Saturday 9 November 2019

RETRO-MEASURE: Klipsch Forté I speaker impedance. (And it looks like Qobuz responds to Amazon...)

Well guys, I'm still working on the Autoranger amplifier test system looking at standardization of the measurement technique and settings which is a pretty big job still, so this week I thought I'd show you some data I grabbed a few months back while working on the speaker impedance measurements series (Part 1, Part 2).

Every once awhile, I'll drag up an old device for a quick measure. Previously it was stuff like my old Sony SACD player, or the old laser disk player. This time, I went over to my dad's place to have a listen and measure his main speakers - the Klipsch Forté (original first version) made back in the 1980's until the early 90's I think.

Front and Rear of Klipsch Forté from the 1980's. Note the large 12" rear passive radiator.
I thought it was good to have a look at this device since there's been buzz lately about the latest model Forte III (Stereophile review, Darko review, Darko bought a pair). Also, a couple years back I ran into this Stereo Review article by Julian Hirsch on these speakers from 1986 as an example of what SR tests looked like back in the day:

While I didn't measure the speakers fully with frequency response, phase, group delay or look at off-axis characteristics (maybe one day if felt to be beneficial), I was intrigued by the highlighted portion about the impedance.

You can check out some history in the Stereo Review Wiki. Remember that SR was known for Julian Hirsch and the objective perspective they focused on when doing product reviews. Admittedly I would have been too young to have cared about hi-fi back when this article came out (I was much more interested in the advent of the home computer then!), but I read a number of the "Hirsch-Houck" reports into the early 1990's (you can read almost all backissues of SR here).

It's hard not to notice the difference in style and content comparing this to typical reviews these days in Stereophile or TAS. No need to make any bold declarations about which is "best" since that probably would be relative, but to me this is like the polar opposite of how TAS reviews products with significant amounts of background information on the company, designers, subjective opinions and general description of features with little actual added independent factual content about performance. As I've said before, it's good that Stereophile aims for a balance between the subjective and objective (even though sometimes I think poor objective results should be better highlighted regardless of what the subjective reviewer believes).

Notice that he jumps right into the technical description from the first sentence. Also notice in the Stereo Review report that there was no hint of a discussion of the music used to listen to the speakers. No paragraphs spent about the reviewer's background, his (typically) ex-girlfriends, what his wife heard in the other room, what coffee blend he drinks, or what alcoholic beverage seemed to go well with the atmosphere. Not a very entertaining review by today's standards although to be clear, there is a treasure trove of data in there we might never see these days for much of the products out there.

I wonder why SR didn't include graphs though until the early 1990's with electronics reviews but not speakers. Perhaps it was difficult getting graphics included in the printing? It would have made results more impactful I think than reading the verbal descriptions...

For example, here is my dad's Forte I's impedance and phase response that could have gone with that highlighted paragraph:

Yup, true to the review, we see quite a mountainous impedance up at 2.15kHz >110Ω correlating with the crossover to the horn tweeter and it dips all the way down to ~5Ω in the bass frequencies as they described. There are a few "crinkles" in the impedance and phase graphs suggesting possible cabinet resonances which could color the sound. Certainly by today's standards with 100+lb speakers made of heavy and stiff materials, the Klipsch does feel a bit "hollow" with the knuckle-to-wood tap.

The other thing we can check is that the two speakers performed equivalently:

A very nice match of the impedance curves between the two speakers. My dad kept the speakers in good shape over the years.

Here's an interesting comparison I think:

That is how the Klipsch Forté I's impedance curve compares to some other speakers I have around here! Visually, I think it's a nice example of just how different speakers can be electrically. The question of course is how this load variation will impact how a specific amplifier will react in terms of impact on distortion, power and frequency response. I don't think it's unreasonable to desire an amplifier that's able to handle large variability with grace (vis-à-vis amplifiers with high damping factor that are more "load invariant").

Looking at the Stereophile review of the Forte III, the electrical characteristics look very different these days. Obviously much has changed in the design over the decades while keeping the name. (Haven't heard the Forte II; I know that Klipsch updated the midrange driver with that model, have not seen measurements.)

So, how do the old Forté I's sound? Well they certainly sound great for projecting an enveloping soundstage set-up in my parents' placed. The speakers are quite capable of strong dynamics with his 25W-into-8Ω Class D amp and his 38W Class AB tube amp (I'll likely measure these in the days ahead :-). However, the speakers do give up precision and resolution in favour of a smoother presentation. Also, I find the sound favoring midrange warmth than deep bass or the sparkle and "presence" of more modern speakers.

Overall the Forté I's are the kind of speakers I'd be happy to listen to for rock, live performances, Top-40's pop perhaps (like say Rolling Stones or Springsteen or even a favorite 80's guilty pleasure like Def Leppard :-). Intimate classical ensembles (like this lovely Schubert Ensemble Fauré recording) didn't seem to have the finesse I was hoping to hear, and pop crossover (eg. Lindsey Stirling's latest Artemis was one I had a listen to) or even EDM/electropop (Armin van Buuren's Balance is fun) just didn't sound adequately exciting with these. My sense is that the treble isn't as extended as most speakers these days and the deeper bass just could not connect viscerally (despite 12" woofer, I think adding a sub would be nice for that lowest octave). I haven't heard the Forte III, so perhaps the new model would sound superior in these regards.

Okay guys and gals. Enough retro discussion of Stereo Review and the Klipsch Forté I for today.


In other news, it certainly did not take long for the Amazon lossless/hi-res streaming announcement in late September with a lower monthly price tag to see that Qobuz has countered with their own moves. Looks like the lossy MP3 tier will be retired and their "Studio Premier" is now the "go to" plan at US$15/month and if you pre-pay for the year at US$150, the price goes down to an equivalent of US$12.50/month - nice!

Seems like a great move for the company and should capture customers who want to support a smaller, specialized service that's not the corporate giant Amazon! If I were living in the US and didn't already have a music streaming plan, I would almost certainly have brought out my credit card to lay down the $150 for the yearly subscription when I heard about this especially since I already am running Roon with Qobuz integration. Come on Qobuz, hope you guys can share the love for us in Canada soon :-).

The obvious question... When is Tidal going to be making their move into the $12.50-$15/month price bracket for the "Hi-Fi" CD-lossless tier? Probably a matter of days in order to keep up with the competition. While they're at it, maybe it's time to also dump unnecessary work and expenses like MQA in this highly competitive atmosphere? Why waste time and money doing data conversion when I assume the labels must be already supplying hi-res FLAC or other lossless equivalent to the other guys? Even if a small nominal amount, I assume Tidal is paying to license the codec. Furthermore, why bother with the development and software support of the proprietary decoder in the various apps? All of that had always been IMO destined to be as meaningless as chasing after the wind.

Have a good Remembrance Day / Armistice Day / Veteran's Day weekend 2019 and hope you're all enjoying the music.


  1. Yes! Great post on the value of speaker measurements. As I hunt for a pair of bookshelf speakers to replace my aging Paradigm Titans, the subjective reviews I come across leave much to be desired in informing me on speakers to consider. There are far too many variables in the reviews for them to be of much use to me. From source and amplification components that are usually much more expensive than my equipment (30-year old Yamaha receiver), to the room that is more than likely a dedicated audio room (mine is not), to the likes of the reviewer. Impedance graphs can help me in deciding if my receiver would be a good match with a set of speakers. Too many manufacturers rate their speakers as 8 ohm nominal, with maybe a minimum impedance rating (e.g.- 3.7 ohm). A graph can tell me if the speaker actually is closer to 4 ohms over a wide portion of the audio band. And frequency/sensitivity graphs are also useful. I may be over generalizing here, but most reviewers seem to value what is described as a "tipped-up" treble response. I despise sibilance and what I consider overly bright treble. I suspect I would prefer a speaker that shows a gentle treble rolloff between 2kHz and 20kHz. I subscribed to Stereo Review in the 80s (it's where I read a review of my Yamaha receiver) ans so wish that style of review would return.

    1. Most loudspeaker reviews, even those with measurements, don't do nearly enough. At the NRC then later at Harman, Drs. Floyd Toole and Sean Olive developed metrics that predict pretty well when a loudspeaker will 'sound good' to most people. These consist of extensive on and off-axis measurements....a 'spinorama' of measurements, preferably in an anaechoic chamber (a method later formalized by the industry in "Standard Method Of Measurement For In-Home Loudspeakers"

      To date there are only 2 places I know of online that generate or compile results of such measurements:



    2. Hey Steven. I am a frequent visitor of the SoundStage! sites. It can be very enlightening comparing their anechoic measurements of speakers they review with the published claims of the manufacturers, especially those that include claimed anechoic data. Thanks for the Speakerdata link. I had not seen that before.

    3. Hi JoethPop and Steve,
      Good discussion and excellent points. Absolutely. Speaker measurements whether it's the electrical side of the equation like the impedance curve or mechanical side with frequency response, time, and off-axis work are essential components that I wish we had more of in reviews.

      Of course knowing these things will not necessarily ensure optimal sound in one's room but they will certainly go a long way to ensure we make good choices when it comes to amplifier matching and adequacy of a product for our needs!

      I do hope in time more audiophiles can come to a place where there is recognition that reviews must change and that the sources where they get their news must do more than just subjective claims without adequate consideration of the myriad variables!

    4. I looked at the graphs on the Speakerdata site and found a graph (SPL/Freq. response) of the KEF R11 that I could compare to the measurements taken by Soundstage at the NRC chamber. Very interesting how even measured data can look very different. I have to assume that something in the measurement technique was different? The Soundstage setup only states that the microphone measuring position was on the tweeter axis, and the grill was off. Their measurements of the R11 show a fairly steep drop off of SPL that begins around 90Hz, falling about 10dB at 60Hz, followed by a sharp 7dB hump that peaks at around 40Hz. The KEF spinorama data on the other hand shows a fairly smooth slope that starts around 80Hz, down about 10dB at 45HZ. Both graphs show an admirable flat frequency response above 100Hz. I think I'm going to pick up a copy of Dr. Toole's book and educate myself some more.

    5. Hi Joe,
      I assume you're looking at this Speakerdata data:

      Yeah, no surprise that given the complexity of speaker measurements there will be many variables that need to be reconciled in order for the results to look similar. There was a comment a number of weeks back about the NRC and limitations with their low frequency measurements I believe.

      Highly recommend Toole's book. A nice reference.

  2. Wholly agreed on the matter of Tidal & MQA. It's a waste of effort and time, since in places with economy strong enough to be a matter of concern, bandwidth is increasing faster than the Pro-MQA forces could get their solution deployed.

    1. Hi mgraves,
      Nice hearing from you. Indeed a waste. Of course, their "solution" was never even a solution that I think audiophiles ever wanted or needed by the end of 2014 when it was introduced to the world :-).

      All I think MQA is to Tidal these days is an albatross around their neck having extra hoops to jump through to convert all that music to this proprietary codec while others around them freely use FLAC and have the ability to scale bitrate all the way to 24/192. (It would suck if they committed themselves to using MQA for a certain number of years.)

  3. As far as Stereo Review not publishing graphs? I think it was mostly a question of space, and the fact that their typical reader was not interested in that much detail. If you desired a full blown speaker review, I mean completely full blown, you'd go to Audio magazine, specifically Richard Heyser's work. Compare his 1986 Klipschorn review.

    1. Thanks for that mp,
      Nice look into what Audio did back in the day. Thorough use of ETC graphs and I hadn't seen those reactance/resistance plots before with the resulting "pig tails". Heyser was truly the rocket scientist of audio! Impressive.

      I believe he died in 1987. Just a year after that Klipschorn article.

  4. Hey, did you measured the real sensitivity of this Forte I? They claims 98db but i'm skeptical.