I've made some new single position measurements to use as a base to try out some different correction windows.
I've reverted to using Fab Filter as a preEQ based on REW EQ settings rather than using REW to combine the impulses. REW always puts the impulse centre at 1s even if the filter has no inherent delay making the latency long even for a minimum phase filter.
I've used REW in File playback mode where the sweep is generated and saved as wav file to be played back through Jriver as I wasn't happy with the Jriver wdm driver method. Too many clicks pops and weirdness. Much happier with the file playback even if it is a bit more complicated per measurement.
This gives me a measured pre EQ'd base.
I also noticed the difference in direct on axis to slightly off axis sound. I have used the arrays setup with some toe in before but not with the drivers facing the listening position. I moved them to facing and measured. This is the difference. 15 cycle FDW 1/6 oct smoothing plus Pre EQ same for both.
In the current setup there is only one listening position so I am interested to see what effect this has. I wondered before if I was making the room sound brighter by EQing an off axis position.
Not much difference really, the good thing is that the speaker seems to be quite consistent across that range of angle, which I always thought it would be but this gives some confirmation of that fact.
I've reverted to using Fab Filter as a preEQ based on REW EQ settings rather than using REW to combine the impulses. REW always puts the impulse centre at 1s even if the filter has no inherent delay making the latency long even for a minimum phase filter.
I've used REW in File playback mode where the sweep is generated and saved as wav file to be played back through Jriver as I wasn't happy with the Jriver wdm driver method. Too many clicks pops and weirdness. Much happier with the file playback even if it is a bit more complicated per measurement.
This gives me a measured pre EQ'd base.
I also noticed the difference in direct on axis to slightly off axis sound. I have used the arrays setup with some toe in before but not with the drivers facing the listening position. I moved them to facing and measured. This is the difference. 15 cycle FDW 1/6 oct smoothing plus Pre EQ same for both.
In the current setup there is only one listening position so I am interested to see what effect this has. I wondered before if I was making the room sound brighter by EQing an off axis position.
Not much difference really, the good thing is that the speaker seems to be quite consistent across that range of angle, which I always thought it would be but this gives some confirmation of that fact.
Attachments
I recently toed my line arrays in to point at the LP. I found I needed about 5 db less HF (high shelf) boost to flatten them and my immediate subjective impression was improved treble - no longer experiencing slightly harsh treble on troublesome tracks. As near as I can tell moving around, this hasn't hurt the response away from the sweet spot.
I've also got improved bass damping and the effect is simply stunning on symphonic music. Ironically, the bass damping comes, not from purposeful room treatments, but apparently from a large number of plastic storage bins that have found their way into the garage.
I've also got improved bass damping and the effect is simply stunning on symphonic music. Ironically, the bass damping comes, not from purposeful room treatments, but apparently from a large number of plastic storage bins that have found their way into the garage.
The combination is correct but on export the IR centre is placed at 1 second. A linear phase filter with 65536 taps usually is somewhere around 0.75 seconds. My minimum phase DRC filters have almost zero delay.
More progress has been made. The last of the damping panels has been completed. This one has the printed fabric picture of wave rock on it. It turned out to be quite a pain to stretch the fabric correctly and I had to re staple it twice, once due to over stuffing the cavity with fibreglass.
There has been a complete overhaul of software and processing which has taken some time. I'm now running Jriver version 26 64bit. Mide Side EQ is using DMG's Equilibrium as it is the only thing I have found that is comparable to PLParEQ that wesayso has been running (that's 32bit only so I needed a native 64bit plugin). I have tried bridging plugins but it has been hit and miss. Pre EQ through Fab Filter ProQ3 and overall house curve in another instance of ProQ3.
With the speakers pointing more or less at the listening position a new EQ curve was needed. DRC template has gone back to longer windows and the older template I was using but with less gain allowed. This has given more headroom back as the amp doesn't have to turned up as loud feeding the nulls
Gmad's templates provided some useful pointers
Looking at the ETC I seem to have got a bit of a natural kicker going on around 28ms. Not quite the same between left and right. The room is not dead sounding but the treatments have made some difference for the better I think.
There has been a complete overhaul of software and processing which has taken some time. I'm now running Jriver version 26 64bit. Mide Side EQ is using DMG's Equilibrium as it is the only thing I have found that is comparable to PLParEQ that wesayso has been running (that's 32bit only so I needed a native 64bit plugin). I have tried bridging plugins but it has been hit and miss. Pre EQ through Fab Filter ProQ3 and overall house curve in another instance of ProQ3.
With the speakers pointing more or less at the listening position a new EQ curve was needed. DRC template has gone back to longer windows and the older template I was using but with less gain allowed. This has given more headroom back as the amp doesn't have to turned up as loud feeding the nulls
Gmad's templates provided some useful pointers Looking at the ETC I seem to have got a bit of a natural kicker going on around 28ms. Not quite the same between left and right. The room is not dead sounding but the treatments have made some difference for the better I think.
Attachments
Nice progress! So JRiver 26 is worth it? Maybe I should try it too, I'm still at version 22 32 bit. The damping panel came out real nice! I do know what trouble it can be to get it right!
I hope you can enjoy this setup for years to come! That ETC looks nice. Any idea what spectral balance belongs to those HAAS like peaks? Filtering the IR with a late start should give you some idea. Maybe it's visible in the wavelet too.
I hope you can enjoy this setup for years to come! That ETC looks nice. Any idea what spectral balance belongs to those HAAS like peaks? Filtering the IR with a late start should give you some idea. Maybe it's visible in the wavelet too.
I had to make a fabric cover strip to go around the edge of the frame to hide all the staples I had to use on the visible edge to get it straight. It's not really obvious in the picture because it was taken head on. I actually prefer it in the end as it's the dark grey fabric of the other panels instead of being bright white
I was using a borrowed version of 22 32bit, as it proved to be worthy I bought a licence for 25, but it was for version 26 too being later in the year.
I quite like the new waveform view that shows the waveform of the track as it goes along instead of a straight bar. Easy to see which tracks are crushed to death, not that you need a waveform view to notice it. You have to reanalyse all your tracks though so it's a good overnight job.
I can't say I have noticed much else so far but I haven't really gone poking under the hood. A lot of the things in the list of changes seemed to be video related so that might be of interest to you. Once they released it on the main download page and there was no rush of crash reports I thought I would give it a try.
I've sent you an email which you might be able to make use of if it gets to you in time
Spectrogram seems to show a fairly defined ridge from 1k upwards in the right channel, it could be there lower down but it's buried in everything else. I'm not quite sure how I will proceed with the ambient channels now.
I was using a borrowed version of 22 32bit, as it proved to be worthy I bought a licence for 25, but it was for version 26 too being later in the year.
I quite like the new waveform view that shows the waveform of the track as it goes along instead of a straight bar. Easy to see which tracks are crushed to death, not that you need a waveform view to notice it. You have to reanalyse all your tracks though so it's a good overnight job.
I can't say I have noticed much else so far but I haven't really gone poking under the hood. A lot of the things in the list of changes seemed to be video related so that might be of interest to you. Once they released it on the main download page and there was no rush of crash reports I thought I would give it a try.
I've sent you an email which you might be able to make use of if it gets to you in time
Spectrogram seems to show a fairly defined ridge from 1k upwards in the right channel, it could be there lower down but it's buried in everything else. I'm not quite sure how I will proceed with the ambient channels now.
Attachments
Got your email in time. I'm quite interested if EQuilibrium is a valid replacement for PLParEQ, as it was holding me back to go to 64 bit.
I've tried 64 bit once, with Fabfilter Pro and JRiver 23 (a license bought that I never used). It just wasn't the same and I just continued using version 22 up til now. Going 64 bit makes sense though, I plan to upgrade towards going all balanced long lines too. That should beat the performance of the long RCA runs I have now.
The subwoofers have a native balanced input, I need a replacement ambient amplifier and get a Pré amp for the Goldmond clone and use it balanced from DAC to pré. The recent Universal Buffer from Tom could be a good solution for me. Very flexible and clean. I'm even thinking about a Modulus 286 for Ambient duties. (to replace that 30+ year old Pioneer A400 I bought to compliment my 757 Mark II)
Seeing your graphs here, It might be fun for you to add ambient channels that fill the 100- 2K making the Haas kicker more complete (keeping timing about the same or slightly earlier), I've never seen a high frequency ridge that sharp arriving that late in time before. Indeed, it is most prominent in that right channel. My ambient channels only go up to 3.5 KHz / 12 dB but I do plan to try it some day with a 'high shelf' cut to try and maybe keep some shimmer up top. Robbintip mentioned that tweak on my thread once and it has been stuck in my mind ever since. Does it ever end?
I've tried 64 bit once, with Fabfilter Pro and JRiver 23 (a license bought that I never used). It just wasn't the same and I just continued using version 22 up til now. Going 64 bit makes sense though, I plan to upgrade towards going all balanced long lines too. That should beat the performance of the long RCA runs I have now.
The subwoofers have a native balanced input, I need a replacement ambient amplifier and get a Pré amp for the Goldmond clone and use it balanced from DAC to pré. The recent Universal Buffer from Tom could be a good solution for me. Very flexible and clean. I'm even thinking about a Modulus 286 for Ambient duties. (to replace that 30+ year old Pioneer A400 I bought to compliment my 757 Mark II)
Seeing your graphs here, It might be fun for you to add ambient channels that fill the 100- 2K making the Haas kicker more complete (keeping timing about the same or slightly earlier), I've never seen a high frequency ridge that sharp arriving that late in time before. Indeed, it is most prominent in that right channel. My ambient channels only go up to 3.5 KHz / 12 dB but I do plan to try it some day with a 'high shelf' cut to try and maybe keep some shimmer up top. Robbintip mentioned that tweak on my thread once and it has been stuck in my mind ever since
. Does it ever end? I'll be interested in your feedback on Equilibrium and the presets I created. I haven't done a full transfer function comparison but I did compare a single Peak and shelf filter to match them up to PLPar. Fabfilter uses a very different gain Q interaction which makes the settings incompatible. Too much effort to reinvent the wheel there.
To me I don't hear anything noticeably different to what I was running in Jriver 22 32bit, but I am not nearly as familiar with my system as you are with yours.
My initial thoughts on the ambient were similar that I would probably need to bring it slightly forward to mask the unintentional room kicker.
I think I'll leave things as they are for a while, changing too much too fast can make it hard to know if you actually made things better. Famous last words
How long are your cables to worry about balanced lines? If you don't have noticeable interference they wouldn't be top of my list for upgrades unless it was a matter of changing the cables only.
To me I don't hear anything noticeably different to what I was running in Jriver 22 32bit, but I am not nearly as familiar with my system as you are with yours.
My initial thoughts on the ambient were similar that I would probably need to bring it slightly forward to mask the unintentional room kicker.
I think I'll leave things as they are for a while, changing too much too fast can make it hard to know if you actually made things better. Famous last words
How long are your cables to worry about balanced lines? If you don't have noticeable interference they wouldn't be top of my list for upgrades unless it was a matter of changing the cables only.
For anyone following along or doing something similar, this is the House/Room curve that I have used to go from flat measured to flat sounding. The values won't work for another speaker and room combination quite the same but the way of setting up the EQ makes it flexible and easily adjusted when listening to dial it in to your preference.
The EQ bands are spaced roughly an octave apart 75Hz, 150, 312.5, 625, 1250, 1750 (This one is an extra tilt), 2500, 5000, 10,000
From 312 onward they are 0.7Q shelves at -0.8dB so you get a slope per octave that corresponds with the attenuation. i.e. 0.8dB per octave.
I added an extra 0.62dB downwards tilt at 1750 which made a surprising difference and fixed a few issues I was having. The two lowest bands have higher Q and more attenuation to bump the bass a bit. Changing these made quite a big difference too, this band was important to the overall balance.
I was inspired by looking at the sound power curves of some well liked Revel speakers, they exhibit a similar shape to that portion of the curve.
The bump in the top end is a peaking filter at 12k, I liked this better than boosting beyond this range with a shelf. This can be quite variable between tracks and is hard to get the right balance for everything.
I tried probably twenty presets I made with different combinations of things to narrow down what I liked and disliked, the curve above morphed out of those
The EQ bands are spaced roughly an octave apart 75Hz, 150, 312.5, 625, 1250, 1750 (This one is an extra tilt), 2500, 5000, 10,000
From 312 onward they are 0.7Q shelves at -0.8dB so you get a slope per octave that corresponds with the attenuation. i.e. 0.8dB per octave.
I added an extra 0.62dB downwards tilt at 1750 which made a surprising difference and fixed a few issues I was having. The two lowest bands have higher Q and more attenuation to bump the bass a bit. Changing these made quite a big difference too, this band was important to the overall balance.
I was inspired by looking at the sound power curves of some well liked Revel speakers, they exhibit a similar shape to that portion of the curve.
The bump in the top end is a peaking filter at 12k, I liked this better than boosting beyond this range with a shelf. This can be quite variable between tracks and is hard to get the right balance for everything.
I tried probably twenty presets I made with different combinations of things to narrow down what I liked and disliked, the curve above morphed out of those
Attachments
I'll be interested in your feedback on Equilibrium and the presets I created. I haven't done a full transfer function comparison but I did compare a single Peak and shelf filter to match them up to PLPar. Fabfilter uses a very different gain Q interaction which makes the settings incompatible. Too much effort to reinvent the wheel there.
To me I don't hear anything noticeably different to what I was running in Jriver 22 32bit, but I am not nearly as familiar with my system as you are with yours.
My initial thoughts on the ambient were similar that I would probably need to bring it slightly forward to mask the unintentional room kicker.
I think I'll leave things as they are for a while, changing too much too fast can make it hard to know if you actually made things better. Famous last words
How long are your cables to worry about balanced lines? If you don't have noticeable interference they wouldn't be top of my list for upgrades unless it was a matter of changing the cables only.
I don't expect an audible change by going 64 bit, it just makes sense to do so since a 64 bit OS has been around for more than a decade
.My RCA runs are 10m (two cables) and 15m (4 cables) long
. Not that many options to run them below my floorboards.There is more to it than going balanced though, but I first have to do some experiments to see if that holds true. This tweak could make an audible difference as opposed to going 64 bit. Reduced noise levels never is a bad thing, even if the room will dominate.
Thanks for outlining your apporach, I might try somthing similar with my miniDSP2x4HD to test. I have a decent baseline at the moment but is not ´perfect´yet
Fluid, your panel looks great !
I figure you've probably seen this chart or certainly similar ones...I've found this one particularly handy. https://www.bobgolds.com/AbsorptionCoefficients.htm
fwiw, rigid compressed-fiberglass panels attached to same size thin luan plywood for wall or ceiling mounting, made for very easy fabric covering.
Also round compressed-fiberglass tubes, capped on the ends with mdf or ply have worked well as corner bass traps. Covered those with fabric too...
Good acoustic stuff at the HVAC trade supply houses...
I figure you've probably seen this chart or certainly similar ones...I've found this one particularly handy. https://www.bobgolds.com/AbsorptionCoefficients.htm
fwiw, rigid compressed-fiberglass panels attached to same size thin luan plywood for wall or ceiling mounting, made for very easy fabric covering.
Also round compressed-fiberglass tubes, capped on the ends with mdf or ply have worked well as corner bass traps. Covered those with fabric too...
Good acoustic stuff at the HVAC trade supply houses...