Do I Just Have Tin Ears?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Just another Moderator
Joined 2003
Paid Member
Yes almost all lenses have a sweet spot. Often between about F6 and F11, outside of these the lens peforms relatively poorly. The better lenses you can get away with it (as the degradation is not that great) but the cheaper lenses (the one I was talking of was a Tamron 28-200 XR) the difference can be quite dramatic! If I try hard enough I could probably dig up the scans I suspect, but this is getting way off topic.

The point of the anaogy was that whilst I could not see anything wrong in the 6 X 4 prints initially (they all looked the same). Once I had seen what to look for (based on the hi res scan of the negatives) I COULD see the difference in those very same 6 X 4 prints. It was there all along I just didn't see it until I knew what to look for.

Same goes for listening Once you know what a flaw sounds like you will be better able to pick up on it.

Attached are images of the first crossover I did for my MTM's and the final one. The difference here is not subtle! Most obvious difference is that I dealt with Baffle step in the new one (which I was originally planning on doing at line level), but the other difference is a change from 2nd order bessel acoustic rolloff to 4th order bessel acoustic rolloff with MUCH better phase tracking.

I'd be very surprised if you could not hear the difference between these two crossovers, regardless of experience :). However I lived with the first one for quite a while (probably about 12 months), because it didn't sound too bad. It was certain tracks like the ones I mentioned that sounded bad that got me looking to fix the problem, and a post from Lynn Olson that started me on the right track. Once I redesigned and implemented the new crossover the difference was very noticeable :)

So I guess where I was going with that was that perhaps the changes you are making whilst they are making the speakers different, they may not be changes that make the speaker substantially better or worse. Just sub optimal in differing ways.

Tony.
 

Attachments

  • old_crossover.png
    old_crossover.png
    37.1 KB · Views: 162
  • new_xover.png
    new_xover.png
    39.9 KB · Views: 160
Yes, Tony's right. For example, I have just co-incidentally today read a test on an expensive new Zeiss prime lens which has its sweet spot at f5.6. Tony, was your scan mentioned earlier of a negative? I'm getting back into film and want a recommendation for a neg. scanner. Apologies for the OT guys.

Sreten, I have that Cowboy Junkies album. Like that you can hear the church atmosphere in the recording. Where did you find the info on how it was recorded?
 
Last edited:
Just another Moderator
Joined 2003
Paid Member
Hi Stuey. Yes it was a negative scanner. A canoscan FS4000 Digital Film Scanners - Canon FS-4000 Scanner Review, Information, and Specifications I doubt that you can still buy them, and I'm sure there must be much faster units available now (that was it's main drawback, it is very slow in 4000dpi mode). Also I think that there are probably some units now with better dynamic range.

One thing I would highly recommend is VueScan which is third party scanning software. It's worth its weight in gold. It's a bit clunky (or at least was when I was using it 8 or more years ago) and takes a while to get used to but is much better than the software that came with my scanner.

BTW fuji provia 100 with a 4000dpi scan I think still beats pretty much any normal format digital for resolution ;) Not that I've shot any for a long while.

Sorry for the OT. Stuey if you want to discuss further PM me :)

Tony.
 
Yes almost all lenses have a sweet spot. Often between about F6 and F11, outside of these the lens peforms relatively poorly. The better lenses you can get away with it (as the degradation is not that great) but the cheaper lenses (the one I was talking of was a Tamron 28-200 XR) the difference can be quite dramatic! If I try hard enough I could probably dig up the scans I suspect, but this is getting way off topic.

The point of the anaogy was that whilst I could not see anything wrong in the 6 X 4 prints initially (they all looked the same). Once I had seen what to look for (based on the hi res scan of the negatives) I COULD see the difference in those very same 6 X 4 prints. It was there all along I just didn't see it until I knew what to look for.

Same goes for listening Once you know what a flaw sounds like you will be better able to pick up on it.

Attached are images of the first crossover I did for my MTM's and the final one. The difference here is not subtle! Most obvious difference is that I dealt with Baffle step in the new one (which I was originally planning on doing at line level), but the other difference is a change from 2nd order bessel acoustic rolloff to 4th order bessel acoustic rolloff with MUCH better phase tracking.

I'd be very surprised if you could not hear the difference between these two crossovers, regardless of experience :). However I lived with the first one for quite a while (probably about 12 months), because it didn't sound too bad. It was certain tracks like the ones I mentioned that sounded bad that got me looking to fix the problem, and a post from Lynne Olson that started me on the right track. Once I redesigned and implemented the new crossover the difference was very noticeable :)

So I guess where I was going with that was that perhaps the changes you are making whilst they are making the speakers different, they may not be changes that make the speaker substantially better or worse. Just sub optimal in differing ways.

Tony.

Ok So I put my foot in my mouth on that one. My only experience with complex lenses is from a Pentax DSLR camera I had that died (smoke came out of it) when it was less than 2 years old. I guess it makes sense that a complex lens would have a sweet spot. My experience is more with telescopes where low f numbers usually mean a sacrifice in edge detail.

Back to speakers, I'm sure I would hear the difference in your two crossovers. Problem is I would not know which one was the better one. I have three vehicles all of which just have the base audio system. They all sound totally different to me but I simply could not say which one is best. If forced to pick one I would choose my pickup truck which is a two seater with no back seat. I can say for sure that none of them sound nearly as good as the B4N system we are discussing here or my other recent build but thats comparing apples to oranges
 
Ok So I put my foot in my mouth on that one. My only experience with complex lenses is from a Pentax DSLR camera I had that died (smoke came out of it) when it was less than 2 years old. I guess it makes sense that a complex lens would have a sweet spot. My experience is more with telescopes where low f numbers usually mean a sacrifice in edge detail.
Telescopes aperture is different to camera lenses as the latter increases light passing through, where the former does not. In telescopes high aperture is required to keep the edge diffraction to a minimum and to increase depth of field. When talking astronomical distances this is critically important.
Back to speakers, I'm sure I would hear the difference in your two crossovers. Problem is I would not know which one was the better one. I have three vehicles all of which just have the base audio system. They all sound totally different to me but I simply could not say which one is best. If forced to pick one I would choose my pickup truck which is a two seater with no back seat. I can say for sure that none of them sound nearly as good as the B4N system we are discussing here or my other recent build but thats comparing apples to oranges

Now you have discovered cabin gain eg the smaller the space you are radiating into the more acoustic gain is seen at low frequencies. To compensate you must create a room curve for each vehicle and EQ to maintain a flat response. < if that's your goal ;)
 
Just another Moderator
Joined 2003
Paid Member
Back to speakers, I'm sure I would hear the difference in your two crossovers. Problem is I would not know which one was the better one.

I believe that you would pick the better one, even without any extended listening.

But as I said earlier in the thread, If you are happy with what you've got, then perhaps it is better to leave it at that, unless you enjoy the quest for getting better and better sound :)

One great test if you have a good quality microphone and mic preamp is to get someone you know well to just talk into the mic (only play one speaker) and see whether it sounds like them! If it does then you have good tonal balance in the critical voice range. If it doesn't then you probably have some work to do :) (note that not all can be fixed with a crossover and there may be inherent colouration in the drivers, but that is another kettle of fish).

Tony.
 
Ok So I put my foot in my mouth on that one. My only experience with complex lenses is from a Pentax DSLR camera I had that died (smoke came out of it) when it was less than 2 years old. I guess it makes sense that a complex lens would have a sweet spot. My experience is more with telescopes where low f numbers usually mean a sacrifice in edge detail.

Back to speakers, I'm sure I would hear the difference in your two crossovers. Problem is I would not know which one was the better one. I have three vehicles all of which just have the base audio system. They all sound totally different to me but I simply could not say which one is best. If forced to pick one I would choose my pickup truck which is a two seater with no back seat. I can say for sure that none of them sound nearly as good as the B4N system we are discussing here or my other recent build but thats comparing apples to oranges

I should warn you that once you can start to hear design faults, you cannot unlearn it ... for the rest of your life. In some ways, ignorance is bliss ... it certainly saves a lot of money on hifi equipment.

But ... since you are determined to progress, all that separates a so-called "golden ear" from anyone else is simply knowing what to listen for. It's not a matter of hearing acuity, but discrimination, and it's certainly not as esoteric as wine-tasting. If you can taste the difference between McDonald's and a $50 steak you can hear the difference between mid-fi and high-fi.

First off, ignore all the hoo-ha from the glossy review magazines. The minimum requirement for a reviewer is knowing how to write, and following the magazine guidelines. Good taste is entirely optional, and is usually absent. So don't trust what you read in the glossies. It's just recycled manufacturer's PR material and extended meditations on the breakfast they ate that morning. If they write about the sound of movie soundtracks (which are entirely synthetic and do not correspond to reality) put the magazine aside.

A very basic test for any loudspeaker is the ability to play pink-noise (look for free downloads) without audible coloration. What does pink-noise sound like? Like falling water, something every human being has heard. If there is an artificial harshness or metallic quality to the sound, that's wrong. Pink-noise should sound smooth, and as you move around while listening, there should be no whistling or phasey-sounding artifacts. 90% of the high-end speakers on the market fail this subjective test, so that makes your job easier.

At a more sophisticated level, spend some quality time listening to live acoustic music (no amplification). You don't have to like it, just get acquainted with the sound of the instruments: the piano, saxophone, violin, cello, and an unamplified singer. High Fidelity literally means "faithful to the source" ... in an age of electronic distortion, modification and synthesis, it's handy to know what reality sounds like (the visual equivalent is a no-Photoshop "straight" photograph). A system that can do a reasonable facsimile of reality can also do a good job with distorted or synthesized music, but the reverse is not true.

Again, you don't have to actually like live, acoustic music; just know what is sounds like. It has the interesting technical characteristic of being the most difficult to reproduce with realism. Hifi systems that come close to reproducing live, acoustic music, along with a sense of the original acoustic space, are very, very rare ... I might hear only one or two at a hifi show with more than a hundred vendors.

If you never listen to live, acoustic music, or are only interested in movie soundtracks, I have no advice to give, except for the pink-noise part. All speakers, regardless of application, should be able to reproduce pink-noise with a minimum of subjective coloration.
 
Last edited:
I should warn you that once you can start to hear design faults, you cannot unlearn it ... for the rest of your life. In some ways, ignorance is bliss ...
I think there's a flaw in this kind of argument. I think it is not pure psychological like that.

A very basic test for any loudspeaker is the ability to play pink-noise (look for free downloads) without audible coloration.
How do you compare the usage and effectiveness of pink noise test versus vocal test? I think I can predict/guesstimate two weaknesses with pink noise test. First, it covers all frequency. Most high end speakers are wideband so not surprising if 98% failed this test as irregularities are normally found at low frequency eg by room resonances. Second, I think pink noise test only relies on the flatness of frequency response. Other effects of phase I think cannot be heard?

spend some quality time listening to live acoustic music (no amplification). You don't have to like it, just get acquainted with the sound of the instruments: the piano, saxophone, violin, cello, and an unamplified singer. High Fidelity literally means "faithful to the source"

But don't forget that what is in the recording is not necessarily what is in live performance.
 
DIY Speaker builder conundrum...

A most interesting topic and one I find akin to my own situation.

When I finish a new design or build, such as the Strads, which I completed in the latter part of last year, I was faced with the same dilemma.

They sounded good from the outset, but was this simply bias for my own creation?
To answer this I did the following,

  • Listened to several tracks that I know well using a good quality set of headphones, using my PC HDMI output.
  • Then doing the same (with headphones) through my main audio amp to get a feel for any differentiation in the sound.
  • Following that I then used a pair of speakers from an earlier build, that do sound good and I know are pretty well run in, noting anything that I think needs to be specifically listened for in the new speakers.
  • Then did the same again with the new speakers, all they while making notes of certain characteristics of some of the tracks.

Be prepared to spend some time doing this. I do this partly as my wife is a retired concert pianist, who has an outstandingly good ear for tonality and coloration in music.
When I build a new set of speakers I spend a lot of time on modelling the crossover before I start the build, as I then form a picture of exactly how to proceed to achieve the end result.

Once all the above is taken care of the speakers get the most important test of all, critical listening from other members of our HI-Fi community. If there's a problem they will soon let you know. I encourage people to bring along an amp and music they are familiar with, as it helps identify flaws.

Using this method usually gets me a result that doesn't need fine tuning to any great extent. The only crossover changes that I made from the finalized crossover design was to increase the values slightly in the compensation for the rising impedance that the tweeters exhibited.

The Strads do double duty as my HT mains and my main HI-Fi speakers. They sounded good from the start and still sound just as good now after the x-over alteration.

Now here's the important part, it really doesn't matter too much about the x-over adjustments. If it sounds good and isn't tiring to listen then its fine.
The difference between my Onkyo HT amp, a Sugden 21se and a Plinius 850 was as different as night and day. Shrillness in the Titanium tweeter isn't there using the Sugden or Plinius. Base was less powerful in the Sugden and warmer and a touch more refined with the Plinius, so what this tells you is that if the sound is good and enjoyable with your setup don't worry about it ... just sit back and enjoy.

Of course I now have a major conundrum, The Strads are by far the best speaker I have built and are considerably better than most speakers I have listened to in the sub $2000000 range, which was the object of the exercise and leaves me wondering what to build next :)
 
I mention pink noise because the BBC Research Labs found back in the Sixties that pink-noise is about 10 dB better than other musical stimuli for revealing "buried resonance" ... the kind of thing that shows up on the CSD display. For example, a "buried resonance" might only cause a 0.5 dB ripple in the frequency response, which is inconsequential, but be audible with musical and pink-noise stimulus (particularly in the 1 to 4 kHz range). Massed choirs (without studio processing) are a close second for assessing resonance, along with audience applause.

I have found this true in my own loudspeaker design experience. Once you start hearing the resonances in the pink noise, sure enough, they're there on the CSD display, and it is worthwhile finding drivers that are free of stored energy and designing crossovers to suppress the magnitude of HF breakup and resonance. The right amount of subjective HF coloration is zero. In the real world, we approach, but don't always get there.

Voice is an excellent cross-check, but unfortunately it is common studio practice to use special "vocal" microphones with a peaked response, and perversely, "de-essers" in the studio chain to remove the resulting sibilant "ess" sound from female singers. So the sibilant sound may be in the recording, thanks to the choice of microphone, potential overload in the mike preamp electronics, and unpredictable use of "de-essing" electronics. Studio engineers cannot resist the temptation to "improve" a singer's voice and make it punch through a dense mix. You can't ever be sure, unless you were at the mix session yourself, and heard what the engineer was doing.

Massed choir is somewhat better, since there isn't a much horsing around with the sound, but many "choirs" in pop music are really just a few singers thickened up with multiple passes of multitracking and electronics processing. The more processing and mucking around with sound, the further we get away from reality, and move towards a studio confection that might sound great but not have much to do with the actual sound of the performers.

I'm all in favor of great studio multitrack pop recordings, but there's no yardstick for "real" any more. The closest you might get would be the exact monitors used for the final 2-track master assessment. If that's what the producer and musicians signed off on, replicating that environment might be the best you could do. Same story for movie soundtracks: replicating the system used for final mastering would probably be the best you could do, since there's no "real" there.

P.S. I concur with "kiwilistener" above. It really helps to have friends who are classical musicians; they can help keep you on the straight and narrow, particularly with the distinctive tonality of acoustic instruments, which so much of high-end audio gets wrong. Good headphones ($300 and up) are also a good cross-check.

P.P.S. I should also add that I have never designed a loudspeaker without aid of instrumentation. When I started this gig back in 1975, all I had was an Altec pink-noise source, an Altec 1/3 octave realtime analyzer, a square-wave generator, a Tektronix scope, and an Altec 1/2" instrumentation microphone. Nowadays, a PC with a good soundcard, ARTA software, and a decent 1/2" electret mike are all you need.
 
Last edited:
I strongly recommend it. Listen for a waterfall sound with hidden mechanical sounds, or edgy harshness, inside it.

It's easy to get ear-fatigued, particularly with loudspeakers that have many uncorrected resonances. Keep the test-listening sessions shorter than 5 minutes, and levels at 70 dB or less. Once you get fatigued, acuity drops very rapidly, and you will feel annoyed and angry. Stop listening before that happens.
 
Hi,

For sure you can find out a shedload of information about
your speakers (especially if they are wrong) (and room) by
using test tracks, like pink noise, pink noise 1/3 octave bands,
warbled 1/3 octave sliding tones, rather than using music.

rgds, sreten.

One good evening was running through my test tracks in a
friends new system that showed he essentially had no basic
problems at all. Unlike his car hifi which had serious problems.
After the latter he understood the simple purpose of the former,
and was delighted his new system showed no basic flaws.
 
Last edited:
I feel incompetent. I felt like a fool starting this thread and admitting that I'm building speakers and not even knowing what they should sound like. Someone voted on this thread and gave it a "2" rating. Hopefully that's an insult towards me and not to the experts here who have been helpful.

My first build, as some of you know, was the Dayton DA175 system . My intent with the B4N system was simply to learn enough to tackle a third and more expensive system and know what to choose and what to expect.

At this point I'm so confused.

"midrange" mentioned listening to just the news or something like that and I did. What I noticed was, I'd hear, on occasion a noticeable "lisp" to voices both male or female. Maybe this is the "ess" that Lynn Olson mentioned. Listening to the DA175 sytsem with just voice I'd still hear these "lisp's" but they were more subdued. For some reason on the DA175 system I hear more pronounced breathy "pops".

I kind of like "winter mules " suggestion of listening to someones voice I know with microphones. I used to have a bunch of older mic's by Astatic and Shure but a few years ago I threw them all out when I was DX'd with cancer. Sounds like something interesting to try so I'll aquire another mic.

I'm also going to try Lynn Olson's suggestion of pink noise. I think my ears would hear the "hidden mechanical sounds, or edgy harshness, inside it."

I dont know how to multi quote but Lynn Olson's wrote:
"I should warn you that once you can start to hear design faults, you cannot unlearn it ... for the rest of your life. In some ways, ignorance is bliss ... it certainly saves a lot of money on hifi equipment."

I'm think there's no turning back now for me. I'm doomed.

And "kiwilistener" wrote :
"The Strads are by far the best speaker I have built and are considerably better than most speakers I have listened to in the sub $2000000 range"

I dont know what to think anymore.:confused: Do you an extra zero or two there? Are there actually speakers that cost $2M or more?

Tony
 
A handful of high-end vendors build $200K/pair loudspeakers as vanity projects; very few are built, and I suspect most end up in the homes of magazine reviewers and industry pundits (opinion makers in the trade). The manufacturer considers this an "extended loan", writes it off as a promotional cost, and after a year or two, the product shows up on Audiogon.

In the higher-volume part of the market, there are plenty of high-end speakers in the $5K to $50K range; even the Ariels would have to sell for more than $10K if they were priced through normal high-end channels.

The industry average is (parts cost + cabinet + assembly and test labor) * 5 = customer price. High-end dealers expect a 40% to 50% profit margin, national or regional reps expect another 8% to 10%, so that can double the price again.

These prices (relative to parts cost) might seem outrageous, but few manufacturers actually make a year-in, year-out profit. The high-end biz combines the worst aspects of the furniture and the fashion industry; high overheads in shipping and storage, combined with a flighty, fad-driven customer base. This makes for rapid obsolescence and difficulty moving out-of-fashion inventory. All it takes is one or two bad (or indifferent) reviews to drive a manufacturer to the brink of bankruptcy.
 
Last edited:
Just another Moderator
Joined 2003
Paid Member
Hi, The ess sound you mention is what is often called sibilance, If you do a search here on sibilance you will find lots of discussion on it! I've seen various reasons given for it's presence. Often just reducing the tweeter level a bit will bring it under control.

It could also be caused by breakup on the woofers not being suppressed enough by the crossover. Some people claim it is much worse with certain amplifiers than with others (for a given speaker).

Now that you have noticed this that is one step forward in your quest! :) you are already starting to hear things that are not natural. If the same person was in the room talking to you I bet you wouldn't hear that same essing! It is a bit tricky in that it the recording itself will be contributing, so you need to listen to a range of things. If it is always present then it is more than likely something you need to deal with. If it is present with some material and not other material, you may need to decide whether or not to make an adjustment.

This was one of the issues I initially was trying to deal with on my MTM's I started out just padding down the tweeter, which helped, but also took too much of the high end out. The rework of the crossover to change the slope to 4th order helped much more, probably indicating it was a combination of the tweeter levels being a bit high and the woofer breakup combining to cause the problem.

Tony.
 
At this point I'm so confused.

Forget about high end boolshits. Many people have been driven away from enjoyment because of these boolshits. They run after something not important and away from something more important. As an example, they get the extra detailed transparent treble but they cannot listen to music for hours because of the listening fatigue.

You have your speaker because you like to listen to music. If you enjoy the experience, you continue. Just like eating food. You like it, eat it. You want something better, go search for it.

As a clue, people enjoy (1) Bass (2) Sweet vocal. Musicality is difficult to understand but you will know.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.