Edifier S350DB review and measurement

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
Rushing this out before 9.9 for the people who are deciding. Thus the review will be coming in parts.

The manufacturer provided their measurement on the box. Variation of 5dB is pretty decent, although the shape is not so decent, particularly the 10kHz peak. This kind of shape is easy to remove with EQ tho, and depending on how the off-axis looks like, we may not even need to remove it.
DwYv3ij.png


Let's see how close we get to this. Note that my own measurements are unlikely to be comparable to a professional's, so any difference needs to be taken with salt. Best case is we get close enough.

My room is too live to do gated measurements, so I will use "Table edge" because that is what I found to be the closest during prior experiments. I can still do gated above 1kHz but the non-gated is close enough for the treble frequencies. Bass reflection is simply impossible to avoid so anything below 500Hz please use your own heuristics to remove the peaks and dips.

1/6 octave smoothing seems close enough to the above graph. Can really see my room murdering the FR.
ESWPxxs.png


The treble peaks I get are actually not as bad as the manufacturer data. Going 1/24 smoothing does not make the 10kHz peak any much worse.

Bass is also largely matching. Subjectively you get usable bass down to ~42Hz and then it fades into the darkness. Note that positioning can change the result by a few Hz.

0 degrees (on-axis) vs 45 degrees
wpYzrgZ.png


Handles off-axis surprisingly well. Actually the off-axis might be even better.

Off-axis
ystHpGr.png
Going off-axis you lose volume in the HF, but that's the same with most speakers anyway. In fact S350DB handles being off-axis very well. Relative speaking vs the on-axis. That treble rolloff in either case tho...

Close mic of tweeter woofer port. Note that close mic can cause woofer bass to appear more than it really is.
aizbALE.png


Pretty high crossover frequency. Treble variation is indeed within 5dB.
The woofer starts to roll off at around 200Hz. But we have a port to do magic. The final result with sub is actually pretty flat.

That zig-zag at 1.2kHz is the cause of the treble peak, which got hidden by smoothing.

Distortion
92jrJCf.png
Doesn't look like any major driver or crossover issue. Note that I test at much lower SPL than most people.

Max vs normal vs min tone adjustments
LhQ30vy.png


Unfortunately, there is no subwoofer volume adjustment. The bass knob is a tone adjustment. There is no way to shift just the subwoofer volume up or down.

The treble control can be useful to tame the treble peak if you're on axis. If you're off-axis the off-axis-ness already takes care of it, but you can add it back with the treble control.

Bass seems unbalanced? The middle setting seems to be above the middle point of the adjustable range. Looks okay at the satellites frequencies tho. Maybe the woofer has compression? Seems to happen awfully early for an 8-inch sub tho.

In fact the 40Hz rated bass is awfully high for an 8-inch sub. Subjectively my bass-boosted Usher S-520 is slightly better than this, and z623 is slightly better than the S-520, so... erm... However I also run the z623 bass at near full if not full setting, versus S350DB which already sounds balanced at middle or slightly lower setting, so the S350DB has way more headroom that you can use to squeeze out another 10Hz via EQ if you want.

Front of subwoofer, mic on floor (to avoid floor bounce)
uwLL3vx.png
 
Last edited:

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
So how does all this translate to real life?
wojXs6o.png


Amazingly pretty decent. Unfortunately it does not have the bass extension I expected from a sub of this size, and the treble extension simply falls off too early. But it does not have a frequency hole between the satellite and woofer, and overall frequency is flat enough at default setting. And you can achieve decently reliable performance while having flexibility with positioning, unlike with most bigger bookshelves. The crossover from satellite to subwoofer is *almost* low enough that I *almost* could not hear bass coming from one side. Which from memory means it is better than other 2.1 I have heard.
 

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
The reason why to me that 10kHz doesn't affect too much is because a normal speaker would drop that much when it's that off-axis anyway.

S-520 on-axis vs 45 degrees
NtS38m4.png


That's also why I said the S350DB actually has pretty consistent off-axis performance.
 

benedium

Banned
Joined
Jan 22, 2005
Messages
12,569
Reaction score
3,782
Wah am sure this is very useful for everyone. So very kind of you wwenze.
 

benedium

Banned
Joined
Jan 22, 2005
Messages
12,569
Reaction score
3,782
I see Lazada price is now $399. wwenze, for close to same budget of around $400, would edifier r1850db plus wharfedale subwoofer be better value and provide higher quality sound? Although it does look like it would take up more floor and desk space.
 

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
Normally the advantage of 2.1 is the sub, and with a normal stereo sub you lose all the control and source functionalities on the speakers. But R1850DB is really an oddity with a sub out. :s13:

Frequency response is hugely in R1850DB's favor. Slightly better treble, but bass goes below 100Hz.
Subwoofer bass seems to be WH-S8E getting the win too. And u get the ability to adjust subwoofer volume.
The only thing that can possibly go wrong is the HPF of the sub-out not optimal matching, but that can be fixed with clever use of the crossover frequency knob on the sub.

The advantage of S350DB and small 2.1 in general is you can throw them anywhere and they will perform reasonably. The performance when placed on top of my Usher is just slightly different from on the table. I can compare the table-edge and listening position measurements to see if it correlates.
Hence set up is easy, and can get almost the sound of a pair of big bookshelves just by throwing things anywhere.
There's also the tablespace advantage.

If I have the space to set up R1850DB I would choose that. While S350DB is a valid choice if I can't.
 

lxXXxl

Master Member
Joined
Dec 7, 2001
Messages
2,616
Reaction score
25
Nice~

S350DB looks like a good all-in-one setup and belongs in the top of the 'computer speakers' category, while R1850DB + sub will be more towards budget hifi setup liao (and requires more real estate).
 

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
If you can put speakers of 127mm width, I think you can squeeze a bit more for 155mm (R1850DB) or 174mm (R2000DB)

pM8WiMj.gif


On the other hand if I use a 27" screen then I can't use 5" speakers, and be forced to use smaller.

This size does look more normal for a computer speaker. Speakers don't usually tower beside the screen nor protrude too much forward from the screen.
The cable connector is too stiff and needs a full finger length behind the speaker. Though this problem is common to all Edifiers using that bi-amp cable.
 

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
Treble sounds a bit harsh but throws a lot of details in your face, noticeably more than other speakers.

And then I EQ-ed away the 10kHz peak and it sounds more normal. The details are still there tho, just lower in volume. But now the satellite sounds slightly muffled. Still, better than having insufficient lower treble like a bad speaker would.

I can live with both versions of the sound, with the 10kHz peak or without the 10kHz peak. I think with the peak it might sound good at first but get irritating in the long run.

There is less lower treble / upper bass than I would have liked. But that is mostly me, because this is the sound I like apparently. However remember we want downward sloping FR when we're in a room.

But I can't find a way to boost the 1kHz region without it sounding weird. Seems like as long as I avoid ~1kHz, I can boost as much as I want using a low shelf filter. Just 1dB removes signs of deficiencies, but you can always go more basshead and absorb both the pros and cons. I still think going for bigger speakers is the better solution for this tho.
Correction: At least 3dB boost is needed to stop my tanks from sounding like tin cans

This is pretty effective. Adjust according to preference. You may want another low shelf attenuation at a lower frequency so you don't boost frequencies that are too low and already receive further boundary gain.
EAsxEi5.png

After the EQ the sound I'm not sure if it sounds slightly abrasive or I remember wrongly.
Avoid boosting 1kHz as makes it sound like gurgling while having a sock in the mouth. But you don't want to start from too low or the sound don't get the warmth, and the sound remains dry. But this is a personal preference tweak to begin with. Too much dB and it becomes bloated.

Overall, the default sound that is pretty decent. I think if I didn't have other things to compare against, or have measurements, I would have accepted the sound as is. But since I do have measurements, there are minor but notable fixes that can be done.

To test whether I can hear the direction of the subwoofer, I positioned my head near the subwoofer. It is possible when the sound from the sub is louder than the satellites by a lot. Else it feels like the sound is coming from the satellites. In practice, I was never bothered by the direction. Not even with the sine sweep. Which is funny, because with sine sweep 100% of the sound should be coming from the sub but I still couldn't hear where it is.

Giving it a score
A few straightforward conclusions can be made -

- Considering S2000MKIII has a comparable bass extension as S350DB, and better performance everywhere else, S2000MKIII outclasses S350DB. Even if you're a bass head, that subwoofer doesn't help since it uses a bass tone control, which S2000MKIII also has. Maybe S350DB has more volume headroom, who knows.
- Against R2000DB it is a toss-up. R2000DB bass extension is almost enough, but not quite. For music, you really want 40Hz. Or at least flat till 50Hz. Things that R2000DB has problem with, but S350DB has no issue with. Deciding between the two is rather difficult and might boil down to whether you need 2.0 vs 2.1, or whether you have enough space on your desk. But you can definitely get more wub wub with the S350DB bass turned up.
- Having to fight against R1850DB + T5 is where it gets interesting huh. Edifier's problems are its own other products. :s13:
- Against other 2.1, well, so far this is the 2.1 that I have the least complaints with. I don't have any close experience with Promedia (Close as in, in my room) so that remains a question mark. A question mark that costs less than USD$200 but somehow SGD$400. I complained about M50W, although now that I am more used to testing speakers, maybe some of those complaints may be rectified. Who has a M50W that needs repair?
- One way of putting it is, it does not feel like I'm listening to a 2.1 speaker. Feels like a 2.0. I wanted to use the word "almost", but I remember there are 5" speakers with even worse midrange performance.
 
Last edited:

meandyou

Master Member
Joined
Jan 12, 2007
Messages
3,465
Reaction score
6
Soon will have another problem..
R1700BT + T5
But we will see R1280DB + T5 too
 
Last edited:

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
The sub-out on R1850DB is the key. Else if connect through sub means lose all the features on the speaker.
 

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
Socking the port
I wanted to find out what that peak in 1.2kHz is, so I plugged the port with a sock

Q6ZoAL9.png


Brown: Port close mic
Purple and green: Port close mic with sock, two measurements

First, plugging the port reduces a lot of port leakage from >800Hz. It also reduces the the port output centered at ~150Hz, which is kind of expected.

However the peak at 1.2kHz remains.

But it's not the port, because it only exists when I'm slightly away from the port. Well I can't be exactly close-mic'ing the port when there is a sock sticking out. I can close-mic the port if I go in diagonally, but I don't measure any 1.2kHz peak. Measuring the cabinet also shows no 1.2kHz peak. Weird. Maybe it's some kind of port-woofer interaction that generates a resonance internally.

Overall in-room response, Red = with port. Differences are minimal. Btw I used heavy smoothing here so you can't see the 1.2kHz peak.
teSIz2x.png


I ended up using Edifier's measurements to boost the 100Hz to 1kHz region, and the result is great. Basically mimicking S-520's FR. Still does not replace the sound from a bigger speaker, but beats a lot of un-EQ'ed bigger speakers. I guess the limit of small speakers goes beyond frequency response.
 
Last edited:

lxXXxl

Master Member
Joined
Dec 7, 2001
Messages
2,616
Reaction score
25
Socking the port
I wanted to find out what that peak in 1.2kHz is, so I plugged the port with a sock

Q6ZoAL9.png


Brown: Port close mic
Purple and green: Port close mic with sock, two measurements

First, plugging the port reduces a lot of port leakage from >800Hz. It also reduces the the port output centered at ~150Hz, which is kind of expected.

However the peak at 1.2kHz remains.

But it's not the port, because it only exists when I'm slightly away from the port. Well I can't be exactly close-mic'ing the port when there is a sock sticking out. I can close-mic the port if I go in diagonally, but I don't measure any 1.2kHz peak. Measuring the cabinet also shows no 1.2kHz peak. Weird. Maybe it's some kind of port-woofer interaction that generates a resonance internally.

Overall in-room response, Red = with port. Differences are minimal. Btw I used heavy smoothing here so you can't see the 1.2kHz peak.
teSIz2x.png


I ended up using Edifier's measurements to boost the 100Hz to 1kHz region, and the result is great. Basically mimicking S-520's FR. Still does not replace the sound from a bigger speaker, but beats a lot of un-EQ'ed bigger speakers. I guess the limit of small speakers goes beyond frequency response.

A 5db cut and boost at 100 and 160 respectively should not stress your speakers too much right?
 

wwenze

Great Supremacy Member
Joined
Dec 2, 2002
Messages
73,386
Reaction score
18,278
Cuts are usually okay. Boosts are questionable since it eats into the headroom for that particular frequency. 5dB is a small amount however, I can run +20dB shelf filter on my Usher if I need the bass.
 
Important Forum Advisory Note
This forum is moderated by volunteer moderators who will react only to members' feedback on posts. Moderators are not employees or representatives of HWZ. Forum members and moderators are responsible for their own posts.

Please refer to our Community Guidelines and Standards, Terms of Service and Member T&Cs for more information.
Top