THE Boutique Bass Cab Shootout

[Foz]

Never would have even thought to set it up this way, but I do have barstools and boom stands on hand...we'll see how it shakes out

That configuration worked very well in that room because of the well baffled 15' cathedral ceiling allowed a long window of data collection before room reflections got in the mic = we could measure accurately lower in the band than any other configuration we tried. Also because the baffle is well off the floor and the ceiling canted and the mids and highs highly directional - the mids and highs can't skip off a near surface into the mic. Just popping the mic front of a cab in normal configuration leads to trying to muff the floor [aint perfect but a slab of 703 in front of the cab will help a lot] or do some kind of ground plane arrangement [means you can't really get centered on the drivers unless you tip the cab just right] to keep floor skip out of the measurement. It's all a series of compromises - keep an open mind and work with what you got - you'll be fine.

 

[MarkA]

When something sounds off......... Always blame the cat.

CYAWAC. Got it...

On second thought, that sounds dangerous.

 

[MDBass]

Look up, look down... unless this is a ceiling-less room you will have a standing wave situation between the floor and ceiling. Even without a ceiling, the floor boundary will insert a response artifact (especially sitting that high off of the floor boundary). No way will the response analysis be accurate, and the coherency values should indicate this.

I wondered why it was taking Andy so long to chime in here

Garey and I traded a few messages, but I never ended up hearing back from him about a loaner or two from Mesa; bummer, as I've heard some REALLY good things about them.

 

[DukeLeJeune]

So you'll hear the distortion on the recorded track when playback volume rises?

In general, I think so.

My understanding is that, unless a distortion is loud enough, it simply is not detected. The same can be said of low-level details; unless they are loud enough, they will not be detected. Some types of distortion are more objectionable than others, and those tend to cross the "objectionable" threshold not too long after they cross the "detectable" threshold.

Like the microphone clipping on Van Morrison's mic on "Moondance", which imo becomes more audible and objectionable as the volume level goes up.

 

[agedhorse]

How do you account for the significant floor boundary at low frequencies and the ~7mSec of delay that the reflection arrives at? It takes more than 5mSec to acquire the low frequency data IME, leading to a low coherency value. 5mSec is a very significant number IMO. There will be a second (lower level) artifact at about 4mSec due to LF leakage from the back side of the cabinet.

When making measurements, and this is a common issue, it's easy to get data that appears good on the surface but really is inaccurate, resulting in inaccurate calculations and results.

I have no dogs in this race (I assume), but when I see a test set-up that is going to generate obvious artifacts yet presented as "scientific", these flaws should be called out for what they are and acknowledged in the results. It's also unfair to judge any of these cabinets based on flawed test methodology.

If you want a proper 1/2-space measurement, at least place the cabinet in an open outside space, on its back on a lawn that will absorb and minimize all of these artifacts (including the ceiling boundary, which will appear somewhat solid due to lack of absorption (low absorption coefficient) at low frequencies of the damping materials on the ceiling. There is also the cavity effect if the room, which will skew the results, but probably not as much as the others.

 

[agedhorse]

In general, I think so.

My understanding is that, unless a distortion is loud enough, it simply is not detected. The same can be said of low-level details; unless they are loud enough, they will not be detected. Some types of distortion are more objectionable than others, and those tend to cross the "objectionable" threshold not too long after they cross the "detectable" threshold.

Like the microphone clipping on Van Morrison's mic on "Moondance", which imo becomes more audible and objectionable as the volume level goes up.

Some distortion elements can increase with falling levels, these can become extremely objectionable at low volumes, and in fact our threshold is crossed earlier as the volume is decreased. (crossover distortion in class B amps is a good example of this)

 

[DukeLeJeune]

Some distortion elements can increase with falling levels, these can become extremely objectionable at low volumes, and in fact our threshold is crossed earlier as the volume is decreased. (crossover distortion in class B amps is a good example of this)

Yes, I absolutely agree! My understanding is that crossover distortion in amplifiers is audible and objectionable WAY out of proportion to what a THD measurement might indicate.

The way it was explained to me by Earl Geddes (who co-authored a two-part AES paper on the subject of distortion perception with Dr. Lydia Lee), the ear's "masking" characteristic plays a big role in distortion perception. Briefly, "masking" is the tendency of the ear/brain system to ignore a low-level signal that is close to a high-level signal, as long as the two occur simultaneously. (Compression algorithms such as MP-3 are based on taking advantage of masking). Low-level noises and distortions are much more likely to be audible and objectionable when there is no masking signal present. At the crossover point, virtually instantaneous though it may be, the noise and distortion are not masked because they are accompanied by relative silence. At least that is my recollection, it's been a long time and I'm really just a hardcore amateur masquerading as whatever, so please correct me if needed.

Getting back into the loudspeaker side of things, diffraction is a distortion that arrives a little bit later in time and therefore tends to not be masked. And diffraction is a type of distortion that follows the perceptual trend that I described above: Inaudible at low level, detectable as we get louder, and finally objectionable. So in other words all those horn tweeters that seem to distort and get harsh at high levels, well the driver is usually not actually distorting (witness how often a tweeter that sounds harsh still lives to fight another day). More often, I think there's significant diffraction within the horn somewhere, of which the ear's PERCEPTION is non-linear (increases disproportionately with increasing volume level).

 

[Lowendtech]

hardcore amateur masquerading as whatever,

I think I’ve seen this band.
You’re way better than them.

 

[MDBass]

How do you account for the significant floor boundary at low frequencies and the ~7mSec of delay that the reflection arrives at? It takes more than 5mSec to acquire the low frequency data IME, leading to a low coherency value. 5mSec is a very significant number IMO. There will be a second (lower level) artifact at about 4mSec due to LF leakage from the back side of the cabinet.

When making measurements, and this is a common issue, it's easy to get data that appears good on the surface but really is inaccurate, resulting in inaccurate calculations and results.

I have no dogs in this race (I assume), but when I see a test set-up that is going to generate obvious artifacts yet presented as "scientific", these flaws should be called out for what they are and acknowledged in the results. It's also unfair to judge any of these cabinets based on flawed test methodology.

If you want a proper 1/2-space measurement, at least place the cabinet in an open outside space, on its back on a lawn that will absorb and minimize all of these artifacts (including the ceiling boundary, which will appear somewhat solid due to lack of absorption (low absorption coefficient) at low frequencies of the damping materials on the ceiling. There is also the cavity effect if the room, which will skew the results, but probably not as much as the others.

I was pretty specific about what I was doing in the OP so as not to inflate the status of the testing to "scientific": a reamp shootout, with a minimum of variables via pre-recorded samples and identical mic placement, using a very high quality/flat response LDC.

Most of what you wrote above is well above my pay-grade (I'm a session musician and recording/mix engineer, not a scientist ) but as to your suggestion of placing the cabs flat on their backs outside...

Seems like that would obviously compromise the designs with rear ports, one of the "series of compromises" someone else mentioned, correct?

It also seems like would also provide something other than a real world listening experience: since the cabs will always be used facing forward, and frequently in spaces as small and/or significantly 'worse' acoustically than my control room, won't reflection and/or boundary issues always be part of the listening experience for both the user and the audience?

 

[MDBass]

Some distortion elements can increase with falling levels, these can become extremely objectionable at low volumes, and in fact our threshold is crossed earlier as the volume is decreased. (crossover distortion in class B amps is a good example of this)

Yes, I absolutely agree! My understanding is that crossover distortion in amplifiers is audible and objectionable WAY out of proportion to what a THD measurement might indicate.

The way it was explained to me by Earl Geddes (who co-authored a two-part AES paper on the subject of distortion perception with Dr. Lydia Lee), the ear's "masking" characteristic plays a big role in distortion perception. Briefly, "masking" is the tendency of the ear/brain system to ignore a low-level signal that is close to a high-level signal, as long as the two occur simultaneously. (Compression algorithms such as MP-3 are based on taking advantage of masking). Low-level noises and distortions are much more likely to be audible and objectionable when there is no masking signal present. At the crossover point, virtually instantaneous though it may be, the noise and distortion are not masked because they are accompanied by relative silence. At least that is my recollection, it's been a long time and I'm really just a hardcore amateur masquerading as whatever, so please correct me if needed.

One of the reasons I really enjoy TB: I'm pretty good at what I do, but I am WAY out of my depth in this particular fish tank.

 

[tombowlus]

Are you doing windowed frequency response data collection or just sound sample files?

Cause if you're just doing sound samples there aint much point in off axis stuff IMO.

My $0.02

We do stepped/gated frequency response measurements, FWIW.

 

[agedhorse]

I was pretty specific about what I was doing in the OP so as not to inflate the status of the testing to "scientific": a reamp shootout, with a minimum of variables via pre-recorded samples and identical mic placement, using a very high quality/flat response LDC.

Most of what you wrote above is well above my pay-grade (I'm a session musician and recording/mix engineer, not a scientist ) but as to your suggestion of placing the cabs flat on their backs outside...

Seems like that would obviously compromise the designs with rear ports, one of the "series of compromises" someone else mentioned, correct?

It also seems like would also provide something other than a real world listening experience: since the cabs will always be used facing forward, and frequently in spaces as small and/or significantly 'worse' acoustically than my control room, won't reflection and/or boundary issues always be part of the listening experience for both the user and the audience?

My comments were addressing some of the folks (not necessarily yours) who were implying that this was THE scientific approach and the results would be definitive (I'm intentionally poking the bear a bit here).

As somebody who has had to deal with inaccurate testing and inaccurate test conditions before, I am pretty sensitive to obvious gross errors when presented as "definitive" or "real world" because it simply introduces too much error to be as such. The more folks know about WHY there may be such errors, and to take the results at they were intended (not as some might prefer), the better it will be down the road as their knowledge has the opportunity to grow.

Testing rear ported cabinets on their backs is indeed a challenge, and because of this there may be no avoidable way of preventing such errors from creeping into the data. This is especially important when comparing these cabinets with front ported cabinets when the test conditions must be different.

 

[MDBass]

My comments were addressing some of the folks (not necessarily yours) who were implying that this was THE scientific approach and the results would be definitive (I'm intentionally poking the bear a bit here).

As somebody who has had to deal with inaccurate testing and inaccurate test conditions before, I am pretty sensitive to obvious gross errors when presented as "definitive" or "real world" because it simply introduces too much error to be as such. The more folks know about WHY there may be such errors, and to take the results at they were intended (not as some might prefer), the better it will be down the road as their knowledge has the opportunity to grow.

Testing rear ported cabinets on their backs is indeed a challenge, and because of this there may be no avoidable way of preventing such errors from creeping into the data. This is especially important when comparing these cabinets with front ported cabinets when the test conditions must be different.

All good

I was in no way presenting the test as scientific: my usage of "definitive" in the title was simply pointing to the deep and well rounded field of cabs, as to my knowledge no one has tested this many high end models together before, and to get more people's ears on some excellent cabs they might otherwise never have a chance to hear before purchasing them.

This thread started with me looking for a high fidelity cab for my own use with some pretty specific parameters, and it quickly snowballed into something bigger. I had not more than 12 hours earlier updated the OP, and the title, to reflect the current state of affairs before someone very strongly objected to my verbiage.

The title was quickly amended to prevent any further confusion

Considering the above "there may be no avoidable way of preventing such errors from creeping into the data", it seems that there is no way (regardless of how stringent) to set the parameters that won't result in some skewing of the results, no matter how slight, but feel free to correct me if that's not what you intended.

 

[MDBass]

My comments were addressing some of the folks (not necessarily yours) who were implying that this was THE scientific approach and the results would be definitive (I'm intentionally poking the bear a bit here).

As somebody who has had to deal with inaccurate testing and inaccurate test conditions before, I am pretty sensitive to obvious gross errors when presented as "definitive" or "real world" because it simply introduces too much error to be as such. The more folks know about WHY there may be such errors, and to take the results at they were intended (not as some might prefer), the better it will be down the road as their knowledge has the opportunity to grow.

Testing rear ported cabinets on their backs is indeed a challenge, and because of this there may be no avoidable way of preventing such errors from creeping into the data. This is especially important when comparing these cabinets with front ported cabinets when the test conditions must be different.

One thing you didn't address in the last post, which I'm genuinely curious about your perspective on:

Since the cabs will always be used facing forward, and frequently in spaces as small and/or significantly 'worse' acoustically than my control room, won't reflection and/or boundary issues always be part of the listening experience for both the user and the audience?

 

[MDBass]

I need to put my head down to actually get to the re-amping (the DI's are done) as well as other real (IE paid) studio work for the next 24 hours, but I'll pop my head up again when there's some stuff for ya'll to listen to

 

[agedhorse]

One thing you didn't address in the last post, which I'm genuinely curious about your perspective on:

Since the cabs will always be used facing forward, and frequently in spaces as small and/or significantly 'worse' acoustically than my control room, won't reflection and/or boundary issues always be part of the listening experience for both the user and the audience?

Correct, the effects will exist but it will be quite different on a case by case basis. This is why unambiguous comparisons are so difficult to do, and why manufacturers often don't bother sharing in these specs as they are easy to interpret out of context, or compare specs made within different conditions.

 

[Lowendtech]

I’m going to offer a little personal analogy here. If you find you’re eyes glazing over then please ignore.

In a previous (not long ago) career, I used to build Supersport and Superbikes for a national factory road race team here in Canada.
In one particular model year, our Supersport bike had an almost (rider chose to play it safe at the last round) unbeaten season. The way the factory engineers had designed the stock bike led me to be able to take advantage of our series rules to produce an engine/electronics package that the other manufacturers couldn’t touch, regarding our horsepower and torque numbers.
The following model year, we had a new model package (and new rules) that we, as engineers, were promised that we would have all the power and all the torque of the old model but with improvements.
Well as the preseason dragged on, I, as engine and fuel management/dyno operator realized after way too much time sniffing leaded race fuel fumes, realized that there was no way I could get the current model’s dyno figures anywhere’s near where I had them the year before. (And boy did I try).
The time came to load the tractor/trailer with our fleet for the first race.
That bike tore the paint off of everything in the field that weekend. On what I would call weak numbers in what we considered “scientific testing”.
We could not (with our existing equipment) factor in the ram air effect of the newer air intake system. Nor, could we anticipate the determination of our rider.
Turns out we had a rocket ship that won several consecutive championships.

So, play bass in the room that’s paying you to do so.

Edit: And do it with gear you like.

Back to the regularly scheduled program.

 

[nutdog]

 

[alaskaleftybass]

I’m going to offer a little personal analogy here. If you find you’re eyes glazing over then please ignore.

In a previous (not long ago) career, I used to build Supersport and Superbikes for a national factory road race team here in Canada.
In one particular model year, our Supersport bike had an almost (rider chose to play it safe at the last round) unbeaten season. The way the factory engineers had designed the stock bike led me to be able to take advantage of our series rules to produce an engine/electronics package that the other manufacturers couldn’t touch, regarding our horsepower and torque numbers.
The following model year, we had a new model package (and new rules) that we, as engineers, were promised that we would have all the power and all the torque of the old model but with improvements.
Well as the preseason dragged on, I, as engine and fuel management/dyno operator realized after way too much time sniffing leaded race fuel fumes, realized that there was no way I could get the current model’s dyno figures anywhere’s near where I had them the year before. (And boy did I try).
The time came to load the tractor/trailer with our fleet for the first race.
That bike tore the paint off of everything in the field that weekend. On what I would call weak numbers in what we considered “scientific testing”.
We could not (with our existing equipment) factor in the ram air effect of the newer air intake system. Nor, could we anticipate the determination of our rider.
Turns out we had a rocket ship that won several consecutive championships.

So, play bass in the room that’s paying you to do so.


Edit: And do it with gear you like.

Back to the regularly scheduled program.

Man. That story kicks arse! Thank you for sharing!

 

[agedhorse]

I’m going to offer a little personal analogy here. If you find you’re eyes glazing over then please ignore.

In a previous (not long ago) career, I used to build Supersport and Superbikes for a national factory road race team here in Canada.
In one particular model year, our Supersport bike had an almost (rider chose to play it safe at the last round) unbeaten season. The way the factory engineers had designed the stock bike led me to be able to take advantage of our series rules to produce an engine/electronics package that the other manufacturers couldn’t touch, regarding our horsepower and torque numbers.
The following model year, we had a new model package (and new rules) that we, as engineers, were promised that we would have all the power and all the torque of the old model but with improvements.
Well as the preseason dragged on, I, as engine and fuel management/dyno operator realized after way too much time sniffing leaded race fuel fumes, realized that there was no way I could get the current model’s dyno figures anywhere’s near where I had them the year before. (And boy did I try).
The time came to load the tractor/trailer with our fleet for the first race.
That bike tore the paint off of everything in the field that weekend. On what I would call weak numbers in what we considered “scientific testing”.
We could not (with our existing equipment) factor in the ram air effect of the newer air intake system. Nor, could we anticipate the determination of our rider.
Turns out we had a rocket ship that won several consecutive championships.

So, play bass in the room that’s paying you to do so.

Edit: And do it with gear you like.

Back to the regularly scheduled program.

Excellent example of the modeling not matching the real world due to skewed data. Now if you were able to revise your tests with the same ram air effect (same manifold air pressure at the same RPM and load), do you think your model would have given different, perhaps more encouraging results?

This is why when modeling speakers, I always recommend testing the finished product against the model to see where the deviations occur and try to understand why they might exist. It's part of making the results of the models more accurate in the future. I would never trust a model or a measurement first time without some additional confidence data to support it, otherwise you have a false sense of security.