• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

Flattest amp & cab set-up?

#41
While I dont have a way to measure beyond my ears, the Acme B2 sounds good enough to use as a home system (if you have 200 to 300 watts a side home stereo amp). I really enjoy the acme cabinets and Andy is great to work with.

I use a class AB power amp with a high skew rate (PA like w lots of power) and a pre amp. The bass preamps can get close to flat but depends on the pre amp. I was not looking for a flat pre, rather ones with tube warmth. The power amps can be heavy and my pre and power amp push 70#. I have not tried a class D on my rig (lighter).
 
#42
alexclaber said:
"Flat" on its own doesn't mean anything. Are we talking on-axis frequency response, off-axis responses, or power response? What is the transient response? What is the dynamic response? What are the various other methods of distortion (harmonic, intermodulation) and how do they manifest?

"Transparent" is a much better word. Or "accurate".

A cab with much flatter on-axis frequency response than another cab be far less transparent or accurate because it isn't as good in the other aspects of accuracy. When listening to any audio equipment it's the weaknesses you will hear most - the weakest link will dominate.
Click to expand...
Nice to see you contributing to the thread, Alex.

In general, I agree that the performance of a loudspeaker cannot be properly distilled down to one chart or measurement.

Compared with 'flat', the problem with 'transparent' or 'accurate' is that these terms are much harder to quantify with reliable measurements. At least on-axis/off-axis frequency response plots give some meaningful basis for comparison, however limited their usefulness might be. How would one reasonably quantify 'accuracy' or 'transparency'?
 
Last edited:
#43
To elaborate further, in my view, ideally a transparent or accurate speaker will have flat frequency response in addition to perfect dynamic response and phase response, zero distortion, etc.... But then properly qualifying how transparent a given speaker might be requires reliable measurements of all of these factors, not just one. In practical terms, it is difficult enough to even get a decent frequency and phase response plot.
 
#44
12bass said:
I'd like to see some decent frequency response plots to support the above claim. The BGM review of the TC112AF shows a significant dip in the upper midrange between 2 - 4 kHz. I also seem to recall reading something about Duke purposely designing it this way. Note, I'm not saying that they TCs are bad in any way, just that they are not studio monitor flat, nor are they "by far the flattest cabs out there".
Click to expand...

Good eyes!

The on-axis dip in the TC112 in the 2-4 kHz region is there for two reasons: To offset excess energy in the vertical plane in that region, and to keep the horn from sounding like a horn.

The horn that I used in that model is too small in the vertical dimension to have good pattern control below about 3.5 kHz or so, so what happens is, the pattern actually flares in the vertical plane in that region. This is called "pattern flip", and it happens with rectangular horns. I'm more interested in the summed in-room response than in the on-axis response, so I compensate for that vertical pattern flare by introducing an on-axis dip.

So I traded off a "flat" measurement in order to more closely approximate a "flat" sound.

Also, the 2-4 kHz region is where the ear is most sensitive, so anomalies in that region tend to be particularly audible. The horn I used is a very gentle waveguide-style device, but it has one design flaw: The round-overs at the mouth are too small in radius for wavelengths of about 4 kHz and below, so they will be a source of diffraction in that region. I can explain why diffraction is particularly audible if you would like (after I get back from an audio show that I will leave for in a few hours), but the solution I chose was to shade back that region just a bit more, so the response actually ends up being down a bit more than was needed to compensate for the "pattern flip" phenomenon described above.

So once again I traded off a "flat" measurement in order to more closely approximate a "flat" sound. In this case the net benefit of the tradeoff is level-dependent (varies with SPL... I'll explain if you're interested when I get back... but have you ever heard a PA cab that sounded pretty good at modest volume levels yet made your ears bleed when cranked? That was level-dependence in action).

Many years ago I built a speaker that measured very close to flat. It was like plus or minus 1.5 dB on-axis from the lower limits of my measurement setup to about 20 kHz or so. I remember vividly tweaking the crossover and getting closer and closer to flat, and the speaker sounded worse and worse. I persevered, having faith that when I got to "flat", everything would snap into place. Well when I finally got there, I had one of the worst sounding speakers I'd ever heard. This experience sent me to the local university library on a quest to find out "what really matters" perceptually.

I'm still on that quest.
 
Last edited:
  • Like
Reactions: Rumblefisher, downlowuponit, dmrogers and 2 others
#45
DukeLeJeune said:
Many years ago I built a speaker that measured very close to flat. It was like plus or minus 1.5 dB on-axis from the lower limits of my measurement setup to about 20 kHz or so. I remember vividly tweaking the crossover and getting closer and closer to flat, and the speaker sounded worse and worse. I persevered, having faith that when I got to "flat", everything would snap into place. Well when I finally got there, I had one of the worst sounding speakers I'd ever heard. This experience sent me to the local university library on a quest to find out "what really matters" perceptually.
Click to expand...

A flat ON-AXIS frequency response will allmost allways result in a bump somewhere around the frequency range that is caused by diffraction. So in that case an on-axis optimized system will not sound ok.
 
Last edited:
#48
DukeLeJeune said:
The on-axis dip in the TC112 in the 2-4 kHz region is there for two reasons: To offset excess energy in the vertical plane in that region, and to keep the horn from sounding like a horn...

....So I traded off a "flat" measurement in order to more closely approximate a "flat" sound
Click to expand...

An excellent example! Our Big Baby 2 and Big Twin 2 also don't measure completely flat on-axis so they sound 'flatter' in the real world.
 
  • Like
Reactions: Mattsochistic
#49
Like I allready said in this thread, there's more to it then the frequency response (especially when one refers to a flat on-axis response).
My 210 doesn't measure flat on-axis, it measures flat 20 degrees off-axis and there's a very valid reason for that.

And like Alex allready said, other important factors are various types of distortion.
One other important one that makes for an accurate sound is a smooth phase response, IME I could only get that with the coaxial unit I use from 400hz and up. The whole system acts as a pointsource. I tried all other options (regular two, - and three-way systems) in the past but couldn't get the cohesiveness I wanted, been there done that....
 
  • Like
Reactions: downlowuponit
#51
Diffraction? You will have it in all transition and border areas on a speaker.
For instance waveform transition across and from surfaces around the source of audio (is that understandable?). Ie sound wave is travelling in a controlled fashion from your mid driver across the baffle, but once it reaches the end of the baffle, there is a shift in energy/vector, resulting in a change of the sound. The shape, size etc of the baffle and cabinet will affect the level of diffraction. Some hifi designers work with bevelled, cylindrical or even spherical shapes to fight diffraction, but then encounter other (often bigger) consequences to deal with.
 
#52
AlexanderB said:
Diffraction? You will have it in all transition and border areas on a speaker.
For instance waveform transition across and from surfaces around the source of audio (is that understandable?). Ie sound wave is travelling in a controlled fashion from your mid driver across the baffle, but once it reaches the end of the baffle, there is a shift in energy/vector, resulting in a change of the sound. The shape, size etc of the baffle and cabinet will affect the level of diffraction. Some hifi designers work with bevelled, cylindrical or even spherical shapes to fight diffraction, but then encounter other (often bigger) consequences to deal with.
Click to expand...
I only understand diffraction from highschool physics in terms of light splitting into component colour and narrow apertures causing wide dispersion at exit.
 
#53
By the way, "flat" electronics is not complicated to build!
"Transparent" electronics is much harder, but doable. (It also involves distortion, noise level, finding optimal impedance levels etc.)
"Flat" speakers"? Not really for MI or PA, unless controlled by a processor. Even the 151566 etc are severly rolled off in the lows - MANY dB around 30Hz.
"Transparent" speakers? Not really available yet, but some specialized HiFi gear is getting closer, as long as you can define the constrains you are working with. 16 - 25 000Hz? 120 dB continous? Below 3% THD 16-50 Hz up to 120 dB? etc...
 
#56
One of the big problems with attempting to get 'flat' response below 100Hz from a bass guitar loudspeaker is that the room has a huge effect on the response at low frequencies. A speaker which has flat response to 30Hz outdoors on the ground (half-space) will have extremely boosted lows in a typical indoor scenario.
 
  • Like
Reactions: downlowuponit
#58
alexclaber said:
One of the big problems with attempting to get 'flat' response below 100Hz from a bass guitar loudspeaker is that the room has a huge effect on the response at low frequencies. A speaker which has flat response to 30Hz outdoors on the ground (half-space) will have extremely boosted lows in a typical indoor scenario.
Click to expand...

True. It is also true that our ears and brain are extremely good at filtering out room anomalies when such are encountered. We can almost always "decode" the sound to its original properties (and enjoy it), as soon as we have "familiarized" oursleves with the room we are in. Music sounds vastly different in a car (even through a good setup) compared to in a living room, but we still decode it to reach the same interpretation.

There are eq knobs on most amps (even "flat" ones) to adjust (in a rough way) the reproduction if room anomalies become intrusive. ;)


By the way, speakers that have frequency gaps/peaks/dips/roll offs in the lows also have other problems in the same range; you can not have a frequency roll off without shift in phase, group delay etc.

A sealed box with its more gentle roll off generally has less pronounced issues of the other kinds as well. A highly tuned bass reflex is really a bad-bad scenario in most parameters except box size and sensititvity.
 

Latest posts

Top Bottom
Back