Swissonic A305

[Weeb Labs]

took some measurements from 3 different A308 active speakers:

Could you possibly post these measurements with a 50dB vertical scale, a 20Hz to 20KHz horizontal scale and 24dB/oct smoothing?

 

[ButterflyFusion]

A-series as well as the X8 use the 1701, but only the A-series can be reprogrammed permanently with SigmaStudio. The three pins to do so are still on the board of the production version. DAC0 and DAC1 are connected to LF as + and -, 2&3 to HF.

Thanks for the info! Could you or someone please let me know or direct me to somewhere i could learn how to do this?

It looks to me like the A308 has a (DSP) LP filter on the top end about 25Khz, id like to move this further up so I can reproduce 20-40k if needed, like many other speakers can play that range out of the box.

Also the standby settings programmed on the speakers act differently on every unit, one forgets to activate standby most of the time, the others seem to respond to differernt volume thresholds etc. i would love to disable the standby / change the timer to 2hrs / change the threshold so it doesnt activate due to self noise, unless the audio interface is OFF.

 

[Weeb Labs]

Thanks for the info! Could you or someone please let me know or direct me to somewhere i could learn how to do this?

It looks to me like the A308 has a (DSP) LP filter on the top end about 25Khz, id like to move this further up so I can reproduce 20-40k if needed, like many other speakers can play that range out of the box.

You can find detailed documentation for Sigma Studio here and some hardware configuration instructions here. The ideal USBi interface is the TinySine unit.

There is no intentional low pass filter. The ADAU1701 is most likely operating at its default sample rate of 48KHz, which will necessarily roll off below 24KHz. You should be aware that there are no audible benefits to increasing the high frequency response beyond 20KHz; only disadvantages.

 

[ButterflyFusion]

Thanks! Ill check that out. Its gonna take me a while. What disadvantages would I possibly encounter if I increased the ADAU1701 sample rate to 96K to match what the interface is outputting?
could the IC handle it? ill check the datasheet to see if its possible.

I tried to export the graph again, but im new to REW and not sure if this is what you requested:

(tried again)

 

[ButterflyFusion]

the chip can theoretically do it:

-28-/56-bit, 50 MIPS digital audio processor
-Sampling rates of up to 192 kHz are supported

I would like it to run 32bit & 96K

 

[VintageFlanker]

took some measurements from 3 different A308 active speakers:

Vertical scale is waaaay too high to observe anything.

By the way, Erin does have both the 306 and 308 to be reviewed soon. The measurements look great so far.

 

[ninetylol]

Vertical scale is waaaay too high to observe anything.

By the way, Erin does have both the 306 and 308 to be reviewed soon. The measurements look great so far.

He said He doesnt have any plans for the 308. My 306 just came in today though! Had some trouble with them at start though, one wasnt outputting any signal, no matter what I tried. After 5 minutes it suddenly worked. Strange.

Also i still dont fully understand the sensitivity. Manual says 320mv. Backside says - 10dbv for unbal and +4dbu balanced. But then theres also a switch to change it (for both bal / unbal i dont understand)

 

[Weeb Labs]

the chip can theoretically do it:

I would like it to run 32bit & 96K

The ADAU1701 is capable of 1024 instructions per frame at a sample rate of 48KHz. Doubling the sample rate to 96KHz halves the available instructions per frame to 512. This effectively halves the DSP resources available for your program, which may or may not present a problem. In addition to this, you are potentially exposing the amplifier and drivers to ultrasonic schmoo which may introduce IMD and spurious signals that worsen performance in the audible band.

The ADCs and DACs in the ADAU1701 exhibit a dynamic range of 16 bits, while DSP computation is carried out at 56 bits. No matter your audio source, the output can never exhibit more than 16 bits of dynamic range.

 

[ButterflyFusion]

Ok thanks for the info. Im not aware of any DSP resources needed apart from the crossover, but it seems a bit pointless to mess with the high end.

The standby mode is definitely worth trying to fix, its a mess

 

[Weeb Labs]

Ok thanks for the info. Im not aware of any DSP resources needed apart from the crossover, but it seems a bit pointless to mess with the high end.

The standby mode is definitely worth trying to fix, its a mess

The automatic standby is a function of the amplifier rather than the DSP. You would need to identify the amplifier IC and check its datasheet. There is usually a pin that can be pulled high to disable the automatic standby.

 

[thewas]

Erin's measurements of the A306 are now online

Thomann/Swissonic A306 - Review & Measurements by Erin

Text data to be updated... PS: Seems like a very, very competitive offering, especially for EU buyers. Can't really see any flaws here: even the compression and THD are more than OK. Honestly, I used to look down at these speakers : Not a name in the industry, probably no experience or...

www.audiosciencereview.com

 

[KLang1]

He said He doesnt have any plans for the 308. My 306 just came in today though! Had some trouble with them at start though, one wasnt outputting any signal, no matter what I tried. After 5 minutes it suddenly worked. Strange.

Also i still dont fully understand the sensitivity. Manual says 320mv. Backside says - 10dbv for unbal and +4dbu balanced. But then theres also a switch to change it (for both bal / unbal i dont understand)

I doubt he has plans for the 308. I sent him the A305, A306 and A203BT.

 

[vesalaasanen]

That V5 directivity looks amazing. I really hope this gets into Klippel's scanner at some point. I'm considering one of these models for studio Atmos setup.

 

[jaakkopetteri]

A-series as well as the X8 use the 1701, but only the A-series can be reprogrammed permanently with SigmaStudio. The three pins to do so are still on the board of the production version. DAC0 and DAC1 are connected to LF as + and -, 2&3 to HF.

Do you use the internal ADC/DAC of the 1701 or separate ones, if I may ask? Great job with the speakers

 

[sweetchaos]

So apparently this woofers sensitivity is increased compared to the sample speaker I used to set up the DSP. I don't like it, but that's what you get when you don't want to pay big dollar for better QC.
Fortunately we added high (2kHz, Q0.35, +-2dB) and low (110Hz, Q0.7, +-2dB) shelf filters to adjust the speaker to ones taste via the dip switches on the back. With the low shelf set to -2dB it is possible to mostly remedy the increased sensitivity of the woofer. The attached measurement is already with this -2dB low shelf filter active. I wish I had set it to +-3dB
Also I have to add another disclaimer: It's really tricky to merge the near- and farfield measurements of a 3-way speaker. The farfield measurement was corrected for groundplane effects and the nearfield measurement of both midrange and woofer was corrected for baffle step. I'd love to see a real anechoic measurement or a nearfield scan extrapolation of the X8, but for now this stitched measurement is the most accurate thing I can provide.

new spin @pierre

 

[dickiefunk]

I’m very interested in the Swissonic monitors and am wondering how they all compare in performance.
I’ve seen Erin’s recent review of the A306’s and they seem to be impressive at their price range. Do any of the other Swissonic monitors outperform the A306’s in flat response, low distortion, directivity etc?

 

[DanielT]

Speaking of Swissonic on a question, justified I think. about why you should even do DIY when they exist:

Other than to reduce the high and largely third-tone-dominated distortion in the middle register, of course. Especially about 3% third tone in combination with fifth tone around 250 Hz already at relatively moderate sound pressure levels will be heard clearly.

faktiskt.io • Visa tråd - Högtalarbygge för mindre pengar....frågor och tanke.

Edit:
It should be added that the comment is made by a professional, ordinary mortal with limited awareness and thus the ability to perceive distortion, it probably doesn't matter.
He has carried out a lot of blind tests over the years so I have no doubt that he can spot this distortion in any case.

 

[testp]

These are the control measurements from the end of the tuning process. I would love to offer something more precise, but ground-plane measurements are the best I can do in our workshop.

from measurements, V7 looks pretty good, if i have my own peq for room, would you say V7 is as good as A306 or better/worse? or are there some hidden benefits with A-series/X8.. (maybe time alignment?)

 

[dannut]

"Other than to reduce the high and largely third-tone-dominated distortion in the middle register, of course. Especially about 3% third tone in combination with fifth tone around 250 Hz already at relatively moderate sound pressure levels will be heard clearly."

Correct, but. This comment needs to be in context of the size- and price-class.

3% and 10% THD in itself doesn't tell us much, electroacoustic systems are usually incapable of generating high-order distortion in their linear operating range - not enough force.
Due to masking, the generated distortion is quite bening with typical program material. Until it isn't. Then linear range is exceeded.

But the 'linear range' seems to be shockingly high for A306, judging by Erins data. (Measurement error perhaps? Wish manufacturers start to publish AES75 max output levels to compare) Different class than the competing JBL306.

By-the-way, best in class (3 and 10% distortion limited) output award goes to JBL705p/i and Neumann KH150 (data from Sound&Recording).
This one rivals them on output, but is not as clean when driven hard. If excursion is kept at check (>80Hz highpass), then probably Le(i) distortion component keeps it from competing with the best. Proper inductance control costs money (copper cap/shorting ring?)
You could hear the issue on loud passages when A-B comparing with a cleaner trancducer as an added 'grunge', and could be annoying with synthetic test signals (who listens to that??).
So not ideal for SOTA sound production use-cases. But for regular listening? No annoying resonances, correct tonality, no output limiting are vastly more important parameters then couple dB higher low-order distortion. All for EUR100? Maybe the port-resonance could be improved in the future? Anyway kudos @KLang1 !

My only gripe is that there is no A304. Neumann KH80/Genelec 8020 needs some low-cost alternatives.

 

[DanielT]

Correct, but. This comment needs to be in context of the size- and price-class.

3% and 10% THD in itself doesn't tell us much, electroacoustic systems are usually incapable of generating high-order distortion in their linear operating range - not enough force.
Due to masking, the generated distortion is quite bening with typical program material. Until it isn't. Then linear range is exceeded.

But the 'linear range' seems to be shockingly high for A306, judging by Erins data. (Measurement error perhaps? Wish manufacturers start to publish AES75 max output levels to compare) Different class than the competing JBL306.

By-the-way, best in class (3 and 10% distortion limited) output award goes to JBL705p/i and Neumann KH150 (data from Sound&Recording).
This one rivals them on output, but is not as clean when driven hard. If excursion is kept at check (>80Hz highpass), then probably Le(i) distortion component keeps it from competing with the best. Proper inductance control costs money (copper cap/shorting ring?)
You could hear the issue on loud passages when A-B comparing with a cleaner trancducer as an added 'grunge', and could be annoying with synthetic test signals (who listens to that??).
So not ideal for SOTA sound production use-cases. But for regular listening? No annoying resonances, correct tonality, no output limiting are vastly more important parameters then couple dB higher low-order distortion. All for EUR100? Maybe the port-resonance could be improved in the future? Anyway kudos @KLang1 !

My only gripe is that there is no A304. Neumann KH80/Genelec 8020 needs some low-cost alternatives.

I agree with you in what you say and I think I-or missed price-performance aspect, which is good.

You don't get SPL monsters that can play at distortion-free high volume. But at lower to normal volume, why not? Then if and when this distortion becomes audible is another matter.It is an individual thing, where people like I-ors with long professional experience working with detecting it are more sensitive. It's an exercise in habit, so to speak.

You and I-or are saying pretty much the same thing. Here's his continuation on the subject of these speakers, and other fairly small inexpensive active speakers:

We've covered this with sound pressure levels many times before. Now let's put the question aside for the last time. There is a difference between average levels and maximum levels.

What most people measure with various sound level meters or software is Leq, i.e. medium level. In addition, A-weighting gives approx. 3 dB lower sound pressure levels than what is obtained linearly (unweighted) for reasonably typical music signals. With a peak factor of 20 dB for a decent dynamic recording, you land with Leq of 80 dBA at maximum sound pressure levels (RMS) of approx. 80+20-3+3 = 100 dB (Leq+peak factor+peak to RMS+A-weighting).

If we assume that this applies at a distance of 1 m, it turns out practically enough that in the listening position for two loudspeakers you get the same sound pressure level in a reasonably typical room (for not too short peaks). Certainly not all signal power is localized to the 250 Hz range, where the most hearing-unfriendly distortion is at its highest in this case, but usually you have some short-term peak of maybe -5 dB relative to the overall level right here. We then land for these speakers in the 250 Hz octave band of approx. 95 dB and approx. 2% THD with dominance of the third tone and fifth tone (very high audibility), which is not only audible but directly disturbing.
Then you should also keep in mind that this is a rather moderately increased volume and that most people are happy to add perhaps another 10 dB higher levels if they have the opportunity. In that case, the distortion would be completely unbearable for anyone (however, the amplifier power is not enough for this here).

This should come as no surprise to anyone who has heard a number of small and inexpensive speakers of this type. They often sound acceptable up to medium sound pressure levels, but in principle always far too distorted when you want to increase the volume a little more.

In addition, active, cheap, small speakers are without exception under-dimensioned in terms of amplifier power and often equipped with over-aggressive protection circuits.

That is the question the distortion would be completely unbearable for anyone. It regardless of practice to detect distortion? Incidentally, one of the most frequent topics/ the threads here on ASR. When does distortion, noise become audible that is.