that close. It depends on the purpose of the measurement. A 1m gated measurement with the speaker elevated well above the ground is a fairly standard way to measure the speaker alone. This is useful for design purposes. For looking at acoustic treatment, speaker placement and bass integration then a listening position measurement is what you want, all room reflections included. If your too close to the speaker, you will be picking up the speakers own inacuracies. I have no idea what that means! A nearfield measurement, with the mic almost touching the cone, is often useful. It's quick way to find out what the driver is doing. You put the mic up to the woofer, then up to the port, you see each without room effects, because the mic is so close and the direct sound swamps all else. You also exclude baffle effects. This probably isn't what you are talking about. Normally you will measure nearfield or at 1m or in the listening position, or ground plane at 2m. Those are all valid depending on what you are aiming for, and they cover most situations. An exception would be a large horn speaker with large vertical horn offsets, in which case you might measure at more than 1m to get the speaker only response.
Arg Posted January 25, 2012 Posted January 25, 2012 My calibrated Behringer mic finally arrived this week, so today i started playing around with REW for the first time, and i have just made my first ever room measurements.....Unfortunately, my mic stand has not yet arrived, so these measurements were taken with the mic supported between two cushions on my couch (!) Just throw the graph away.... On the other hand, the waterfall plot seems to show nasty room ringing ... 1.5 seconds seems wayyy long, .... Is there a piano in the room? Was the door closed? Any rattly windows?
Paul Spencer Posted January 25, 2012 Posted January 25, 2012 Richard, try muting the audio then run a measurement. Then unmute and measure again, changing nothing else. This will reveal the noise floor. You need to rule that out first.
Sir Rab of Everest Posted January 29, 2012 Author Posted January 29, 2012 OK, so my mic stand arrived and i did my first measurements with the mic at ear height as if i was seated on the couch. Due to my living circumstances, i must keep my speakers set back on either side of a large sideboard most of the time. Only when i get the house to myself do i move the speakers out into the room, closer than the front of the sideboard. The first thing i did was a measurement of the speakers back vs out (incidentally, the earlier measurement was also with the speakers back). Here's the comparison: [ATTACH=CONFIG]40570[/ATTACH] The first thing you'll notice is that the sharp nulls at around 70 and 83 Hz disappear when the speakers are moved out, but the rather wide dip at around 36 Hz remains. There is, however, a rather wide trough between 45-65 Hz. I then did measurements with only the left speaker, only the right speaker, and neither speaker: [ATTACH=CONFIG]40571[/ATTACH] What you will notice here is that most of the variation originates from the left speaker (mustard coloured line); the right speaker (purple line) is relatively well behaved, except for the sharp null at ~ 220 Hz. In contrast, the left speaker has nulls at 38, 52, 80, and 170 Hz. As my room is highly asymmetrical, the difference is perhaps not altogether surprising. Waterfalls: left, then right: [ATTACH=CONFIG]40572[/ATTACH][ATTACH=CONFIG]40573[/ATTACH] I need to stress here that the speaker and listening positions are rather arbitrary, and i have not yet had a chance to do any experimentation whatsoever in this regard. I think my next step will be to do some (i) listening position variations to get the best response, and then (ii) vary the speaker positions - in particular the left! - to see if i can get some overall smoother responses... only once i have found quasi-optimal and practical listening and speaker positions will i start investigating which surfaces/corners are responsible for what and then how i can get some pressure traps into the room etc etc. As always, i welcome any insights you folks may wish to make, but it seems to me i have a fair bit more research homework to do before i start adding traps etc (if i can get them past my wife!) Due to work commitments, it will probably be at least a couple of weeks before i can get back to this...
Drizt Posted February 14, 2012 Posted February 14, 2012 All very interesting rab, thanks for posting the info. I look forward to following your progress on this front. ... i start investigating which surfaces/corners are responsible for what and then how i can get some pressure traps into the room etc etc. This is the part that will be very interesting.
Recommended Posts