Products
Acoustics Advice
Consultancy
Education
Contact
Home
Customer Reviews / Photos

© Blue Frog Audio 2015           Tel : 07800 952 474            Email : bluefrogaudio@hotmail.co.uk

PayPal: View Cart
Blue Frog Audio Acoustic Treatment Logo

Free UK Delivery on Orders over £800.00

Room testing :

To find out what the lowest frequency issues are in a given room, and how severe they are, the best way to be 100% certain is to test the room. Room testing is very straight forward once set up, and is well worth the effort both for room treatment and for finding the ideal location in a room for speakers and listening position (which can make large improvements to the sound heard at the listening spot, and is a free improvement).

To test a room you’ll need :

- an omni directional mic. A low cost version will work fine as we’re looking for big variations in the room’s response. If you don’t own an omni directional mic   the Behringer ECM 8000 is an affordable option that’s adequate for room testing duties. Up from that as a good quality calibrated test mic is the   Beyerdynamic MM1.

- software such as RoomEqWizard, which is free to use and is a great package for room testing, or Fuzzmeasure, which has a 2 weeks trial.

- a mic pre-amp to plug the mic into. Soundcards / interfaces with built in mic-pre’s are fine

- an a/d converter / soundcard to feed the mic pre-amp’s output into, or just plug the mic into mic pre on your soundcard if available, and route the signal   internally so that the room testing software can see it.


How the tests work :

The room testing software plays back a swept sine wave through your studio monitors / speakers, which ‘excites’ the room - showing up its peaks and dips in frequency response, decay times (just as important as the frequency response), and distinct mid and high frequency reflections (destructive early reflections need to be damped with absorption or re-directed for accurate / pleasing sound reproduction).

The sound of the room is then captured by the microphone, which is feeding the room testing software - the software decodes this capture of the room’s acoustic response and generates various plots that show the acoustic behaviour of the space. From the plots generated everything needed to determine what type of treatment a room needs to bring it within acceptable specs for critical listening is available.

Where to position the Mic and what tests should be run :

The test mic should be positioned at ear height in the middle of the usual listening position, ie centrally between speakers.

3 sets of measurements should be taken for a stereo set up - 1 set with both speakers playing back the test tone, 1 set with just left speaker, and 1 set with just right speaker.

If using a sub do 1 test with all 3 speakers, then 1 with left and sub, and 1 with right and sub.

These tests with the mic at the listening position will give the room’s response at the listening position only.

If you have a seating area for clients at the back of the room tests can also be run there, but ensuring the ‘sweet spot’ after treatment is located at the main listening area is most important.

The final test is to focus on the low end of the room and capture the response, which gives a broader picture of the bass response of the space (especially important if also tracking in the control room / listening space). To do this place 1 speaker in for example the front right corner on the floor, then place the test mic in rear left wall / ceiling corner (ie speaker and mic in opposite diagonal corners to eachother). Because bass naturally builds up along room boundaries (and especially in corners) running this test will excite the room modes, giving a broader picture of low freq behaviour than just running tests at the listening position (although listening position tests are the main tests to run).


Calibrating playback volume of the test tone :

The volume that the test tone is played back at in the room is very important, so that average listening levels are simulated when running the tests. If the test tone is too quiet then the room won’t be ‘driven’ to realistic / normal listening levels, so room modes won’t be as severe and mid and high frequency reflections will be quieter than under normal listening conditions.

The recommended level to set playback of the test tone is around 79db C-Weighted Slow on an SPL (sound pressure level) Meter.

To set playback levels a burst of pink noise is usually played back through speakers (via the room testing software being used) and the SPL meter is held at ear height at the listening position, where the reading is taken for the SPL of the test tone.

If the level on the SPL meter is too low raise the output level using the software mixer on the computer running the testing software, or turn speaker volume up. If the level is too high lower the output level.

Once the level calibration is complete proper testing can take place.

The frequency range should be set from 20hz to 20khz for a full range test of the room.



Interpreting The Results :

The main plots to look at are :

• Frequency Response

• Decay Times / Waterfall Plot

• Energy Time Curve


Frequency Response :

Below is the frequency response taken during a bass only test (ie.. Speaker in corner to drive the room’s modes) testing the below 150hz range of a fairly large untreated studio control room.

From this plot it’s obvious that the most severe problems are in the bass region (which is usually the case, and the reason why getting the bass trapping right is so important).

There is a 7.6db peak at 38hz, then the 20db dips at both 75hz and 83hz. During listening tests noticeable boomy bass around a D#1 could be heard (ie.. Near the 38hz peak), but the main issue was the lack of bass between C2 and F#2 (65hz to 90hz) caused by the severe 20db dips.

The locations in the room where the bass energy was building at 38hz, 75hz and 83hz was found using steady sine waves and an spl meter. Due to layout constraints in the room only 2 of 4 vertical corners could be treated, so we treated behind each speaker floor to ceiling using BF-1200 corner units (which are the most effective low frequency corner traps in our range).

Focusing purely on the low bass response the corner trapping gave this improved frequency response :

As can be seen there’s a marked improvement in the 65hz to 90hz range, with the both the 75hz and 83hz dips being reduced by 17db or more. Also the 38hz peak, which is a very difficult area to treat effectively due to wavelength, has been reduced by around 5db to within 3.5db of flat.

This result was obtained by treating just 2 corners floor to ceiling with large corner traps - if treating the other 2 vertical corners had been possible the result would have been further improved, but overall the low bass was much tighter and more controlled, and the lack of bass between 65 and 90hz was almost completely resolved.

Please see the plots below from another real world room testing example of a small studio control room. This was a full range test (ie. 20hz to 20khz), and the room was treated minimally using BF-612 corner chunks, BF-175 traps on rear wall, 1 x BF-175 ceiling cloud, and 2 BF-125 bass traps at each side wall reflection point.

The main issue here is clearly the 20+db peak at 55hz and the wide and the almost 20db dip between 100hz to 140hz. The 55hz peak was giving a lot of boomyness at A1, and there was a total lack of bass between A2 and C#3.

The after treatment result shows a 7db improvement at 55hz, and over 13db improvement in the 100hz to 140hz dips. The mid and high frequencies also show less severe peaks and dips. This result would have been further improved via using BF-850 corner traps instead of BF-612s due to the severity of the 55hz problem, but the layout of the room wouldn’t allow for it. Overall though a vastly improved result both on paper and audibly was achieved.

Decay Times / Waterfall Plots

Decay times and Waterfall plots show how long certain notes / frequencies take to decay in a room. Achieving an even decay time in a room provides tight / punchy bass and smooth / controlled mid and high frequencies.

Most critical listening spaces ideally want to have decay times in the range of 0.2s to 0.4s depending on room size across all frequency bands.

The first plot below shows the before treatment decay times for the bass test in the first frequency response plot above. The Waterfall plot for this test is also shown.

As can be seen from this plot the bass decay times below 300hz without any treatment are far longer at around 0.7s than for the mid and high frequencies, which is due to modal ringing.

The ‘after’ treatment plots shows that in the 40 to 90hz range the decay time has been brought down by 35ms, reducing the boomyness at 38hz and generally smoothing out and shortening the decay response. This resulted in noticeably tighter bass and more controlled mid and high frequencies.

The improvement in frequency response and overall decay times shown in the after treatment Waterfall should also be noted - it shows a generally flatter and more controlled frequency and decay time with the corner units in place vs the untreated room plots above.

Energy Time Curve (ETC)

The Energy Time Curve plot shows in detail all reflections as they arrive at the listening position.in relation to the direct speaker signal, both in time and level.

Good specifications for critical listening is for there to be no loud reflections arriving in the first 20ms after the direct speaker signal reaches the listeners ears.Loud in this case means no reflections within at minimum (for hi-fi listening rooms) -15db, and ideally more like -20 or -25db for excellent imaging and accuracy in mid and high frequencies (best for studio control rooms).

Treating reflection points (left, right, rear wall and ceiling, as per the example plans) will reduce these loud reflections to a good degree, and often enough to achieve the no reflections over -15 or 20db within the first 20ms criteria. However the ETC can be used to locate and treat these individual reflections, and depending on room size and geometry it may be that some 3” or 5” spot treatment panels may be needed to deal with all reflections in this time frame. This ‘reflection hunting’ can be done using a small 2ft x 2ft piece of sound absorber and trial and error by holding the panel on different sides of the test mic to narrow down where the reflection is coming from. Or the path of the reflection can be calculated based on 1ft of travel takes around 1ms.

Below is an ETC of a clients’ room before treatment, which shows several loud / destructive reflections above the -20db point within the first 20ms.

And finally below is the ‘after’ treatment ETC showing that these reflections have been brought under the -20db mark, which subjectively gave far better imaging and accuracy in mid and high frequencies. The basic idea with this is that you want the direct signal to reach the ears without any loud reflections interfering with it.

The use of room testing and proper analysis the 3 main plots outlined on this page (freq response, decay times and ETC) aid hugely towards ensuring that the correct type of treatments are used in the best places in a room in order to give the optimum post treatment results.

These improvements shown in the plots aren’t subtle! Taking the frequency response plot as an example.. If you put an eq on your main output and boost 55hz by 22db (which is what the room was doing in the plot above) it’s going to sound pretty awful! Likewise cutting specific frequencies by 15 to 20db (which is a very common natural variation in a lot of rooms) means that certain bass notes can’t be heard at all.

In simple terms rooms behave like a built in eq (with an unhelpful boomy bass dial), which you have no control over other than via using properly designed and targeted acoustic treatments, and spending some time on speaker and listening spot locations.

Room correction software does have it’s uses, but is best used for final tweaks once a room has been properly treated. Room correction has limitations in terms of dealing with decay times in low frequencies, and dealing with dips in frequency response (in fact some types of room correction don’t deal with dips at all).