Its hard to say what exactly is going on as the two sets of measurements cannot be easily compared.
We can observe differences, but it is difficult to determine which are a result of room changes and which are the result of mic and speaker position changes.
Our ability to comment on the room is also further restricted as we do not see what the ‘native’ response is in the room without treatment. Thus, we cannot tell if the treatment that has been applied was necessary or done optimally. …an issue that is important in a small room where we want to preserve as much specular energy as we can while managing it in a productive manner. So, we are almost necessarily limited to noting what issues remain despite treatment.
So let’s step through a few observations – being careful to avoid jumping to actual judgments of room behavior due to the lack of 1:1 correspondence.
One particularly curious note is the presence of low level high Q resonances that increase with time! Cue Twilight Zone theme…
And contrary to my earlier admonition that we cannot definitively determine many aspects, one that we can is the presence of a 60 Hz electrical issue that is evident in the waterfall.
Overlaying the 2 sets of ETC measurements shows a significant difference, which cannot easily be explained by the corner traps and most likely is attributable to a repositioning of the speakers and mic between measurements.
In fact, in the later set of measurements, it looks like the entire room response was damped with the exception of the prominent sparse early arriving reflections In other words, that the low level reflections have diminished in gain while the higher gain sparse reflections have not – something the corner traps simply cannot do, especially considering that specular energy is not incident on the traps. It is quite curious that the prominent sparse reflections remain at about the same levels while the associated lower level reflections are significantly reduced!
Another concern is the assignment of T=0 to the initial arrival of the impulse response.
The Preferences->Analysis panel is set correctly but there remains a problem in that sound does not propagate in zero time. Something is amiss in the hardware loopback configuration. Perhaps more significantly, it means that you must determine the vector resolution of the early arriving high gain sparse ‘spikes’ by using the blocking method, as you cannot accurately determine the total time of flight and thus cannot calculate the actual corresponding distance.
Another issue that we can determine from the measurements is a big one and you won’t like it – but neither will you like the practical impact of the problem that is created.
One thing that is apparent in both sets of measurements is the very significant very early arrivals (0-~0.4ms as well as the sparse reflection at ~1.45ms) which is almost certainly due to speaker mounting and reflections off the work surface. This NEEDS to be addressed as it is a serious detriment to the response. Get the speakers properly mounted on stands (whose base does NOT protrude beyond the edge of the speaker cabinet!) located further behind the work surface. And tilt the work surface to minimize the upward reflection of any reflections. This can be done relatively easily by simply moving the desk out when you are actively working and pushing it back when not in use. I would recommend marking the floor in order to facilitate the proper reproducible positioning of the desk.
Additionally, you might realize a further benefit from the corner traps by adding a 6+mil barrier to the front beneath the cloth face. This will reflect any incident specular energy above ~600 Hz which may be useful to help ‘flesh out’ the early arriving low energy density. But before going to the trouble, you might measure with the barrier applied to the outer surface to first to determine if it would add any detrimental energy returns...and the additional returns will be dependent upon the amount of mid or high energy sourced from the speakers…but it’s hard to tell from the angles of the speakers in the picture.
The next observations are dangerous in the sense that I do not know what the response of the system is with9out treatment. We only see how the system behaves in the room with the presence of some treatment, so the actual nature and source of the issues in not easily determined without more investigation. (May seriously suggest, if it is not TOO outrageous a chore, to make measurements at least without absorbent panels. Then we can make definitive determinations…)
Since those reading this come from many application points of view, let me quickly walk through a few considerations applicable to all, and not merely to this specific environment.
I am assuming that the primary use of this space is for recording and production work – recording, critical listening and mixing of an electronic (as opposed to an acoustic) source (the keyboard).
Thus the early reflections in the ISD up to the first significant energy return from the rear of the room, should be reduced with the surgical use of absorption by a minimum of about 15 dB. This will insure an analytical – effectively anechoic - arrival of the very early signal rendering the localization in the mix and the imaging to be very precise and tightly defined. In other words, you will hear exactly what is being recorded and reproduced in the speakers, prior to the augmentation of the sound by the room. Your goal is to optimize what is recorded independent of the contribution by the space.
Thus you will want to identify the vector paths of the few sparse early arriving reflections in what is the ISD region from ~1.5-7ms prior to the first significant energy return. And seeing as how the early arriving soundfield in the ISD is very sparse, you would do well to also identify the points of incidence corresponding to the few reflections whose gain is just below the 15 dB threshold and to mitigate them as well. If they were part of a more spatially and temporally well mixed set of reflections I would not worry, but their sparseness is a potential problem.
In this particular instance a few such sparse reflections persist and should be addressed.
Also, the later arriving soundfield from ~7-30+ms. is characterized by many sparse reflections whose gain is up to twice that of their neighboring reflections. You will want to smooth the slope of the decaying soundfield by lowering the gain of the sparse reflections, while simultaneously increasing the spatial and temporal density of the soundfield. In other words, you want to make the response look more like a thick dense mat of reflections of similar gain levels without individual spikes exceeding the average gain level of the surrounding spikes – and the soundfield should decay in an orderly manner.
Were this a casual listening room and if one preferred a larger but less well defined localization and imaging, a higher degree of diffuse early reflections would be acceptable up to a level of 10 dB below the direct signal. But it should be stressed that such a soundfield would necessarily be temporally and spatially diffuse and without any exceptional higher gain reflection(s). Any higher gain reflections would need to be treated through the surgical application of absorption.
In either case, aside from identifying and addressing the very early high gain sparse early reflections with absorption, the remaining later arriving soundfield would benefit from the use of some diffusion in order to remediate the isolated sparse relative high gain reflections and would reduce their gain and create a more uniformly spatially and temporally dense early and later arriving soundfields.
And in this case there are quite a few sparse relatively high gain reflections that would benefit from diffusion which would both reduce their gain while increasing the temporal and spatial density of the soundfield.
But it’s hard to tell the specific effects of any existing treatment without good /consistent before and after measurements that would allow us to note the specific effect existing treatment has had on actual reflections, as well as how possible adjustments of existing treatments may be made, or how the addition of additional new treatments can most optimally be utilized to create the optimal environment through the preservation and ‘re-purposing’ of existing energy.
When we talk we can delve into other factors about which I cannot make an informed comment without additional information.
I guess this is my opportunity to get into IR and RT60, spec analysis, etc.
