I've settled on the gear I want in my room. All analog, all vintage, all rebuilt (by me). Taken a few years to get it all done, but I'm done. Or at least I'm satisfied with what I have. It all sounds great, but there's not much left to do with it besides listen to music. So it's on to room correction to see if better sound can be found.
I've got it all in a small and odd shaped room:
It was originally my home office, but became the full time "audio room" once I retired. Odd shaped, asymmetrical and small. Far from an ideal audio room, but we all work with what we've got...right?
I figured if I was going to get as much out of it as I could, I needed to get a little more "scientific" about it than random chance placements and my old, damaged hearing.
So I bought an Eversolo EM-01 and downloaded REW. Almost half the cost of the Umix-1 and fine for my limited uses. After a few days of reading and playing around with it, I tried my first sweeps.
I must admit that starting off from nothing, it was a little overwhelming. Not so much the process of getting set up and shooting some sweeps, but interpreting the data into some kind of understanding of what I was looking at and just what was going on.
I had a general idea of what I was looking at, but being new at it, I just wasn't sure. So I tried leveraging ChatGPT to confirm what I thought I was looking at was actually what I thought it was. I didn't use it so much to do the work, but as a confirmation I was going in the right direction with interpreting the data. It seemed to work out fairly well and seemed to increase my learning rate. I basically used ChatGPT as a tool to learn a little quicker by confirming what I thought, not to do the actual work. Although it did occasionally cause me to think about options I wouldn't have without the "cue".
I started off with an absolutely horrible chart on the first few sweeps. Nulls and dips everywhere. A complete mess. The funny thing is: I didn't think the system sounded that bad at all before! Seems to support that old saying: "the best you know is the best you've heard".
I ended up handling most of the low freq nulls and dips by experimenting with listening position and sub positioning. I ended up with a near-field sub, right under my lh armrest. It was literally sensitive down to a inch in difference fore and aft in "helping" cancel out some nulls. Lateral moves were less sensitive, but it was also fairly picky about that too, down to a 2-3" inch move in either direction being worse. Small rooms are just fussy: go figure.
The seating/listening position ended up less than ideal for the space by "conventional wisdom", but it was a necessary compromise between nulls and being able to actually use the chair and still be able to move about the room without wanting to throw the chair through the window. I still have a null at 50-51Hz, but it's narrow, a smooth transition into it and strong recovery at about 52-55 Hz. I'm calling that "manageable". At least for now with what I have at hand.
SBIR is a fairly deep/wide null around 130Hz, but positioning helped make the entry smooth and shallow and the recovery the same. I can pretty much make it go away with moving the towers to 3" from the wall behind them, but it introduces even worse compromises elsewhere and imaging just goes completely to pot. So I'm accepting the 130Hz null where it is for now. I'm not even sure i can "hear" it but REW says it's there so I'll get around to dealing with it in other ways later down the road.
While not ideal for analysis, I had ChatGPT combine my adjustments and overlay the general graphs into an image (this is for speaker movement and toe in, not the sub moves):
ChatGPT obviously has some problems rendering text! LOL!
The good news is the sound stage is absolutely fantastic on the listening test. It's a little on the narrow side because the "triangle" is a little smaller than conventional wisdom says it should be on the base measurement. But that's fine as I listen to a lot of detailed stuff anyways and the width is just fine for my preferences, or at least I'm perfectly happy with what I'm hearing. I followed the numbers, not the generally accepted "rules of thumb". The left/right balance is really good, even with the room asymmetry. Listening impression is also a perception of clarity, detail and delicate separation of the instruments. The "phantom speaker" is dead on the money as is Neil Peart's drum rolls transiting across the sound stage, the left/right crashes and his snare hits and kicks dead center stage. Just wonderful.
I still have a little problem with the rh tower pointing into a corner, but careful positioning minimized the bad stuff happening in there and I'm hoping a trap will help tame it a bit more on further REW measurements. As it stands, I honestly can't hear any issues coming from the rh corner right now, but it's probably that "the best you know" thing again. Maybe I'll hear a change when the traps go in, maybe not. I'm under no misconceptions that the SBIR it will go away with trapping because it's an "unchangeable" geometry issue. I'll be happy if it just tames it a bit on the measurements.
I think I'm pretty much at the limit of what I can do with positioning as I'm pretty sure I've just hit the limits of the physics of the room.
Next steps are probably adding a second sub to see if I can't excite that 51Hz null a bit more with an additional/differing position and some absorption schemes to try and smooth out the 130Hz issue.
First will be a couple "traps" behind the speakers (2" space, 6" Rockwool safe n sound and another 4" of space). I'm limited to 47" of height due to switches and other wall fixings. That's still a good 4-5" taller than the towers. I'll do a couple sweeps to see what it does and then on to ceiling clouds.
I had ChatGPT do a render (with a photo of my room) on the ceiling cloud concept I'd like to use:
I'm looking at 6" of absorption (Rockwool safe n' sound) and a 4" air gap. I still have to think about if I want that much drop from the ceiling though. Then more sweeps to see what that does. I'll play with the air gap a bit and see what returns optimal measurements. The shape was picked for aesthetics as well as absorption. After all, I do have to enjoy seeing/being in the space as well as listing to it.
At that point, I'll pretty much be out of options for room corrections as the side walls are not candidates for absorption, nor do I really want to change much of anything there, or on the rear wall with is a display wall. There's always the soffit area, but I'm not really willing to go through all that and the aesthetic impact it will have on the room. There 's actually limited space there, so I'm not sure it would really do much more than just make the room look more "closed in".
I'm ok with just achieving "better". Chasing "perfect" isn't on my dance card for this room because I know this room just won't ever get there because of it's physics.
That's where my room, is now: "better". It's not a night and day change (after REW work) from what it was before, but there is an audible improvement. Or perhaps I just couldn't recognize it as "not good" before. Heck, I can' even discount "the placebo effect' that I want it to sound better. But the measurements sure seem to be going in the right direction.
Well worth the 130 bucks it cost for the EM-01. Probably the biggest "bang for the buck" improvement so far. REW being free to download is a big part of that "bang" though. If you had to pay $300+ for it like most others, the ratio of bang to buck would go down pretty fast for a "casual" user.
REW sure is a steep learning curve and I've barely scratched the surface. Totally worth it though!
Besides, at 130 bucks for a measurement mic, can't really get hurt a lot if it doesn't work out.
I've got it all in a small and odd shaped room:
It was originally my home office, but became the full time "audio room" once I retired. Odd shaped, asymmetrical and small. Far from an ideal audio room, but we all work with what we've got...right?
I figured if I was going to get as much out of it as I could, I needed to get a little more "scientific" about it than random chance placements and my old, damaged hearing.
So I bought an Eversolo EM-01 and downloaded REW. Almost half the cost of the Umix-1 and fine for my limited uses. After a few days of reading and playing around with it, I tried my first sweeps.
I must admit that starting off from nothing, it was a little overwhelming. Not so much the process of getting set up and shooting some sweeps, but interpreting the data into some kind of understanding of what I was looking at and just what was going on.
I had a general idea of what I was looking at, but being new at it, I just wasn't sure. So I tried leveraging ChatGPT to confirm what I thought I was looking at was actually what I thought it was. I didn't use it so much to do the work, but as a confirmation I was going in the right direction with interpreting the data. It seemed to work out fairly well and seemed to increase my learning rate. I basically used ChatGPT as a tool to learn a little quicker by confirming what I thought, not to do the actual work. Although it did occasionally cause me to think about options I wouldn't have without the "cue".
I started off with an absolutely horrible chart on the first few sweeps. Nulls and dips everywhere. A complete mess. The funny thing is: I didn't think the system sounded that bad at all before! Seems to support that old saying: "the best you know is the best you've heard".
I ended up handling most of the low freq nulls and dips by experimenting with listening position and sub positioning. I ended up with a near-field sub, right under my lh armrest. It was literally sensitive down to a inch in difference fore and aft in "helping" cancel out some nulls. Lateral moves were less sensitive, but it was also fairly picky about that too, down to a 2-3" inch move in either direction being worse. Small rooms are just fussy: go figure.
The seating/listening position ended up less than ideal for the space by "conventional wisdom", but it was a necessary compromise between nulls and being able to actually use the chair and still be able to move about the room without wanting to throw the chair through the window. I still have a null at 50-51Hz, but it's narrow, a smooth transition into it and strong recovery at about 52-55 Hz. I'm calling that "manageable". At least for now with what I have at hand.
SBIR is a fairly deep/wide null around 130Hz, but positioning helped make the entry smooth and shallow and the recovery the same. I can pretty much make it go away with moving the towers to 3" from the wall behind them, but it introduces even worse compromises elsewhere and imaging just goes completely to pot. So I'm accepting the 130Hz null where it is for now. I'm not even sure i can "hear" it but REW says it's there so I'll get around to dealing with it in other ways later down the road.
While not ideal for analysis, I had ChatGPT combine my adjustments and overlay the general graphs into an image (this is for speaker movement and toe in, not the sub moves):
ChatGPT obviously has some problems rendering text! LOL!
The good news is the sound stage is absolutely fantastic on the listening test. It's a little on the narrow side because the "triangle" is a little smaller than conventional wisdom says it should be on the base measurement. But that's fine as I listen to a lot of detailed stuff anyways and the width is just fine for my preferences, or at least I'm perfectly happy with what I'm hearing. I followed the numbers, not the generally accepted "rules of thumb". The left/right balance is really good, even with the room asymmetry. Listening impression is also a perception of clarity, detail and delicate separation of the instruments. The "phantom speaker" is dead on the money as is Neil Peart's drum rolls transiting across the sound stage, the left/right crashes and his snare hits and kicks dead center stage. Just wonderful.
I still have a little problem with the rh tower pointing into a corner, but careful positioning minimized the bad stuff happening in there and I'm hoping a trap will help tame it a bit more on further REW measurements. As it stands, I honestly can't hear any issues coming from the rh corner right now, but it's probably that "the best you know" thing again. Maybe I'll hear a change when the traps go in, maybe not. I'm under no misconceptions that the SBIR it will go away with trapping because it's an "unchangeable" geometry issue. I'll be happy if it just tames it a bit on the measurements.
I think I'm pretty much at the limit of what I can do with positioning as I'm pretty sure I've just hit the limits of the physics of the room.
Next steps are probably adding a second sub to see if I can't excite that 51Hz null a bit more with an additional/differing position and some absorption schemes to try and smooth out the 130Hz issue.
First will be a couple "traps" behind the speakers (2" space, 6" Rockwool safe n sound and another 4" of space). I'm limited to 47" of height due to switches and other wall fixings. That's still a good 4-5" taller than the towers. I'll do a couple sweeps to see what it does and then on to ceiling clouds.
I had ChatGPT do a render (with a photo of my room) on the ceiling cloud concept I'd like to use:
I'm looking at 6" of absorption (Rockwool safe n' sound) and a 4" air gap. I still have to think about if I want that much drop from the ceiling though. Then more sweeps to see what that does. I'll play with the air gap a bit and see what returns optimal measurements. The shape was picked for aesthetics as well as absorption. After all, I do have to enjoy seeing/being in the space as well as listing to it.
At that point, I'll pretty much be out of options for room corrections as the side walls are not candidates for absorption, nor do I really want to change much of anything there, or on the rear wall with is a display wall. There's always the soffit area, but I'm not really willing to go through all that and the aesthetic impact it will have on the room. There 's actually limited space there, so I'm not sure it would really do much more than just make the room look more "closed in".
I'm ok with just achieving "better". Chasing "perfect" isn't on my dance card for this room because I know this room just won't ever get there because of it's physics.
That's where my room, is now: "better". It's not a night and day change (after REW work) from what it was before, but there is an audible improvement. Or perhaps I just couldn't recognize it as "not good" before. Heck, I can' even discount "the placebo effect' that I want it to sound better. But the measurements sure seem to be going in the right direction.
Well worth the 130 bucks it cost for the EM-01. Probably the biggest "bang for the buck" improvement so far. REW being free to download is a big part of that "bang" though. If you had to pay $300+ for it like most others, the ratio of bang to buck would go down pretty fast for a "casual" user.
REW sure is a steep learning curve and I've barely scratched the surface. Totally worth it though!
Besides, at 130 bucks for a measurement mic, can't really get hurt a lot if it doesn't work out.
Last edited:
