Soundman2020 - Studio Design Forum

Really nice work!

I downloaded your latest mdat and had a quick look - maybe it's just my inadequacy but I am not seeing a major improvement in the low end from before you did the alterations - can you point out what exactly has improved in the measurements?

looking forward to the new overall room and LR. before after soffits (i took smoothing off).

Reposting as I realise I was comparing the wrong measurements. I didn't read the titles of each measurement properly and wrongly assumed they were arranged top down: least recent - most recent. But now I see the middle set of measurements are the most recent, which makes sense since they show the biggest improvement!

So, what was added to the room between the May and August measurements being taken?

Hi
Sorry for the delay. I took the measurements this morning and my summary is below. These are all unsmoothed.

Firstly, let's compare the end of phase 1 (basically where I got to before this year's forum posts, before I got advice from Gulfo), to where we are now. Same mic position.



Very happy with the improvements.

Then, showing the effects of the panels only:



I have already moved my mix position forwards approx 6", so here's the comparison of different mix positions.





You can see there is a phase shift or something causing quite a different response at the sofa (well it was a quick test, so in fact the mic was just in front of the sofa). You can also see that in the front bit of the room there's a boost around 200Hz. I had spotted this before and also the boost around 30Hz, so figured these might be able to be EQ'd down.

I'm happy with where I'm at from a treatment point of view, so I had a go at using REW's EQ pane to suggest settings for my Behringer FBQ2496. I just tried to affect things under 200Hz.



With 6 filter moves at the revised mix position it was easy to get it to read flat, but move the mic and it all changes. I quickly unplugged the FBQ and realised I needed more research or more advice!



I think I already know that I need to do the walking mic test to establish the best position. My question is, do I have to move the desk with the mic? I'm guessing so. I'm also wondering if there's any more that can be done with treatment to reduce the 200Hz hump at my mix position.

Any thoughts appreciated.

The two REW files are below - the one called 'Gareth Oct22 tweaks tuning' shows the effects of the FBQ box, whereas the other has all of the steps getting up to this point.

the higher the frequencies, the less movement it can take to alter response (objectively). yes, moving the listening position and desk as a unit. the sofa level is 10db higher than the desk? on the 200hz maybe a few low Q attenuations - 2-3db - 110, 175, 200, 225, 250

Thanks Glenn. I'll try some settings on the FBQ as you suggest.

The 10dB higher for the sofa - are you looking at the bottom graph (green line)? In which case, this is the sofa position with the FBQ filters engaged. There are some high peaks compared to the other positions and I'm assuming this demonstrates why boosting EQ to 'fix' a null at one position can cause issues elsewhere (the FBQ had peaks of +9dB at 90 and 130).

yeah eq'ing to fix the room is fraught with danger but done gently, it can help. keep in mind that the human hearing is 1/3 octave by its very mechanics - so subjective measures and smoothing are valid as well as purely unsmoothed/filtered raw readings. esp when you do have problems. one excellent test is a 10 octave piano note - a MIDI scaling - 1 note per 3 seconds second of 1/2 second note duration. A0-C8. you can then hear very clearly what notes are reinforced or not. somewhere i have a .mid file which also accounts for F-M curve so it's not purely a single level played. i'll look for it and share it.

unzip for type 0 midi file. i use a synth with the osc set to sine wave, 0 attack, 100% decay 100% sustain, 0 release. many systems will not reproduce the lowest frequencies. i use a 1khz sine @ whatever you're monitoring at to set level. this midi was defined around 80db.

of course you could also find one of these types of files as well but the midi gives you more options to tweak things...

Thanks Glenn, that’s a good idea. Will run the midi test and see what frequencies poke out the most, and report back.

Cheers

Digital room correction of this precision is not something I have attempted. But if I was to attempt it, I'd take on board the following from Rod Gervais "Home Recording Studio: Build it like the Pros" to get meaningful results. As someone once said, we don't mix with our heads in a vice!

p183 "Repeatable, Precise, Accurate, and Relevant Measurements"

"To measure the room response as a whole, multiple measurements must be taken. This normally consists of placing a mic at one location, taking a measurement, moving the microphone, and then taking another measurement. The correct number of averages must also be used to ensure that there are enough averages to ensure repeatability of the experiment for one who uses slightly different mic locations in the same setting."

p184 "Measuring LF Response"
A single measurement for low-frequency response can show lots of parasitic effects from room reflections... These random parasitic effects can be overcome by taking many measurements and averaging them. This makes resonances more obvious"

p186 "People often want to measure a room-speaker power response over the full audible range. This requires many measurements and combines the effect of the shape of the room along with the nature of the loudspeaker response. Many mic locations are used. In this particular test case, there were 32 spread over a 2' wide x 6' long x 3' high area (36 cubic feet)

In a small control room, an appropriate area might be something like a two-foot cube centred at the normal position of the mixing engineer. The spacing between the microphone positions determines the lowest frequency for which this measurement is repeatable. In this series of measurements, the microphone positions were more than one foot apart. The measurement is highly repeatable if the fractional octave smoothing is 1/3 (0.333) octave or lower for frequencies above a minimum wavelength of four times the mic spacing distance. The experimental result below is repeatable for the frequency range above about 250Hz; below 250Hz the results would be slightly different if the experiment were repeated with slightly different mic positions."

"How Many Meausrements Are Needed?"

A single, or only a few, measurements would indicate a random nature in the data gathered. Due to this, no real conclusions would be able to be drawn."

Thanks Endorka. Does anyone know if there's a way to average results in REW, or is it just with one's eyes!?!

Going to try the MIDI file and ears approach in a bit, rather than the REW and eyes approach.

Cheers
Gareth

Not at the computer just now so please forgive imprecision. Answer is yes - it's in the All SPL display. The controls panel on top right. You can choose a vector or SPL average of all selected.

Cheers,
Jennifer

and the smoothing setting as well.

I just tried averaging some measurements I made with the walking mic technique a while back. One set were front to back, the other set was left to right. Of course this will only really be offering utility in one dimension at a time, but the "parasitic effect reducing" was really quite impressive.

If you have any historical walking mic measurements stored you might get some insight from them by doing this.

Cheers,
Jennifer