4-way instead of 3-way?

[hifijim]

I think Andy's point was that the statement "Everything you need to know" is a bit of hyperbole. the videos do not actually instruct the audience on how to design a speaker.

 

[Hydrogen Alex]

Well in my opinion its by far the most comprehensive video series I have seen on designing a modern DSP / active monitor.
Whilst I keenly await to see how Andys designs are superior, maybe you Jim can enlighten me to what is missing and what you would add to "fill in" the missing sections?
Thanks in advance...

 

[hifijim]

I had seen parts 1,2,3 soon after they were produced. I had not seen part 4 with the new Purifi driver. I love the enthusiasm of these guys, and I admire their tenacity in trying many many variations to get it right.

As you say, it may be "by far the most comprehensive video series I have seen on designing a modern DSP / active monitor."... but that does not make it complete. And it is not everything you need to know.

These are not "how to" videos, and they are not intended to be. Someone just starting in the process of designing a "modern DSP / active monitor" could gain some inspiration from these guys, and pick up a few interesting bits of knowledge, but they really don't touch upon the design process at all. I suspect there was a very large effort behind the scenes that we (the YouTube audience) are not shown. In reality, these are infomercials, and that is not a bad thing. For potential buyers, it is a good thing. After watching the 4 videos, the potential buyer has enough information to decide whether or not the product might be right for them, and whether they should try to get an audition. But it is far from a how-to guide. Saying that these videos are all you need to know would be the same as saying that I could design a high performance sports car because I watched a 45 minute video about how Porsche designed their latest toy.

A real How-To design document/video would give detailed coverage of such things as
(1) how to model/simulate the bass response for vented box and sealed box (to start with), and this would include the mathematics, the theoretical basis, and examples of how to use simulation software
(2) how to model/simulate the baffle layout, meaning the size and shape of the baffle, the positions of the drivers and their spacing relative to each other, and the shape of the edge (radius, beveled, or hard-edged for example)
(3) providing a detailed and deep understanding of the appropriate design targets. For the most part, this is the work of F. Toole and S. Olive, as they have described in detail what makes a good loudspeaker good, and what makes a poor loudspeaker poor. It would define the appropriate design targets for sound power, directivity, early reflections response, on-axis response, bass extension, mas SPL, distortion, etc.
(4) Digital signal processing theory, so the designer has an understanding of sampling rate, bit depth, biquads, IIR and FIR processing.
(5) Analog signal processing, so the designer understands how and why crossovers work, including including the effects of driver spacing and positioning.
(6) etc
(7) more etc
(8) the list goes on and on... cabinet design, driver selection criteria, how to optimize within a budget constraint, how to optimize within a size/weight constraint, etc, etc,

@jheoaustin - you are contemplating some expensive drivers for your first speaker design. You need to go into this process with your eyes wide open, and fully aware of the possible results. You will learn a great deal while doing this project, and it is very likely that when you are done, you will know enough to recognize numerous deficiencies that will make you want to start over. This is not to say that your result will sound bad, but that you will see areas where you could have made a better decision and gotten a better result. This is normal. It is virtually impossible for someone to get it right the first time. Speaker design is an Art and a Science. It takes practice, and we can't be afraid of failure. If you do not have the time or patience for this process, and you want to just build a set of speakers one time and be done, then the best option is to find a good kit.

Also - be very cautious about subscribing to a 1-dimensional design philosophy that promises sonic excellence by focusing on a particular issue. There are some ideas floating around that claim there is a "secret formula" to good sound. The truth is that everything about a speaker matters, although some things matter a lot and somethings matter just a tiny bit. All the competing priorities must be balanced with each other using judgment.

j.

 

[jheoaustin]

I presume you meant vm752/753 since the vm527 is a 2"?

JH> Actually no. I meant VM527, as there were advices of 3" dome being too large for the upper end coverage, as well as being too small for the lower side. Sorry for not being clear.

I think people that would consider paying £1320 for a pair of 3" midrange drivers in support of a speaker DIY hobby interest have such different values to myself I'm not sure my comments will be of much relevance.

JH> That is true. Another reason for me being reluctant to professional drivers is that we audiophiles don't really get many chances to listen to perfectly or near-perfectly tuned professional studio monitor systems. My only connection in the industry moved to east coast, so I don't have any more opportunity.
Having said that, I think I at least have some ear for correct or reasonable sound, after attending SF symphony concerts at Davies Hall dozens of times. So, I don't think I am too subjective, either. My first system, I'd call half DIY'ed, measures pretty flat, and sound great. I worked at Audyssey Laboratory as a DSP engineer, learned a lot there, and completely agree that the sound individuals feel ideal are not too much different, supported by pscychoacoustics research.

Like the ATC 3" it is a decades old design that made more engineering sense when it first appeared compared to where evolution has moved speaker design and manufacturing today. The niche it seems to address is premium audiophile retro designs with basic cabinets and crossovers (e.g. monitor boxes). It is too big to be appropriate for small monitors where a 3" midrange may be a good choice. To make best use of it given its small cone area and huge motor suggests a large waveguide and sitting a reasonable distance away given the less than ideal distance between tweeter and midrange. It offers good performance (within the limitations of a soft dome) but possibly a touch below the ATC in terms of useable passband, distortion,... although a more considered judgement would require more knowledge than I possess having little interest in either driver. What is not in doubt though is the poor value for money. This is not a ripoff but a reflection of how the very tiny numbers are inefficiently manufactured given the miniscule demand.

JH> I am being motivated by many ATC bashing here, though I still think systems built with them sound great. I also agree that it is not a cost-effective one. But my experience so far was that it in the core of the best sounding system I've heard. So, cost-not-so-much-object, ATC dome was a safe choice back in 2005. Thanks to you and other people(no sarcasm, seriously), I now became willing to look at other options. Mentioning that Satori midrange is a bit of deviation from my bias toward ATC. Good that it costs a lot less!

A transformer is the straightforward way to handle the low impedance but there are alternatives. Cannot comment on the audiophile appeal of the alternatives since even more than 3" soft domes the appeal of a ribbon is an audiophile one rather than a technical performance one. Just like valve amplifiers, record players, tape machines, etc... the value that some find in things other than technical performance and engineering (real that is rather than "over-engineering") can be substantial and is perfectly valid for a hobby.

Since Raal 70-20 sounded just great, I think I have to abandon the prejudice on the ribbons with transformers. But as I said, Raal was only one of 2 exceptions that I liked among ribbons. I have MG 1.7i and I learned that I wouldn't be a big fan of ribbons anyway, lol.

 

[jheoaustin]

These tutorial videos are superb, everything you need to know about how to design a modern high performance loudspeaker system... Highly recommended!
Part one, intro and crossovers:

Part two, drivers:

Part three, cabinet:

Update -

I've just watched the part 1, and it is not sufficient to create a design so far, but interesting pieces of information. Let me continue watching them and post replies later.

 

[jheoaustin]

Interesting certainly w.r.t. where things are going but tutorials? Everything you need to know? The youtubers know little about the technical side of speakers and, probably wisely, keep away from measurements, technical reasoning, engineering and that sort of thing sticking with assertions, associations and the usual subjective playbook. If you were to take a poll on this forum of the most fancied tweeter, midrange and woofer that is what they have stuck in a box with a flat baffle. The review highlighted they were not on top of how to design a crossover, there are no plots of the radiation pattern, an 8" woofer in a sealed box is insufficient for stereo use as main/midfield studio monitors without subs (subs are visible in some of the pics) and even then is likely borderline with the Dayton driver version (a multichannel setup with 7, 9, 11 or whatever 8" drivers could be fine if setup with understanding).

The price is £5-6.8k a pair depending on options. This is the price of a well engineered and fully supported modern studio monitor like a Neumann KH 420 or Genelec equivalent. Now it's early days and to expect technical performance and support to be on par with the products of large established studio monitor companies is perhaps unreasonable. Indeed they may have little-to-no interest in serving the conventional professional market and are looking to the audiophile market and whatever the prosumer market sector is called (i.e. people with studios in their homes largely looking to youtube and forums for knowledge and guidance).

Will be interesting to know what the OP as a subjective audiophile thinks of the product?

Let me post my review of the video clips after I watch through all 4. What do you mean by 'the product' at the last line? Do you mean the speakers the video is describing?

BTW, I consider myself an audiophile who's quite more objective than most of 'audiophiles'. I wrote a bit about my background in a post earlier.

 

[tmuikku]

It takes practice, and we can't be afraid of failure.

This is one thing that gets learned doing first project, stumbling on things but it's a good one to learn in life in general, don't let failures beat you down but take em as learning opportunity instead. I've simplified this to a thought: the more you fail the faster you learn. And here is why one wants to do some prototypes, fail cheap and end product will be so much better, all failures you were able to notice have a possibility to be factored out. This is actually leads to my simplified definition whay makes good playback system, which is problem free system.

 

[jheoaustin]

@jheoaustin - you are contemplating some expensive drivers for your first speaker design. You need to go into this process with your eyes wide open, and fully aware of the possible results. You will learn a great deal while doing this project, and it is very likely that when you are done, you will know enough to recognize numerous deficiencies that will make you want to start over. This is not to say that your result will sound bad, but that you will see areas where you could have made a better decision and gotten a better result. This is normal. It is virtually impossible for someone to get it right the first time. Speaker design is an Art and a Science. It takes practice, and we can't be afraid of failure. If you do not have the time or patience for this process, and you want to just build a set of speakers one time and be done, then the best option is to find a good kit.

Also - be very cautious about subscribing to a 1-dimensional design philosophy that promises sonic excellence by focusing on a particular issue. There are some ideas floating around that claim there is a "secret formula" to good sound. The truth is that everything about a speaker matters, although some things matter a lot and somethings matter just a tiny bit. All the competing priorities must be balanced with each other using judgment.

j.

This is actually not about my first design, if I can call my current one my design(maybe half). Mine was custom-built by Wilmslow Audio back in 2005 through some discussions with the experts there, mainly because I suck in handcraft and building hardwares. I chose the drivers, and built the DSP crossover S/W twice by modifying existing commercial products. Wilmslow later took the design and improved it(better tweeter and updated woofer?) and named it 'Encore' as an off-the-shelf design for their kit. Though the speakers were the first of this kind for me, I am using the 2nd generation DSP crossover which is much better than the 1st.

I think my current system sounds fine at least, and one of my new friend, who I consider quite objective audiophile or with neutral taste, said he wanted to buy my system when he heard it first, as a compliment. I don't think my next project will sound bad, or worse than my current. And, yes, I am asking experts' advice here to do a better job than the last one. And at this time, I want to do more myself, or try not to rely on Wilmslow.

 

[jheoaustin]

This is one thing that gets learned doing first project, stumbling on things but it's a good one to learn in life in general, don't let failures beat you down but take em as learning opportunity instead. I've simplified this to a thought: the more you fail the faster you learn. And here is why one wants to do some prototypes, fail cheap and end product will be so much better, all failures you were able to notice have a possibility to be factored out. This is actually leads to my simplified definition whay makes good playback system, which is problem free system.

I am afraid I am in trouble now because I haven't failed with my first DIY(?) speakers and 2 DIY DSP crossovers. ;=)

 

[jheoaustin]

I would still urge you to use 4 eight inch per side as you instantly half all distortion plus get the benefit of a pseudo line array (true line array must be over 90% of floor to ceiling height) bass system ie You will eliminate floor bounce, plus get much more even dispersion (this shows as reduction of peaks / troughs when measuring) and longer "throw" into your room.
This is an important point, especially in larger rooms, not so important in Pro studio near field monitors though:
Line array SPL reduces by just 3 dB per doubling of distance Vs point source reduction of 6dB ... That is actually a huge factor ie at a 4 meter listening position your twin driver bass/mid is reduced from 109 dB at 1 meter to 97 dB at the listening position Vs 103 dB for the 4 driver array... Thats 6 db is the difference between using 50 watts on the 4 driver array VS 200 watts (!) on the twin drivers... This is a massive (50%) reduction in all distortion... Imagine the huge media fan fare if some high end company just patented a new way to reduce all distortion by 50%.... Well we all can do this, just double the Sd... Size matters!!
Cheers
A.

I am afraid 4 woofers are still a bit too much. If they are all front-mounted, the bass module will be pretty high, leaving less room for other upper-band drivers. I was momentarily thinking about isobaric ones, so 2 mounted on the baffle, and 2 in the box. But that's an unknown area to me, and I'd be scared until I learn more from experts' advice.

 

[tmuikku]

I am afraid I am in trouble now because I haven't failed with my first DIY(?) speakers and 2 DIY DSP crossovers. ;=)

yeah it's just a philosophy that one can utilize as needed, like foster more philosophy. Here is my favourite: the system is only as good as listening skills

 

[mayhem13]

Thank you for your advice. Though there are some successful and popular implementations of x.5 way speakers, I am reluctant, and afraid of the combing effect coming from the overlapped band. Since I will do active crossover(DSP), I guess there is one less reason to do '.5' way which I think is difficult to tune.
But yes, I like narrow baffles for many reasons. Isobaric enclosures are quite interesting, too. Unfortunately, I haven't listened to one yet. I hope I have a chance soon to get the idea how they sound.

A .5 coil isn’t at all difficult…..one of simplest circuits and a true 1st order so minimal phase and it actually eliminates combing as the drivers don’t overlapp except where the wavelengths are as long as the nearfield boundaries so no issues whatsoever.

For your intention, it’s actually the perfect solution for 2 8” woofers. A pair of 8’s with an ATC dome and a SS tweeter sounds like an excellent solution. I agree and would not use a ribbon with the ATC….the dispersion patterns are too conflicting…..the ribbon will call too much attention to itself in near to midfield listening. A few of the custom monitor install builds I’ve seen using the ATC mid have the dome tweeter flange machined to get the drivers as close together as possible to keep the upper lobe as fwd and small as possible.

There is another direction to take and that’s NOT using the ATC mid dome but instead, a larger midbass driver crossed low to a robust dome in the 1.5-2khz range depending on the tweeter. This requires a waveguide solution for the tweeter to ‘match’ the directivity and keep it somewhat narrow in the upper freq range which eliminates or at least greatly reduces room interaction. You would need to decide what makes sense for you given you room and placement limitations or intention…THIS IS BEAUTY OF DIY in that it allows your design to match the environment……something off the shelf speakers simply cannot do. I’d suggest that THESE EFFORTS are better undertaken than selecting expensive or boutique drive units. I can tell you from experience and a huge fan of the ATC dome that given the choice today, I’d choose the B&C 8pe21 midrange all day over the ATC as the most dynamic and detailed mid range driver available today regardless of cost. I would not have believed it myself until I first bought a pair and mounted them open baffle by themselves and just listened for a few weeks…..remarkable what they’ve accomplished with the cone material and breakup eliminating ridge.

 

[Hydrogen Alex]

I had seen parts 1,2,3 soon after they were produced. I had not seen part 4 with the new Purifi driver. I love the enthusiasm of these guys, and I admire their tenacity in trying many many variations to get it right.

For sure, they are music lovers and very passionate guys.

As you say, it may be "by far the most comprehensive video series I have seen on designing a modern DSP / active monitor."... but that does not make it complete. And it is not everything you need to know.

Well, its about 90% once we go through your list, see below.

These are not "how to" videos, and they are not intended to be. Someone just starting in the process of designing a "modern DSP / active monitor" could gain some inspiration from these guys, and pick up a few interesting bits of knowledge, but they really don't touch upon the design process at all.

If a DIY novice watched the videos all he needs to do is clone the build... Forget the 100 year old history of passive crossovers and ported boxes from Tool!
Looking at the RRP parts cost at retail, PDP charge about £1,200 to assemble and test the speakers, so you save that cash and have peace of mind knowing that "your design" is the result of a team of world class audio engineers and a years worth of their R&D.
Plus you get to choose any paint or finish you want... Just order a clone cabinet from Wilmslow Audio or similar DIY cabinet making service... Bolt in the drivers and follow the intuitive videos on how to Eq and room correct.... A million home cinema users do that every day!. I suspect there was a very large effort behind the scenes that we (the YouTube audience) are not shown.

In reality, these are infomercials, and that is not a bad thing.

Yes, they are giving away 90% plus of the finished build... How many high end speaker companies would show this level of detail on their builds...

 

[Hydrogen Alex]

After watching the 4 videos, the potential buyer has enough information to decide whether or not the product might be right for them, and whether they should try to get an audition.

Why? The vast majority of DIY kits are bought without auditioning them... They sell on the reputation of the designers... PDP are world class audio engineers and their only way to earn a living is if their results are the best...

A real How-To design document/video would give detailed coverage of such things as
(1) how to model/simulate the bass response for vented box and sealed box (to start with), and this would include the mathematics, the theoretical basis, and examples of how to use simulation software
(2) how to model/simulate the baffle layout, meaning the size and shape of the baffle, the positions of the drivers and their spacing relative to each other, and the shape of the edge (radius, beveled, or hard-edged for example)
(3) providing a detailed and deep understanding of the appropriate design targets. For the most part, this is the work of F. Toole and S. Olive, as they have described in detail what makes a good loudspeaker good, and what makes a poor loudspeaker poor. It would define the appropriate design targets for sound power, directivity, early reflections response, on-axis response, bass extension, mas SPL, distortion, etc.
(4) Digital signal processing theory, so the designer has an understanding of sampling rate, bit depth, biquads, IIR and FIR processing.
(5) Analog signal processing, so the designer understands how and why crossovers work, including including the effects of driver spacing and positioning.
(6) etc
(7) more etc
(8) the list goes on and on... cabinet design, driver selection criteria, how to optimize within a budget constraint, how to optimize within a size/weight constraint, etc, etc,

Sorry Jim, this is not just wrong, it is actively damaging the DIY community, not your intent of course but this mis information would deter lots of new DIY enthusiasts from ever daring to build speakers... Point by point breakdown to follow.

 

[andy19191]

I meant VM527, as there were advices of 3" dome being too large for the upper end coverage, as well as being too small for the lower side.

The 2" is a newish design and will have been created with a niche in mind and likely a customer but because of it's size it isn't clear to me what that niche might be. Anyone?

JH> I am being motivated by many ATC bashing here, though I still think systems built with them sound great.

Bashing? Their drivers offer good performance and if a speaker design works with the pros and cons it can deliver very good results. Other companies like K&H likely did this better than ATC themselves in terms of full speakers. ATC also properly support and maintain the drivers over their life which is important when used professionally. The drivers ceased to be available several years ago which makes their pros and cons somewhat academic today.

I also agree that it is not a cost-effective one. But my experience so far was that it in the core of the best sounding system I've heard. So, cost-not-so-much-object, ATC dome was a safe choice back in 2005. Thanks to you and other people(no sarcasm, seriously), I now became willing to look at other options. Mentioning that Satori midrange is a bit of deviation from my bias toward ATC. Good that it costs a lot less!

The 2.5" soft dome has been designed as a filler driver rather than a conventional midrange driver. Picking drivers first rather than a speaker configuration and then seeking drivers to fit can be a less than optimum way to go about things. But it's a hobby!

What do you mean by 'the product' at the last line? Do you mean the speakers the video is describing?

Luxury products consist of an object with a technical performance/function and a set of intangibles associated with that object. The value of both can vary greatly from person to person depending on their knowledge and beliefs. A £250k audiophile speaker for example will often have a lower technical performance than a £25k one but purchasers can still be perfectly happy with the product because of how they perceive value. So I mean the speakers, the marketing in the videos and everything else that may add value for an audiophile.

 

[Hydrogen Alex]

A real How-To design document/video would give detailed coverage of such things as
(1) how to model/simulate the bass response for vented box and sealed box (to start with), and this would include the mathematics, the theoretical basis, and examples of how to use simulation software

Why? All this has been done by the PDP team... didnt you watch the videos?!
I use AJ Designer driver simulation, its free, it takes less than 30 minutes to understand and I dont have a clue about any of the maths or algorithms in its engine.... One does not need to understand how something works to use it correctly!!

(2) how to model/simulate the baffle layout, meaning the size and shape of the baffle, the positions of the drivers and their spacing relative to each other, and the shape of the edge (radius, beveled, or hard-edged for example)

Once again Jim, this has already been done by PDP....! IF one wants to start from scratch and build their own design, driver location knowledge is required but not when you buy a fully finished design which has had all the R&D done by experts!

(3) providing a detailed and deep understanding of the appropriate design targets. For the most part, this is the work of F. Toole and S. Olive, as they have described in detail what makes a good loudspeaker good, and what makes a poor loudspeaker poor. It would define the appropriate design targets for sound power, directivity, early reflections response, on-axis response, bass extension, mas SPL, distortion, etc.

Sorry Jim, this is getting repetitive, but of course PDP have a "detailed and deep understanding of the appropriate design targets" and have used them to create the finished product!

(4) Digital signal processing theory, so the designer has an understanding of sampling rate, bit depth, biquads, IIR and FIR processing.

Why? As I said above - I use AJ Designer driver simulation, its free, it takes less than 30 minutes to understand and I dont have a clue about any of the maths or algorithms in its engine.... One does not need to understand how something works to use it correctly!! They used various software packages and they have refined the design... The buyer does not need to understand how a Mini DSP, or Hypex DSP works, just plug it in, follow the instructions and you are done... As in Home cinema. Floyd Toole and Co lived in a Pre (historic) DSP age.... DSP replaces maths tables and calculus.

(5) Analog signal processing, so the designer understands how and why crossovers work, including including the effects of driver spacing and positioning.

Come on Jim, you are really struggling here... Its already done! Understanding is not required... Copying is all thats required.

(6) etc
(7) more etc

Ok, now you have won... Without high quality etc and more etc life on earth ends...!!

(8) the list goes on and on... cabinet design, driver selection criteria, how to optimize within a budget constraint, how to optimize within a size/weight constraint, etc, etc,

OMG... The list goes no for sure, PDP have designed the cabinet, selected the drivers, chosen a budget (prices are public domain) and they decided how big they wanted it to be... All done nicely for you in advance... You are welcome!

 

[hifijim]

Oh I am sorry, @Hydrogen Alex, I misunderstood you. I thought you made the claim that

These tutorial videos are superb, everything you need to know about how to design a modern high performance loudspeaker system...

To me, this means "starting from scratch" as you say, and designing a system. Any system.

but what you actually meant was

If a DIY novice watched the videos all he needs to do is clone the build.

Learning to clone an existing design is a good first step in becoming a designer. But it is a first step.

Sorry Jim, this is not just wrong, it is actively damaging the DIY community, not your intent of course but this mis information would deter lots of new DIY enthusiasts from ever daring to build speakers

Wow.

 

[Robh3606]

One does not need to understand how something works to use it correctly!!

LOL carry on !!!! The blind leading the blind.

I would add a response but the statement is so ridiculous it doesn't merit one.

Rob

 

[hifijim]

This is actually not about my first design, if I can call my current one my design(maybe half). ..
.... And, yes, I am asking experts' advice here to do a better job than the last one. And at this time, I want to do more myself, or try not to rely on Wilmslow.

It sounds like you have reasonable expectations and a good frame of mind for this project.

 

[hifijim]

For your intention, it’s actually the perfect solution for 2 8” woofers. A pair of 8’s with an ATC dome and a SS tweeter sounds like an excellent solution.

I agree.

There is another direction to take and that’s NOT using the ATC mid dome but instead, a larger midbass driver crossed low to a robust dome in the 1.5-2khz range depending on the tweeter. This requires a waveguide solution for the tweeter to ‘match’ the directivity and keep it somewhat narrow in the upper freq range which eliminates or at least greatly reduces room interaction. You would need to decide what makes sense for you given you room and placement limitations or intention

I agree again. I will add that the Bliesma T34B can be crossed amazingly low without the use of a waveguide, but the baffle will need to be carefully designed and profiled to get a good directivity/power response.