Samsung MS650 Soundbar In-Room Measurements

[LevityProject]

I've started to have some fun measuring my speakers all over the house. The following are measurements I took of a Samsung Soundbar that I've got in my living room. I measured this flat at first, but it has no bass at all and the HF roll off made them too dark, requiring very large boosts in both extreme ends to sound somewhat neutral. So I used the simple EQ on the unit to boost LF and HF to +4 each. This gave more even measurement results to start working from.

Next I plan to try measuring my many bluetooth speakers which I've collected over the years. These measurements are mostly just a way for me to see what's happening with my speakers, and I'm not yet concerned with the room interfering with the measurements as I want to know specifically what these speakers are doing in the room they are in since I'm not able to treat the room more than it is. So I'm working with what I've got for now until I can figure out a better way to do this.

A couple of precautions I took:

• No desk reflections. I made sure the capturing microphone (just a macbook pro microphone) was placed on the edge of a little desk - so at least the desk wouldn't be as much of factor in the capture.
• Levels - I captured at a level that was line level before it hit the speaker. I chose line level (-18dbFS RMS) for the signal as Electronically, my understanding is that line level signals ensure optimal performance of the electronics in the chain - so that whatever the electronics do (good or bad), they'll at least be doing it in a somewhat optimized way.

Once again, this is not a professional speaker measurement, so it shouldn't be taken too literally as "the response of this speaker". Though I think it's still somewhat revealing, and using AutoEQ in rew made these sound a lot better. I wondered about that possibility, and was happy at how well it worked with very little effort.

All in all, this took me 30 minutes from start to listening with the new EQ settings. What I hope that can show is how easy it is to do this, in any room, without expensive equipment other than a laptop, to improve the sound of speakers connected via laptop (and routed through Soundsource for Mac or whatever global EQ programs exists for windows). Doing this, although not perfect, is so much better and easier than trying to guess what will sound "good".

I realized today that REW doesn't export graphic EQ settings like the ones wavelet uses, so that is unfortunate. Edit: I figured out a way to get REW Eq settings to wavelet, by uploading the REW EQ text file to chat gpt and feeding it the contraints of wavelet. It took forever to figure out but it finally worked.

Hopefully this give some insight into "consumer level measuring", regardless of this only being one speaker measured by a measurement noob in a long rectangular room that is only treated with area rugs and furniture like couches and chairs (aka, a typical apartment living room).

Here's the measured Frequency Response of the MS650 (no sub) + The Room

What I noticed right away was the extreme HF roll-off. I've always felt this soundbar doesn't have sparkle, and I was happy to see the measurement confirmed my long standing suspicions.

At first, I thought that maybe the laptop microphone just can't capture that high, or maybe I set REW to not measure above 14khz. After checking settings, I had the software set up right. I also know from my measurements of an old pair of Alesis M1 Actives Mk2 monitors that if the speaker can reproduce it, my Macbook Pro's microphone can measure it up to 20khz. So I believe cavernous HF drop off is a physical limitation of the MS650, or perhaps I've blown a tweeter.

What I see in the specs is a frequency range from 40hz to 20khz. But this can't be, at least not with my unit.

I've always thought this soundbar was a bit on the muddy side as well. I was pleased to see this confirmation in the measurement. Sometimes I think my ears are playing tricks on me, and after sitting with the soundbar for a while, I inevitably stop noticing the imperfections as my ears adjust. But I know I've got a good set of ears on me that I go to great lengths to take care of...

Here is the Frequency Response after applying REW's auto EQ and making a couple manual adjustments.

Distortion Levels. Distortion at 40hz looks bad. As I've listened, I can't hear the distortion so it might not be a practical problem.

Group Delay - what's happening at 68hz?? I don't completely understand what GD is and am still learning, but from other reviews I think this would be a "don't EQ" situation.

Listening Post EQ
Pleasantly surprised! It sounds full with good midrange presence, and has a wee bit of sparkle thanks to the filters evening out the response from 1khz to ~14khz. It still bugs me that I can't hear any significant sparkle with these. I really notice it in snares, hi-hats and cymbals. Cymbals and hi hats lack shimmer, and snares sound dull but at least they have some brightness now that feels correct. I also like what the sub bass is doing down to about 60hz when listening straight on, though I miss the 40hz rumble.

I went to a Luke Combs double header this last weekend at the Gorge. I wasn't a huge Combs fan before and didn't know any of his music, but after seeing him play two nights, I'm a convert so I used my new favourite song of his as my test track for Pre and Post Measurement listening.

I've attached the EQ txt file and the REW measurement and EQ files.

.