Determining 'skyline' diffuser frequency range.

[HookedOnHardware]

Hi slutz.

I was asked to post my email below as a new thread due to the following answer:

"Hey man,
Thanks for the PM, but I can only give personal help to our customers. If not I would go broke from never getting to my paid work.
If you would like to start a thread I will try to help you out. That way it may answer the question for someone else also.
Thanks for your understanding. "

Of course I understand!

Originally Posted by HookedOnHardware
Hello Glen.

As I get closer and closer to opening my new studio, specific questions arise.
I will get the new room (s; plural. But I this case I’m talking about the control room) measured and tuned by people who do it for a living but I was wondering about how to calculated the effective frequency range of a skyline type diffuser. The same people should know, however I’m an autodidact and wish to know the gist of things.

I know how to calculate. I just don’t know what to base the calculations on.
Let’s say I have a response graph (which I don’t right now). Do I just make the frequency range of the diffusers as wide as I can manage? (I will build and place them similar to what Darius did in his mastering studio. 6 in total. The whole rear wall will be a bass trap with a minimum of 40cm of an air gap behind it. The skyline diffusers can go as deep as the bass trap will be).

Right now I have them calculated on about 860 Hz to 3440 kHz. This range gives me round numbers for the pieces I’ll have to cut. 0cm, 5cm, 10cm, 15cm, and 20cm respectively.

My second (sub) question pertains to (two) skyline diffusers on the front wall. My control room is a good 6.20meter deep (+-20') so at the 38% mark I’m a nice 2.35m from the front wall (minus 20cm due to another, smaller, bass trap in the front). I often read that Ethan prefers a good 3m between the wall and the listener. Others have said 2m. In case it’s possible. Is this desirable (for increased imaging etc)?

As this is a professional recording studio, the ‘I’ll hire people to do it for me’ aspect is always needed to some extent. At the other end, I just want to know this stuff in depth.

I’m eager to hear your thoughts.

Sincerely yours,

W.Gollings

Anything you can share Glen? Ethan?

Thanks.

EDIT: I am playing with the numbers right now. It seems that the ground plate and column width figures change if I make the high freq number anything greater than 3440kHz. So for now. Let's keep that as the max. I (due to my basstrap depth of 40cm) can get the low freq number as low as 430Hz (the longest columns are now 40cm instead of 20)

Glen, Ethan.
Is 'go as low as you can go' appropriate? Or is the return value too low to consider?

Thanks!

[Ethan Winer]

If you design the diffusor to work from maybe 800 Hz to 3 or 4 KHz, that should be fine. Going much lower requires a lot of diffusor depth, and diffusing bass frequencies may not be so useful anyway.

--Ethan

[Glenn Kuras]

Thanks for taking this off of PM and on to a thread. Ethan gave you some good info so I am going to email Jason our diffusion expert to see if he has anything to add.

Glenn

[Jason Jones]

I really would not make it deeper than 20 cm (8"). A diffuser will typically be effective around one octave below the design frequency. So your design will work down to +/-500hz. Diffusing below that is usually pretty dubious.

On the HF, if you make the wells 4-5cm you will get up to around 7 Khz.

Your position of 2 meters from the front wall diffusers is no problem at all with a 2d diffuser design.

Jason

[SaSi_SiDi]

And there is also the issue that diffusers work well (create a diffuse field) at a distance of at least 2-3 wavelengths of the diffused frequency. According to this, the lower the frequency diffused, the farther away the listener needs to be from the diffuser hence a larger space is needed.

Do I have it right?

[HookedOnHardware]

Originally Posted by Ethan Winer

If you design the diffusor to work from maybe 800 Hz to 3 or 4 KHz, that should be fine. Going much lower requires a lot of diffusor depth, and diffusing bass frequencies may not be so useful anyway.

--Ethan

Thanks a lot Ethan.
(Btw, I read some stuff on your personal site. Now I knew you were an experienced man but DAMN you have done a lot of stuff in your musical life!thumbsup)

[HookedOnHardware]

Originally Posted by Glenn Kuras

Thanks for taking this off of PM and on to a thread. Ethan gave you some good info so I am going to email Jason our diffusion expert to see if he has anything to add.

Glenn

Thanks Glenn!

[HookedOnHardware]

Originally Posted by Jason Jones

I really would not make it deeper than 20 cm (8"). A diffuser will typically be effective around one octave below the design frequency. So your design will work down to +/-500hz. Diffusing below that is usually pretty dubious.

On the HF, if you make the wells 4-5cm you will get up to around 7 Khz.

Your position of 2 meters from the front wall diffusers is no problem at all with a 2d diffuser design.

Jason

Thanks, this is what I wanted to know. I have the depth to do so, but if it doesn't add any value (knowing that it's still "effective around one octave below the design frequency") then I'll just keep it at a max of 20cm. I'll save some trees in the process.

In my original question I stated the following:
"I know how to calculate. I just don’t know what to base the calculations on."
What is it that is going to tell me (let's include the hypothetical response graph again) which hf max I should use in my calculations? Around 3 4kHz. makes sense to me but what would the advantage be of let's say a hf max of 7kHz?
Last question for now (I promise):
Would the design for the front wall diffusers (i.e. freq range) differ from the rear wall ones?

Thanks again Ethan, Glenn, and Jason.

[jayfrigo]

Regarding diffusion and bass, in the typical control room or similar acoustically-small room, diffusing bass is indeed not usually called for. Absorption (trapping) is best for bass in most sealed small rooms. However, in large rooms, (halls, theaters etc.) diffusion of somewhat lower frequencies becomes an option, of course depending on the room's intended purpose.

In small rooms, consider using diffusion above the cutoff frequency - where modal density starts to even out - and trapping below. This is 300-500 Hz in many control rooms, and it's no mistake that this range is most often cited as the place below which you need trapping, and also no surprise that it is the lower effective limit of the original RPG 734 QRDs that we have seen and continue to see in so many recording studios.

The average skyline doesn't go quite as low, so you may want to consider if the gap in coverage of that part of the spectrum is important if you go with low-only traps (typically not soft-front) and shallower 2D diffusors.

[jayfrigo]

Originally Posted by HookedOnHardware

What is it that is going to tell me (let's include the hypothetical response graph again) which hf max I should use in my calculations? Around 3 4kHz. makes sense to me but what would the advantage be of let's say a hf max of 7kHz?

One part is that you're looking to diffuse frequencies for which strong specular reflections will cause the most trouble with imaging, obtrusive comb filtering, or noticeable variations in response. Consider both the most sensitive range of human hearing, and frequencies at which we get many of these localization cues and intelligibility etc. Next, consider that at the higher frequencies, things tend to scatter around more easily on their own, and reflections get more dense (lots of little waves bouncing around all over the place), at the wavelengths involved, even the slightest head movement changes things considerably, and this range also gets absorbed more easily due to a number of factors. For all these reasons, and a couple others, not installing diffusion for high frequencies just won't cause as much trouble.

However, a reason one might purposely design a diffusor with a higher range is if the wall it will be placed on is somewhat close to the receiver (listener or mic). To get the desired 3 wavelengths away, you may need to adjust the lower limit to avoid causing as many problems as you are trying to fix. In that case, even though the range may extend higher, you're really moving the bandwidth higher to avoid combing, uneveness, and possibly 1/4 wave well absorption problems (not so much with skylines), rather than actually wanting it to grab lots of 8 kHz.

Lastly, you may be simply trying to treat flutter without using absorption, and in this case you'd design a smaller, high frequency diffusor as well. This kind of diffusion may also be useful in some recording spaces. It is also often added to traditional QRDs as a second, small prime sequence inside each well. In this case it's called a diffractal, and is essentially a dual band diffusor. This is generally to create a smooth, diffuse environment, rather than being aimed specifically at imaging and all the rest listed above. It's a great second step if you already have the first step covered.

[HookedOnHardware]

Originally Posted by jayfrigo

One part is that you're looking to diffuse frequencies for which strong specular reflections will cause the most trouble with imaging, obtrusive comb filtering, or noticeable variations in response. Consider both the most sensitive range of human hearing, and frequencies at which we get many of these localization cues and intelligibility etc. Next, consider that at the higher frequencies, things tend to scatter around more easily on their own, and reflections get more dense (lots of little waves bouncing around all over the place), at the wavelengths involved, even the slightest head movement changes things considerably, and this range also gets absorbed more easily due to a number of factors. For all these reasons, and a couple others, not installing diffusion for high frequencies just won't cause as much trouble.

However, a reason one might purposely design a diffusor with a higher range is if the wall it will be placed on is somewhat close to the receiver (listener or mic). To get the desired 3 wavelengths away, you may need to adjust the lower limit to avoid causing as many problems as you are trying to fix. In that case, even though the range may extend higher, you're really moving the bandwidth higher to avoid combing, uneveness, and possibly 1/4 wave well absorption problems (not so much with skylines), rather than actually wanting it to grab lots of 8 kHz.

Lastly, you may be simply trying to treat flutter without using absorption, and in this case you'd design a smaller, high frequency diffusor as well. This kind of diffusion may also be useful in some recording spaces. It is also often added to tradition QRDs as a second, small prime sequence inside each well. In this case it's called a diffractal, and is essentially a dual band diffusor. This is generally to create a smooth, diffuse environment, rather than being aimed specifically at imaging and all the rest listed above. It's a great second step if you already have the first step covered.

Thanks a lot for your extensive answer.

I am going to put diffusers at the rear end (see rendering which I use as my avatar) but I didn't originally plan to put some on the front wall. As I read more about (front wall) diffusion every now and again a comment/experience would pop up talking about increased imaging/depth.
This sounds interesting. However, I'm sure the tuned room and monitors (Barefoot) will give me plenty of information on it's own.
Come to think of it, the diffusion might have been over the mix position or on the side walls... I'm not sure now..

Anyway. Thanks again.

[barefoot]

Whether you use diffusers or absorbers, it's always a good idea to position treatment at the first refection points in your room. If you play around with my Wall Bounce Calculator you will see how these early reflections can significantly affect the response at the listening position. Notice that not just the side wall, but the front wall also causes a lot of response anomalies even up into the midrange. If you want to isolate the effect of either the side or front wall, just make the other wall distance very large.

If you have a mildly functional drawing program, you can determine the first reflection points using a graphical mirror image technique. Make a scale drawing of your room. Mirror the speaker positions across each wall (or ceiling). Then draw lines from the mirrored speakers to the real listening position. The first reflection points are where these lines intersect the walls. Here's an example:



The red dots show the first reflection points. The further the mirrored speaker is away from the real listening position the less problematic its effect. For example, the reflection of right speaker from the right wall will cause larger response anomalies than the left speaker reflection from the right wall. If your walls are angled, like in this example, you may have to do some strategic rotation of the image in order to get the mirror images correct. Remember, you need to mirror the image across the walls, not just mirror them relative to up, down, left or right of the drawing.

Hope this helps!
Thomas

[HookedOnHardware]

Originally Posted by barefoot

Whether you use diffusers or absorbers, it's always a good idea to position treatment at the first refection points in your room. If you play around with my Wall Bounce Calculator you will see how these early reflections can significantly affect the response at the listening position. Notice that not just the side wall, but the front wall also causes a lot of response anomalies even up into the midrange. If you want to isolate the effect of either the side or front wall, just make the other wall distance very large.

If you have a mildly functional drawing program, you can determine the first reflection points using a graphical mirror image technique. Make a scale drawing of your room. Mirror the speaker positions across each wall (or ceiling). Then draw lines from the mirrored speakers to the real listening position. The first reflection points are where these lines intersect the walls. Here's an example:



The red dots show the first reflection points. The further the mirrored speaker is away from the real listening position the less problematic its effect. For example, the reflection of right speaker from the right wall will cause larger response anomalies than the left speaker reflection from the right wall. If your walls are angled, like in this example, you may have to do some strategic rotation of the image in order to get the mirror images correct. Remember, you need to mirror the image across the walls, not just mirror them relative to up, down, left or right of the drawing.

Hope this helps!
Thomas

Hi Thomas,

As always you give awesome answers! However, I wasn't, in this case, asking a 'how do I treat my room' question. It simply pertained to the science of calculating the frequency range of a skyline diffuser (not how to calculate the necessary depths of the wells but what to base the specific calculations on).

As I wrote in my original post, the room(s) will be designed and tuned between myself and a professional studio acoustics advisor. I was just wondering about specific info so my contribution and understanding (of diffuser calculations) would be useful in the process.

I'm sure you found this post by searching for your name (I am very fond of your product and am going to purchase it via msv.nl for the new studio! Can't wait!!)

I'm sending you a PM right now regarding something else.

Thanks again Thomas (and others who have and/or might chime in)!

thumbsup

EDIT: Thanks for the wall bounce calculator. It looks quite interesting!