wafflesomd,
Thank you for sharing your experience and observations—it’s clear you’ve put considerable time into analyzing and modifying the Pit Vipers, and your technical curiosity and commitment to better sound are commendable.
That said, I’d like to offer a more nuanced perspective in response to your critique—one that takes into account the many complexities involved in speaker design, measurement, and ultimately, listener preference.
First and foremost, speaker voicing is inherently subjective. What sounds “right” or “neutral” to one person may sound dull, bright, recessed, or forward to another. Designers like Paul Carmody often develop their own sonic signature—an intentional voicing based on their personal taste, listening environment, and chosen design philosophy. That isn’t a flaw; it’s part of the artistry of speaker design. The use of specific drivers, crossover slopes, baffle dimensions, and porting are all tuned toward a goal that balances measured performance with listening experience. Changing any one of these elements—let alone multiple—will inherently shift that balance.
Regarding your point about the midbass dip and the differences from Paul's simulation, it’s important to recognize the limitations of simulation tools. Software like XSim and PCD can give highly accurate predictions if the inputs (FRD/ZMA files, acoustic offsets, and driver alignment) are perfectly captured. But in reality, even slight differences in mic placement, box stuffing, cabinet diffraction, or measurement techniques can lead to discrepancies of several dB. It’s not uncommon for DIY builders to see variances in the 3–6 dB range depending on their build precision and measurement setup. A 10 dB midbass null, while definitely significant, could also stem from room interaction or even mic distance—not necessarily an error in the original design.
You mentioned swapping the tweeter and running the system active—which is great if that got you closer to your preferred sound. But it also means the result is no longer representative of the original design. Different tweeters have different sensitivities, dispersion characteristics, and baffle interactions. Naturally, a crossover that was optimized for one dome will not work optimally with another. Likewise, moving the port, changing its diameter and location, or altering cabinet dimensions will affect both tuning frequency and cabinet resonances. These changes may well yield better results for your room and ears, but that doesn’t necessarily mean the original design was poor—just that it wasn’t tailored to your specific use case.
As for the Overnight Sensations and PE’s marketing graphs—while I agree that transparency matters, we also have to remember that the OS is a wildly popular entry-level kit, created over a decade ago, and never intended to be a measurement reference. It was designed as an affordable and great-sounding first build for many hobbyists. And for that purpose, it has succeeded immensely. Expectations should be aligned with that context.
Finally, while you’re right that designing your own speaker is more accessible than ever—thanks to measurement tools, software, and shared knowledge—it’s still far from trivial. A successful design balances dozens of variables: driver compatibility, impedance behavior, power response, lobing, phase alignment, distortion characteristics, cost, and buildability. Many builders fall into the trap of measuring flat on-axis response but overlooking off-axis behavior or crossover phase nulls that only show up in polar plots or summed impedance curves.
So in conclusion: your critiques are not without merit, and your desire for transparency and better design rigor is completely valid. But it’s also important to approach these topics with an understanding that design choices often involve compromise and artistic intent, not just raw data. What you might consider a flaw may have been a deliberate tuning choice. Measurement discrepancies can come from dozens of sources, and sound quality is ultimately personal. That’s why even among commercial speakers, you see such a wide range of tonal balances and voicing philosophies.
Respectfully, I think the best takeaway here is the one you already reached: build, measure, tweak, and share your work. That’s the real beauty of this hobby—learning from others, adding your own twist, and eventually developing your own voice in speaker design.
Looking forward to seeing your Amiga variant when it’s ready!
Thank you for sharing your experience and observations—it’s clear you’ve put considerable time into analyzing and modifying the Pit Vipers, and your technical curiosity and commitment to better sound are commendable.
That said, I’d like to offer a more nuanced perspective in response to your critique—one that takes into account the many complexities involved in speaker design, measurement, and ultimately, listener preference.
First and foremost, speaker voicing is inherently subjective. What sounds “right” or “neutral” to one person may sound dull, bright, recessed, or forward to another. Designers like Paul Carmody often develop their own sonic signature—an intentional voicing based on their personal taste, listening environment, and chosen design philosophy. That isn’t a flaw; it’s part of the artistry of speaker design. The use of specific drivers, crossover slopes, baffle dimensions, and porting are all tuned toward a goal that balances measured performance with listening experience. Changing any one of these elements—let alone multiple—will inherently shift that balance.
Regarding your point about the midbass dip and the differences from Paul's simulation, it’s important to recognize the limitations of simulation tools. Software like XSim and PCD can give highly accurate predictions if the inputs (FRD/ZMA files, acoustic offsets, and driver alignment) are perfectly captured. But in reality, even slight differences in mic placement, box stuffing, cabinet diffraction, or measurement techniques can lead to discrepancies of several dB. It’s not uncommon for DIY builders to see variances in the 3–6 dB range depending on their build precision and measurement setup. A 10 dB midbass null, while definitely significant, could also stem from room interaction or even mic distance—not necessarily an error in the original design.
You mentioned swapping the tweeter and running the system active—which is great if that got you closer to your preferred sound. But it also means the result is no longer representative of the original design. Different tweeters have different sensitivities, dispersion characteristics, and baffle interactions. Naturally, a crossover that was optimized for one dome will not work optimally with another. Likewise, moving the port, changing its diameter and location, or altering cabinet dimensions will affect both tuning frequency and cabinet resonances. These changes may well yield better results for your room and ears, but that doesn’t necessarily mean the original design was poor—just that it wasn’t tailored to your specific use case.
As for the Overnight Sensations and PE’s marketing graphs—while I agree that transparency matters, we also have to remember that the OS is a wildly popular entry-level kit, created over a decade ago, and never intended to be a measurement reference. It was designed as an affordable and great-sounding first build for many hobbyists. And for that purpose, it has succeeded immensely. Expectations should be aligned with that context.
Finally, while you’re right that designing your own speaker is more accessible than ever—thanks to measurement tools, software, and shared knowledge—it’s still far from trivial. A successful design balances dozens of variables: driver compatibility, impedance behavior, power response, lobing, phase alignment, distortion characteristics, cost, and buildability. Many builders fall into the trap of measuring flat on-axis response but overlooking off-axis behavior or crossover phase nulls that only show up in polar plots or summed impedance curves.
So in conclusion: your critiques are not without merit, and your desire for transparency and better design rigor is completely valid. But it’s also important to approach these topics with an understanding that design choices often involve compromise and artistic intent, not just raw data. What you might consider a flaw may have been a deliberate tuning choice. Measurement discrepancies can come from dozens of sources, and sound quality is ultimately personal. That’s why even among commercial speakers, you see such a wide range of tonal balances and voicing philosophies.
Respectfully, I think the best takeaway here is the one you already reached: build, measure, tweak, and share your work. That’s the real beauty of this hobby—learning from others, adding your own twist, and eventually developing your own voice in speaker design.
Looking forward to seeing your Amiga variant when it’s ready!
Last edited by a moderator:
If you enjoy watching building videos, check out Cameron's build log here:
He also measures the pit vipers after completion and gets a much closer measured response to Paul's than yours.
He also measures the pit vipers after completion and gets a much closer measured response to Paul's than yours.
That is awesome, I hope you get a response. I was going to suggest this, as I think a few of the designs he has published were originally commissioned projects.
Maybe somebody also active on the PE forum could also make such an offer???
I tend to Bi-Amp such endeavours tho
I tend to Bi-Amp such endeavours tho
Something you may try if you have an account on PE tech talk is start a thread titled "Paul Carmody Design Commission" or something of that nature and explain your idea and ask if anyone else is interested in the idea and comment if they are. This way everyone interested, myself included, could comment that they are interested in such a "flagship" or "endgame" large tower design and also if they're willing to help fund the project. It would only be a matter of time before Paul seen the thread and comments himself. After seeing interest in such a design, he just might take on such a project. This obviously doesn't mean he will, however you have a better shot if a lot of people are interested and willing to help fund the project.
If you do, tag me. I will happily be the first comment and willing to help fund such a project.
If you do, tag me. I will happily be the first comment and willing to help fund such a project.
Last edited:
That excludes me but I think I could afford a $100-Aud [ $70-Usd] kicker to see such a project to fruition.
How much would be needed to see something HUGE started?
How much would be needed to see something HUGE started?
Last edited:
What do you mean that it excludes you? That would be plenty! Just think even if only 20 people donated $70 usd that would be $1400. That would be plenty for Paul to design some brilliant. He’s rather good at picking high performing affordable drivers.
I meant that I am not active on the PE forum and don't ever intend to be really.
I've just assembled a stack of drivers and passive parts for my own huge build, something I've been planning for a few years now. Hoping Paul would do something bigger than my own dual 12" pipedreams using Jaycar Carbon fibre woofers
I've just assembled a stack of drivers and passive parts for my own huge build, something I've been planning for a few years now. Hoping Paul would do something bigger than my own dual 12" pipedreams using Jaycar Carbon fibre woofers
Oh I gotcha, that makes sense.
It's hard to say an exact price because it ultimately depends on how he designs the speakers. What drivers he decides to use, the crossover parts needed, how big of an enclosure it requires, even wood is getting pricey here. If I had to guess $800-1k per pair all in.
I don't see Paul going any bigger in design than dual large woofers (probably 12inch) and MTM design for mid/high per cabinet. Like a scaled up version of the pit vipers. I can't speak for everyone's result but his pit vipers that I built plays extremely loud and has tons of bass and low end extension. My amp does 300wpc into 4 ohms. I'd imagine he would utilize a passive crossover and the complexity gets pretty out of hand with design past 3-way.
It's hard to say an exact price because it ultimately depends on how he designs the speakers. What drivers he decides to use, the crossover parts needed, how big of an enclosure it requires, even wood is getting pricey here. If I had to guess $800-1k per pair all in.
I don't see Paul going any bigger in design than dual large woofers (probably 12inch) and MTM design for mid/high per cabinet. Like a scaled up version of the pit vipers. I can't speak for everyone's result but his pit vipers that I built plays extremely loud and has tons of bass and low end extension. My amp does 300wpc into 4 ohms. I'd imagine he would utilize a passive crossover and the complexity gets pretty out of hand with design past 3-way.
If you do start such a post keep us all in the loop please; and I've not got a reply to my personal email but he probably gets hundreds so that is no surprise.
I have a rough plan for a 15" based 3 way based on drivers we can easily get in the Antipodes designed for a friend, but he raided my driver stores and looks like using the older JBL drivers I have spare.
If there was sufficient interest, I might resurrect it. The thing that puts me off publishing my projects is the endless picking at details by the peanut gallery.
Comment mainly directed at Moondog.
If there was sufficient interest, I might resurrect it. The thing that puts me off publishing my projects is the endless picking at details by the peanut gallery.
Comment mainly directed at Moondog.
Brett the only 15" woofers I have in my stash are old Jaycar party woofers then need refoaming and a singleton Vega154 in a big sealed box.
Many thinks for the thoughts tho
Many thinks for the thoughts tho
It actually doesn't. His measured response shows the very issues I've pointed out. You have to take into account that his graph is smushed and visually misleading.
Here it is with a proper scaling. Focusing on everything above 200hz, there is an 8db difference between 250hz and 2k. The overall response is just frankly poor with wide tonality swings. The tweeter appears to be 5db down for whatever reason. This builder might actually have an issue with his tweeters crossover.
We can compare this with Paul's sim. As we can see, this simulation shows no low mid valley. For whatever reason Paul left out the dB numbers on his graph... I assumed 5db increments. We can again ignore the low end here.
Some more comments on this low mid valley, it's worth noting my data was captured with the speaker elevated at least 6' off of the ground so we can rule out room interactions causing the dip. There is also another user on the tech talk forum who is a musician/engineer who has built the pit viper and also noted that subjectively they could hear this dip and corrected it with EQ. I myself am an active mix engineer and musician. I have 15+ years of experience mixing music profressionally in live and studio environements, my preference is generally as neutral as you can get with excellent dispersion. Genelec and Neumann speakers are common for me to use and deviations from that sort of linearity are quite obvious for me at this point and bothersome.
This discrepancy between sim and real world response has nothing to do with voicing. What is shows is that the Paul did not capture data properly for his filter design. This brings into question the accuracy of the rest of the filter tasks. Subjectively, this dip has been pointed out by a few people as problematic. I am very well aware of the process of analyzing a speaker and developing filters both active and passive. My process is honestly quite a bit more in depth than Paul's as I utilize off axis data and he does not. The projects I have designed speaker for themselves as they measure very neutral on axis and offer good dispersion. Paul is admittedly a bit behind in this aspect.
The measurements I posted are of the stock crossover and tweeter. Frankly my updated version was a huge improvement in just about every aspect not only objectively but subjectively as well. I did not use the stock crossover with a different tweeter, a new filter set was designed to accommodate the new tweeter.
As I mentioned previously, I did my measurements outside with the speaker elevated. The dip is not from room interactions. While real world performance can differ from sims, it should not differ this much. If it does, one needs to find out why and fix their analysis process. None of my sims have ever differed nearly this much from real world performance.
I would agree with this, 20 years ago. Today though, not so much. Great sounding is highly debatable as anyone with any sort of reference of good audio reproduction finds them to sound quite poor. Unfortunately beginner builders are not well expereinced in audio reproduction with speakers so it's hard to put much weight on their impressions. Considering the c-notes exist, measure and sound much better, and even cost less I see no reason why one should accept the false information provided in the data for OS. I don't really care how little one is spending on a speaker, they should be getting what is advertised. I will be messaging PE to talk to them about this discrepency, they have modified product listings in the past based off my feedback so hopefully this will be addressed.
I'm sorry to say, but I feel that you've simply made a lot of excuses for the Pit Viper which is simply just a poorly executed speaker with a lazy design process behind it, all in the name of "art" as you say. There is more than enough evidence here to show that something went wrong along the way. I very much feel like what I see here is an example of cult of personality elevating a designer beyond their actual output. You covered that an excellent speaker design takes into many performance metrics, none of which Paul seems to take into account. While I was personally disappointed in the Pit Vipers, it did have an upside. I no longer take into account designs with limited performance data.
Last edited:
@wafflesomd I'm always amazed when people build a published design and there isn't a divergence in the measurements! I don't know how driver manufacturers manage to keep the level of consistency they do. But equally, I know that it's hard to get even a pair of drivers where the parameters aren't a few percent diifferent, and there are plenty of documented cases of driver models which vary considerably over time, even from the best manufacturers. Also I know how easy it is for a crossover to produce significant differences in response when there's a small difference in parameters.
You can see my point I'm sure. The valley you've measured is pretty much centred on the bass/mid XO frequency, and the mountain at 2 kHz is pretty much centred on the mid/treble XO frequency. Your comments on the designer are very harsh, but is it possible your drivers are just not a great match for the ones in the original design?
You can see my point I'm sure. The valley you've measured is pretty much centred on the bass/mid XO frequency, and the mountain at 2 kHz is pretty much centred on the mid/treble XO frequency. Your comments on the designer are very harsh, but is it possible your drivers are just not a great match for the ones in the original design?
Here's a thought.
I've not built them and definitely not listened to them, but is it possible they actually sound like all the big stuff we used to listen to when we were 17 and 18, all bass and top end because rock&roll lyrics are often unintelligible anyway.
Just an idle thought in passing
I've not built them and definitely not listened to them, but is it possible they actually sound like all the big stuff we used to listen to when we were 17 and 18, all bass and top end because rock&roll lyrics are often unintelligible anyway.
Just an idle thought in passing
? All the shared measurements ARE using original drivers. I had one dc130 replaced as it was a little out of spec but everything else was well within spec and matched well. I've shared data from myself and others that show the issues I've described. I don't really get what is so hard to understand about my criticisms all of which are backed up by evidence. Were just talking about speakers here people, it's ok to call out problems if they exist and I would expect the same level of scrutiny from any design that I've published.
I'd argue Paul is just from a different time. These days it's not unheard of for one to have their first design utilize most of the modern design performance metrics and deliver an excellent speaker. There's simply way more info online and design tools than what was available 10-20 years ago. I find it concerning that despite all the evidence both objective and subjective, it seems nearly impossible to convince anyone that this speaker has some issues. It seems like most audio design forums are entirely driven by anecdote rather than evidence and I find that troublesome. In the world of designs like mechano23, speakers from heissmann acoustics or mtg-designs, to me it really makes no sense to buy into the anecdote based designs.
I just received an email form Paul asking me for my thoughts on a possible project; so something might happen