Salk WoW1 Bookshelf Speaker Review

[Zvu]

If it was me, i'd use SB12CACS25 and Morel CAT308 for a two way such as this. Bass port on the back and optimised to be listened on speaker stands. You can always put a sock in the bass port if you get too close to the wall but you can't make more bass if you optimise the crossover for loudspeaker to be used with heavy boundary reinforcement.

Manual that goes with this should be detailed leaving no place for assumptions about the way it is used.

 

[q3cpma]

I'd agree that luck can be involved in whether or not a speaker has significant port resonances. They're apparently hard to predict from modeling software; often not obvious in testing; often not obvious in listening; and fixing them can require lots of experimentation, prototyping, trial and error, etc.

Sometimes marketing sets a schedule and a price point, and fixing such problems takes too much time or money. Sometimes the companies aren't even aware their speaker has such a problem (not every speaker company has access to high-resolution measurements of the midrange that one gets from the Klippel or from an anechoic chamber).

Some companies have the resources to work hard to remove them intentionally, but many (most?) do not. So when speakers do not exhibit such issues, yeah, a bit of luck is/was often involved - or a lot of work.

Well, we agree on that point that bass reflex needs lots of R&D to be made optimal. Reason why I buy Genelec, really. But still, two things can be said about this specific model:
* Modelisation is hard, but testing isn't that much; that dip could be seen with a simplistic on-axis measurement.
* Front porting makes things even harder, as midrange leakage possibly leading to cancellation is now a thing.
* Not about the model per se, but lack of resources is an explanation, not an excuse.

I continue to say that the wise position is to start with level scales and for every bad model (be it due to design or QC), tip them on the bad side. The hard part is that it requires only good models to get on the good side of my scales, since it's the only way to show consistency in R&D, design and QC.

 

[MZKM]

Which cables were used to test this speaker?

If Amir isn’t using Nordost or MasterBuilt cables, why bother?

 

[MZKM]

@jsalk send in more speakers to be tested. This will solve all of these problems. That fact is that whether justly or otherwise, you have a bad review on a site that is now very popular and growing

His speakers take 3-9 months to buiid. Maybe he has showroom models, but I doubt he wants to pay out of his own pocket for shipping.

The review isn’t even bad, the only real issues are the port response and the lower than advertised sensitivity (kind of obvious a 4.5in “woofer“ where even the manufacturer recommends using it only down to 60Hz will be bass-lite with lots of distortion).

 

[ROOSKIE]

His speakers take 3-9 months to buiid. Maybe he has showroom models, but I doubt he wants to pay out of his own pocket for shipping.

The review isn’t even bad, the only real issues are the port response and the lower than advertised sensitivity (kind of obvious a 4.5in “woofer“ where even the manufacturer recommends using it only down to 60Hz will be bass-lite with lots of distortion).

Howdy. The review is very bad. Yes the comments alleviate much of it but the headless panther and the 1st post, the actual review are very bad. Most people do not read all of these comments.
In any case if it is true that other models are excellent and judging Salk Sound based on this one oddball monitor is unfair, solve the problem by sending some others in. And with all due respect, yes pay for shipping, why that would be an obstacle for a manufacturer I have no good idea.

Plus this way Salk determines the next speaker tested and can pick the best one, rather than a customer who may choose another mixed bag.

 

[EchoChamber]

With all due respect, there is no perfect product for everybody.
That basically sounds like an Apple advertisement. This is the classic split exemplified by Android vs Apple. One has many choices(Android) which please some folks very much and confuses others, the other has very limited choices and that pleases some but seems overly generic and catered to other folks.
I think it is great that there are companies that make 1 or 2 speakers, and others that make a wide range.
That said I will back up the premise that any $1600 speaker needs to be excellent sounding and well engineered.

Yes, I’ve been accused before of being an Apple fanboy... ;-)

My point was that D&D is focused on getting a class reference product for their line up. Their masterpiece, what they think the “perfect” speaker for home and for demanding audiophiles should be. So they spend the proper amount of time and resources on it before moving to the next challenge. They establish a North Star and focus on getting there. They won’t get there, but will get closer and closer by perfecting every detail.

That approach doesn’t garantie success, but it is good practice to have a chance at succeeding, i.e., do your homework and take the design process seriously. In my professional experience, something good will come out of it.

Having a very large catalogue with products completely dissimilar and carrying legacy products is likely to distract a small R&D team from focusing on their North Star.

BTW, the open source nature of Android is great. It provides a healthy dose of competition for Apple among other things.

 

[Bachemar]

One bad review tarnishes the whole line. However throw in a couple of stellar reviews and it changes the conversation. The suggestion above to send a speaker in for testing makes sense. @jsalk maybe a middle of the line speaker like the Song3 or BeAT. Although, I wonder how the low positioned woofers of the Song3 series would work with Kippel’s near field mic placement.

 

[MZKM]

One bad review tarnishes the whole line. However throw in a couple of stellar reviews and it changes the conversation. The suggestion above to send a speaker in for testing makes sense. @jsalk maybe a middle of the line speaker like the Song3 or BeAT. Although, I wonder how the low positioned woofers of the Song3 series would work with Kippel’s near field mic placement.

I would really love to see the SS 7M measured, as it’s basically the bigger brother to the BMR:

At ~$5000 I would love to see what improvements there are (and while @hardisj measured the BMR, Amir would have to too for good measure ).

 

[pjug]

On rear porting this kind of mini bookshelf speaker: the three main reasons I chose this speaker are the height, front port, and veneer cabinet. Most likely these will go in a bookshelf, or else on a mantle up against a wall. So I would just be plugging the plug the port of a rear ported design.

 

[CDMC]

Why can't people understand the concept of burden of proof? If you have trustworthy measurements of only one model of a brand showing bad design (and not just low cost optimisation), it's just basic logic to consider the whole brand as bad until more data is made by the same measurer. Keyword being consider and until, the level of "set in stone" just depends on the number of measured models.

Though I personally shun any manufacturer not providing at least basic data about frequency response, horizontal dispersion and distorsion/SPL in favour of those who do. Even if they can do good products, voting with your wallet for the more honest brands is important.

You are misunderstanding and misapplying the concept of burden of proof. The burden of proof is the party in action that has the burden of proving their case to the burden required for the type of case. By way of example, in a criminal action in the United States, the State has the burden of proving, beyond a reasonable doubt, that the defendant committed the crime for which they are charged. In civil matters there may be a presumption that a defendant is liable, shifting the burden to the defendant that they did not do the act for which they have been sued. An example would be financial elder abuse, where upon showing of certain facts by a plaintiff, the burden shifts to the defendant to show the acts they engaged in were not harmful to the plaintiff.

Important is that in a legal proceeding, you cannot extrapolate evidence of other acts to meet the burden of proof that a defendant committed the present act that they are being sued upon (or named a criminal defendant). The example I always give is just because someone has stolen 20 cars in the past, does not mean they stole the vehicle for which they are now a defendant. The time the evidence of the person had stolen 20 cars previously would come after conviction, at the sentencing phase. In a civil action, it would come in as part of a punitive damages claim (e.g. person has defrauded others in the past).

Neither basic logic, nor science, assume that when you have a sample size of one, that you should extrapolate the data to an entire group. All you have is a sample of one. Get to three or four and you may start to see a trend. This is especially true here where you have a manufacture that produces multiple designs with different components, not merely one design scaled up and down. Using your "basic logic" if you stayed at a hotel an the maid forgot to empty the trash in your room when cleaning, then no trash in any room was empty, the trash in the lobby is piling up, and the restaurant is a disgusting mess full of garbage.

 

[CDMC]

One bad review tarnishes the whole line. However throw in a couple of stellar reviews and it changes the conversation. The suggestion above to send a speaker in for testing makes sense. @jsalk maybe a middle of the line speaker like the Song3 or BeAT. Although, I wonder how the low positioned woofers of the Song3 series would work with Kippel’s near field mic placement.

Further up in the thread someone mentioned that a pair of BeATs being built for a member here are going to be shipped to Amir to be measured. Having had a pair of Song 3s and comparing their measurements to my F208s, they measured similarly in room, except the lack of a low end bass boost (which I really like). If not for the Song 3s having more "accurate" in room bass, they would still be here. Similarly, I posted my measurements of the Song Surround v. WOW1 v. Dynaudio X12 in my desktop system. Those show the same dip as Amir's, but also show the the less expensive and larger Song Surround measure better. I think the WOW1 is just at the limit of what can be done size wise.

 

[q3cpma]

You are misunderstanding and misapplying the concept of burden of proof. The burden of proof is the party in action that has the burden of proving their case to the burden required for the type of case. By way of example, in a criminal action in the United States, the State has the burden of proving, beyond a reasonable doubt, that the defendant committed the crime for which they are charged. In civil matters there may be a presumption that a defendant is liable, shifting the burden to the defendant that they did not do the act for which they have been sued. An example would be financial elder abuse, where upon showing of certain facts by a plaintiff, the burden shifts to the defendant to show the acts they engaged in were not harmful to the plaintiff.

Important is that in a legal proceeding, you cannot extrapolate evidence of other acts to meet the burden of proof that a defendant committed the present act that they are being sued upon (or named a criminal defendant). The example I always give is just because someone has stolen 20 cars in the past, does not mean they stole the vehicle for which they are now a defendant. The time the evidence of the person had stolen 20 cars previously would come after conviction, at the sentencing phase. In a civil action, it would come in as part of a punitive damages claim (e.g. person has defrauded others in the past).

The burden of proof isn't only a legal term. I'll use the closely related Occam's razor (since it's used to select the default hypothesis/where the burden of proof lies) to make it simple: we see a badly performing speaker and can't find any excuses for the main issue, so it makes perfect sense to ascribe this issue to a design fault. Which then makes it perfectly normal to view with suspicion other speakers designed by the same brand (which is responsible for the design and individual speaker QC).
One might also talk about the directivity mismatch, but I seem to remember that the designer cares more about directivity width than matching (which is a bit of a shame, honestly).

Neither basic logic, nor science, assume that when you have a sample size of one, that you should extrapolate the data to an entire group. All you have is a sample of one. Get to three or four and you may start to see a trend.

This is only true in processes where you try to prove the link between cause and consequence, not when you know it (good designer and good QC -> no speaker with such a blatant mistake would be left as "finished").

Using your "basic logic" if you stayed at a hotel an the maid forgot to empty the trash in your room when cleaning, then no trash in any room was empty, the trash in the lobby is piling up, and the restaurant is a disgusting mess full of garbage.

This analogy is obviously wrong because we know there's only one maid and because the problem isn't "forgetting" here, an engineer doesn't "forget" to measure his end result, that's essential to his role. The hotel with an only maid leaving a pool of vomit in the center of a room would be more fitting.

The only thing the sample size here tells me is to not write off the brand completely and wait for more to possibly reach "redemption" in my eyes. But I certainly won't consider it until that when there are brands like Genelec that got validated by multiple self and third-party measurements.

 

[snapsc]

It would be interesting for every poster on this thread to reveal how many Salk models they have heard... and whether at a show or in a home.

I’ve heard three models... all in a
home environment... two towers, one bookshelf. All three sounded really good in the system/room/application they were used in.

I’d be happy to own any of them... I do currently own the Salk Veracity ST which is in a 2 channel system...very happy with sound and no expectations of making a change any time soon. And yes, the cabinets are a work of art... which in itself provides a lot of smiles.

 

[CDMC]

The burden of proof isn't only a legal term. I'll use the closely related Occam's razor to make it simple: we see a badly performing speaker and can't find any excuses for the main issue, so it makes perfect sense to ascribe this issue to a design fault. Which then makes it perfectly normal to view with suspicion other speakers designed by the same brand (which is responsible for the design and individual speaker QC).


This is only true in processes where you try to prove the link between cause and consequence, not when you know it (good designer and good QC -> no speaker with such a blatant mistake would be left as "finished").


This analogy is obviously wrong because we know there's only one maid and because the problem isn't "forgetting" here, an engineer doesn't "forget" to measure his end result, that's essential to his role. The hotel with an only maid leaving a pool of vomit in the center of a room would be more fitting.

The only thing the sample size here tells me is to not write off the brand completely and wait for more to possibly reach "redemption" in my eyes. But I certainly won't consider it until that when there are brands like Genelec that got validated by multiple self and third-party measurements.

Occam's Razor only says when you have multiple alternative solutions, choose the simplest one. Again, you have taken a sample of one, decided that issue is an engineering problem, and then jumped to the conclusion that all of the products by the manufacture are poorly engineered. There are huge gaps in your logic, not to mention your conclusions.

P.S.- I am not sure if English is your first language, but if not, it is excellent.

 

[q3cpma]

Occam's Razor only says when you have multiple alternative solutions, choose the simplest one. Again, you have taken a sample of one, decided that issue is an engineering problem

This is where Occam's razor is applied.

and then jumped to the conclusion that all of the products by the manufacture are poorly engineered.

Absolutely not, this is where the chosen axiom intervenes: combine it with the fact that good designer and good QC => no speaker with such a blatant mistake would be left as "finished", you apply a modus tollens on this, you get not good designer and good QC <=> bad designer or bad QC. After that, it's just applying the knowledge that the same people are doing all of the Salk speakers to the situation.

There are huge gaps in your logic, not to mention your conclusions.

Well, point them out. The only "gap" isn't big enough for me to consider it right now: the only possible excuse would be damage after the sample left the factory.

P.S.- I am not sure if English is your first language, but if not, it is excellent.

Thanks.

 

[Rick Sykora]

The body of work produced by Salk has some very nice designs and many are generally highly regarded. I felt it might help to add that I never fully appreciated what can happen to equipment after it is shipped. Until I started shipping speakers to Amir, I barely worried about whether they would get knocked around in transport...

In the past, I might have glued a crossover component, but rarely did a tie-wrap it. I have never really cared for anchoring damping materials in place and avoided using adhesive-backed foam, but now I do. I have posted elsewhere that I am often putting more effort into designing the shipping box than I did for building the speaker. While this applies not just to Salk, but more broadly, please remain open to the fact that unanticipated stress may affect the speaker that just arrived on your doorstep! For that matter, few have Klippel and not every product that gets shipped, gets the scrutiny that you think it does/should.

Just for starters, when I do my pre-shipment testing, my FR test resolution is about 300 Hz. I have said it before, I cannot hope that has a much information as Amir's (Klippel resolution at .7 Hz). That said, he is going to find stuff that I never would. Does that make the build bad or me a bad builder, maybe, but I appreciate the chance to respond. Until we fully understand the Klippel measurements and acknowledge that the speaker may have been roughed up after it shipped, please let's give Salk (and any other supplier) the benefit of the doubt and hold final judgement until they have a chance to fully respond.

 

[Krusty09]

Well not sure about if kali speaker was damaged at the factory or not but it was damaged and after that whole thing happened kali got a second review and it seemed to do well. Not saying this is gonna be the case here but crazy stuff happens. And if my memory is right Kali seemed to be very professional through it and Jim seems to be as well so far.

Time will tell.

 

[MattHooper]

This is where Occam's razor is applied.

Absolutely not, this is where the chosen axiom intervenes: combine it with the fact that good designer and good QC => no speaker with such a blatant mistake would be left as "finished", you apply a modus tollens on this, you get not good designer and good QC <=> bad designer or bad QC. After that, it's just applying the knowledge that the same people are doing all of the Salk speakers to the situation.

Well, point them out. The only "gap" isn't big enough for me to consider it right now: the only possible excuse would be damage after the sample left the factory.

Thanks.

I think your application of the Burden Of Proof is proving a bit awkward.

The burden of proof simply stated puts the onus on the one making the claim to justify the claim.

The question then is who is making a claim about what, and how is it being justified? You mentioned burden of proof, and then made a claim.
You claimed that, in the example of the Salk WOW1, we ought to conclude from one "bad" set of measurements that "it's just basic logic to consider the whole brand as bad until more data is made by the same measurer."

Well, that's a fine start. You made a claim then gave your justification. And then you invoked Occam's Razor, which tells us to select the the simplest explanation that accounts for the phenomenon to be explained, without multiplying entities unnecessarily. Using this rule, I understand you to be saying that we ought to accept your (provisional) conclusion that "the whole brand is bad" because the simplest conclusion to draw from one poorly measuring speaker is best explained as being the result of a poor speaker designer. (It could also be just a basic inductive inference too IF the WOW1 were your only measured reference).

The problem is that all sorts of priors can't be ignored in an explanation. There's all sorts of data attached to Salk speakers that have to be accounted for in this "simplest explanation." For instance: Dennis Murphy is known as an excellent designer, particularly with crossovers, and we have seen even on this site, measuring his BMR monitors, that they display very competent speaker design.

And yet your explanation makes a claim about the ENTIRE BRAND.

Remember in Occam's razor you have to have an explanation that accounts for all the relevant data, and part of that Data is that someone who has produced very competent designs works on many of the speakers in that brand. So how does the explanation of a poor speaker designer account for THAT?

It doesn't. At this point.

 

[q3cpma]

I think your application of the Burden Of Proof is proving a bit awkward.

The burden of proof simply stated puts the onus on the one making the claim to justify the claim.

The question then is who is making a claim about what, and how is it being justified? You mentioned burden of proof, and then made a claim.
You claimed that, in the example of the Salk WOW1, we ought to conclude from one "bad" set of measurements that "it's just basic logic to consider the whole brand as bad until more data is made by the same measurer."

Yeah, you're right, my usage of the expression "burden of proof" was quite clumsy. Meant that the burden of proof lies with future speakers to disprove this hypothesis, but I shouldn't have invoked this, it made my whole post less intelligible.

Well, that's a fine start. You made a claim then gave your justification. And then you invoked Occam's Razor, which tells us to select the the simplest explanation that accounts for the phenomenon to be explained, without multiplying entities unnecessarily. Using this rule, I understand you to be saying that we ought to accept your (provisional) conclusion that "the whole brand is bad" because the simplest conclusion to draw from one poorly measuring speaker is best explained as being the result of a poor speaker designer. (It could also be just a basic inductive inference too IF the WOW1 were your only measured reference).

Please notice that I also included bad QC as a possibility.

The problem is that all sorts of priors can't be ignored in an explanation. There's all sorts of data attached to Salk speakers that have to be accounted for in this "simplest explanation." For instance: Dennis Murphy is known as an excellent designer, particularly with crossovers, and we have seen even on this site, measuring his BMR monitors, that they display very competent speaker design.

This is indeed something to consider, but the crossover doesn't seem to be the problem and porting, especially front porting, is something that can easily go wrong without the necessary R&D. Honestly, I'd rather put this on QC's head, but it must be pretty hard to botch a port during manufacture.

And yet your explanation makes a claim about the ENTIRE BRAND.

Remember in Occam's razor you have to have an explanation that accounts for all the relevant data, and part of that Data is that someone who has produced very competent designs works on many of the speakers in that brand. So how does the explanation of a poor speaker designer account for THAT?

It doesn't. At this point.

You usually choose the result of Occam's razor to formulate a complete (albeit easily changeable) position when you have nothing left and must take a decision - which is the case in this very competitive market.
I mean, if anyone has a better explanation, I'm attentive, but right now, I don't see many possibilities.

 

[Xulonn]

When I see that, I know they are taking the time to perfect each model.

Really? I know that Burchart speakers are quite good, but no loudspeaker is "perfect". Indeed, based on Amir's measurements and review of the Burchardt S400, I would say they are heading in a good direction, but still far from perfection.

Technically, perfection would be a speaker that is ±0.0db from 20Hz - 20kHz across a listening area. Unfortunately, acoustics in any space will not permit that, even if transducers could produce perfect output.) Perfection means no further improvement is possible, and "perfection" is not possible in the design of loudspeakers. Engineering and design staff can only balance variables with compromises that make a range of sonic "flavors" available. Even if 85% of random groups with verified good hearing who are selected for testing by Harman and others prefer a particular sonic profile, that does not mean that the likely range of "preferences" associated with the other 15% are based on "imperfections" and therefore invalid.

Quite frankly I subjectively see Burchardt speakers, in spite of their good performance, as unattractive, sharp-edged monkey coffins that will appeal only to those who like modern styles and interior decor - which is a lot of people. However, they will likely not appeal to those who prefer more organic or traditional decor. OTOH, I consider Salk speakers with their exquisite veneer finishes to be absolutely beautiful. If my budget were not so limited, I very well might have ordered a pair of Salk Song Towers and center speaker with a custom veneer.

It should be obvious from my comments that I love beautiful wood as much as I like excellent music from my audio system. Below is a picture of "Oasis", my 2011 award-winning African bubinga and quilted big-leaf quilted maple woodturning piece. That turning sits on my media stand between my Wharfedale D320s with their faux rosewood enclosures with their rounded edges and corners.)