John Sayers' Recording Studio Design Forum Archive

You could make the desk a little larger out to the sides, yes (extending the "wings" a bit) without too much acoustic penalty, but it might start getting in the way of the access path through the doors if yu go too far. And you could extend it towards the window exactly 1" Why? Because that's how far you'll be moving it backwards to compensate for the soffit issues!

I would suggest not going further than that towards the window: we don't want reflections from the desk surface hitting your head and we don't want to create acoustic zones above and below the desk surface that are wildly different in response.

I also have some ideas about how to reduce the total surface area of the desk faces, especially the legs, so when we are done with the soffits, cloud, and slot walls above the door, I'll do a bit of re-design on the desk to see how big I can make it, and also to optimize other aspects.

- Stuart -

Cool,

I want as much desktop as I can get. I was also looking at ways to get increased rack space as time goes on. Not needing to add it right away, but having the option to do so later. Maybe in between the legs under the racks or adding side modules later. There's a decent amount of space before you get to the midpoint of the sliders.

I'm optimistic....

Also, this might be interesting for some people, regarding the importance of correctly and accurately positioning your speakers and mix position, and also the measurement mic. It might be an eye-opener, too, since most people wouldn't expect such large changes in response from such small changes in position, but here it is:

Backstory: Frank noticed that some of the wood he had used for the soffit framing had warped, and there were also some additional inaccuracies in the construction, that prevented him from getting the speakers flat, level, plumb and correctly aimed at the mix position while also getting the baffle on straight. Not huge differences (just a couple of inches, in total), but big enough to be of concern. So I explained to Frank how to do a series of high precision REW tests at several carefully selected spots around where the mix position was originally supposed to be, to see if we could find a new acceptable spot without needing to physically modify the soffits (or worse still: tear them down and re-build them! ). So Frank ran those tests, while getting bored to tears (its a whole bunch of tests, one after the other, with minor moves of the mic in between), and I analyzed them in various ways. Here's one of the results:
That's a series of graphs of the DIFFERENCE in level between the Left and Right speakers. Wherever a curve is above the 0 dB line, the left speaker is louder for that frequency in that location in the room. And wherever the curve is below the 0 dB line, the right speaker is louder. Each curve shows the difference in level at a different physical location in the room. As you can see, there's clearly a pattern here as the mic moves from the D-4 position (pink curve) to the D+4 position (dark purple curve). If you take a close look at the area around 48 Hz, and also at around 180 Hz and several other spots too, you'll see that there's a progression in the curve getting smoother and smoother, then getting rougher and rougher again. The light green line marked D-0 is obviously the best, and smoothest, but also clearly not as smooth as it could be. You can extrapolate mentally, and expect that if there were ANOTHER line in there, with the mic somewhere in between the D-0 and D+2 locations, then the line would be flatter still. And that's the reason why Frank is going to move his desk! To get his head on that flattest part of the line, he can either spend days re-building his soffits to correct the warping and construction errors... or he can move the mix position! So guess which one we decided to do!

Now, you might think that those various curves were taken with the mic at points that are a long way apart, because the differences are clearly very large: you might suspect that the mic moved a couple of feet between tests. And you'd be wrong! Each step is just a couple of INCHES. Between the pink curve and the light blue curve, for example, the mic moved exactly two inches. This is why you often see me harping on the importance of precision, and being extremely careful to always get your mic in the exact same spot in the room when doing REW tests while you are building. I'm sure some folks think I'm nuts when I tell then that the mic absolutely must be in the precise same spot, accurate to with 1/4", for all REW tests, because if not, the readings are not validly comparable. Here you can see that I'm NOT nuts, and there really IS a huge difference in curves from moving the mic just a small amount. Imagine how wrong you would be if you did a REW test with the mic in the "pink" position, then put up some acoustic treatment panel that you just built and measured again, but with the mic 2" off from where it should be, and you saw the light green curve: then you thought to yourself "Cool! my panel worked fantastic! It flatted everything perfectly!". Oh, how wrong you would be!

So always make sure that your mic is at the EXACT same point in the room for every test.

In this case, I specifically WANTED the mic at several different spots, but that's an unusual case. Normally you DON'T want the mic to move at all.

And just to clarify before someone gets confused here: these curves are NOT normal SPL curves. These curves here show the DIFFERENCE in performance between the two speakers.

- Stuart -

Success! Frank has just completed the very successful "floating" of his speakers in the soffits! They are now sitting happily in there, fully decoupled from the structure, and just waiting for him to put the baffles on, in a few days.

Here's some REW graphs, to show where we are now. The graphs are posted in pairs, with the first one in each pair being the "before" graph from the previous tests (right when the hangers went in to the tops of the soffits), then the "after" graph, with the speakers floated. For the "before" graphs, the speakers were just sitting directly on the soffit shelf, and the "after" graph they are floated.

First (and least important) the frequency response graphs.

BEFORE:
AFTER:
---
Then the same as above but with the phase graph added: This is a lot more interesting! You can clearly see how the phase is now rotating much more nearly the way it should be across the entire low end:

BEFORE: AFTER:
--------
And finally the spectrograph plots, once again showing how the low end is smoothing out more.

BEFORE:
AFTER: We'll try to get some photos of the soffits tomorrow, after Frank finishes celebrating! Getting the speakers floated in the soffits is an important milestone, and it tool a while longer to do, thanks to the unexpected error with the Sorbothane calculator, but we figured it out in the end, and everything is floating very nicely now. So I expect Frank is having a party right now, to congratulate himself for getting to this point. Things should go a bit faster now.

- Stuart -

Indeed this has been hard! No partying here, but the fire is back and I'm pushing 12 hour days now! So spirits are definitely high.

To be clear, that was absolutely no fun at all! I shot some pics over to the man for approval

Peace out

As the saying goes: "Pics, or it didn't happen"...

Thus... :
It DID happen!

Looking very nice, too. Fully floated speakers, perfectly aimed, perfectly symmetrical...

Nice work, Frank!


- Stuart -

Some shots of progress for morale sake.


We have the lower hangars, and baffle boards in production!

Looking good!


- Stuart -

Frank has been a busy boy!


Speakers are fully floated, soffit interiors are complete, with hangers above and below, insulation loosely packed around the speaker enclosures, and on the baffle extension....

So next up will be the baffles! He's working on those right now, so stay tuned for the next update...

It looks EXCELLENT Frank! And the latest REW data also looks great.

- Stuart -

Another day another milestone.......Test fitting the Baffle before finishing it.

Both baffles are up, and there's some new REW data, for those interested in understanding what baffles do, and how the room is progressing with that, acoustically.

First, the graph that everyone looks at but isn't really the most important: Frequency Response:
The turquoise color curve is before the baffles went on, and the purple curve is right after the baffles went on. You can clearly see them doing their magic, boosting the power below about 600 Hz by about 5 dB and smoothing the response in some key areas (notably around 500 Hz). There's also a slight lift and smoothing in the highs.

Then the waterfall for the bottom end of the spectrum (lows and low-mids). In this case, I overlaid "before" and "after" onto the same graph. The turquoise curve is before the baffles went on, and the purple curve is right after they went on: You can see how modal things have smoothed out a bit in the time domain. For example, at the spot highlighted by the cursor at 113 Hz, there was some modal ringing going on there that is now attenuated. Ditto around 54 Hz, 59 Hz, 84 Hz, and a few other places. But the really interesting thing is in the low mids, where we got that 5 dB power boost, there was no change in modal stuff! So that's a pure intensity increase (from fixing the power imbalance inherent in all small speakers), with no modal penalty.

Then the really nice one that shows some great precision workmanship by Frank. This shows the DIFFERENCE between the left and right channels, smoothed to 1/3 octave, which is how the specs say it should be measured. Wherever that gold-colored curve rises above the "0 dB" line, the left speaker is louder at that frequency, and wherever it falls below 0, the right speaker is louder. As you can see, the line is pretty flat! That implies that both channels are producing the exact same thing, with no differences.

But that line is so flat that we have to zoom to see the details... Notice the scale up the left edge of the graph: Instead of show a range of +/- 30 dB, it is now zoomed in to show only +/- 7 dB, and the line is STILL flat! So we need to zoom in even more, and take some actual measurements of that curve.


Sorry that this image is so big that it blows the margins of the post, but it's the only way to see the level of detail necessary to understand what Frank has achieved here... Note the scale: each division vertically is now only 2 dB. I have highlighted the entire spectrum with that lavender-colored measuring tool, and if you scroll across to the right, you'll see the actual measurement numbers. It says that the range of variation is about 1.2 dB. Thus, the variation is is about +/- 0.6 dB. The most stringent specs for control rooms call for this to be about 1 dB, but Frank has gone way better than that, and is getting just 0.6 dB. There is no way that your ear could ever discern such tiny variations: the minimum change that you can hear is about 1 dB, if your hearing is in perfect shape, and under optimum conditions. So even if you have perfect hearing, and you were to go sit in Frank's chair, you would be totally unable to hear even a vague hint of any difference between the channels, no matter how hard you try.

In other words, Frank's room has near-perfect symmetry, and is producing a perfect stereo image, with the cleanest possible sound stage, exceeding the toughest specs for control rooms in this aspect.

But we ain't done yet!

There's still quite a bit that we plan to do to this room.

The next big thing will be the desk, and of course that will mess up some of what we have achieved here so far, simply because it is a large solid, reflective object in a small room, but there's some tricks up my sleeve still for designing the desk to minimize problems, and then to treat the remaining untreated parts of the room to help with that...

Congratulations, Frank! You've done a pretty darn good job of getting that room symmetrical.


- Stuart -

A pic of both baffles after testing... Now I gotta finish em...

So here is the rough cut of my desk for testing purposes. This will prove to be the final tug of war between Acoustic Engineer and Sound Engineer... Form vs. function....Will I have to comprise awesome work flow for accurate sound????? Time will tell....

Having left the realm of sensibility with Stuarts desk design, morale is holding out for now as I craft my dream desk design. But how long can it last and how much of my shenanigans will Stuart allow????

Stay tuned as the man likes to say!!!

Today's episode. Can't wait to see it all trimmed out.....

Hi Stuart.....

Ready for more action...