Zaph Audio ZA5.2 DIY Kit Speaker Review

[DDF]

What have you done?

Electrical Engineer, I worked at Bell Northern Research/Nortel for years (Canada's version of Bell Labs at the time) as an audio design lead
designing sound rooms, running blind psychoacoustic experiments, designing audio electronics, DSP audio algorithms, audio measurement methodologies for standards bodies, mic/speaker integration etc (also inventor/designer of audio for VoIP in its first years). Designed numerous home speakers over the years since the late 70s, on and off. This was written in 2005:
http://www.helarc.com/guru/gurusfarra.htm (list of others here http://www.helarc.com/ )

Been in fiber optics for years but still play in audio design/theory/research for kicks. Its too addictive.

 

[DeruDog]

I find it interesting how well this speaker scored, but how poorly it was reviewed, or how bad it subjectively sounded. Is there a problem with the Preference Score algorithm that we need to adjust? Or in a blind test would it have done better?

 

[GelbeMusik]

I find it interesting how well this speaker scored, but how poorly it was reviewed, or how bad it subjectively sounded. Is there a problem with the Preference Score algorithm that we need to adjust? Or in a blind test would it have done better?

I explain the result to me with the demand for output level. A 5" bass can't do it. In DIY it is quite common to overestimate the capabilities of small speakers. The tuning is often too low as to support the tiny warrior against cruel physics. Even with top grade 7" woofers a real full range peak SPL of about 84db@ listening position, stereo are hard to reach. If some quality shall be maintained.

The exaggerated presence around 3kHz in sound power doesn't help either. But, different from OEMs a DIYer rarely would give up linear on axis response for a smoother over-all sound field.

 

[DeruDog]

I explain the result to me with the demand for output level. A 5" bass can't do it. In DIY it is quite common to overestimate the capabilities of small speakers. The tuning is often too low as to support the tiny warrior against cruel physics. Even with top grade 7" woofers a real full range peak SPL of about 84db@ listening position, stereo are hard to reach. If some quality shall be maintained.

The exaggerated presence around 3kHz in sound power doesn't help either. But, different from OEMs a DIYer rarely would give up linear on axis response for a smoother over-all sound field.

Is this an argument to adjust the Preference Score, then? Maybe weighting the low end output level in some way? It does seem that the rest of the speakers that score well have 6" or more woofers in them

 

[GelbeMusik]

Is this an argument to adjust the Preference Score, then? Maybe weighting the low end output level in some way? It does seem that the rest of the speakers that score well have 6" or more woofers in them

Always quoted, Dr. Toole says, that bass extension and level of output level were a matter of cost. You have to adjust. By this quote one might think, that the price could be used as a second criterion? Another scalar number. So, together with the spinorama preference evaluation figure the choice becomes two dimensional

For instance--to this I was referring to with mentioning the common DIY mistake--a deep bass can only be heard, let alone experienced to the fullest, if some elevated level is available at the same time. The well known Fletcher/Munson curve implies a compromise of bass extension and volume capability in the bass. It has to always have both, level and deep to be successful. Especially deep without level is wasted effort.

I only suspect that this is the case here.

 

[Maiky76]

Hi,

Here is the EQ design for the ZA5.2.
This most probably not applicable to the rest of the ZA5 family.


Score with No EQ: 5.69

Score with EQ: 6.38

Since the main grief was bass extension I decided to add a HP filter to boost the 45-120Hz by up to 2.75dB.
The dynamic range of the speaker will therefore be decreased by the same amount.
So caution applies there...
On the other hand, the HP filter will somewhat protect the speaker at very LF.

EQ design:

Type         freq       Gain        Q
HighPass     49.5,     0.00,   1.30,...
PEQ         192.0,    -1.50,   0.88,...
PEQ         491.0,    -1.40,   22.0,... % to be verified with DUT
PEQ         873.0,    -2.52,   5.50,...
PEQ           1418.0,    -1.05,   4.56,...
PEQ          2651.0,    -1.84,   3.60,...
PEQ       14760.0,     0.80,   1.75,...

Experimenting with a BBC dip the score rises to 6.62 but LW/ON compromised to tame the off-axis excess of energy.
Listening test required, data not shown:
PEQ        4551.0,    -1.50,   1.60,...

The rest of the analysis is enclosed.

 

[Kustomize]

I am always comparing data of all these speakers to my own Zaph speakers. I might be biased, but I still want to understand what I am missing. This is a comparison of JBL 705P measured by Amir here, which was well received by him. The Zaphs I have are towers, with dual woofers, but the 5.2 here that Amir tested I guess the major issue seemed to have been low to no sub bass - mid bass. But I was comparing and found there to around 1db (or even less at times) variation amongst the whole frequency spectrum.


I am not bashing Amir's subjective OR objective testing and results, I really appreciate all the work he puts in them(thankyou!), I just want to understand better what I am missing in this comparison to learn how to compare speakers better next time!

 

[Ericglo]

The other issue seemed to be the low sensitivity. Despite those two, it still measured quite well. It would be interesting to see the difference in bass and distortion with multiple woofers.

 

[Kustomize]

The other issue seemed to be the low sensitivity. Despite those two, it still measured quite well. It would be interesting to see the difference in bass and distortion with multiple woofers.

Thankyou for that. Answers like this is exactly why I posted. So I noticed the difference in distortion (in bass). Definitely there. I have 2 mids, my design is the Zaph ZA5.3T

http://zaphaudio.com/ZA5/

 

[Ericglo]

I believe I stayed out of this discussion. The criticism may be warranted but it is a decade old design. Plus when Zaph introduced the speaker design and woofer, he said he was emphasizing the midrange.

You compared it to the 305, which is a good inexpensive speaker that I own. Ten years ago I don't recall anything that would have been close to this performance for the price.

It looks like the sensitivity of the 5.3 is 86 or 87. I am sure that would go a long way to eliminating the complaints about sensitivity

 

[Kustomize]

I believe I stayed out of this discussion. The criticism may be warranted but it is a decade old design. Plus when Zaph introduced the speaker design and woofer, he said he was emphasizing the midrange.

You compared it to the 305, which is a good inexpensive speaker that I own. Ten years ago I don't recall anything that would have been close to this performance for the price.

It looks like the sensitivity of the 5.3 is 86 or 87. I am sure that would go a long way to eliminating the complaints about sensitivity

Not sure if you know but there is another design done by Michael Chua using the same drivers as these. Its got distortion plots and others. I wonder if this would improve on what Zaph worked out years ago!

https://ampslab.com/blog/2020/04/02/swift-25/

would it be worth changing crossovers? Probably cost around 180 dollars.

 

[Ericglo]

I am not sure how much or if any improvement there will be. You might be better off having Amir or Erin measure your speaker to see how it compares.

 

[ta240]

Is this an argument to adjust the Preference Score, then? Maybe weighting the low end output level in some way? It does seem that the rest of the speakers that score well have 6" or more woofers in them

The other issue seemed to be the low sensitivity. Despite those two, it still measured quite well. It would be interesting to see the difference in bass and distortion with multiple woofers.

I think it is largely a matter of knowing what you are getting. If you are expecting lots of bass and high efficiency then you'll be let down. If you are expecting clear midrange and it is enough efficiency for your listening level then you'll be fine. For listening to dialog in TV and movies these are super clear. I used to have to turn up the volume when people were talking and back down when the explosions started in movies but I no longer find myself going "what did they say?" if I don't turn up the volume. In another setup with the reduced baffle step crossover and the speaker up against a wall I get plenty of bass for me. But that setup isn't a 'feel the music' or 'share with the neighbors' type setup either.

If speaker listening tests can be impacted by equalization of the input signal then shouldn't they also be able to be impacted by adding a subwoofer and a receiver with a built in crossover? Audioholics says always set your speakers as small when setting up your surround receiver so it will send all the bass to your subwoofers and it makes me wonder 'why buy big speakers with lots of bass if you aren't going to send the bass to them anyways?'. So just cross these over at 100 or 120 and it won't task them with struggling to produce frequencies they aren't made to do.

 

[H-713]

IMO, there should be some consideration in the weighting system about what a particular speaker was designed for. If you read the design page on these, he emphasized midrange performance over bass. He also noted that they will sound good full range at low to medium levels. Furthermore, there is significant time spent discussing subwoofer integration, implying that he expects these to be used with a subwoofer for any sort of HT application.

I suspect that the review would have looked very different if a pair of these were tested at medium volume levels in a relatively small room, or supplemented by a subwoofer in a large room. Is it really fair to criticize a product for not doing something it wasn't designed to do? If I were to review a Honda CR-V, would it be fair for me to dock points because it can't tow my 4,000 pound boat?

Sometimes having limited bass capabilities isn't a bad thing, especially if it's going in a small room with issues at those frequencies. When I stayed at one of the dorms at my university, which is of new construction, a pair of 5" 2-ways (admittedly with slightly above average bass capabilities) were enough to be overwhelming, even at moderate volume levels. This would likely be a pretty good speaker for that sort of application.

 

[Kustomize]

In all honesty, i compared a lot of frequency responses on ASR and preference scores and everything. I never understood why this Zaph wasn’t recommended. Measured well. Takes plenty of power before bottoming out. I have the floorstanding with 2 5” so I can’t complain much about not getting loud enough as it does for me. This system states on website that subs are recommended for higher levels of output. I think if one could simply EQ this speaker at 125hz by like 2db. This could be improved. We always recommend speakers and their recommended EQ. Why couldn’t we with this? Am i missing something again?

 

[moonlight rainbow dream]

It's been stated already, but there's simply a mismatch with expectations here. Zaph expressly states in his design goals, these sacrifice bass output and extension for smooth and low distortion midrange and are intended to be used with subwoofer. Amir, OTOH, wants standmounts to deliver satisfying bass response on their own. He has also stated in other reviews that he believes subwoofers are too difficult to integrate for the layman and so leans toward full-range solutions.

 

[ta240]

View attachment 54616

What am I missing with the above impedance measurement and the below chart from his site? Different types of measurements?

 

[maverickronin]

What am I missing with the above impedance measurement and the below chart from his site? Different types of measurements?

It's mostly just very different scaling.

 

[ta240]

That is funny, the look crazy different but you are right.

 

[RickS]

What am I missing with the above impedance measurement and the below chart from his site? Different types of measurements?

View attachment 128140

Scaling is important, but it is not the whole story. You will get differences in impedance measurements from these too:

1. Differences in measurement equipment (different calibration resistors may be used for example)
2. Differences in drive voltages during measurement (see Erin's impedance measures taken at different levels)
3. Differences in the amount of damping or type of damping material used or how it is applied
4. Differences due to other cabinet losses (leaks around joints, screws, drivers)
5. Differences between types of cabinet materials or other sources of resonance (usually show up as a blip in the impedance curve)
6. Differences due to crossover component tolerances/defects
7. Differences due to drivers (can be both batch specific and/or other changes over time)

Sure I may have forgotten something, but in BR tuning, it is more essential that the peaks and valleys occur close to the intended frequencies. The magnitude of the impedances it less critical. For upper frequencies, major differences may indicate a driver or a crossover issue.