Building a proper control room – final tweaks using EQ, primary diffusion & installation of a sub-woofer

Hi there and welcome back!

We are slowly nearing the end of this series.
Today we will have a look at what EQ correction can do as a final tool. Furthermore, after a little bit of research on the web, I decided to tune the bass region a little more by installing and tuning a subwoofer. Meanwhile I also installed a primary diffusor on the back wall of the control room to diffuse the reflections there. But let’s proceed step by step.

If you ever worked with REW you will have noticed that it provides a module to generate correction curves for different types of EQ. In a first step I decided to simply use an EQ plugin that I had at my disposal, DMG Audio Equilibrium. So I let REW calculate a “generic” EQ curve, fine tuned the settings a little bit. The following graph shows what REW came up with.

EQ_correction_REW

Calculated EQ correction curve (REW)

I then took these settings to Equilibrium and saved them. The only problem now was that I wanted this EQ to correct everything that is played back on my system, not just my DAW. A possible solution to that problem is to use a loop back on your audio interface (RME Raydat in my case), send that to some plugin container where you insert the EQ and then send the signal back to the monitor output of your interface. As a plugin container I am using AULab on Mac, but there are equally usable containers for Windows out there.

Now let’s look at a measurement of the EQ’d frequency response. What we see is that the EQ certainly makes a difference but probably not as much as one would have expected. Well, you know you can’t remove things like comb filtering, since that’s coming from first reflections. An EQ would only shift the frequency of this comb filtering, but wouldn’t remove it. Moreover, the settings in REW may have been a bit to restrictive as far as the number of bands and settings of bands goes. I will show you next time what a dedicated calibration software (Sonarworks Reference 3) can achieve here.

fg_eq_ohne_diff

Frequency response after EQ correction without sub

rt60_eq_ohne_diff

EDT after EQ correction without sub

What made a lot more difference for my control room was bringing in a sub woofer – the Presonus Temblor T10 in my case – and tune it such to improve the frequency response up to about 300Hz. This took at least 50 measurements to get it correctly. However, it was well worthwhile the hassle as you can see from the comparison measurements below.

fg_sub_vergleich

Comparison measurements with and without subwoofer – frequency response

The bass response has been tamed quite a lot and is much flatter now, with less highs and dips. Very nice. Taming the bass response is probably not what you would expect from a subwoofer, but this is how you should use subs actually. The thought behind it is, that you have a lot more freedom in moving the sub arround in the room, thus finding a position that has shows interference with walls as speakers that are positioned near the front wall. And you can even take care of the phase between sub and speakers and use that as a tool to smooth out the frequency response. That would even work better with a dual sub setup but that is out of my financial reach at the moment.

rt60_sub

EDT after installing and tuning the subwoofer

etc_subdiff

ETC diagram after installing and tuning the subwoofer

waterfall_subdiff

Waterfall diagram after installing and tuning the subwoofer

What we see from the ETC diagram is, that there are still reflections on the table that we should care about for the sake of image clarity. So I  decided to rearrange my desk by angling it’s surface to reflect away from the listening sweet spot and move from the Qcon control surfaces to a touch based setup. More on this desk setup as well as a closer look on Sonarworks Reference and the according comparison measurements will be the topic of next weeks post.

See you then and take care,
Markus

Advertisements

Building a proper control room -installing the studio desk

Hi again!

It’s time to install the studio desk.

After thinking about how much money I would be willing to spend on a desk, I came to the conclusion that none of the professional offers are within my budget. So I came up with my own sketches, some more others less elaborate. In the end I decided to keep the construction as simple as possible. I had already built a 19″ rack out of glued wooden boards that incidentally had the perfect height to place a tabletop on top of it. So I just built a second identical 19″ rack and placed a wooden kitchen worktop (hevea; oiled afterwards) on top of these two racks. Finally I placed one glued wooden board in between the two racks for stiffening purposes and everything was finished. Building this desk took less than 3 hours, including cutting the wood and oiling the worktop.

Meanwhile I had also built some monitor stands. These are again just glued wooden boards, screwed together, sealed along the edges and then filled with dried sand. Each of these stands weighs about 80kg. They work perfectly in tightening the bass response of the speakers according to my measurements (which I either didn’t save or lost unfortunately; sorry for that). Then my 3 PC screens were mounted on supporting arms screwed to the wall and placed in a way to seamlessly integrate with the Icon Qcon controllers that I used at that time. The controllers themselves are angled a little bit – but not enough as we will see in a moment. All that and a lot more is shwon on the pictures below.

This slideshow requires JavaScript.

One drawback of a desk – especially if it is not angled sufficiently – is that it inevitably introduces comb filtering into the perceived frequency response. This happens due to reflections on the tabletop. Furthermore, in the setup you see above, even the edges of the PC screens introduce some comb filtering since the speakers are placed behind the screens. Naturally, the frequency response is now worse than before.

In the meantime I sold the Qcon controllers and switched to a touchscreen, altering the studio desk to a more transparent design. But more on that in a future article.

tisch_fg

Frequency response with the studio desk installed (gotten worse)

The EDT curve shows no real change as compared to the last measurement. However, I also start plotting the Topt curve from now on, since in my rather small control room RT60 is not necessarily the best concept due to a lack in volume of the room. REW tries to correct for that with the Topt option in the EDT plot. There you can see that we are down to around 200ms of decay time for the whole frequency spectrum, which is quite good. For some folks this may already be on the dry side, but I personally like the way it sounds. It’s certainly not sounding dull, since the 200ms are achieved quiet uniformly across the whole frequency spectrum. Remember that dullness mainly comes from overdamping the high frequencies.

tisch_rt60

EDT curve with studio desk installed

tisch_waterfall

Waterfall diagramm with studio desk installed

tisch_spectrogram

Spectrogram with studio desk installed

In the ETC diagram we now see the nasty reflections on the screens and the desk. Not that good actually, but as I said above I already changed the desk setup in the meantime to successfully get rid of most of those peaks.

tisch_etc

ETC showing more peaks due to reflections on the screens and the desk

That’s it for now. Next time we will continue with installing a sub woofer and tuning it such that the overemphasis of bass frequencies is removed (!) and the dips at 80Hz and 120Hz are filled a little bit. Furthermore, we will have a look at primary diffusion and at what some final tweaks using an EQ plugin can do for the room.

Hope to see you then.

Best,
Markus

Building a proper control room -construction of a cloud absorber

Hi there!

We are moving forward with our control room series. Today we install a cloud absorber, i.e., a broadband absorber on the ceiling between speakers and listening sweet spot to tackle early reflections on the ceiling. Of course, we are going to analyze what difference that makes to the measurements.

Since the cloud absorber is hanging from the ceiling on simple chains mounted to screw hooks I decided to make it as lightweight as possible. Therefore, I used BASF Basotect which is an open cell foam made out of melamine resin. With a thickness of 10cm it’s so lightweight that I decided to glue 3 pieces (1 x 0.5m) directly together and glue those onto a frame made of thin wooden slabs in order to be able to screw the hooks into these slabs. You can see that construction in the first picture.

dscn0001-3

Construction of the cloud absorber as seen from the back

And the second picture shows how the cloud absorber looks like when mounted to the ceiling. Due to the special bass absorber construction that we talked about earlier, the cloud absorber has also been tilted upwards. This leads to the positive side effect that all frequencies which are not absorbed completely, are reflected away from the listening position.

dscn0001

Cloud absorber installed in the control room

When looking at the frequency response graph now one sees that the region from 300Hz to 5kHz has been further smoothed out due to the cloud absorber. Moreover, the different dips between 100 and 200Hz have been eliminated leaving just one single dip at around 150Hz, while the 280Hz dip that we saw in Building a proper control room – construction of early reflection absorbers has been shifted upwards in frequency to about 320Hz.

cloud_fg

Frequency response with cloud absorber installed

The EDT curve, waterfall and spectrogram show another decrease of the early decay time apart from the region below 90Hz that is clearly not affected by the cloud absorber as we could have imagined.

cloud_edt

EDT curve with cloud absorber installed

cloud_wf

Waterfall diagram after installation of the cloud absorber

cloud_spectrogram

Spectrogram after installation of the cloud absorber

Fortunately, the peaks in the ETC are now also down to approx. -14db. We will try and lower them to -20db, which is no easy task taking into account that a studio desk has to be installed as well. This in turn will probably lead to further early reflections on the surface of the desk.

cloud_etc

ETC diagram after installation of the cloud absorber

That’s it for now. Thanks for following that blog post series and see you next time, when we will install the desk and talk about further tweaks using equalization.

Best,
Markus

 

Building a proper control room – construction of early reflection absorbers

Hi there!

In our control room series we now start to battle the early reflections.
This is done by the use of early reflection absorbers that need to be able to absorb frequencies as broad as possible. Therefore, one should not use any reflecting materials like foils in front of such elements as these would reflect higher frequencies.

In order to prevent the release of fibers I decided to leave mineral wool aside and used Caruso Isobond WLG35 instead. Isobond is a material that’s woven out of polyester fibers and that looks and feels similar to a mattress. With a specific flow resistance of 12 kPa*s/m^2 Isobond WLG35 is ideal for a thinner early reflection absorber. Moreover, it’s completely free of chemical additives. For each of my two absorbers I used 3 pieces of Isobond with a thickness of 10cm, leading to a total dimension of 1 x 1.5m of absorbent material mounted inside of a wooden frame that had been previously mounted directly to the wall. At the end everything has been covered with the same fabric as the bass absorbers in the previous articles.

The correct points to mount these early or rather first reflection absorbers are on the side-walls in between your (near-field) monitors and listening sweetspot, such that they cover the area of first reflections on the walls, see e.g., the sketch here.

And this is how a finished absorber looks like.

dscn1836-1

Finished early reflection absorber on one wall of my control room

Furthermore, I needed a place for my clients to sit down and relax when listening to mixes or discussing recordings. I opted for a big fabric couch and placed that at the back wall of the control room. As it is 240cm long and 100cm deep I expected that it would have quit some impact on the room acoustics, even for lower frequencies.

dscn0001-2

Big couch for clients to sit down and listen to mixes

So let’s see how the measurements look like after that changes.

When taking a closer look at the frequency response graph we see that some of the bigger minima and maxima have been smoothed out quiet a bit as compared to the graph in Building a proper control room – recovery of mid and high frequency response by the use of wooden boards. Most noticeably the pronounced dip at 90Hz and the maximum at around 130Hz. However, we also see a new dip at approx. 270Hz and a second one that appeared between 100 and 200Hz. No win without loss, you know. We will see what happens to the frequency response when we proceed in the construction process.

er_fg

Frequency response with ER absorbers and couch installed

The EDT curve, waterfall and spectrogram show a significant improvement now. You can clearly see that we succeeded in bringing down the Early Decay Time to below approx. 300ms nearly everywhere in the spectrum, apart from the region below 90Hz, which is due to room modes obviously. But even there it is down to 450ms already, which is a really good value for such a small room.

er_rt60

EDT curve with ER absorbers and couch installed

er_wf

Waterfall diagram after installation of ER absorbers and couch

er_spectrogram

Spectrogram after ER treatment

Interestingly enough, the ETC shows now even more peaks. We will se how that develops with the next measurements.

er_etc

In the next article I will describe how the ER absorber on the ceiling, i.e., the cloud absorber has been constructed and how mounting it changed the measurements again.

Thanks for sticking with me here and all the best,
Markus

 

Building a proper control room – recovery of mid and high frequency response by the use of wooden boards

Hi there!

It’s time for the fifth article in the control room series.
This time we’re gonna talk about how to recover mid and high frequencies without affecting the bass frequency absorption too much. I have done this by using wooden boards that are mounted onto the bass absorbers with certain slits in between them. The boards are about 19.5 cm wide while the slits are approximately 2.5 cm wide. In theory this construction should reflect all frequencies above approx. 1760 Hz while everything below that threshold is diffracted around the wooden boards and then absorbed by the bass absorbers behind the boards.

This is how the construction looks like in the finished control room (anno 2015; desk changed again already – more on that in a later article).

However, be aware that the measurements shown here have been done without the furniture, i.e., without desk, screens, racks, couch, etc. That’s very important since everything you put into the room affects the measurements. So in order to see what the effect of the wooden boards is we need to just add those and must not change anything else in the room. That’s exactly what I did.

When looking at the frequency response we can see that it changed quite a bit compared to the measurements in Building a proper control room – mounting fabric and difference measurements. You can also clearly notice the deeper dip at about 160Hz which might be some kind of resonance effect due to the panelling.

vertaefelung_fg

Frequency response after mounting the wooden boards

However, EDT graph, waterfall and spectrogram all clearly show that the goal of recovering mid and high frequencies has been perfectly achieved.  What one sees however, is that the reflective region of the panelling already starts at significantly lower frequencies than the theory would suggest. Remember we calculated approx. 1760 Hz for the lower frequency limit while the measurement shows significant differences down to 900 Hz. This suggests that simply calculating the limiting frequency by a “diffraction model” doesn’t capture the acoustical phenomenons well enough. However, that’s no real problem here as we verify every step of the construction by measurements anyhow.

vertaefelung_rt60
Early Decay Time (EDT) curve after mounting the wooden boards
vertaefelung_wf

Waterfall diagram after panelling

vertaefelung_spectrogram

Spectrogram after panelling

The ETC diagram is now worse than before with more and stronger early reflections. However, that is also perfectly reasonable since the panelling significantly increased the area of reflective surfaces.

vertaefelung_etc

Now that we have recovered the mid and high frequency response the next task is to build some early reflection absorbers and put a big couch for clients in the rear of the control room. These steps will significantly lower the mid and high frequency response again and will finally bring them into reasonable boundaries.

The main point here is the following: If one would refrain from using panelling or any other reasonable method to “rescue” mid and high frequencies and instead proceed with building early reflection absorbers the high frequencies would have been damped way to much in the finished control room. This in turn would lead to a muffled sound that no-one needs in a control room.

That’s it for now. We will see what the early reflection absorbers and the couch can do for us in the next chapter of this series. See you then.

All the best,
Markus

Lake People v4 (1176 clone) short review

Hi there!

Here is the next video upload for you. This time I’m talking about the Lake People v4 compressor which is kind of a clone of the famous Universal Audio 1176. It has been manufactured by german company Lake People in the late 1980s and early 1990s and I was able to get my hands on serial # 2 on the used market recently.

If you have any comments or questions please leave them at YouTube and I guarantee you to answer there. I am really looking forward to get into a discussion with all of you.
Please also consider subscribing to my YouTube channel for more to come.

Thanks and all the best,
Markus

Mix insight video

Hi there!

I recently uploaded a short video to my YouTube channel that shows a little bit of how I approached a mix I did for Herbert Guschlewski.

I’d be glad if you consider subscribing to my YouTube channel and am looking forward to your comments.

Best,
Markus