CR Treatment and Tuning---The home stretch!!

OK...I think we're onto something with the tilt here...Check it out. I like the feel and look too. Gonna keep it either way. file sent... Also I edited the post above with a pic of the sub...

There is an improvement, yes, but the reflection is still there. Slightly lower intensity, and slightly earlier in time (very slightly!). Here's what it looks like: Gold-colored curve is without the desk tilt, purple curve is with the desk tilt. You can see that the level came down a little and the spike moved earlier by a few microseconds (about 70 us). It came down from about 46% to 39% in energy, which is nice, but not enough. It's a useful change, yes, but the issue didn't go away completely... It would have to get below the 10% line for me to consider it fully resolved. The fact that it moved earlier in time means that it took a slightly shorter path, which suggests that the reflection point moved higher up in distance above the floor.... so it is probably now located more towards the back of the desk (I assume you lifted the rear of the desk?) It also smoothed out from being two spikes to now being just one, so I'll go out on a limb and say that previously the reflection point was straddling the desk surface and the rebated "console tray" section, but now it is all on the top surface, none on the tray. That puts it quite far back on the desk, in that triangular shaped area next to the tilted rack section, and in a direct line from the speaker to the mic. Try putting a patch of insulation back there on the "triangle", and see what happens. Dinner-plate sized piece of insulation... But if you like your desk like that, then you can keep your desk like that! :) However, we do still need to find the source of that reflection, and knock it down a bit more. Under ten % is the goal! So it's time to get out your string and masking tape! - Stuart -

oh alright then..... :roll: Quick question though...Hows about I put some insulation in the picture frame to keep it from resonating?

Hows about I put some insulation in the picture frame to keep it from resonating?

I was actually going to suggest that if I saw it misbehaving in the REW data.... But sure, no problem! Go ahead and do it preemptively: it will save you extra work taking it down and putting it back up, if it does turn out to be necessary... - Stuart -

I'm having a hard time understanding your string directions. I don't understand the how marking a spot with a sharpie comes into play. I also don't understand the concept of moving the slack around the room. I'm sure this is difficult to convey in messages so I'm sorry that I'm not picking up what you're throwing down. LOL I do understand the idea of taping the string to the acoustic Center and taping the string to the mic and the slack hanging down. I'm not sure how I'm supposed to move that slack around the room effectively if it's attached to two fixed points. How does the Sharpie come into play if I'm attaching the string to two fixed points? Maybe I'm not?

The sharpie just marks the exact length on the string. The distance that you are trying to trace. If you don't like sharpies, then you can also use a pen, pencil, nail polish, paint, or a quill dipped in ink... :lol: 8) Anything that leaves a clear mark on the string at the two points you need. So, first lay out a long piece of string on the floor, then a couple of inches from one end, mark the string with the sharpie (or quill). That "couple of inches" is to give you enough extra string so you can tape it securely to the speaker. The the sharpie/quill mark is the spot that just touches the front panel of the speaker. But with the string still on the floor (before you pick it up to tape it to the speaker), measure from that first sharpie mark the distance that I specified above: that is, the distance from the tip of the mic to the acoustic center of the speaker (the "axis" in the model: just below the tweeter), PLUS the extra "bounce distance", which is 33 cm EXTRA in this case. Make another mark on the string at that exact spot (mic-speaker distance + 33 cm) , and that second sharpie mark is the spot that has to end up at the very tip of your mic. A couple of inches beyond that second sharpie mark, you can cut the string. You now have a piece of string that is a few inches longer than the total bounce distance, with tow marks on it, and a but extra beyond each mark, for taping. So, your string has two sharpie marks on it: one of those you position exactly at the tip of the mic and tape in place, and the other you position exactly at the acoustic center of the speaker and tape in place. The sharpie marks are just there so you can be accurate with the taping up of the string. 1ms is a very short delay, so you do need to be very accurate here, thus the sharpie marks: to give you accurate points for taping up the string. OK; so now you have your string taped up, with one mark where the string meets the speaker, and the other mark where the string mics the mic, and the middle of the string is hanging down, but in a natural curved shape: Sound does not travel along curved lines: it travels in STRAIGHT lines, (at least for high frequencies in air....). So you need to use your finger to push down in the middle of that loop of slack, so that the string is stretched taught on either side of your finger into two straight segments. On one side of your finger the string now runs in a straight line to the speaker, and on the other side in a straight line to the mic. As you run your finger along the string like that, at some point it will touch the desk surface: the point where it touches is the center of the specular reflection that is causing the problem. You might find that you can move the string left and right a bit, with your finger still pressing to keep it taught, and it still touches the desk surface. For example, you might find that there is one spot you can reach on the desk surface, and another on the "console tray" surface, and another on the vertical side of the angled rack section. All of the spots that you can reach with your finger like that, keeping the string taught in both directions, are likely part of the reflection. Put small pieces of masking tape at every spot you can reach with your finger pressing on the string. You might need help here: as you press down, the mic stand will want to topple over, or the tape might come loose at the speaker. So if you can find some helpful volunteers to hold those to spots steady, that would help! In this first case, there's only enough extra loop that it can touch the desk surface: the loop isn't long enough to reach other surfaces. But for the OTHER ones that I mentioned, the bounce is quite a bit longer (several ms), so the string will be quite a bit longer too (roughly one foot longer for every ms), and could potentially reach many surfaces. For example, if the loop hangs down to the floor, you might find that you can also stretch it out sideways and touch the wall in a few spots, as well as stretching it upwards to touch the cloud in a few spots. If that happens, then mark ALL of those points with a bit of masking tape. If you take a look at Steve's thread, in the most recent photos you can see some colored markers on the front walls, floor, and soffit wings where he did this exact same process, and found several potential reflection points, which I am about to deal with. The different colored markers are for different string lengths = different flight times. That's the end result of this process: a bunch of markers on the desk and around the room that show where your strongest reflections are coming from. Finally, not all of the points that you can reach with your finger really are reflection points: To be CERTAIN that it really is a reflection point, the "angle of incidence will be equal to the angle of reflection". So look closely at how the stretched-taught string touches the surface: the angle that it makes on one side of your finger (heading towards the speaker) between the string and the surface should be about the same as the angle of the string on the other side (heading towards the mic). If the angles are clearly very different, then that's probably not a refection point, but if they are similar, it probably is. In other words if the sound wave (string) is hitting the surface at an angle of 60°, but the angle from there to the mic is only 20°, then that's not a reflection point beause sound always bounces off a flat surface at the "same but opposite" angle: if it came in at 60°, then it will leave at 60° the other way. But if the second angle in this example (string to mic), is also about 60°, then it is a reflection point. The angles might be anything: both 20°, both 8°, both 41.297°, both 81.6°, or whatever), but they will be equal for all spots that really are causing that reflection to hit the mic (your head) Not sure if that cleared it up a bit! Or maybe just confused you even more... - Stuart -

I'm going to digest that in a sec.....In the meantime I did found the 1ms issue.. I'll post a pic in just a minute. I'm fine tuning the spot with insulation now. I used a little rug and a towel to get the rough idea after moving them around the desk. back shortly...

:thu: Read it over again: I just edited a bit, to make it clearer... maybe! - Stuart -

OK...I sent you another file..AS you can see in the pic, The right side needs a bit more coverage. Also, the monitor controller is affecting that side. I could mount a keyboard shelf underneath the center to put it there along with the keyboard of course...Man! This is quite finicky!!! :shock: I did note that the LRS did not show a big change but the L/R measurement did? The cummulative affect seems to still be substantial despite a big improvement in the L/R measurements.

Some art for the CR! :mrgreen:

Very cool Frank. I really like the color scheme you chose for the room, and I LOVE the LED lights. And the painting fits in very neatly too! But please send me some REW tests! I'm itching to see how that latest set of digital tuning settings is turning out... - Stuart -

Only pay attention to the last email I sent....much was discovered as I chased my tail all night!!! :horse: :shot:

Well, Things are chugging along. Stuart is working away on the digital tuning and I'm getting all of the details in order. Cabling, finish, trim,etc....Gonna let Stuart post graphs on the results of course. In the meantime here's a pic of things as they are..Not a bad place to be. The 2 glass Orbs on the desk are dealing with bounce issues inline with the speakers, but they sure look cool. Love it when you can just go to the store and grab something quick that still looks dope! :mrgreen: And of course the sound is world class!

Looking very swanky Frank! (see what I did there?) When it's finished you'll have to take some pics from all angles.

But it only looks good from that angle!!!! :ahh:

Hey Frank, Looks great man ! :?: QUESTION: Where did you get the molding material to make the rounded (semicircle) edges to your desk ?

Actually, I just ripped a standard coat hangar pole in half..In my case it was a 1 5/8" dia...hanging around the lumber yard, no pun intended! :D

Keyboard tray!!

Got a hold up on the CR tuning...gonna try to switch to LR treatment for a bit. Just gotta get some plans..... :shot:

Got a hold up on the CR tuning...

Yup. We ran into a very strange issue with the sub, where the input gain control on the sub itself appears to affect only part of the spectrum: it does not reduce all frequencies the same way. Frank has contacted the manufacturer to see if there is a work-around, and we are working on a "plan B" using another device to control the sub level externally, but it won't arrive for several days.... So in the meantime, we are "on hold" with the CR tuning... But do stay tuned if you are following this thread! We should be back on track with that by next week. In the meantime, here's some tantalizing hints of where we are in the final precision tuning process right now, as reported by REW. First, frequency response, from 11 Hz to 500 Hz, smoothed to 1/6 octave: Yes, that graph really does start at 11 Hz, and yes, the response really is flat down to 16 Hz. ... And yes, the frequency response of the entire low end really is within +/- 1.5 dB at 1/6 octave smoothing! (and if you really must know, it is still within +/-3 dB UNSMOOTHED , but we aren't quite ready to show that graph yet...., until the process is complete... :) ) I'm fairly pleased with the results so far, in terms of frequency response. I think Frank is, too. Now for the waterfall, smoothed to 1/3 octave: This time it goes all the way from 10 Hz up to 1 kHz. In other words, half of the entire audio spectrum. I think you could say that this is pretty smooth.... Frank reports that the low end is very tight, very clean, and very smooth... but we ain't done yet! There's still a few minor details that we are working on, before we show the final results at higher resolution. Now, do bear in mind that some studio designers only ever show their results at 1/3 octave, claiming that "this is what the specifications require", and they flat refuse to show their results at higher resolution... I wonder why? Compared to what they show, I think this is pretty decent... And some other designers refuse to show their results AT ALL! Claiming that the client won't let them, or whatever. But in my experience, when you get results this good, the client is damn EAGER to show the results in public, in order to attract business for his amazing studio! But whatever... I'm not the kind of guy to hide the results of the rooms I design, so you will be seeing higher resolution of these graphs, once the process is completed. And I'm pretty sure that Frank is proud to show off such amazing results from his very hard work, and very careful construction, with great attention to detail. It's only thanks to that, that we can get such flat response: it isn't possible to get this type of graph from haphazard building. See my signature block at the end of every post, to understand why... And do bear in mind that this is a very small room: Total floor area is a tad over 144 square feet (14.4 m2), and total room volume is 1380 cubic feet (39.1 m3). That's just 67% of the smallest recommended size for a critical listening room. It goes to show what can be accomplished with careful design and careful construction. I'm not saying that to "blow my own horn", as the saying goes, but rather to encourage others with small spaces that it is possible to achieve decent results. That said, I would NOT encourage you to shrink a room that COULD be larger: it's always best to make the room as large as possible. It's not simple at all to get these results, so unless you have a good understanding of acoustics, don't try to make your room this small, if it can be larger. keep it above 180 ft2, at least, and over 220 wherever possible. I should also point out that these graphs do not show the final "house curve" that I will be applying right at the end. Right now our aim is to get the response as close to flat as we can, then we'll apply the house curve that I have developed, which is based on the famous B&K curve from several decades back, with some modifications suggested by Floyd Toole in some of his papers and books, based on decades of research into the way people perceive sound, speakers, and room acoustics. So for now you are seeing only flat, but that will be livened up a bit with the final tweaks. There will be more graphs, and in greater resolution, once we are done, along with more detailed descriptions, and Frank will add his own comments on the final outcome, as an experienced mix engineer, but for now at least I did want to post this update. Just in time for Christmas! - Stuart -

We ran into a very strange issue with the sub, where the input gain control on the sub itself appears to affect only part of the spectrum: it does not reduce all frequencies the same way.

Is this due to a faulty unit or how the sub was designed? The specs and reviews of this sub are amazing, so I'd hope this is a faulty unit or that there is indeed a fix. I hope one day to own one! Greg

Hey Greg, As of yet, I don't know....My sweetwater guy is going to put in a message to Neumann, but I don't expect a reply until after the holidays...I haven't run this by Stuart yet, but the idea got floated that it might be the prm-1 mic... What do you think Stuart? I can say that this sub is AMAZING!!! I can live with this issue by adding a filter to cut just above 20hz if need be...You won't be disappointed.

.I haven't run this by Stuart yet, but the idea got floated that it might be the prm-1 mic...

I don't see how that could be the case at all. If the mic was faulty and did not have flat response, then we would have seen that already, in the other tests, both with and without the Tannoy subs, and the Dynaudio sub. Also, we can reduce that level very easily with the "magic wand"... ( :) ), along with all the other levels, and the mic shows those changes... it's really hard to see how the PRM1 would be able to do that selectively! Besides, you pulled the input gain level down by about 10 dB, I think, and the 20 Hz area didn't budge at all: if it was a mic problem, then the 20 Hz region would have dropped as well, just staying above the rest. Mics are not selective like that: the response curve is more or less fixed, and doesn't vary much by SPL, except at very high levels. 90 dBC is not a very high level for that mic: it can take 132 dB, according to the specs! I find it very hard to believe that it is the mic! But if you want to check that theory, then just plug in any old mic in its place, and repeat the same tests we already did. The curves will be different, of course, but the overall behavior will reveal if the PRM1 has a problem or not. Even a good old SM-58 would do just fine as a sanity check for this. The other issue I was thinking is that it could be some type of strange resonance in "something", but that seems very unlikely too: it isn't a mode (your room doesn't have any modes that low), but it might be something vibrating or resonating on / around / in / under the desk... however, it doesn't behave like that either: the level decays normally, and the decay isn't that long anyway. I think you would have noticed if something in the desk was vibrating wildly at 20 Hz. The other possibility is that it is ringing that I have induced somehow... but it doesn't behave like that either! We could check that theory by using the Tannoys one last time: Just sit them down close to where the KH805 is, and plug them in instead of the 805, then run a test like that. They don't really go down that low, but they do still put out enough energy down there that the effect should be evident. Turn them up until things are almost clipping, and see what you get. Then turn them down, and see if 20 Hz goes down like the rest of the low end. - Stuart -

... continued... Frank, I just sent you a test you can try, to see if it might be related to the mic. Let me know how that goes. - Stuart -

Wow Guys !! Amazing results. The newest graphs are insane :shock: Really impressed! and man that desk is gorgeous :thu: //Jakob