PBK-1 mic about to record some subwoofer beeps and boops... |
Alas, the Christmas & New Years holiday season is behind us. Each year, I'm amazed at just how quick everything goes by... Never enough time to enjoy the music among the rest of life's other demands.
I didn't get much time off but I did manage to try out some room measurements in the evenings when the kids were in bed.
For all the time I spend thinking about audio components, it's unfortunate that I've spent too little on the room itself. With the new house and room, I hope to change this. Remember folks, as much as "we" (audiophiles) might obsess over the quality of our DACs or (some) might sweat over things like (supposed) cable differences, without question, the quality of the transducer/room interactions trump essentially any of the other factors in the audio system given the quality of electronics these days. Frequency variations in the audio room can fluctuate wildly, on the order of magnitudes of difference compared to the rather trivial differences between DACs for example. It might not be cool or sexy to speak about bass traps, diffusers and absorbers versus DSD or fancy cables made of Unobtanium, but this is what's ultimately going to make huge differences.
Of course, room measurements are not a new topic... Many excellent web pages have been devoted to this. For room treatments, check out Ethan Winer's RealTraps site. Also, Mitchco's excellent room calibration article not long ago documented his procedure nicely.
I'll have the years ahead to add hardware around the room, but as a first step, I'm just going to try the simplest of "room corrections"; EQ'ing frequency response using DSPs.
As of this writing, the current version of the PBK software (2.01) is not compatible with Windows 8 (even with compatibility settings)! I had to dual boot the old laptop used for measurements with Windows 7 and run the software through that... Paradigm, please update the program!
The software was easy to run and essentially self-explanatory. A USB cable connects the included PBK microphone (shown above) to the PC, then another USB connects the PC to the subwoofer. Each microphone has been calibrated and identified by serial number in the software. The program is capable of measuring multiple locations (max. 10) in the room to smooth out the bass response. Since I'm most concerned about the "sweet spot" and want to limit the potential of suboptimal calibration, I took 5 readings all around the central seat and the 2 adjacent seats. The program then will run thorough the calibration algorithm and show a screen that looks something like this...
You have the option to adjust the DSP crossover frequency point which in the graph above I've set to 160Hz (default is 250Hz) as well as how steep the filter should be. This gives you some customization options (not much).
The red curve was what I got in the room. As you can see, I have quite a dip at around 65Hz (red) which was correctable to some extent (purple). The peaks (eg. around 30Hz correlated to the calculated lateral mode for the room size at 29.1Hz using this online calculator) were easier to correct, and it's nice to see good frequency response down to 20Hz with this sub. (Check this link out for the room mode math calculations.)
The program will automatically upload the new settings to the subwoofer and away you go... Very simple calibration to do.
About 3 years ago, I purchased a calibrated Behringer ECM8000 mic (I see they don't sell these calibrated any more at this site). This microphone has served me well over the years and put to good use here again (note that I actually measured it a little lower at ear level sitting on the couch than the picture below):
Using the EMU 0404USB as measuring ADC, here's the raw room response with just the PBK settings in place for the subwoofer integrated with the main front speakers (1/6 octave smoothing) using the TEAC UD-501 DAC:
Although I don't know if I fully trust the Behringer mic below 30Hz and above 15kHz, it's good to see frequency response down to 15Hz. Again we see the room mode around 29Hz. The deep blue line represents the eventual target curve we're aiming for based on the default REW house curve (for those looking for the excitement of a bass-induced thrill ride, try this target curve). For those looking for more based on home theater wisdom, check out this link on House Curves and more!
Letting REW perform its own EQ from 20-200Hz plus a few small adjustments on my end resulted in this mathematical prediction of room response:
Much more controlled on the low end using 7 parametric EQ settings (you can see the numbers above) plus 2 settings at 3kHz and ~11kHz to roll off the top. Total of 9 EQ settings were programmed into the Behringer DEQ2496 applying the adjustments in the digital domain and looped back to the Transporter for DAC duties.
Since some frequency boosting is involved, I reduced the DEQ2496 digital output levels by 6dB and double checked with some really LOUD music to make sure the EQ settings did not lead to clipping. The 1997 Iggy Pop insane remaster of The Stooges' track "Your Pretty Face Is Going To Hell" off Raw Power is a good one - average dynamic range value of 1dB for the album! If the DSP processing doesn't clip with that track, it's probably not an issue with >99% of my music.
Since it's always good to confirm that the EQs are actually doing what they're supposed to, here's the actual measured room response with & without the Behringer DEQ2496 played through the Transporter as DAC (measured at a higher level on a separate day):
Before REW EQ:
After REW EQ:
OK. Nice real-world confirmation that the EQs are doing what they're supposed to. The predicted results checks out even with a different DAC (measured with TEAC, confirmed with Behringer & Transporter - I knew from previous measurements that the Transporter is very close in frequency response to the TEAC)!
Regarding room treatments, there's much to do! Here's the measured waterfall spectral decay plot in REW:
With EQs in place, the decay time of course remains high:
Plotted at 500ms duration down to 40dB with 1/6 octave smoothing, I'd really love to see more uniform steeper decay (<<300ms). Bass traps in the corner and absorption panels to the sides at the first reflection points could do the trick. Hmmm, maybe this will be a project for Spring Break - whip out the saw, stapler gun, make some wood frames, grab fabric and a stash of Roxul Safe'n'Sound :-).
There is also the issue with time alignment as addressed by mitchco using Acourate and convolution filters. That's another level of tuning I'll have to leave for another day! So much to do, so little time...
For now, the subjective sound quality has improved. There's already notably better control to the bass notes from Rebecca Pidgeon's "Spanish Harlem" (to use a well known audiophile favourite). Time to just sit back, relax, and enjoy some tunes! I've got a couple of ideas for tests coming up.
Musical selections recently:
- Daft Punk's TRON: Legacy soundtrack (2010) on Blu-Ray was stunning! I missed the movie in the theaters and rented the 3D Blu-Ray the other day... The movie itself was OK but the surround effects and techo score really made the movie an audio feast.
- The Eagles. Love 'em or hate 'em, I reacquainted myself with the multichannel DTS version of Hell Freezes Over (1997) the other night. IMO another fantastic multichannel release from the earlier days of surround sound when DTS was releasing their DTS-CD's (this was also my first concert DVD). That live ambiance really shines through as if you're sitting in the audience that night. The guitar work and percussion sound great in the new system on "Hotel California" - especially the bass impact. I saw them in concert about 3 years back and that too was a blast.
- Speaking of bass... I don't often buy modern pop recordings but I did enjoy listening to the recent album by Lorde - Pure Heroine. Fantastic job by the 16-year-old from New Zealand. The current Top-40 'hit' "Royals" gives a nice taste of the cavernous bass found throughout the album ("400 Lux" is another to check out). Have a listen to this album through a system with clean bass down to 20Hz and see whether you think you need a subwoofer :-).
Addendum:
In the "Answer To What Question?" and "What Were They Thinking Of?" files... I was looking at the recent CES2014 announcements in the usual audiophile watering holes and found this:
Esoteric Grandioso D1 Monoblock DAC
Given the level of performance of even modest DACs these days, I really can't imagine what would be the reason to go monoblock with a DAC. Seems like doing this could make things worse (channel desynchronization? need for DAC matching?) and at a significant expense ($22,000 each, not to mention all those extra cables!). Small price to pay to feel grandioso I suppose. As usual, would love to see the measurements for a pair...
I'll have the years ahead to add hardware around the room, but as a first step, I'm just going to try the simplest of "room corrections"; EQ'ing frequency response using DSPs.
I. Paradigm PBK-1 (Perfect Bass Kit)
Paradigm, Anthem, and Martin-Logan are owned by the same parent company and share similar technologies in terms of room correction. The Paradigm Signature and Martin-Logan subwoofers have built-in DSP units that can be programmed using a PC with the PBK kits; based off of the Anthem Room Correction algorithms. As the name implies, this kit is used for bass correction only as it applies to the subwoofer. Audio levels still should be checked with the rest of the system in order to make sure the sub integrates well.As of this writing, the current version of the PBK software (2.01) is not compatible with Windows 8 (even with compatibility settings)! I had to dual boot the old laptop used for measurements with Windows 7 and run the software through that... Paradigm, please update the program!
The software was easy to run and essentially self-explanatory. A USB cable connects the included PBK microphone (shown above) to the PC, then another USB connects the PC to the subwoofer. Each microphone has been calibrated and identified by serial number in the software. The program is capable of measuring multiple locations (max. 10) in the room to smooth out the bass response. Since I'm most concerned about the "sweet spot" and want to limit the potential of suboptimal calibration, I took 5 readings all around the central seat and the 2 adjacent seats. The program then will run thorough the calibration algorithm and show a screen that looks something like this...
You have the option to adjust the DSP crossover frequency point which in the graph above I've set to 160Hz (default is 250Hz) as well as how steep the filter should be. This gives you some customization options (not much).
The red curve was what I got in the room. As you can see, I have quite a dip at around 65Hz (red) which was correctable to some extent (purple). The peaks (eg. around 30Hz correlated to the calculated lateral mode for the room size at 29.1Hz using this online calculator) were easier to correct, and it's nice to see good frequency response down to 20Hz with this sub. (Check this link out for the room mode math calculations.)
The program will automatically upload the new settings to the subwoofer and away you go... Very simple calibration to do.
II. Room EQ Wizard (a.k.a. REW)
REW (5.01beta) can be downloaded free off the Home Theater Shack website - just need to register. It's just an amazing piece of Java code for the audio enthusiast.About 3 years ago, I purchased a calibrated Behringer ECM8000 mic (I see they don't sell these calibrated any more at this site). This microphone has served me well over the years and put to good use here again (note that I actually measured it a little lower at ear level sitting on the couch than the picture below):
Although I don't know if I fully trust the Behringer mic below 30Hz and above 15kHz, it's good to see frequency response down to 15Hz. Again we see the room mode around 29Hz. The deep blue line represents the eventual target curve we're aiming for based on the default REW house curve (for those looking for the excitement of a bass-induced thrill ride, try this target curve). For those looking for more based on home theater wisdom, check out this link on House Curves and more!
Letting REW perform its own EQ from 20-200Hz plus a few small adjustments on my end resulted in this mathematical prediction of room response:
Much more controlled on the low end using 7 parametric EQ settings (you can see the numbers above) plus 2 settings at 3kHz and ~11kHz to roll off the top. Total of 9 EQ settings were programmed into the Behringer DEQ2496 applying the adjustments in the digital domain and looped back to the Transporter for DAC duties.
Since some frequency boosting is involved, I reduced the DEQ2496 digital output levels by 6dB and double checked with some really LOUD music to make sure the EQ settings did not lead to clipping. The 1997 Iggy Pop insane remaster of The Stooges' track "Your Pretty Face Is Going To Hell" off Raw Power is a good one - average dynamic range value of 1dB for the album! If the DSP processing doesn't clip with that track, it's probably not an issue with >99% of my music.
Since it's always good to confirm that the EQs are actually doing what they're supposed to, here's the actual measured room response with & without the Behringer DEQ2496 played through the Transporter as DAC (measured at a higher level on a separate day):
Before REW EQ:
After REW EQ:
OK. Nice real-world confirmation that the EQs are doing what they're supposed to. The predicted results checks out even with a different DAC (measured with TEAC, confirmed with Behringer & Transporter - I knew from previous measurements that the Transporter is very close in frequency response to the TEAC)!
Looking Ahead...
Like I said above, I have many days ahead to make this room "work" better acoustically. EQ'ing so far is just the "quick and dirty" first step at this point focused essentially just on volume equalization (although I guess the PBK might be doing more in the algorithm). I haven't even begun to try using the Audyssey MultEQ XT for multichannel yet (in the Onkyo receiver), nor room treatments...Regarding room treatments, there's much to do! Here's the measured waterfall spectral decay plot in REW:
15Hz - 20kHz |
20-200Hz |
15Hz - 20kHz |
20-200Hz |
Plotted at 500ms duration down to 40dB with 1/6 octave smoothing, I'd really love to see more uniform steeper decay (<<300ms). Bass traps in the corner and absorption panels to the sides at the first reflection points could do the trick. Hmmm, maybe this will be a project for Spring Break - whip out the saw, stapler gun, make some wood frames, grab fabric and a stash of Roxul Safe'n'Sound :-).
There is also the issue with time alignment as addressed by mitchco using Acourate and convolution filters. That's another level of tuning I'll have to leave for another day! So much to do, so little time...
For now, the subjective sound quality has improved. There's already notably better control to the bass notes from Rebecca Pidgeon's "Spanish Harlem" (to use a well known audiophile favourite). Time to just sit back, relax, and enjoy some tunes! I've got a couple of ideas for tests coming up.
Musical selections recently:
- Daft Punk's TRON: Legacy soundtrack (2010) on Blu-Ray was stunning! I missed the movie in the theaters and rented the 3D Blu-Ray the other day... The movie itself was OK but the surround effects and techo score really made the movie an audio feast.
- The Eagles. Love 'em or hate 'em, I reacquainted myself with the multichannel DTS version of Hell Freezes Over (1997) the other night. IMO another fantastic multichannel release from the earlier days of surround sound when DTS was releasing their DTS-CD's (this was also my first concert DVD). That live ambiance really shines through as if you're sitting in the audience that night. The guitar work and percussion sound great in the new system on "Hotel California" - especially the bass impact. I saw them in concert about 3 years back and that too was a blast.
- Speaking of bass... I don't often buy modern pop recordings but I did enjoy listening to the recent album by Lorde - Pure Heroine. Fantastic job by the 16-year-old from New Zealand. The current Top-40 'hit' "Royals" gives a nice taste of the cavernous bass found throughout the album ("400 Lux" is another to check out). Have a listen to this album through a system with clean bass down to 20Hz and see whether you think you need a subwoofer :-).
Addendum:
In the "Answer To What Question?" and "What Were They Thinking Of?" files... I was looking at the recent CES2014 announcements in the usual audiophile watering holes and found this:
Esoteric Grandioso D1 Monoblock DAC
Given the level of performance of even modest DACs these days, I really can't imagine what would be the reason to go monoblock with a DAC. Seems like doing this could make things worse (channel desynchronization? need for DAC matching?) and at a significant expense ($22,000 each, not to mention all those extra cables!). Small price to pay to feel grandioso I suppose. As usual, would love to see the measurements for a pair...
hi archimago, I have a question for you. I'm a complete noob abaout all this sample rate stuff. okay, so I downloaded a Daft Punk album, sampled at 24/88.2. So logically, it will not play on my PC stock soundcard that only support 16 bits 41/48/96 and 24 bits 41/48/96/192 right? But why I still can play it, using foobar2000? I tested every sample rate and it plays just fine. Does it has anything to do with resampling?
ReplyDeleteAnother question is do we need to match the output format (sample rate) of our system to the sample rate of the files to achieve highest quality possible? Many thanks. Nice blog, I'm going to recommend this blog to my other beginner audiophile friends.
Hi Muhammed, sorry for the delay in response.
DeleteYes, the 24/88 album is being resampled. Unless you specify drivers like the ASIO driver which may have limitations to sample rate and bit depth, Windows will do it all automatically through the built-in Windows Mixer. It's not the best resampling algorithm, but will get the job done.
In general, it's best to match the bit depth and sample rate so the DAC converts it natively - this basically guarantees (as much as possible!) that the data isn't being massaged along the path. However, Windows could still be massaging the data if the volume isn't set to 100%.
Best thing to do to make sure it's 100% best quality - set Foobar to use ASIO or WASAPI interface (you need to download the ASIO/WASAPI plug-in). Set the playback to your sound card's ASIO/WASAPI interface.
hey, many thanks man. sorry for not responding early though. I'll definitely try the ASIO/WASAPI plug-in. I noticed that the stock soundcard doesn't have the 88200Hz sampling rate, just the CD quality 16.44100 up until 24.19200. Anyway, thank you so much Archimango.
DeleteMở bán an bình city hà đông
ReplyDeleteMở bán an bình city hà đông
Mở bán an binh city hà đông
Dự án chung cư an bình city
Dự án chungcuanbinh.com.vn
Chủ đầu tư an binh city phạm văn đồng
Giới thiệu chung cư five star cầu giấy cổ nhuế
Mở bán chung cư river park hà đông
Chủ đầu tư an binh city phạm văn đồng