As I mentioned last time, I changed the tweeter in one of my speakers resulting in a significant adjustment to the frequency response in the 2-5kHz range. This means it's time to update my AudioVero ((acourate)) (current version 1.9.12, 286€) room correction filters for late 2019. Looking back, I haven't posted on this since 2016! My, how time flies.
Given that it has been awhile, picking up Acourate again for some measurements required that I dig out my old notes and review the previous procedures again. As such I figured that it was time to write an update for a reasonably "quick and dirty" measurement which basically took me an afternoon to perform with excellent results. If you've tried Acourate, you'll know that this program is very powerful but can take a bit of time to figure out the interface and get comfortable with the process.
Okay then, let's get going with a quick but detailed summary while it's fresh in my mind and provide some pointers the next time I might have to do this again maybe in a few years :-). We'll then end this post with a few other related subtopics that came out during the measurements...
1. First, as per the general walkthru instructions from Mitch Barnett a few years ago, let's reduce the number of objects that can potentially obstruct or reflect sound waves in the room between the speakers and microphone. For me, this meant moving my leather sofa back a few feet (against the white Ikea Kallaxes on the left where my LPs are), moved items like the music stand and guitars out of the room, and likewise the padded ottoman/coffee table out of the way:
 |
| Before |
 |
| After |
Ideally, the leather sofa should be out of the room since the measurement is still susceptible to rear reflections so close to the mic and that corner seat to the left of the microphone should likewise be moved out. Since I wanted to get the job done without too much hassle, this was good enough.
2. Prepare the audio hardware and get the microphone calibration file ready.
As you can see on the floor up front by the turntable, this is my Microsoft Surface 3 Pro computer lying at an angle. I'm using the Focusrite Forte USB audio interface with power supply plugged in for the 48V phantom power to microphone (connected to mic input 2) and the unbalanced line outs are sent to my Emotiva XSP-1 pre-amp. Remember that the program only supports ASIO drivers used for the DAC/ADC.
The Dayton Audio mic comes with individual calibration data - go to their website, enter your serial number and download the text file.
Now import the .txt calibration file for the microphone and convert to an inverted calibration WAV:
Import it as standard 65536 samples, minimum phase. Now apply the inverse function into "Series 2":
You should now see something like the image below. Then save the Series 2 waveform with File --> Save Mono WAV out as a 64-bit file to somewhere easily accessible - this is what you'll calibrate with when doing the room sweep (Step 5):
3. Make sure the measurement microphone is positioned centrally between the two speakers using LogSweep --> Microphone Alignment tool.
This is a very cool tool that will send little right and left "clicks" out to each speaker which the microphone will pick up and the program will check for speaker distance based on the timing. It'll point to the direction to move the microphone and you'll see a green "0" when the speakers are equidistant. Also you can check the little waveform box for alignment:
If the speaker volume is too low, the program may not pick up each "click" very well and you might see a situation where the waveform for one of the speakers isn't being displayed consistently.
4. Set the File --> Project Workspace Definition directory.
This is the directory where all the subsequent measurement sweeps will be saved, room target curves live here, as well as the subsequent test convolution directory we'll see later on.
5. Time to record the frequency sweep - use LogSweep --> LogSweep Recorder.
Click on "Active Curve" number 1. Now run the LogSweep Recorder:
48kHz works well (and has been said to be preferable), and logically set the "Sweep End" to 24000. "Target Length" at 65536 works well although if you want, 131072 is also available for more "extreme" filters (this will increase latency). Note the "Mic Calibration" is pointing to the calibration file we created in Step 2.
Use the "Recording Level [dB]" scale to check to make sure it's not clipping. I typically turn the volume up to a peak of around 80dB SPL, C weighted, checked with my old SPL meter.
After the recording is done, you'll see the IACC (Interaural Coherence Coefficient) calculated and displayed:
As one would expect, IACC will vary quite a bit depending on the room and is also dependent on the furniture and other reflective surfaces in the listening space. (We'll talk a little more about this below.)
6. Room Macro 1: Amplitude Preparation.
Now that we have the sweep recorded, it's time to do some psychoacoustic smoothing out of the waveform in preparation for filter creation. Here's what I used with Room --> Room Macro 1: Amplitude Preparation:
Noticed that I used 30/30 in the "Window" parameter (default is 15/15). The first value refers to the number of cycles before, and the second number is logically the number of cycles after the frequency component being analyzed. By increasing this value, it reduces the effective amount of smoothing applied by incorporating more of the pre- and post- cycles around the frequency component. You can read more about Frequency Dependent Windowing here. FDW is a key concept, and part of the "magic" on how this all works.
7. Room Macro 2: Target Curve Design.
This is where objective meets creative subjectivity :-). It's time to decide what you want your room "house curve" to look and sound like. Feel free to design to subjective taste! Let's use a design like this:
Clicking and dragging the left "Basic" setting (greenish square) allows you to set what level you want the curve to sit at. Dips in the frequency plot below the target curve line will either not be corrected or at most mildly boosted up depending on your filter generation settings below (Step 9). Notice I've set this curve to flat from 10 to 500Hz, then dips to -6dB into 24kHz. That's a pretty good sounding, classic target based on the Brüel & Kjær "ideal speaker curve" research from back in 1974 (see paper):
Remember there are many other curve variants out there. Experiment.
Save Target as "Target.dbl" in your working directory and Exit.
8. Room Macro 3: Inversion.
Time to invert the frequency response based on the room curve to calculate the "correction" that needs to be applied. Run the Inversion macro.
We see that the room curve will attenuate the signal by about -3.86dB in this example. Remember that since we're primarily correcting peaks, there should be an attenuation of negative a few dB's.
9. Room Macro 4: Filter Generation.
Create the various filters at selected samplerates:
You can play with the various settings like "Excessphase window" and "Pre-ringing comp.". In my room, 4 for "Excessphase" and 2 for "Pre-ringing comp." work well based on test convolutions shown in the next step. Notice that I don't mind letting the program do a little "correction gain" by 2dB. Make sure to keep this low, <6dB recommended. I've checked off filter creation for 44, 48, 88, 96, 176, and 192kHz. No real need for 352 and 384kHz unless you routinely upsample and DSP at such high samplerates on your system.
As you can see, the program is reporting a -1.9dB attenuation with the created filters. With the potential for 2dB "correction gain" applied when needed, the attenuation is less than -3.86dB reported in Step 8.
10. Room Macro 5: Test Convolution.
Running the test convolution helps to confirm that the filters are achieving the goal of improving frequency and time-domain performance.
Notice that the IACC has improved overall. Make sure the little time domain step response tracing at the bottom looks good. If we zoom into this we see that it's reasonably symmetrical in each R+L channel and there's no pre-ringing for what should be a clean minimum phase "step".
If some of the settings are incorrect (especially the "Pre-ringing Comp."), what you'll often see is abnormal ringing. For example:
And we can get rid of that pre-ringing with setting "Pre-ringing comp" to 2/2:
Notice in the 0/0 image above, it's only the right channel (green) affected by ringing. Therefore one could just use "0/2" to address it; the two channel settings do not have to be the same.
You can also use the test convolution to have a look at the frequency response with the filter applied. What you do is start Room Macro 1 but point the setting to <Project Workspace Definition>/TestConvolution/Pulse48L.dbl like so:
After the macro is run, you'll see the psychoacoustically smoothed corrected filter response (make sure the "Frequency Dependent Window" parameters are the same as in Step 6 above). Now load into Curve 3 the left smoothed uncorrected sweep data "Pulse48Lmp.dbl" and Curve 4 the right smoothed uncorrected data "Pulse48Rmp.dbl" from the default "Workspace" (not from the TestConvolution directory). At this point you'll see the 2 stereo sets of curves overlaid. What we can then do is apply -20dB attenuation with TD-Functions/Gain to Curve 1 and Curve 2 (the corrected frequency responses) to separate the two stereo pairs and see the difference that the DSP has made:
Notice the (expected) effect that the "room curve" has had on the corrected frequency response dipping -6dB into 24kHz. You can further confirm the filtering effect by using Room EQ Wizard to measure like I did previously. Or just have a listen which is more fun. :-)
Here's another measurement to check out. Load up the <Project Workspace Definition>/TestConvolution/Pulse48L.dbl into Curve 1 and <Project Workspace Definition>/TestConvolution/Pulse48R.dbl into Curve 2. These are the unsmoothed test convolution data, and have a look at TD-Functions --> RT60: Reverberation Time to make sure your reverb time looks decent post-filtering.
This is the amount of time it takes for a 60dB decay across the frequency spectrum in your room:
The DIN18041 tolerance outline is the standard for "Acoustic quality in small and middle-sized rooms" - see more here. Looks like my room is less reverberant ("dryer") below 200Hz than the usual DIN18041, and more in line with the EBU3276 studio recommendation.
11. If you need to retry with new filter settings (yes, you probably will!)...
Suppose the test convolution isn't looking good (eg. pre-ringing and the symmetry of the two channels look poor). Go back to Step 9 above and restart Macro 4 making sure that it's pointing to your "Pulse48L.dbl" file in the correct work directory if all you want to do is change some parameters for the filter generation. This is the fastest way to try out the different parameters and run a test convolution after each iteration to get things right.
Because the "Excessphase window" parameters can result in unexpected outcomes and the values are not particularly intuitive, I wonder if there is a way to automate the process somehow in future versions of the software. For example, maybe letting the machine iteratively test out the parameters a step at a time for each value storing the resulting filters for the right and left sides. Then run test convolutions comparing the sides and report on the IACC results so the user can have a better look at the best calculated outcome? Given the possible combinations, this could take awhile but these days we have powerful computers that can probably provide interesting results over a couple hours of number crunching.
12. I use Roon mainly for listening in the soundroom these days. Acourate now has an option to ZIP up the FIR filters (up to 192kHz samplerate) for import into Roon which is handy.
I can then name the ZIP file to something like "Paradigm S8 - -6dB Taper to 24kHz.zip" and import it into Roon's convolution filter for the device / zone I want to apply the DSP to:
 |
| Import the filters into Roon. Feel free to enable "Headroom Management" if you're worried about any clipping especially if you've asked Acourate to add "correction gain" during filter creation. |
13. You're done! Congrats...
Now go listen to some music and see if the correction filters have "tuned" your speakers / room to achieve subjectively better sound.
Let's now talk about a few subtopics that came to mind during the procedure...
--------------------
Interaural Coherence Coefficient (IACC) and room...
The IACC value (obtained in Step 5 with the LogSweep Recorder) will change with the room and all the surfaces the sound bounces off. For fun, let me show you the IACC values I get when I record the LogSweep in the cleared room:
 |
| Cleared room. |
And now in my "typical room" layout:
 |
| Typical room. |
Despite the lower IACC, I have used Acourate to record the sweep and create filters with the bottom room configuration for comparison. While it does still sound good, I agree with taking the extra time to clear out the room a bit and measuring the inherent sound of the space. The ear/mind has an ability to adjust when it sees the furniture in place, sound reflections off of which would affect higher frequencies (>200-300Hz).
--------------------
Remember that when we look at step response graphs in Stereophile, typically, only about the first 10 milliseconds are being shown (usually measured 50"/4-feet away, at tweeter axis). Depending on the room, beyond 10ms, we're seeing room characteristics encroaching more significantly into the measurement.
Let's have a look at the first 10ms of the step response in my example above with the "cleared room". Remember that the microphone is around 9' away and the height is at ear level which for these speakers is below the tweeter axis perhaps by up to 12". Here's the "raw" recording without DSP applied, then what it looks like with DSP applied:
Let's have a look at the first 10ms of the step response in my example above with the "cleared room". Remember that the microphone is around 9' away and the height is at ear level which for these speakers is below the tweeter axis perhaps by up to 12". Here's the "raw" recording without DSP applied, then what it looks like with DSP applied:
Much improved step response with the driver components better time-aligned.
When time domain performance is good, my perception is that the speakers will perform the proverbial "disappearing act" and you'll notice that instruments and vocals become more focused in the soundstage. To me, sometimes I get the impression that the soundstage seems less "cluttered" as individual instruments and voices can be more easily picked out in discreet space. Subjectively, this is a pleasing effect, I feel less distracted, and ultimately less fatiguing.
Remember that it's still "garbage in, garbage out". Not all recordings are good or have the ability to project a convincing soundstage. Some albums simply do not benefit much from high-fidelity playback.
When time domain performance is good, my perception is that the speakers will perform the proverbial "disappearing act" and you'll notice that instruments and vocals become more focused in the soundstage. To me, sometimes I get the impression that the soundstage seems less "cluttered" as individual instruments and voices can be more easily picked out in discreet space. Subjectively, this is a pleasing effect, I feel less distracted, and ultimately less fatiguing.
Remember that it's still "garbage in, garbage out". Not all recordings are good or have the ability to project a convincing soundstage. Some albums simply do not benefit much from high-fidelity playback.
-------------------
Let's talk about room/house curves.
I like options. With my room correction filters I've created over the years back to 2016, I typically will construct 3 sets that I can switch between depending on mood (this is a way bigger difference than just some digital filter setting on a DAC!)...
Here are the 3 variants I have available to switch between in Roon:
Here are the 3 variants I have available to switch between in Roon:
As I said above in Step 7, there's a certain level of subjectivity we can bring into the room curve design. The "Harshness Dip + B&K" curve for example has been one I have been listening to over the last few nights. It adds a little parametric dip centered at 4kHz straddling the upper mid / low "presence" frequencies. It's a cousin of the controversial "Gundry" or "BBC" Dip. While it won't make very harsh recordings or strongly sibilant vocals pleasant, I like how it can dampen a bit of high frequency harshness / fatigue during longer listening sessions.
Here's the pre- and post-correction frequency response for the "Harshness Dip + B&K" target to show the effect (-20dB applied to the "After" tracings for separation):
Well audiophiles, the take home message is that it's good to have control over the kind of sound you want to achieve. Computer-based FIR filters like this provide a much more powerful tool than tone controls or basic EQ knobs and sliders. From what I have heard, I do not believe that we need to be afraid of deviating from the "purist" audiophile playbook of avoiding ways that could enhance the sound in our rooms. Remember that this isn't analogue audio where sticking something like an EQ box into the playback chain will inevitably add some noise and loss of resolution.
Looking at the bigger picture, IMO, the audiophile pursuit is not about spending money on products in the hopes that the system you put together sounds good (especially when good money is spent on likely sonically meaningless products like tiny stick-on "resonators", or expensive cables, even those that actually do act as "tone controls"). I think it's good for the audiophile's soul to act proactively and optimize the sound of your existing system based on your preferences.
Remember that audio hardware designers do not live in your home, have not heard your room(s), and are unlikely to own the gear you have especially component systems made up of various brands. This means ultimately it's up to oneself to tweak the final sound rather than depend just on the product designer's "ear" and certainly be careful with putting too much weight into purely subjective reviewer impressions.
-------------------
One last subtopic that I should mention for my system is that I also have DSP correction for my subwoofer. The Paradigm SUB 1 in the corner by the left speaker has its own "Perfect Bass Kit" microphone and software to program the settings inside via USB cable connected to a Windows machine.
Unfortunately the SUB 1 doesn't seem to work with the new ARC Genesis software Paradigm is using with their latest generation products. But the older PerfectBassKit-v2.10 program that came with my kit still works fine for me although I have to turn off Windows 10 driver signature enforcement (easiest is to hold SHIFT while Restart and go into Troubleshoot/Advanced Options - silly hassle) when installing the software and the USB microphone driver.
Unlike in the past where I ran Acourate with both main speakers and sub as a single unit, this time around, I've just used Acourate for the main speakers. PBK correction is applied to the sub and I'll just use the volume knob on the sub to adjust to taste how much bass I like. The bass is already quite good with my main speakers. Typically, listening to Roon from my Oppo UDP-205 for 2-channel music, the sub is set to around 20% just to add a bit more heft to the low end. I'll certainly pump the bass up for multichannel movies.
PBK software settings:
I know some folks like to have a bass boost for the room curve such as something like this:
In my system, turning up the sub volume would have that kind of effect (with better frequency response down to 20Hz and lower). And I can use the analog sub cut-off control at the back to adjust how high I want the sub to bump frequencies up to (typically ~80Hz to reduce/remove localization at higher frequencies).
-------------------
As I have expressed in the past, I do like the sound of my room with DSP filters active. I know some people have issues with "room correction" but I think anxieties are a bit exaggerated.
While I love sharing music with friends and family, I have no issue with admitting that for the most part, when I listen to my "hi-fi", it's a solo affair. I usually have the sofa and room to myself. If I have 2 or more people in the room, unless it's a demo and I have the guest in the sweet spot, the atmosphere will be more social and we'll generally be chatting rather than paying close attention to music. Already, as audiophiles, we position our speakers around the "sweet spot" and perform adjustments like toeing in to focus on that central sitting position. Some panel speakers especially are known to have narrow sweet spots. Room correction in my mind is just an extension of that idea which is central to optimizing 2-channel audio regardless. Remember that even though the DSP optimizes the sound in that central sitting position, this doesn't mean that off-center the sound is somehow suddenly any worse than usual playback without the DSP. One can easily measure just off center and double check that the results don't look too terrible (it doesn't measure too badly for me 6" or a foot away from the sweet spot along the sofa).
These days, our DACs are capable of fantastic resolution (which can handle digital processing and volume normalization without audible deterioration especially with 32 and 64-bit precision algorithms) and computers have plenty of processing cycles that can be put to use to improve perceived sound quality. Speaker quality and room interactions as I discussed recently are really the main arbiters of whether a system delivers high-fidelity. They always were the limiting factors in sound systems but have become even more so as a result of very affordable digital audio and amplifier technology. I advocate doing whatever you can to maximize your playback quality with well-controlled frequency response, good time-domain performance, and of course low distortion.
Remember that there are a number of ways to achieve room correction with varying quality outcomes. On the "free"/donation side, we have Room EQ Wizard and DRC Designer (I wrote about DRC Designer here). Check out this great how-to for REW and Roon if you haven't seen it. On the commercial side, we have Acourate (286€) as discussed here, Audiolense (165-390€ depending on version, see Mitch's review here) and the relatively inexpensive MathAudio RoomEQ (~$100, I have not tried this). Of course there are other hardware/software products out there like Sonarworks, miniDSP, DEQX, Trinnov, Lyngdorf, and TacT for 2+ channels. It goes without saying that multichannel systems often require individual level adjustments for each speaker and likely EQ adjustments to optimize tonality. Many commercial products use Dirac or Audyssey.
Obviously I cannot vouch for all the ways of getting the job done and I'm not sure how well each product handles both frequency and time domains. What I can say is that with Acourate, the difference in my experience when I turn on the DSP is well worth the time and effort to implement. The software also has provisions for higher complexity such as setting digital crossovers as discussed by Mitch previously and also addressed in his book.
Listening is the final arbiter so freely experiment for yourself.
In other news... The other day, ahofer contacted me about his blind testing idea. I think he has some great thoughts for "crowdsourcing a blind test" posted on the Audio Science Review forum as a fun way to test out claims and money can go to charity. I see that PS Audio's PerfectWave DirectStream DAC "Windom" FPGA firmware upgrade being audibly different despite lack of measurable effect has been promoted as a potential candidate for a blind test like this. Sounds good to me and I'm happy to chip in some money for charity or help in any way I can. Maybe Paul McGowan or Ted Smith (word on the Net is that he can hear jitter) can step up to take the blind test and prove that this Windom upgrade is significant - as claimed in the press release:
Obviously I cannot vouch for all the ways of getting the job done and I'm not sure how well each product handles both frequency and time domains. What I can say is that with Acourate, the difference in my experience when I turn on the DSP is well worth the time and effort to implement. The software also has provisions for higher complexity such as setting digital crossovers as discussed by Mitch previously and also addressed in his book.
Listening is the final arbiter so freely experiment for yourself.
--------------------
In other news... The other day, ahofer contacted me about his blind testing idea. I think he has some great thoughts for "crowdsourcing a blind test" posted on the Audio Science Review forum as a fun way to test out claims and money can go to charity. I see that PS Audio's PerfectWave DirectStream DAC "Windom" FPGA firmware upgrade being audibly different despite lack of measurable effect has been promoted as a potential candidate for a blind test like this. Sounds good to me and I'm happy to chip in some money for charity or help in any way I can. Maybe Paul McGowan or Ted Smith (word on the Net is that he can hear jitter) can step up to take the blind test and prove that this Windom upgrade is significant - as claimed in the press release:
"Windom reveals richness and natural detail within the music at levels formerly unattainable in any digital to analog converter we have auditioned. With its improved retrieval of reverb and ambience, Windom communicates with a glorious sense of openness and effortlessness, adding a sense of immediacy and ‘rightness’ without sacrificing honesty or transparency."
Wow... That sounds impressive! Surely this is "99% True" and shouldn't be too hard to blind test based on that description, right?
Music of the week for me has been Grimes' Visions (2012, DR8) which I only heard recently. "Grimes" is Canadian artist Claire Boucher from right here in Vancouver. Congrats on the song "Oblivion" off this album making it to #2 on Pitchfork's "200 Best Songs of the 2010s" list!
Also trying to get into Wilco's new Ode To Joy (2019, DR8), one of their slower, more reflective albums - I typically need a few listens before I can get into Wilco.
Happy Canadian Thanksgiving weekend to everyone, eh?! As you can see, we here in Canada say "thanks" earlier than our neighbours to the south, and without any Black Friday shopping event needed :-).
Hope you're enjoying the music...
Addendum - October 13, 2019
Thanks to mastercheif on the comments. Indeed, I could save each setting as a separate Roon DSP Preset for example here when I play the Oppo UDP-205 in my sound room:
And yup, these presets can be accessed with the Roon app on my phone / tablet while sitting and listening.
Hope you're enjoying the music...
Addendum - October 13, 2019
Thanks to mastercheif on the comments. Indeed, I could save each setting as a separate Roon DSP Preset for example here when I play the Oppo UDP-205 in my sound room:
And yup, these presets can be accessed with the Roon app on my phone / tablet while sitting and listening.
So, I take you are in the camp that says correction above the room's Schroeder frequency is good. I've tried both with my 5.2 system (using Audyssey and its neat tablet/smartphone app) and have come down on the 'correct the bass only' side....for now!
ReplyDeleteHi Steven,
DeleteYeah, I'm in the camp that's okay with correction above Schroeder just like I'd be completely okay with someone using EQ to tune down treble frequencies when needed.
I have not personally come across issues with doing this and I do feel that the gentle slope of the B&K curve is psychoacoustically pleasant. Whatever it takes, man :-).
I hear ya! Dr. Toole himself endorses broadband EQ -- tone controls like the 'treble' and 'bass' knobs no longer found on receivers -- for correcting imbalances in the recording, on a recording-to-recording* basis. But again, that's after we've bought speakers with 'neutral direct sound, and well behaved off-axis ', and then paid proper placement/treatment/mode taming attention to the room. The deficiencies we are correcting with EQ after that, aren't constant ones.
DeleteArchimago, thanks for a great post! How do you feel about Toole's statement that good speakers do not need correction above the bass region, and that any correction applied above the transition frequency effectively degrade speaker's flat free-field response?
ReplyDeleteFrom my personal experience, I start with placement of speakers and toe-in to ensure that the phantom center and imaging are realistic. ETC graphs are a great tool for spotting early strong reflections, and various forms of shaped pink noise and "pings" can help checking the quality of imaging subjectively. After you get this right, very minimal correction of speaker response is required--a couple of PEQ filters can do this job.
I agree that such a procedure is a more tedious process than just going through correction software steps, but my understanding is that it's better to treat the root of the problem (unwanted speaker-room coupling) rather than the symptoms (uneven frequency response). I actually use both Acourate and REW, but more like diagnostic tools and for corrections in the bass region.
Thanks for the note Mikhail,
DeleteGood tip about using the Energy Time Curve for checking out those reflections. Looks like you're well seasoned in the hobby! Nice.
Certainly, philosophically I don't think anyone would disagree that it's best to get down to the root of the speaker-room coupling issue and if that can be done without too much pain, by all means do it.
I'm sure doing adjustments like this will affect the free-field response of the speakers (whether the speaker's response is actually "flat" is not a given!). As a pragmatist, the question is whether the change is ultimately "worth" it from the context of the listener in his/her environment. More specifically, if it "sound better" in that sweet spot, should there be any concern?
My sense is "no", not if the listener prefers the change.
To summarize, I don't think it's unreasonable to say...
1. Get good speakers you like the sound of so you have the ability to achieve the kind of frequency response naturally. Of course we also need a decent room and capable equipment.
2. Start with the physical layout, placement, toe-in, room treatments. Use REW to get a sense of the untreated FR, RT60, etc. to help with optimization and gauge satisfaction with the results.
3. Play with what you already have on your gear to optimize sound - eg. tone controls, EQ if already available, DSP subwoofer settings.
4. If still not satisfied or want to experiment, try out digital room correction and see if it suits your taste! One might even find that this exceeds one's expectations!
A couple of blind test results I would love to see if anyone has them:
1. What is the audibility of aligned time-domain speaker performance?
This will help us decide whether time-domain correction is something the perfectionist audiophile should seek out (whether by DSP or hardware design).
2. Audiophile preference between a level-matched uncorrected speaker vs. DRC "corrected".
This will help us decide how many % of people like the effect from DRC. And also help us understand if the DSP processing added any kind of distortion that could be unpleasant or experienced as "artificial".
Important caveat: Toole's recommendation *only* applies to loudspeakers that are quite well-behaved on and off axis. For all others, EQ of mid and treble could be beneficial.
DeleteSounds very reasonable.
DeleteBTW: Here's an interesting article off the Linkwitz website regarding the importance of time domain performance and speakers:
Deletehttp://www.linkwitzlab.com/Attributes_Of_Linear_Phase_Loudspeakers.pdf
Some interesting papers in there although it would still be great to see some actual blind tests of audiophile speakers and real music. :-)
The only DRC preference research I know if is Sean Olive's
Deletehttp://seanolive.blogspot.com/2009/11/subjective-and-objective-evaluation-of.html
AES presentation deck here:
https://drive.google.com/file/d/0B97zTRsdcJTfY2U4ODhiZmUtNDEyNC00ZDcyLWEzZTAtMGJiODQ1ZTUxMGQ4/view?hl=en
Folks thanks for a great discussion and links! Archimago, special thanks for a detailed reply. I've re-read your post from 2015 on Acourate and comments to it, and fished out an excellent presentation by JJ Johnston on DRC--http://www.aes-media.org/sections/pnw/pnwrecaps/2008/jj_jan08/. I think, discussions on this topic will never end :)
Delete2. Audiophile preference between a level-matched uncorrected speaker vs. DRC "corrected".
DeleteToole answered that in his book and he's posted the actual data online on one of the forums before, somewhat recently, though I'd really have to dig to find. May have been in that mega AVS thread iirc.
Bottom line, in blind listening tests, listeners preferred the uncorrected response, despite the DSP'd one looking pretty to the eyes like the B&K "target" one.
2 ears/bran can interpret the direct onset and the subsequent reflected sound very unlike the pressure mic you are using. Read JJs work carefully as well, he's saying the same.
You're not removing expectation biases et al in your DRC listening. Strange things can happen when you do ;-).
Hi AJ, thanks for a good pointer. I've found some good posts by Dr. Toole on DRC at AVSForum:
Deletehttps://www.avsforum.com/forum/89-speakers/3038828-how-choose-loudspeaker-what-science-shows-162.html#post58538632
https://www.avsforum.com/forum/89-speakers/710918-revel-owners-thread-506.html#post58216342
https://www.avsforum.com/forum/89-speakers/3038828-how-choose-loudspeaker-what-science-shows-34.html#post57498070
But I haven't found anything about preferring uncorrected response. Although, his personal opinions seem to be against excessive DSP adjustments based on in-room measurement as they throw you back into the "circle of confusion"--adjusting the target curve and filters based on your taste while listening to favorite programs.
I may have misremebered/mispoke. The post Im refering showed an anechoicly "corrected" speaker, vs the same speaker in room with Archimago style "room correction". The measured in room response of the anechoicly corrected speaker looked worse than the smooth "target curve" "room corrected" one...but it won in the listening test. Blind of course. I think the curves shown in post were the 2 in room measured responses.
DeleteAnyway, this paper covers well why the kind of EQ'ing Arch is doing is rather misguided http://www.aes.org/tmpFiles/elib/20191015/17839.pdf
>500 hz eq/correction should be done anechoic (or quasi) judiciously, with careful attention to on/off axis, not just the on axis. The only "in room" EQ should be below 500hz (+/-). Cut peaks, don't boost holes, take multiple measurements around LP.
AJ
Soundfield Audio
What Archimago needs to do now is a 1m (vertically centered between tweer/mid) 0/30/60 degree off axis measurement of the speakers *WITH* and without the "correction" filters active, to see what it has done to the speakers native response.
DeleteUnsure what differences there are based on "V3" version, but the originals had pretty good native responses https://www.soundstagenetwork.com/measurements/paradigm_signature_s8/
Hi AJ,
DeleteThanks for the links. Will check out the discussions from Toole as well Mikhail...
I think it would be very interesting compiling the data on this since I'm not "sold" that generally correcting those higher frequencies is that much of a problem based on what I'm hearing. For a "typical" speaker like these Paradigms, I get the sense that the change in time domain has been beneficial for soundstage. Instruments appear to be more well spaced, and for electronic music like the Boris Blank Electrified I mentioned last time, spatial effects sound more "surround" with the DSP on (eg. the simulation of sounds coming from behind the head better accomplished).
Also for situations where speakers might be poorly balanced such as when the frequency response was off on the tweeter as mentioned last week, using the DSP these last few years masked the anomaly reminding me that at least it can correct deficits like this.
Maybe ultimately it's gonna be down to subjective preferences... If a person buys a pair of speakers and just loves the sound in the room already, certainly putting a DSP on will change the character of those speakers and that change, even if more "accurate" or even "neutral" may not be up to the person's liking. I'm sure there will be many subjective (psychological) elements at play; not the least of which if a person spent $50,000 on a pair of speakers, to think they could use a $1000 DSP hardware/software probably at some level may "feel" wrong.
Yeah, sounds good AJ, when I have time, I'll try that 1m 0/30/60° measurement to see. The Signature V3 does appear to be a very different beast even based on the impedance graph!
BTW AJ,
DeleteI'm not seeing that PDF file from AES. Is there a title I can search up?
Mikhail,
I really like this quote from Toole recently:
I know that there are other digital equalizers with problems. All originate with clever math/DSP engineers doing things that may make academic sense, but that pay insufficient attention to the peculiarities of human perception. At professional audio gatherings I have had extended discussions/arguments with some of their engineers. It has always come down to opinion, not fact, and the opinions are inclined to enhance the customers' perceived value in the product. It is part of a mighty struggle to be different or distinctive in a product that delivers something that nowadays many people can do for themselves with off-the-shelf DSP, free measurement software and a $100 mic.
None of these processes are supported by published double-blind subjective evaluations. Tell me if I am wrong.
The universal availability of "room EQ" is now a kind of "disease" in audio. People place trust in these devices that is misplaced. As I have stated several times, when the operating manual of the "calibration" device basically states that if the customer does not like the sound from the default target curve, then change the target curve. At this point it becomes a subjectively guided tone-control exercise, not a calibration. The "circle of confusion" for whatever program used during the tweaking is now permanently installed in the system.
Well, looks like double-blind subjective evaluations are not available. And maybe at the end of the day, unlike say DAC measurements or amp measurements where we can "simply" use science to look for flat responses, low noise, and lack of phase anomalies, it's OK at this time in history to just have fun and be "more subjective" with the speaker-room interface in the home environment... Leaving "more science" to the speaker designers with their anechoic chambers.
Sorry, here is the page link http://www.aes.org/e-lib/browse.cfm?elib=17839
DeletePlease keep in mind that the program is doing several things, so beware ascribing subjective X to "time domain". It seems based on the described process that there might be greater *inter-channel* matching (rather than inter-driver), so yes, I would expect this to have perceived benefits with spatial rendering. That is well known and a key point in JJs papers.
Unfortunate the NRC measurements are for the V1. However your 1m quasi anechoic measurements should suffice to show the changes involved with the FRs of interest, >500hz
Also sorry, missed your Toole quote above. Yes, couldn't agree more...
DeleteOn one hand Floyd says room correction is a disease... OTOH Floyd says, "For perspective it is worth remembering that about 30% of the factors influencing our judgement of sound quality relate to bass performance - and this is dominated by the room, only correctable in-situ. So, ANY loudspeaker can sound better after room EQ, so long as it competently addresses the bass frequencies - this is not a guarantee, but really is not difficult for at least the prime listener."
DeleteFloyd also says, "Most commercial algorithms these days appear to improve the low frequencies - for a single listener at least. That is about 30% of the factor weighting in sound quality ratings. Above the transition frequency it is the "Wild West". Some algorithms use broadband "tone control" adjustments (good) while others pride themselves on flattening even small irregularities (probably bad)."
So room correction below the transition frequency = good. Above transition frequency, broad tone control adjustments = good. Acourate and other good DSP DRC packages can do this so = good.
What gets lost in translation is the great work Sean Olive and team have done in correlating user preference with objective measurements to show that there is a preferred target frequency response for both loudspeakers and headphones (and DRC for that matter, Steven in this thread has already linked to above). Check out:
Perception and Measurement of Headphone Sound Quality: Is There a Preferred Target Response.
The Perception and Measurement of Headphone Sound Quality: Do Listeners Agree on What Makes A Headphone Sound Good?
While it mentions headphones, it is based on the same approach they used for loudspeakers and is mentioned a number of times, how they did it, what the results are, that it is repeatable and that most participant listeners agree on a preferred target frequency response for both headphones and loudspeakers, which turns out to be the same, but translated for headphones. i.e. a straight line tilt from 20 Hz to -10 dB at 20 kHz at the LP for loudspeakers. This was also the preferred response for Sean's DRC paper again linked by Steven above. Lots of good references to AES papers if you want to dig deep.
This has been my anecdotal findings as well, use room eq for below transition and broad "tilting tone control" above transition if required. Some speakers already hit the target right out of the box. Of course, given who performed the work it should come as no surprise that Harman Revels and modern JBL monitors hit the target with no tone controls required. I have only had a couple of speakers through my room that have hit the target with no tone control applied (D&D 8c and KEF LS50).
So break out the room correction, use it within the guidelines to optimise your investment of gear in your room and enjoy some accurate sound reproduction! As Floyd Toole says accurate and preferred are synonymous.
Excellent comment, thanks Mitch...
DeleteLots to digest!
By far the most important thing for you to digest, is that a single pressure microphone at the LP, in no way represents binaural hearing/brain.
DeleteThat sure looks like a single pressure mic at the LP in both your and Mitch's setups ;-).
Toole tends to mention this in his AES paper and elsewhere....for those who will heed.
cheers,
AJ
Arch, I'll add a few more links before signing off. Unfortunately, this blog format makes it difficult to post very relevant info, graphs, pics, etc.
DeleteAs such, I think I'll start a related thread on HA, the only online forum not run by shysters ;-). Pretty sure both you and Mitch are members.
1) The Olive "target curve", from this link https://seanolive.blogspot.com/2009/11/subjective-and-objective-evaluation-of.html
His *very first comment*
"In truth, the optimal in-room target curve may depend on the loudspeaker directivity and reflectivity of the listening room. If the room is acoustically dead with few reflections and/or the directivity of the loudspeaker is quite high, the in-room response will represent a higher proportion of the direct sound, which should be flat. Using a target curve with large downward tilt will make the loudspeaker sound too dull."
2) The Harman room. Look at the measured seat responses in that rather large room, from almost 3m away http://www.aes.org/tmpFiles/elib/20191016/14873.pdf
3) The highest rated speaker according to Harman, real in room response
https://www.stereophile.com/content/revel-ultima-salon2-loudspeaker-measurements
Do you see anything resembling a "target response", Acourate or otherwise?
;-)
cheers,
AJ
From: The Subjective and Objective Evaluation of Room Correction Products. Conclusions : “A flat in-room target response is clearly not the optimal target curve for room equalization. The preferred room corrections have a target response that has a smooth downward slope with increasing frequency.” Looking at slide 24, the most preferred correction is 20 Hz to -10 dB at 20 kHz. “When done well, room correction can significantly improve the quality of sound production”.
DeleteFrom: Perception and Measurement of Headphone Sound Quality: Is There a Preferred Target Response. Preferred in room loudspeaker target response methodology: “Choose an accurate loudspeaker and equalize it to a flat in-room response in a reference listening room. Listeners adjust the bass and treble levels to their preferred level (single versus two parameters)” See preferred in-room loudspeaker response on page 43. Conclusions: “Preferred in-room loudspeaker response is not flat but has a bass boost about 6.6 dB @ 105 Hz and treble cut of -2.4 dB above 2.5 kHz (i.e. 20 Hz to -9 dB at 20 kHz). The general shape of the in-room target response approximates the sound power or predicted in-room response of a well-designed loudspeaker above 200 Hz. The preferred headphone target response closely approximates the preferred in-room loudspeaker response with about 2 dB less bass and treble.”
From Development of Harman Headphone Target Curve : See Slide 10 Results: “The green dotted curve is response of the loudspeaker equalized to a flat in-room curve. Listeners did not like this baseline curve and adjusted the bass 6.6 dB higher and the treble -2.4 dB lower. More evidence that the in-room loudspeaker target should have a 9-10 dB downward slope from 20-20 kHz”
See slide 11: “A flat in-room target curve (green curve) is not preferred; to achieve the preferred target (the black curve). The preferred in-room target has a response with a ~ 10 dB downward slope from 20Hz to 20kHz.”
Re: 3) The highest rated speaker according to Harman, real in room response
https://www.stereophile.com/content/revel-ultima-salon2-loudspeaker-measurements
Do you see anything resembling a "target response", Acourate or otherwise?
Yes, absolutely! Again, it is 20 Hz to -10 dB at 20 kHz = the preferred in-room target. You can read it right off of JA’s in-room measurement!
I could pull more references, but what is the point? You just like to argue. Re: "Start a thread at HA, the only online forum not run by shysters." Really man? Don't you mean it is the only forum you have not been banned from ;-) Anyway, all in good fun. Have a great day!
Cheers,
Mitch
(Updating this, I got the DSP wrong the first time ) Arch, what's extra-special about that Toole quote (https://www.avsforum.com/forum/89-speakers/3038828-how-choose-loudspeaker-what-science-shows-162.html#post58538632) is that for much of it he appears to be talking about 'EQ issues' within JBL's own SDP-75 processor, whose room EQ is Trinnov's + Harman research parameters. I think perhaps he had to have it customized to limit processing to bandwidth above transition, but I'm not sure.
DeleteAlso, here's another forum where Dr. Toole wrote (and replied in comments) about small room EQ....
Deletehttps://www.audioholics.com/room-acoustics/history-of-multi-sub-sfm
"Yes, absolutely! Again, it is 20 Hz to -10 dB at 20 kHz = the preferred in-room target. You can read it right off of JA’s in-room measurement!
DeleteI could pull more references, but what is the point? You just like to argue. Re: "Start a thread at HA, the only online forum not run by shysters." Really man? Don't you mean it is the only forum you have not been banned from ;-) Anyway, all in good fun. Have a great day!"
Yes Mitch, as I anticipated above, a forum where posting actual graphs etc would be needed to counter misdirection, misreading, misunderstanding, etc.
This blog format is not conducive.
Ask and ye shall receive. ;-)
You have a great day also. See ya'll there.
Ooop, the link!! https://hydrogenaud.io/index.php?topic=118314.msg976409;topicseen#new
DeleteSteven, the reason I liked the text from Toole was as a reminder that much of this (ie. DSP adjustment for room acoustics) is not necessarily based on firm science nor well tested (ie. no DBT for all that we're talking about here).
DeleteWhat I can say is that in my system, I like the sound of what I'm hearing through the Acourate filter, and those I've invited over seem to concur. I'll for sure measure the speakers at 1m-0/30/60° as suggested by AJ but even then I'm not sure whether the change from un-DSP'ed is necessarily all that bad unless I truly see something horrible! I don't listen in an anechoic chamber nor sit 1m from the speakers.
While we can and should use "objective methods", and I can confirm with my mic that around the listening sweet spot, the frequency response looks good and time-domain characteristics likewise suggests "time coherence", I'm okay also with taking off my "more objective" stance and aim for what sounds good to me at times like this. After all, I'm not suggesting anyone pray to the God of Audiophilia, buy voodoo cables, or select irrational room curves.
And as I noted in my post, my expectations are not necessarily lofty! All I care about is that central sweet spot. If this one location sounds better, more "flat" especially of the bass, targeted for a pleasant psychoacoustic experience, better soundstage characteristics, I'm happy because that's where I'll be seated >95% of the time when I listen for such things. And thus far, I don't hear anything I'm dissatisfied about. If anything, I've been listening to a few more favourite albums this week and tweaking the target curve with the "harshness dip" trying out a few settings.
[Aside: while I'm no fan of TAS, I do enjoy their recent article on Siegfried Linkwitz and his discussion of the 3-4kHz, 4dB(!) dip. http://www.theabsolutesound.com/articles/in-memoriam-siegfried-linkwitz-19352018/ ]
Bottom line... I'll definitely measure the speakers at 1m at some point for another post to show what the DSP did there to the quasi-anechoic response. But I'm still having fun enjoying the clear difference the DSP is making and appreciate the power of being able to target the room curve to my liking. I trust my ears/room/gear can be quite different from everyone else's and thus we can all choose what we like! How's that for an objectivist accepting the subjectivist position at times? :-)
As Toole says regarding changes to the target curve: "At this point it becomes a subjectively guided tone-control exercise, not a calibration." Absolutely! I'm not calibrating a studio control room here. IMO, there's a time for "subjective guidance".
"I don't listen in an anechoic chamber nor sit 1m from the speakers."
DeleteArch, you have some fundamental misconceptions of what is being discussed. You, like Mitch, have this totally back-asswards;-).
The "preferred curve" is not some silly visually attractive summed response target at a single pressure mic at LP where 2 ears/brain would be. It's the polar response of a Salon2 et al!! *THAT* is what your 1m 0-60 measurements should try to "target". *THAT* is what listeners prefer in loudspeaker tests (not "Room correction" tests). Then, whatever curve >500hz that results in IN ROOM AT LP (like what I show over on HA for the salon2), *IS* "the target". Not the other way around!! LOL
Over and out.
Okay AJ,
DeleteI can certainly agree that the Salon2 sounds good. And those Stereophile measurements for them are also good. But at the same time, there are other speakers that sound good as well with significant variation in the results. Let me just add a thought which I hope is practical and what as a hobbyist, would make a difference...
You mentioned Olive's quotes above and the Harman rating for the Salon2; is this part of a publication where the Salon2 is compared to other speakers or are we talking internal data? From what I can gather, the rating must be a subjective response from a group of listeners in their standardized room which certainly might also reflect JA's Figure 8 room response curve from furnishings/absorption (as mentioned by Mitch).
I can certainly accept that it would be great to have speaker mids and tweeters measure similar to the Salon2 and use room treatments to optimize the sound rather than DSP-type devices. Given that this is unlikely to happen, what is then important to know is whether there is evidence of "harm" here by using DSP above somewhere between 200-500Hz.
For example, if one were to take 2 different but well built speakers, identical source and amplification, same room, have a few people take turns at the sweet spot listening. Suppose one of these speakers is clearly preferred, we then measure the room response curve at the listener for that speaker. Now apply an approximate frequency response curve to the less preferred speaker using say Acourate or similar full-spectrum EQ (a test like this could also allow us to compare DSP algorithms as well).
If we were then to blind test the same listeners using the 2 speakers volume-matched, one with "correction" to make it similar to the other one, what would we find? Would the 2 speakers sound similar to listeners at the sweet spot regardless of various details unaccounted for (eg. directivity)? Would there still be a clear preference for one vs. the other? More importantly, would listeners hone in on the DSP and be able to pick out distortions, artificiality or other anomalies?
Anyone tried something like this?
Mitch's KEF LS50 vs. JBL 4722 test is similar but imagine if we used a pair of Salon2's (without any DSP to potentially mess up anything) and compared this with something less expensive but still good like B&W 703 S2's with DSP emulating what the measurement mic heard at the listening position of those Salon2's.
If a practical test like this suggests that DSP like Acourate can get us the experience we desire for much less expense, then from the perspective of the consumer, it's a step in the right direction for hobbyists to experiment, potentially achieve great sound with lower cost, wouldn't it?
Hi Arch, your post wanders a bit too much for a concise response. Yes, the Salon **Due to it's smooth polar response** measured anechoically, is highly rated in the Harman tests. Yes, its possible other speakers **With similar smooth polars measured anechoically** are same/more/less preferred. Harman doesn't always publish that sort of thing. There was one recently, a 228 vs a Magico, that was published. My understanding is if you have any doubt, you can bring your speakers there for a comparison.
DeleteOnce again, I think you are entirely missing the point. It's clear you believe that the far field, in room "blind" curve (to quote the CA article) is the main determinant of SQ and should be used as such.
You need to contact Toole to get what you're seeking, as I stated before - the blind listening test os an anechoically EQ'd speaker vs *the same* speaker "room" EQ'd, including both measured "in room".
You are however correct. These "Automated EQ for dummies" will (have!) win out in the mass consumer market, vs telling someone to learn some perceptual science, EQ the direct native speaker polar response very carefully, *then* the room response only below 500hz. No question.
If you want to use as tone control **FOR recordings variability** as Toole is saying (and being grossly misrepresented) and NOT a permanent signature, sure.
Hi Arch,
DeleteRe: what is then important to know is whether there is evidence of "harm" here by using DSP above somewhere between 200-500Hz.
There is no evidence of harm as can be seen by AJ’s response. Just more appeal to authority.
One does not need to drag speakers to some facility in the US to compare, you can do it in your own living room, even using different speakers that are eq’d the same and not even using room correction :-) For example, I reviewed the Kii THREE and D&D 8c’s. Using nothing more than the on-board eq controls supplied by both speakers, we can see here that the measured in-room frequency response at the LP is virtually identical
Martijn Mensink, designer of the D&D 8c says, “I've had the Kii's and the 8c's side by side in my living room for a while. The Kii's too are remarkably good speakers. With just some subtle EQ the two could be made to sound very similar on most program material - to the extent that I might not be able to distinguish them in a proper blind test. I'm still amazed sometimes by the extent to which differences in sound can be explained by frequency response.”
Kudo’s to Martijn, even comparing to a competitor, he can honestly tell it like it really is.
Btw, the Kii THREE’s measured +4 dB hotter from 3 kHz on up to 20 kHz as compared to the D&D 8c’s out of the box. To my ears, the D&D 8c’s sounded neutral in the top end and the THREE’s overly bright. So using the Kii THREE’s contour control, introduced a -4 dB shelf at 3 kHz, which then brought the treble balance down to match the D&D 8c’s. And matching the 20 Hz to -10 dB 20 kHz preferred target ;-0 No harm in that and in fact, standard operating procedure and why the speaker designers put in the eq controls in the first place.
There is no evidence of harm as can be seen by AJ’s response. Just more appeal to authority
DeleteThe evidence is the test Toole has posted showing harm in a blind test, unlike your believer "faith based" subjectivist uncontrolled listening posited as evidence.
None of your, or any faith based "Room EQ" listening is ever done blind, as Toole likes to point out when he interacts with your ilk.
The Harman tests used a terrible speaker, not a well engineered one that both Olive and Toole keep telling you Mitch, need no EQ >500hz.
But I understand you have "enticing marketing story" science fiction books to sell
"I believe every piece of audio equipment has its own sonic signature. E.g. CD transports, cartridges, tone arms, turntables, preamps, amps, cross overs, speakers, interconnects, basically every component, part, and wire in (and around, e.g. power supplies) the audio signal path will have its own sonic signature, whether designed or not."
Deletehttps://audiophilestyle.com/blogs/entry/206-sonic-signatures-the-art-and-the-science/
No surprise then when Mitch falls for enticing marketing stories and "hears stuff" when seeing measurements.
Floyd Toole
Sep 30, 2019#565
Spocko said:
Wait, how does one "copy" another's EQ curve, with 2 completely different rooms, speakers, etc.??
It cannot work as advertised, but that doesn't stop people from doing it. It has its origin in the decades old notion that steady-state room curves are definitive of sound quality. We now know that it is simply not true above a few hundred Hz
Seemingly endless promotion of "room EQ" algorithms - a for-profit exercise - is partially responsible, aided by human nature which is inclined to believe a good story. It is an ingredient in "faith based" audio - if you believe it, you just might hear it.
Arch,
ReplyDeleteGreat article. I agree with the thought that we ought to optimize the system that we already own. Also agree that applying DSP filters is much more efficient (and powerful) than using fancy cables. If I understood correctly, you mentioned that you use different convolution filters. Are you able to switch to these filters “on the go” within Roon or do you have to search and install the new filters every time you want to make a switch? Thanks!
Hi nicoff,
DeleteIt's not hard to switch filters in Roon although you'd need to start and stop the song so it's not like an immediate switch.
So what I do is have the 3 filter ZIP files in the same directory. In Roon, click on the DSP setting for the device/zone and select which of the 3 ZIP files it should use for the convolution filter.
I don't think the Roon phone/tablet apps can do it (doesn't work on my Android tablet), but from my listening sofa, I can Remote Desktop using a laptop into my computer where Roon Core is running and get it done without leaving the chair...
You can make each convolution filter a “DSP Preset” so that you can easily switch between them, even on mobile.
DeleteAlso allows for different headroom management settings per filter, which is often necessary with Acourate.
Thanks mastercheif,
DeleteI guess you must be a Halo fan :-).
Great tip and should have thought of that! Indeed I see I can change DSP presets with the phone/tablet app. Thanks.
Archimago
ReplyDeleteAs you are doubtless aware, the controversy remains over the audible benefits or not of time/phase coherence for speakers. I've seen arguments for both sides (and by people who have done tests). So I sure don't know the answer.
That said, my personal, anecdotal experience aligns with the description you gave about at tidying up of imaging. I've own a great many speakers of different types (everything from panels, to omnis, to much in between). Among them have been the Thiel CS6, and later the Thiel 3.7 and 2.7 speakers (time/phase coherent). If there was one overriding characteristics of the Thiel speakers vs all others I've compared them with, it is the sense of imaging precision and density; the sense that "all the sonic information coming from a voice or instrument has been lined up properly" rather than blurred or spread out. They tend to make other speakers sound somewhat vague and messy by comparison.
I don't know if this comes from the time/phase coherent part of the design, but it sure does seem to be a distinguishing feature (and one noted constantly in Thiel reviews, and by Thiel owners...and often in reviews of other time/phase coherent speaker designs).
As usual very good subject matter presentation. I am confused regarding the step to move furniture etc. Room correction is by definition DSP correction to the system (room, speaker and amplifier).
ReplyDeleteNicely done Arch! Good looking responses there. Bet you it sounds pretty good :) Thanks for linking to my DSP articles and book.
ReplyDeleteI have tried a level matched, blind test between time domain and non-time domain correction. The frequency responses were virtually identical, with the only difference being the time domain correction, which included both time alignment of drivers and excess phase correction. To my ears, it is a subtle but audible difference and similar to what you describe. I find the depth of the sound field increases and makes the difference between 2D and 3D sound. I would love to see double blind testing performed in this area given the advancements of powerful DSP software that was not available some years ago.
Room correction using these powerful DSP packages like Acourate is most often understood. These packages use frequency dependent windowing both for the analysis and correction. What this means is even with a full range correction applied, only below transition or Schroeder is the room included with the direct sound and above the transition frequency (let’s say 500 Hz), then it is only the direct sound that is being eq’d like using a “tilting tone control” that applies a broad tonal correction if the speakers are a bit too bright or dull depending on one’s speaker and room. This follows JJ’s excellent article as linked to above indicating it is the direct sound timbre our ears hear first. Frequency dependent windowing is fully and independently adjustable for both the frequency and time domain
As far as moving objects out of the way when measuring and correction speakers/rooms, the idea is that the sound field will be more natural without trying to correct for reflections from objects placed in the way of the speakers and listener. I have experimented with both approaches, several times, and to my ears, the correction always sounds more natural without any objects in the way of the measurement and then placed back after the measurement.
Finally and perhaps a preview of a future article, folks should know that the “Standard Method of Measurement for In-Home Loudspeakers (ANSI/CTA-2034-A)” is available as a free download. Anyone interested in how speakers should be measured and spec’d should take a moment and download the article as this is the defacto standard.
The article has an excellent section on how the anechoic measures are used to estimate the in-room frequency response. One can see on page 37, Fig 11 a comparison of the predicted in-room response and the actual measured in-room response. Of course, one can’t fully predict the effect of room modes below transition/Schroeder, but does a pretty good job. Above Schroeder we can see that the measured in-room response is virtually a perfect match to the predicted in-room response based on anechoic data.
As consumers, I hope we put more pressure on speaker designers and manufactures to follow the standard and produce the reports that include the predicted in-room response. This would really benefit the consumer as the report would convey a reasonably good representation of how it may sound in a room based on its off-axis response and how this response affects the consumer’s experience.
Keep up the great writings Arch!
Mitch
Mitch, thanks a lot for all writings and links! I re-read the section "Acoustics and Psychoacoustics of Room Correction" from your book on Acourate. So if I understand correctly, room correction software that employs FDW produces the same correction as if were have magically removed all the reflecting surfaces from the room. And it is also the same as if the speaker manufacturer has added DRC to their speaker while it was in an anechoic chamber (as it's done for NTi TalkBox for example). Thus, applying correction with Acourate or Audiolense doesn't actually contradict Toole's principles. That sounds re-assuring!
DeleteAnd another question. I see that REW can also apply FDW (via "IR Windows" dialog). Is this the same as Acourate's FDW? Did you have a chance to compare them?
DeleteMikhail, thanks for purchasing my book! The section on "Acoustics and Psychoacoustics of Room Correction" was peer reviewed by James (JJ) Johnston whose presentation you linked to :-) Wrt FDW, conceptually this is the idea as indicated in your post. To learn more about FDW, here is how DRC implements frequency dependant windowing It can get a little complicated, but the graphs help.
DeleteYes, REW and Acourate and others use the same FDW algorithm but different software packages offer more or less control over FDW. REW has a single variable where one can adjust the FDW width in cycles. Acourate and others allow FDW width controls in low and high frequencies both in the frequency domain and time domain (i.e. excess phase correction). So if you use the FDW controls in both REW and Acourate (only from the TD-Functions menu bar), you will see the same result. However, using Acourate Macro's you won't as Acourate also applies a proprietary psychoacoustic filter during filter generation. It's close, but there is a difference.
"James (JJ) Johnston began the presentation with some basic acoustic and psychoacoustic issues regarding room correction. He described the acoustic characteristics of a room, then some of the problems. If you want to correct the room-speaker combination for good imaging and timbre, exactly what do you correct and how? Long-term or short-term frequency response, or some combination? What about interchannel matching?
DeleteThe precedence effect, where the brain discounts arrival of signals after the first, helps. The first arrivals also are the primary cue for binaural localization, with the uncorrelated late arrivals sensed as "envelopment."
The upshot is to smoothly equalize the direct arrivals at high frequencies, remove the worse peaks in the lows, and not touch deep low frequency dips. Broad dips can be equalized. **Do not correct overall phase and magnitude with the processing** or you will cause unpredictable pre-echo. You need to equalize gain and delay from each speaker just for the first arrival, which locks in the spatial cues. Frequency response correction for the first arrival should be applied - exact for better speakers and relative (i.e. between channels) for sets of poor speakers. Correct the worst problems with the speakers or room, and leave other problems alone."
There must be some serious misunderstanding of what is being said by JJ vs what DRC is being done here by Acourate et al with **LP "target curves" **.
The only thing that should be touched >500hz +/- is the direct response, best done with quasi-anechoic systems at 1m, or 2m if one has an actual anechoic chamber. NOT IN ROOM at LP. Only peaks in bass (in room) should be flattened. Judiciously.
Of course I've said all this already ;-)
Re: The only thing that should be touched >500hz +/- is the direct response...
DeleteYes my man, that is exactly what is happening using Acourate et al frequency dependant windowing. It is a fancy way of saying measuring a quasi-anechoic response. As frequency increases, the time window decreases so that beyond 500 Hz you are only measuring the direct sound from the loudspeaker. No room reflections! Click on the frequency dependant windowing link I posted above your response to learn more.
There is no way it's doing that at the LP Mitch. I'll wait for Archs 1m measurements to see what it's actually done to the direct sound.
DeleteRegarding you "time alignment" link above. It sure looks like the FRs are some kind of summed response, once again at the LP. That's a huge no no also.
Thanks for all the information Mitch!
DeleteI think I see what AJ's point is. From the studies, and as seen on the Revel US2 example, flat on-axis anechoic response and smooth directivity "translates" into "preferred" target curve in-room at LP (due to absorption of HF by room furniture and other objects). If with FDW we can do "quasi-anechoic" measurements, why not to equalize the speaker to flat response above 500 Hz when measuring on-axis from 1 meter? Surely, one can't correct issues with directivity this way, but this can be dealt with by varying toe-in.
DeleteAnd actually, toe-in of the speakers is what can compromise the EQ done from the LP measurement. As the ANSI-CTA-2034-A standard quotes, "as shown in the Devantier (2002) survey, over half of those investigated had the prime listening position 10º to 20º off axis". Which means, EQ from LP will not be correcting on-axis response but something else. I see from the photos of Mitch's room that he is listening on axis, perhaps that's why EQ from LP works great for him. On Archimago's photos it seems that his speakers toed-in only slightly, if any.
So perhaps the most universal approach for EQ is first to perform time and frequency correction above the room transition frequency by doing on-axis individual speaker measurements from 1 meter distance, then put them into their final positions, check spatially averaged measurements around the LP, toe-in and/or apply broadband tilt if necessary, then perform relative correction to compensate for any room/furniture assymetries, and then correct below transition frequency.
"So perhaps the most universal approach for EQ is first to perform time and frequency correction above the room transition frequency by doing on-axis individual speaker measurements from 1 meter distance, then put them into their final positions, check spatially averaged measurements around the LP, toe-in and/or apply broadband tilt if necessary, then perform relative correction to compensate for any room/furniture assymetries, and then correct below transition frequency."
DeleteClose. The 1m +/- quasi measurements must be out to at least 60 off axis. The EQ must then be done carefully, if the deviations are linear, which the should be with a *well designed* speaker like an S8, i.e. a dip on axis, corresponds with a dip off axis. Vs a dip on/peak off or vice versa. There one would have to be careful. 30-40yrs of measuring would come in handy ;-).
Btw, it not what "AJ" is saying, its what the science/evidence shows so far.
Mikhail, check out the chapters on driver linearization and time alignment in my book as that is more or less the same approach as what you describe. But with FDW, I find I can achieve virtually the same results measuring at the LP, so why bother with the extra steps. Btw, I should point out that the speakers used in the book are employing constant directivity waveguides beyond 500 Hz. Given an eq on axis, we already know what the off-axis frequency response is going to be with a "constant directivity" device :-) The off-axis results can be seen in the filter design verification section of the book.
DeleteIt must be said that AJ is a loudspeaker manufacture. What loudspeaker manufactures don’t want to acknowledge is if their speaker is eq’d the same as speaker B, then they pretty much sound the same. You can listen for yourself with binaural recordings I made of eq’ing a KEF LS50 the same as a large JBL Cinema loudspeaker and comparing the two Other than the directivity differences, which is about as different as one can get with these two speakers, they sound remarkably similar in the binaural recording from a frequency response perspective. The big difference being how much room sound is mixed in with the direct sound. It is audible on the binaural recordings switching back and forth every 10 seconds when listening to the track comparing the two loudspeakers.
I have had dozen speakers through my room over the last 6 years, some have been reviewed with measurements on Audiophile Style, with no eq ;-) and then eq’d to the same target for comparison purposes. Aside from the heavy lifting below the rooms transition frequency, some required no eq (i.e. direct sound tone control) beyond 500 Hz, but most required a tone control correction as too much top end. This is because the speaker was designed using a single on-axis response in an anechoic chamber which translates to about +5 dB too much high frequency energy at the LP when comparing to a neutral target. This is in contrast to the CTA 2034 standard of 70 off axis measurements which as stated in the CTA spec to, “convey a reasonably good representation of how it may sound in a room based on its off-axis response and how this response affects the consumer’s experience.”
There are a group of loudspeakers I measured that fall into this category of too much HF at the LP. The Dutch and Dutch 8c and KEF LS50 required no HF tonal correction and hit the target out of the box. I.e. neutral sound. As pointed out in the subjective listening tests and objective measurements referenced from Sean Olive’s studies, Harman has found out what a good speaker sounds like in a typical listening room based on listener’s preferences. That’s why when you look at all of the Revel/JBL spinoramas, they are all pretty close. And had they included the predicted in-room response as outlined in the CTA 2034 A spec, then we would see that. But there is a reason why it is not included.
If you are listening to the same/similar frequency response, then the speakers pretty much sounds the same, as one can hear in the binaural recording comparison I linked to above. No speaker manufacture wants to hear that unvarnished truth.
Thanks folks! That's a very educating discussion.
DeleteI definitely noted that manufacturers are unsurprisingly biased towards the importance of their own products--after all, that's why they have started their business. Speaker makers say that of course it's the speakers that matter the most. Authors of DRC packages say that everything can be corrected using DSP. RealTraps founder Ethan Winer says that "Acoustic treatment is the most important thing you can have to achieve realism in a home listening environment". Microphone manufacturers remind that any microphone has its own transfer function and produces a different TD impulse response, so it's better not to use a cheap one. Which of course makes sense, because with DRC we are attempting to correct in time domain, too. And so on.
https://www.audiosciencereview.com/forum/index.php?threads/an-enticing-marketing-story-theory-without-measurement.7127/page-29
DeleteSep 30, 2019#563
watchnerd said:
This seems to be a common internet phenomenon.....people going full EQ overkill even above 500Hz to hammer everything super flat, then sharing the graph online, getting praised for it, others copying it, etc.
Wait, how does one "copy" another's EQ curve, with 2 completely different rooms, speakers, etc.??
Floyd Toole
Active Member
Audio Luminary
Technical Expert
Industry Insider
Sep 30, 2019#565
Spocko said:
Wait, how does one "copy" another's EQ curve, with 2 completely different rooms, speakers, etc.??
It cannot work as advertised, but that doesn't stop people from doing it. It has its origin in the decades old notion that steady-state room curves are definitive of sound quality. We now know that it is simply not true above a few hundred Hz, but it has validity at low frequencies, which account for about 30% of the factor weighting in subjective evaluations of sound quality. So, because of that "room EQ" might be beneficial. The unfortunate part is that in going full bandwidth it is possible to degrade the performance of well designed loudspeakers.
Toole, F. E. (2015). “The Measurement and Calibration of Sound Reproducing Systems”, J. Audio Eng. Soc., vol. 63, pp.512-541. This is an open-access paper available to non-members at www.aes.org http://www.aes.org/e-lib/browse.cfm?elib=17839
Seemingly endless promotion of "room EQ" algorithms - a for-profit exercise - is partially responsible, aided by human nature which is inclined to believe a good story. It is an ingredient in "faith based" audio - if you believe it, you just might hear it. Even though some EQ exercises "sound similar" does not mean that any are as good as they could be - perhaps the important similarity is at low frequencies."
LOL at comparing speakers over ones headphones with a binaural recording. I guess it's one step up from Youtube I suppose ;-).
Maybe confused with Binaural Scanning?? http://seanolive.blogspot.com/2009/03/binaural-room-scanning-powerful-tool.html
Not the same!!
I guess if I had a book to sell....
BTW Mikhail, I'm more of an AES/Toole/Salmi/JJ et al papers reader, so you'll have to forgive if I've never read much of Mitch's work before, including what he linked above https://audiophilestyle.com/ca/reviews/kef-ls50-david-versus-jbl-4722-cinema-goliath-speaker-comparison-with-binaural-recordings-r768/.
DeleteI did briefly listen to the binaural recording for a chuckle. I've now had time to read it in entirety. Including the comments, so I see Arch has too.
Did you notice where Mitch placed the mic when referencing against the KEF NRC measurements?
(Hint, it wasn't at 3m with FDW ;-) ). Hmmm.
Oh and this gem of a comment Pg2:
"As Floyd Toole and Sean Olive have often said, one cannot eq directvity. So both the measurement and correction software are "blind" to directivity, as we are measuring and correcting the sound power in the room (i.e. steady state response)."
Can you guess who said this? ;-)
AJ
Soundfield Audio
Sorry for a late reply, I've done some experiments measuring my LXmini speakers using Acourate and REW, and comparing the results between them and the free field measurements presented by SL here: https://www.linkwitzlab.com/LXmini/Design.htm
DeleteWhat I have realized is that although FDW indeed gets rid of the reflection components very well, the measured and processed response still remains an in-room response, not an anechoic response, not even "quasi-anechoic".
The visual difference is that the true anechoic response of a good speaker is flat. And an in-room response of the same speaker is sloped down as per "target curve" recommendations. This is described in the ANSI-CTA-2034-A, and Toole's works and elsewhere. Also, an in-room response includes low frequency anomalies due to room modes while an ideal anechoic response doesn't (I'm saying "ideal" because real anechoic chambers are only "anechoic" down to a certain frequency).
So I must correct my idea from the "17 October 2019 at 21:59" comment that one could try equalizing a speaker to an ideal response using FDW-treated measurements. Not really. What one can do with them, is only to adjust the speaker response in the room where it is being measured. Bring this speaker to a different room, and it will measure differently even above the transition frequency.
My conclusion from this is--yes, digital room correction products can help making speakers to sound better in one's room, but the corrections made don't really improve the speaker itself, they merely try to mask the deficiencies of the speaker. It's possible to make an "universal" correction of a speaker to some extent with DSP, but this requires making real anechoic or "quasi-anechoic" measurements, not in-room measurements.
Mikhail, cool! It would be great to get more details on what you did in your experiment. What FDW settings were used? You can see how to calculate the milliseconds here: https://www.audiovero.de/acourateforum/viewtopic.php?f=11&t=14
DeleteIf the FDW above the transition frequency is too many cycles, then for sure the room is in the picture. Cycle values of 1 or less should be used in the high frequency FDW. Of course, we want the room below transition to definitely be in the picture, as Floyd Toole agrees that eq below transition is not only recommended, but mandatory. I also noted elsewhere that the FDW used in Acourate, in Macro4 is not the same FDW used by REW. You need to use Acourate's FDW from the TD-Functions menu and then it will be the same as REW, except REW does not allow for different FDW settings like Acourate where one can use different values for low frequencies versus high frequencies and therefore it is still not an apples to apples comparison.
As you know, there are several approaches to measurement/correction. One approach is to linearize the drivers to be flat nearfield as you can find that procedure in my book. Also, one can take "quasi-anechoic measurements with Acourate and in-room. See the chapter on Beamforming in my book that uses this paper on Quasi-Anechoic Measurement of Loudspeakers Using Beamforming Method: https://www.researchgate.net/publication/272498144_Quasi-Anechoic_Measurement_of_Loudspeakers_Using_Beamforming_Method
You can also see the results of that procedure and what it did to the response of my speakers in that chapter.
Since we are talking ANSI-CTA-2034A that folks can download for free at: https://www.cta.tech/Research-Standards/Standards.aspx?search=2034
Checkout the predicted versus measured in-room response in Figure 11 on page 37. They are identical...
It's a pleasant surprise to get a reply :) You must be subscribed to comments here, Mitch.
DeleteI've published the graphs and setup description here: https://melp242.blogspot.com/p/experiment-measuring-lxmini-with.html
I didn't use FDW in REW, and used default FDW parameters for Acourate.
My goal is to try to linearize LXmini as a whole (not just drivers) after some changes to the crossover that I'm experimenting with, and do that as if I had an anechoic chamber, that is--universally, not in situ as DRC assumes.
As I understood from Vanderkooy & Lipshitz paper, at least above the transition frequency gated measurement should be quasi-anechoic. I will check out the beamforming method too, thanks for the link!
Mikhail, thanks for sharing your experiment. Unfortunately there are two issues that make this an apples versus oranges comparison. As mentioned before, you cannot use the Room Macro's for the type of measurement/correction you want as disclosed in your post above. Just like "Excel" macros where functions are combined, so it is in Acourate. Macro 1 "amplitude preparation" is actually applying these two functions together sequentially: TD-Functions/Psychoacoustics and TD-Functions/Frequency Dependent Window. So as mentioned above, don't use Macro 1, but rather just TD-Functions/FDW.
DeleteThe 2nd issue is the FDW window size. Since you used the default 15/15 cycles, means that the window at 1 kHz is open for 15 milliseconds, which is of course letting room reflections in. You need to reduce the size of both windows so that (much) less of the room is in the measurement. I provided a link in my previous reply to the FDW math so that you can work out the window sizes to "window out" the reflections based on your rooms geometry relative to the speaker location and mic position. That should get you much closer to SL's free field measurements. Good luck!
Thanks Mitch! I agree, attempting to use Macro 1 instead of elementary TD functions was a mistake. Glad you have catched it!
DeleteThanks a lot for your time and suggestions!
Mitchco18 October 2019 at 09:14
Delete"It must be said that AJ is a loudspeaker manufacture. What loudspeaker manufactures don’t want to acknowledge.."
LOL
https://accuratesound.ca/
I wonder what must be acknowledged now?? ;-)
Very interesting document and discussion !
ReplyDeleteI'm running a 3 way active homemade speakers using a DEQX digital crossover with EQ room correction. This makes possible to digitally correct each driver's measured frequency response as well as giving them perfect phase and time alignment with Linear Phase filters with slopes up to 300 db/octave !
I call this the 21st century 3.0 speaker design: 1.0 being the old passive crossovers (with all it's inaccuracies), 2.0 being the active analog crossover like the Marchand XM-9 (which I had, a step better !) and 3.0, the "space age" active digital with all its corrections possibilities done in the digital domain without degrading the audio. What you did wit Accourate is quite similar to what I can get with my DEQX and it worth the effort !
Based on the result I'm having with my DIY speakers (which I would rate as much better than many very expensive commercial speakers, but yes, I know, I'm not neutral !) I firmly believe that active DSP speakers is the way to go: you can calibrate them to better fit your room, the last step that is so much important !
Didn't understand if Acourate is managing multichannel (i.e. DTS or 5.1) audio.
ReplyDeleteAnyone?
Acourate only deals with 2 channels, but there is a way to work with multi-channel setup by aligning channels in pairs, see https://www.audiovero.de/acourate-wiki/doku.php?id=en:wiki:anhang:anleitungen:the_repeated_stereo_pairs_method
DeleteIn acourate you design the number of channel you want may be 2...6...100...
ReplyDeleteI use acourate and acourateconvolver now for many years.
I use acourate for stereo and for surround, 6 channels (5.1). Its the same setup. If the file is stereo it plays stereo. If the file is surround it plays surround.
The filters are designed with acourate for 16 channels.
-Front 2x: 4-way crossover Linkwitz LX521.4
-Side: 4 Channels (2x2)
-Headphones: 4 channels
A total of 16 channels. all 16 channels are filtered (convolution) with acourateconvolver in realtime with a PC Intel-I7.
Output is to the audiointerface RME UFX and RME UFX II.
With acourate you can, at least I know and use:
1) measure 2channels at a time with multiway crossover (means for example a 4-way speaker you measure in one shot 8 channels).
2) Filter design, EQ etc.
3) crossover design (any even complicate ones, Horbach Keele etc.)
-Time align speakers
-RoomEQ
For measurement I use the same audiointerface (RME UFX, RME UFX II).
The acourate filters can run in acourateconvolver, JRiver, Roon etc.
Acourateconvolver is a convolution engine which runs on a PC an can run 3x9=27 filtersets. A filterset is a set of n-channels, my setup has 16 channels. With a click you can switch all 16 channels to a different filterset!
Now whole procedure of my setup:
1) Desing filters, crossover, RoomEQ, Timealign in acourate. Measure to verify in acourate.
2) Load the filters in acourateconvolver.
My signalflow for musireproduction:
1) Harddisk (flac-files, stereo or surround)
2) Jriver as musiclibrary and player
3) JRiver outputs to acourateconvolver
4) acourateconvolver filters all the channels in realtime
5) over USB to the audiointerface RME UFX, UFX II
6) from RME UFX, UFX II to poweramps
7) Poweramps to speaker
8) Speaker to listener
Now how to measure a multiway speaker in one shot?
Acourate together with acourateconvolver lets yo measure and run all filters of a multiway speaker in once.
Design the crossover filter in acourate. Then load the filter in acourate (for LX521.4 these are2x4 filters).
Then in acourate select acourateconvolver as output.
The signal then is generated by acourate (Logsweeprecorder), then sent through acourateconvolver filters , RME, poweramps speakers and the acourate mic pics the sound up.
I do not know if acourateconvolver has a channel limit, however up to now I didn‘t run into a limitation.
Hi Archimago,
ReplyDeleteGreat posted. Just the posted i needed. I'm curious what your observation/results are with Mathaudio Room Eq if possible in comparison with orher software you use. I'm using it and it realy changed my listning experience quite a lot. Other software as REW i find it to complex one of the big advantage of Mathaudio Room Eq.
Regards Frans
Yeah, sounds good AJ, when I have time, I'll try that 1m 0/30/60° measurement to see. The Signature V3 does appear to be a very different beast even based on the impedance graph!
ReplyDelete- Still no 1m "speaker" measurements to show the damage done >500hz by "Room Correction"???