Attic Shaped Studio

[gullfo]

so - the thick material in the box - good.
on the buzzing - the noise is related to the speaker output and not the air flow?
if the space around all of that is filled with insulation (which is will be evenutally), i'd suggest in most cases this eliminates the problem.
one other option is to make a "hard duct" using the duct board to create the ducting (a couple of right angles), as this tends to be more damped and less prone to vibration the resonant frequency will be much higher and thus more easily attenuated. a short section of flex to connect to the bottom of the speaker box connector.

[endorka]

Thanks Glenn. Yes, definitely related to the speaker output, it's fine with airflow.

Cheers!
Jennifer

[endorka]

I removed the "plastic barrier between the duct and the glass wool". It's that brittle crinkly type plastic, really noisy when it moves, like a crisp (potato chips for USA?) packet. I replaced it with food wrap (cling film) round the duct and that is silent. Problem solved.

Cheers!
Jennifer

[gullfo]

eggsellent! yes, we call "crisps" = "chips". handy reference: https://www.mashed.com/279089/is-there- ... nch-fries/

[endorka]

Good article. Also in Scotland, "A fish supper" is what you ask for when you'd like "fish and chips"

[gullfo]

good to know as me thinks our daughter wants to do some touring of the UK

[endorka]

I stuck the cooker hood filter material to the inside of the soffit speaker box with contact spray glue. Unfortunately there wasn't as much room at the top so I couldn't fit the duct liner there as I'd hoped to do.

Duct flanges on the plenum and bottom of the speaker box.

I've added some more horizontal framing to the soffit structure to make it stronger and more rigid. I'll do something similar round the sides soon. Nice to be able to re-use some old table legs I've been keeping for years in case I found a use for them one day

I wanted to make sure the horizonal soffit structure beams did not pull the soffit face inwards or push it outwards when it was screwed into them. So I attached these beams from behind with these pocket holes when the baffle face was in place

Soffit on the other side done the same way, and also has some of the flex duct attached. The top flex duct comes out the back of the speaker box and will emerge on the side of the soffit.

[gullfo]

very nice! the extra rigidity should help as it raises the resonant frequencies which are more easily damped when you layer in MLV and/or fill up the space with insulation.

[endorka]

Time for an update!

I decided to replace the buzzy acoustic flex hose on the speaker ventilation with standard flex hose. When surrounded by the insulated outer layer there is no buzzing. Hallelujah! Also shown is the side panel of the soffit. It's made from one layer of 18mm MDF and another layer of 18tmm OSB with two layers of MLV sandwiched between. Total surface density about 31kg/m2. Since this is the narrower of the two soffits I had to fit the OSB round the stud structure. The flex hose will attach to a flange on the round hole.

Test fit of the above. An egg crate grill will go into the square hole cut into the MDF.

The slightly wider soffit required some spacers to get the side panel flush fitting.

Fill them up with low density (10 kg/m3) loft roll insulation.

Side panels back on and we've back in business

[gullfo]

about friggin' time! LOL (actually it was rumoured we'd never hear from you again once you got the soffit operating and the studio was now complete!)

looking good and glad the smaller problems are getting eliminated. any significant change in the audio?

[endorka]

LOL!

My subjective notes from listening to reference tracks after fitting flush mount side panels and filling structures with insulation

- ambience e.g. snare room sound, reverbs, sparkly effects (Lose Yourself To Dance) even more present.
- bass note that was too much in "Motherboard" tamed
- bass less overpowering and tighter. Frequency balance seems correct to my ears.
- The attack on transients like handclaps very precise sounding and more present than before.
- phantom centre stronger.
- Cleaner sounding midrange. No longer hearing the need to clear up midrange blur / slightly dodgy sound in recent drum overhead mic recordings. Presumably it was the mix room, not the recording, all along.

I did some quick REW sweeps. Impulse plots show a couple of prominent reflections in the first 30 milliseconds.

The peak at 9ms is a reflection from the side wall opposite the speaker. It's easily tamed with a 100mm rockwool panel on the door.

I think the peak at 3.5ms comes from the 2.5 hedral corner at the ceiling in front of the speaker. I expect a cloud or angled hard panel could deal with this.

SPL for the right speaker: There

Still a bit lumpy for my liking, but a big improvement over the previous setup. Mixing is much more enjoyable, I hardly ever need to check the bass on headphones, and no longer feel the need for any fixative EQ or "house curve" trickery in the monitor path.

There's still some cosmetic work to do on these soffits, and fitting the ventilation grilles. After that it's time to address the height and width treatment, which is mostly just sparse 50mm deep absorber panels that will be having no significant effect on these low frequencies. Lots of scope for improvement there that will hopefully address the lumpiness somewhat.

Cheers!
Jennifer

[gullfo]

so the edge of the baffle and the ceiling join - fairly sharp (from photo) and likely some edge diffraction there. in general, soffit assemblies like to have as flat as possible overall surface between the speakers - so the window inset depth and the corners may be playing there. if you had some pipe insulation (the black plasticy kind that comes in a split tube) you could try that on the edges to see if it changes. it's one of the reasons your speakers have rounded edges. if it is the edges, then crafting a rounded moulding (somewhat flatter on the baffle plate, and wrapped and rounded towards the side wall) over the edge to cause a more smooth transition to the side wall into the window space could do it.
also you notes the wall over top, possibly an absorber down to the top edge of the baffle plate, likely 2" thick, might be good if that is also a source.

[endorka]

That's a good call Glenn thank you, and a simple and effective way to test for it. I'm sure I have some of that pipe insulation downstairs in the garage.

If it turns out to be an issue I could probably route a rounded edge on the sharp edge of the soffit.

Cheers!
Jennifer

[endorka]

I've tried to make a fair comparison between before and after the speaker flush mounts were installed. Here's an SPL chart of red (before) and black (after).Var smoothing to give a readable presentation of highs while preserving detail in the lows.

In my opinion the substantially more even response from 100 to 1000 Hz is massive improvement. The previous dips in such a crucial part of the midrange are far from ideal in a mix room.

The slight rise at 7kHz is not a concern of the room, it's because of the off axis response of the speakers. The frequency response graphs in the Genelec manual say this. I imagine it would be less of a concern when one is listening with two ears more directly in line with the tweeter than the single measurement mic.

Below 100Hz the "after" peaks and nulls are more pronounced. Two possible reasons for this:

1) I strongly suspect I had the speaker low frequency reduction engaged in the "before" measurement. I don't feel the need for it any more to get a balanced response, so it's not engaged in "after".

2) There are far fewer broadband traps in the "after" measurement.

I think the first of those is most likely.

For perspective on how much less absorption is now in the room have a look at these "before" photos. All but one of the original four diagonal corner straddle traps have been removed. Better results with fewer absorbing surfaces can't be bad! The room being a bit more live (and evenly so) also helps with mixing I think. And there is plenty scope for re-introducing low frequency absorption in a better way.

[gullfo]

yeah, the 70hz dip is likely your ceiling. so, some deeper absorption overhead should help some or possible consider some tuned membranes. eq'ing the peaks down is a valid approach to the 50, 100, and 250hz.