Jump to content
  • entries
    19
  • comments
    286
  • views
    15845

What is the tone quality of your audiophile system?


mitchco

Using the chart below, what subjective terms would you use to describe the tone quality (a.k.a. tone color or tone balance or timbre http://en.wikipedia.org/wiki/Timbre) of your sound system at the listening position?

 

Frequencyrange.jpg

 

 

 

This chart, used by permission from Bob Katz, Mastering Engineer extraordinaire, http://www.digido.com/ shows the subjective terms we use to describe excess or deficiency of the various frequency ranges.

 

As an aside, there is an important underlying concept here that needs some explanation. That concept is that audio is both an art and a science and that there is correlation between the two. In the case above, the art is the subjective descriptors and the science is the frequency range that the subjective descriptors map to.

 

It is my belief and experience (more on that later) that shows there is a direct correlation between the art and the science of audio. Put another way, how does it sound using our ears can be mapped to what is being measured and vice versa. I would suggest that this is a balanced view of our hobby where art and science come together and it's not just one or the other extreme.

 

Looking at the chart also provides us with common terminology in describing our sound systems or the sound of a particular song. So when we say the song sounds "bright" we know that the frequency range for this is likely 3KHz to 10Khz and has been lifted 1 or more decibels relative to the other frequencies on the scale. This can be measured using a real time spectrum analyzer VST plugin http://en.wikipedia.org/wiki/Virtual_Studio_Technology on either the Mac or PC (if your music player software supports VST plugins) using Blue Cat's free FeqAnalyst for example: http://www.bluecataudio.com/Products/Product_FreqAnalyst/

 

BlueCatFeqAnalyst.jpg

 

Listening to various songs and watching the spectrum analyzer at the same time, you will soon start correlating songs that sound bright with songs that sound warm for example.

 

Let's get back to tonal quality. How do I train my ears to understand the tonal quality of my sound system? Nothing like experimenting to help assist with what is going on. As mentioned above, if your playback software supports VST plugin's there are literally thousands of plugins available for both the Mac and PC platforms: http://www.kvraudio.com/allpluginsononepage.php Note that most modern recordings and masters have been processed through a Digital Audio Workstation (DAW) and uses the same VST plugin technology. This has been going on since the mid 90's http://en.wikipedia.org/wiki/Digital_audio_workstation and is very likely that most of the music you listen to has been processed through a DAW, with several VST digital plugins and analog processing chain.

 

With the advent of unprecedented computer processing power for cheap, software designers of music players http://www.jriver.com/audiophile.html can take advantage of 64bit processing that is way beyond our hardware output capability of 24 bits and therefore has 0 impact on sound quality. You should have no fear in using any of these VST plugins as they will not affect the sound quality of the audio signal passing through them.

 

Here is a free parametric equalizer that is available on the Mac and PC that you can download and install. http://sonimus.com/site/page/downloads/ The beauty is that you can play with the controls while at the same time listening to the sound and hearing the effect in real time. You can then correlate the sound you hear at the frequency range and the corresponding subjective terms as described in Bob's chart above.

 

eq.jpg

 

With this eq, you can roll off the extreme frequency ranges and listen to the effect on your speakers or headphones. Note the center slider. What I like to do is turn up the boost and sweep the frequency range while listening to the tonal differences in real time. It is a real ear opening experience. Then you can start correlating the sound of your audio system not only with subjective terms but exactly at what frequencies. You can also correlate by flipping back and forth from the eq to the frequency analyst and correlate with what you are adjusting to what you are seeing to what you are hearing all at the same time.

 

Where is this leading to? Well, the frequency response of your sound system at the listening position not only describes tonal quality (or timbre) but also directly correlates to the sound stage presented at the listening position. This is an important concept to understand. For example if there is too much high frequency arriving at the listening position, not only is it a bit bright sounding or at the frequency extreme more "air", it correlates to the soundstage being "upfront" or too forward or lacking depth. Conversely, if the high frequency roll off (or slope or shelf) is too much, then not only does it sound "dull" but the soundstage is too far back or "distant".

 

Is there an optimum tonal balance or timbre or frequency response at the listening position that also has the right soundstage depth? Yes there is. Here is an excellent paper, complete with subjective descriptions and scientific measurements describing such a frequency response we should strive for at the listening position: http://www.bksv.com/doc/17-197.pdf

 

Specifically:

 

bandk.jpg

 

It has been my experience that this target frequency response curve measured at the listening position, not only provides the best tone quality from my sound system (not too bright or dark, but just right) but also the perfect soundstage (not too forward or too far back, but just right).

 

Sean Olive is another double blind test. (Thanks hulkss for the link!) Interestingly enough, the virtually same target curve as the B&K curve above is the preferred spectral response at the listening position. It is the top curve in this comparison:

 

Harman.jpg

 

And here is the measured frequency response of my system at the listening position:

 

jan29leftspeakerwithaudiolensedrc.jpg

 

Using the B&K target curve as reference, my system is easily within +-3db from 20Hz to 20KHz at the listening position. I have experimented with dozens of target curves, including flat, which is etch a sketch bright with the soundstage in our face. I always end up coming back to the B&K curve.

 

Note all three graphs are virtually identical. This is no coincidence. If you read Bob Katz's excellent book, Mastering Audio - The Art and The Science, you will note the frequency response of his monitors in his mastering studio also exhibit the same target curve. Finally, when I was working as a recording/mixing engineer at several studios, there was always at least one set of monitors "calibrated" to the B&K curve above in each control room.

 

Point is, if you want the best tone quality and soundstage from your audiophile system, calibrating your speaker to room interface to a reference target, like those described above, will give you the best possible result.

 

Most people that hear my rock and roll sound system http://www.computeraudiophile.com/blogs/Rock-n-Roller-s-Guide-Designing-Audiophile-Sound-System are surprised at the soundstage. Casual listeners comment on how they can easily hear the different layers of sound (i.e. the mix) whereas they cannot hear that on their own sound system. That is the first comment I get before, wow, does it sound clean, or punchy or whatever other subjective terms are used.

 

I wrote a series of six articles detailing on how I went about achieving perfect timbre in my system starting with: http://www.computeraudiophile.com/blogs/Hear-music-way-it-was-intended-be-reproduced-part-1 Given that the speaker to room interface has the biggest impact on tonal quality of your sound system, the best investment you can make to optimizing your existing system is to measure its frequency response at the listening position and compare it to the optimum frequency response as described in the B&K article.

 

How to do this? For software, REW is fantastic: http://www.hometheatershack.com/roomeq/ and well supported. All you need is a calibrated microphone. It must be calibrated. One such mic is http://www.parts-express.com/pe/showdetl.cfm?Partnumber=390-801 and another is kit so you don't have to fuss with phantom power: http://www.content.ibf-acoustic.com/catalog/product_info.php?cPath=30&products_id=35 I use the latter and have had excellent results.

 

What's my point in all of this? Using free tools (save the measurement mic) you can experiment with tone quality to see what frequencies you may have too or too little of and correlate that with what you are seeing in the way of spectrum analysis and frequency response graphs and compare to a known standard. That way you can achieve the best tonal quality of your existing sound investment. Given that the speaker to room interface effects timbre and soundstage the most, with little effort as described in this article, can produce huge returns on your existing sound investment.

 

Happy Listening!

 

Mitch

17 Comments


Recommended Comments

I think it is very dependent on what loudness level one is listening to what recording. I find exaggerated highs are very tiring but more of a problem is the raised low frequencies in many recent recordings. I my view the best compromise for most recordings is a level frequency response with a slight reduction of highs of about 4db at 20kHz. Besides it really depends on the room accoustics!

Link to comment

I often use the parametric EQ built into my Metric Halo DAC to tame bright classical recordings.

 

 

 

I find that a dip of 1 to 3 dB centered at 3 KHz (sometimes 4 KHz) usually does the trick. By not attenuating the top octave, you don't sacrifice air.

Link to comment

Mac OS X includes parametric EQ (and various other) audio plugins, but to access them you need a music player program that can load AU (audio unit) plugins, such as Pure Music, Decibel or Fidelia.

 

 

 

Some audiophiles recommend the FabFilter Volcano ($150) plugin as having better sound quality than the parametric EQ built into OS X. I don't have an opinion because I use the parametric EQ built into my Metric Halo DAC.

Link to comment

I just checked my parametric EQ presets. I most commonly use a dip centered at 4 KHz, almost never 3 KHz as I erroneously stated in my 2nd preceding post.

Link to comment

aloha monteverdi

 

 

 

Re: I think it is very dependent on what loudness level one is listening to what recording.

 

 

 

85 db SPL C weighted is commonly stated as the level which our ears hear the flattest: http://www.independentrecording.net/irn/columns/gstep/index.php?id=84

 

 

 

Therefore, it is no coincidence that most mixing and mastering engineers (and audiophiles) critically listen at 85 db SPL. One prominent mastering engineer, Bob Katz, has proposed a formalized way to do this: http://www.digido.com/honor-roll.html

 

 

 

Re: Besides it really depends on the room acoustics!

 

 

 

Oh yes, absolutely the # 1 factor on tone quality is indeed room acoustics! If you look closely at the frequency response chart I supplied above, you will note I use Audiolense Digital Room Correction (DRC) software.

 

 

 

Here is an overly of the measured uncorrected frequency response (purple) and the corrected frequency response (blue) at the listening position:

 

 

 

jan29rawanddrcfreqrespoverlay.jpg

 

 

 

It is an ear opening experience to measure the frequency response of your speakers at the listening position to see the very audible impact room acoustics has on tone quality. As mentioned in my article, I wrote a series of 6 blogs posts on room acoustics and how to achieve the best possible sound one can get from their existing room.

Link to comment

I have an unusual listening room with 20' ceilings in the central area and open/attached dining and entry areas plus a loft. I would like to analyze my listening area and make corrections via the LIO8 which I use as a ADC/DAC. I am totally new to this and would like guidance on hardware/software needed and methodologies to do it. I'm just a hobbyist so would like to keep costs reasonable... free is good. I looked at the XTZ room analyzer as reviewed in Stereophile but it's Windows only and I'd like to stay in the Mac family. Any guidance, links, personal experience is welcome. Thanks, Rod

Link to comment

I believe REW runs on the Mac: http://www.hometheatershack.com/roomeq/ Scroll down to the download section near the bottom of the page.

 

 

 

Just under the download section is the help files, which provides tutorials on how to take measurements.

 

 

 

There is also the REW forum to ask questions and get assistance. Great group of folks on the forum: http://www.hometheatershack.com/forums/rew-forum/ On the forum, is another link with a walkthrough guide: http://www.hometheatershack.com/forums/rew-forum/11707-room-eq-wizard-rew-information-index-links-guides-technical-articles-please-read.html

 

 

 

REW is free, works fantastic, and easy to use.

 

 

 

Aside from a mic stand (a camera tripod works too if you have one) and some cabling, a calibrated measurement microphone is required.

 

 

 

I use http://www.content.ibf-acoustic.com/catalog/product_info.php?cPath=30&products_id=35 and had excellent results. You could also try http://www.parts-express.com/pe/showdetl.cfm?Partnumber=390-801 but you will need a mic preamp with phantom power... I like the IBF Acoustic "kit" as both the mic and the mic preamp that comes with it are calibrated.

 

 

 

With the REW software and calibrated mic, you can now take acoustic measurements. I like to correlate what I hear with what I measure and vice versa. It is great to experiment by moving the speakers/listening position around and listening and taking measurements. It is pretty quick to find the best spots in the room that sound and measure the best. That is if you have the flexibility to move the gear around.

 

 

 

From there it is a matter of how far you want to take your hobby with respect to adding acoustic treatment and/or digital room correction to further smoothen the frequency response of your speaker to room interface.

 

 

 

Others may chime in that have specific experience on the Mac.

 

 

 

Cheers, Mitch

Link to comment

You can get away with using a Radio Shack sound meter with the REW software as long as you're using it for the lower frequencies only. The people at HometheatreShack say it's fine for that. And I can verify, as I've used it, along with REW and an EQ for my subs and my system sounds much better.

 

 

 

The REW software is very good btw, and free. And it comes with a thorough downloadable manual. You will however need something like a usb dac that can record, or a similar sound card. Fortunately they can be had fairly cheaply, less than $100. I have a Cakewalk model which is just fine.

 

 

 

-Chris

Link to comment

the Rad Shack mic, make sure you install the mic calibration file from: http://www.hometheatershack.com/forums/downloads-area/19-downloads-page.html

 

 

 

From your comment, looks like you already have an ADC, so you could plug the Rad Shack mic directly into your LIO-8 with the right connector.

 

 

 

However, if you want to measure the overall tonal quality of your system, you would be better off with a Behringer ECM8000 or Dayton EMM-6 or one of the measurement mics I mentioned earlier.

 

 

 

If your LIO-8 has the mic pre option, then you won't need a seperate mic preamp/phantom power.

 

 

 

Cheers, Mitch

Link to comment

Thanks for the perfect reply... sorry for the slow response (been out of town). The Dayton mic looks like a perfect fit, especially since it covers the full frequency spectrum and is individually calibrated. Question... the LIO-8 uses the DB25 connector for the mic inputs. I have had a DB25 to RCA cable made to accommodate my turntable feed. The LIO-8 has phantom power but with my existing set up I would have to use a XLR/RCA adaptor to connect to the mic input. Would this disable the phantom power? Do I have to have a special DB25 to XLR cable made? I need a long mic cable to do the measurements so what's the best approach? Perhaps Bob Stern could chime in since it seems like he's done the same thing with the same equipment.

 

 

 

I use an iMac Intel Core2 duo as the dedicated music server (with attached NAS [Drobo] via USB).  The iMac is always attached to the LIO-8 (via firewire) but wondered if playing test tones and taking measurements at the same time would be possible? The measurement response would be traveling in opposite directions on the same firewire simultaneously? To do the initial SPL calibration I'm assuming that I can use the pink noise test tone on the computer, output (via firewire) through the LIO-8 to the speakers, at the same time as I record the measurements via the LIO-8 to REW on the same computer?? Sorry I can't explain my confusion better. Kinda seems like sending water both directions thru the same hose.

 

 

 

The REW software looks like it will work perfectly. Have also been reading about Fuzzmeasure and IK Multimedia ARC... any familiarity with these? Pros/cons? I wish there was a "audio measurement for dummys" book. I've been reading online and following links but it's all so new. I'm slow but I'll get there.

 

 

 

Cheers and thanks for all your help.

 

Rod

Link to comment

Rod, keep going you are almost there.

 

 

 

I am not too familiar with the LIO-8, but it sounds like you will require a DB25 to XLR connector for the mic. This link points to Redco that should know how to to do this: http://www.computeraudiophile.com/content/RCA-out-instead-XLR-LIO-8-possible-or-even-recommended-no-loss-Sound-Quality

 

 

 

You should have no problem using REW to output pink noise and swept sine wave, while at the same time input the response from the mic. This is should be a routing setup in LIO-8 mixer. Have a look at: http://www.hometheatershack.com/forums/spl-meters-mics-calibration-sound-cards/10001-rew-cabling-connection-basics.html

 

 

 

I don't know about the other software you mention, I have had excellent results with REW

 

 

 

Let us know how it goes.

 

 

 

Cheers, Mitch

Link to comment

The two TRS jacks at the right edge of the front panel are identical to (connected in parallel with) the Line 1 and Line 2 inputs on the rear panel DB-25 connector.

 

 

 

I have not connected a mic, so for instructions on turning phantom power on or off you'll have to read the MIO user guide, which you can download from the MH website.

Link to comment

Though speakers are not very flat in their response, but hopefully everything else is. I never use any form of tone control, equalizer, room correction etc, as I assume the recording studio knows what it is doing. And even if it does not, how can I tell that my guesses, or even microphone measurements, to try and make it 'correct' make the sound what they intended but did not have the skill to achieve?

Link to comment

Thanks Mitch. I had not been active on CA for a while when I joined this thread and I wasn't familiar with your blog. I have now read it in its entirety (as well as your posts to other threads) and want to say kudos to you for all your hard work/time and the benefit it has for us neophytes. Very much appreciated!!

 

 

 

In the past couple days I've spent a lot of time doing reading/research. I confirmed with Metric Halo that a DB25 to XLR adaptor cable will be needed. I have used Redco before and will be contacting them for a new adaptor.

 

 

 

I thought I had finalized my choice of mic but have delayed after further research. Initially I thought the Dayton EMM-6 would be the best choice because of their individual mic calibration. Then I noticed that the Behringer ECM8000 had a lower freq. extension and might be a better fit for my speakers, which the manual says FR is from 16 Hz - 45 kHz +/- 2dB. Checking with Behringer, the ECM8000s are not individually calibrated and the website cross-spectrum.com, which does individual mic calibration (for about $30 more than the going online price) shows a graph with multiple overlaid FR curves that show the FR tracings all over the place. I can appreciate that the better the linearity and calibration of the mic leads to better results within the REW software... but how important is it? Many folks use RadioShack SPL meters with good results. Is it worth the extra $30 to have the mic individually calibrated? Then the question of how to enter that info into REW? It seems the calibration data can come in different formats... FDR, .CSV, Excel spreadsheets, or perhaps just a graph on paper. How does that data get into REW? I hope not manually!! Which brings me to the mic cable... are the ultra low noise cables needed or recommended? To much info and not enough understanding!

 

 

 

Lastly, I find that I prefer the sound/soundstage when the tops of my speaker panels (Infinity Beta) are tipped forward. The speakers in the vertical position raise my ideal listening height and shrink the sound stage. I can imagine this creates havoc with the time domain. Can REW correct time domain problems or do I need something like Audiolense?

 

 

 

Thanks again for all the guidance/info... I'm slow but I'll get there!

 

Cheers, Rod

Link to comment

MH confirms I'll need a DB-25 to XLR adaptor to use the mic in of the LIO-8. When I get to the stage of constructing my EQ within MIO I hope you can share your experience. Thanks!

 

 

 

Cheers, Rod

Link to comment

Hey Rod, thanks for your kind words.

 

 

 

A calibrated mic is indeed critical. In my speaker to room interface frequency response tests, I have found a few dB up or down makes a substantial impact on the overall tone quality (i.e. timbre) and soundstage.

 

 

 

Relative to a "target" frequency response curve like either the B&K target or the similar (i.e. almost identical) to Harman target (both are in the article above), a few dB up slope or down slope makes the sound either too bright or too dull. Small difference make big changes.

 

 

 

I think if you were just tuning the low end and used a Rad Shack meter and corresponding calibration file would work, but not ideal. If you are interested in the full range tonal balance, then a calibrated mic, from 20Hz to 20KHz is the ticket.

 

 

 

No worries on the calibration file format. There is a "standard" format for this and REW accepts the format as simple as loading the file. Since it is standard, any of the folks doing mic calibration will supply that file to you np.

 

 

 

Don't worry about the quality of the mic cable and don't spend a fortune. I just used regular mic cable and all good.

 

 

 

Yes, REW can help figuring out time alignment, but that is another topic. First step is to measure the frequency response to see how well your nice Infinity's interface with your room. I have another post coming out shortly that walks through the steps for what you are about to embark on.

 

 

 

Edit: Here is that walkthrough: http://www.computeraudiophile.com/entries/173-Speaker-to-Room-Calibration-Walkthrough

 

 

Cheers, Mitch

Link to comment



×
×
  • Create New...