Speakers distortion

[RayDunzl]

One more thing - I will repeat the same measurements in a couple of days. If I get similar results only then I would be willing to trust them.

Do a sweep, see if the harmonics rise out of the noise floor... The lower the speaker level the less likely that will occur.

When mine don't, I'm not too sure about the results, other than to say the distortion levels must at least be "low enough".

The RTA window seems good for examining single frequencies, though, again, only if the harmonics rise out of the ambient/measurement noise.

 

[Blumlein 88]

Let me add specs of my mic. It's cheap but I believe it can do the job well.

I've got one of those. It compares will with the umik. Heck of a deal for measurements.

 

[Krunok]

Do a sweep, see if the harmonics rise out of the noise floor... The lower the speaker level the less likely that will occur.

View attachment 19635

When mine don't, I'm not too sure about the results, other than to say the distortion levels must at least be "low enough".

The RTA window seems good for examining single frequencies, though, again, only if the harmonics rise out of the ambient/measurement noise.

They do. I dont' trust sweep so I'm not taking it into account..This is RTA measurement of 42Hz tone and you can clearly see them:



And this is "silence". You will notice refrigerator and the heating pump singing at 50Hz and sounds from fireworks coming from outside.

 

[Krunok]

I've got one of those. It compares will with the umik. Heck of a deal for measurements.

Yes, it's cheap but it does the job well.

Btw, I would very much appreciate if you can take some time to measure distortion of your Revels. While they probably can't compare with Ray's MLs these speakers are really fine piece of equipment and I am extremely interested to see how they would perform.

 

[andreasmaaan]

So no big difference in the two distances.

Gotta say, I beg to differ slightly

There are 2- to 5-fold differences in distortion factor between the two measurements. There are also up to 2-fold differences between channels at a given distance, which is unlikely to be caused by the speaker (you’d hope).

I think, as @RayDunzl said, these kinds of sweeps are very helpful as a way to ensure things are “good enough” ie that nothing is seriously wrong. But they don’t tell us much about whether a speaker is just pretty good, or exceptional, which is what I thought the OP was about?

 

[Krunok]

Gotta say, I beg to differ slightly

There are 2- to 5-fold differences in distortion factor between the two measurements. There are also up to 2-fold differences between channels at a given distance, which is unlikely to be caused by the speaker (you’d hope).

I think, as @RayDunzl said, these kinds of sweeps are very helpful as a way to ensure things are “good enough” ie that nothing is seriously wrong. But they don’t tell us much about whether a speaker is just pretty good, or exceptional, which is what I thought the OP was about?

I cannot get sweep based distortion measurement to give any meaningful results, so I'm relying on RTA as they give figures that are consistent (that can be repeated).

And no, this thread is not about proving that my speakers are exceptional, as I know they aren't. This thread is about establishing distortion figures among various speakers, from budget to excpetional. For my Castles I would say they are pretty good, but nothing more than that. To give you an idea of the cathegory they fall in I will say I paid them 1000GBP 13 years ago. At that time I believe it was app $1300, so midrange class.

I certainly have no idea what can I expect in terms of tolerance between my left and right speaker.

One more thing - you really have to be quite carefull even with RTA distortion measurements if you want them to be consistent ie not to significantly differ. As you can see my room is quite noisy at this time of the evening and you can't really expect lab alike precision at home environment.

This is the data from the control mesaurement at 42Hz I posted above:

65536-point spectrum using Blackman-Harris 7 window and 16 averages
Input RMS 80,4 dB, 78,5 dBC, 47,4 dBA
Distortion at 42,0 Hz, -22,2 dBFS based on 8 harmonics:
THD: 0,23 %
N: 4,53 %
THD+N: 4,54 %
2nd harmonic 0,17%
3rd harmonic 0,094%
4th harmonic 0,074%
5th harmonic 0,056%
6th harmonic 0,042%
7th harmonic 0,036%
8th harmonic 0,037%
9th harmonic 0,029%

 

[Krunok]

I think it's time for everyone with the capability to do some you close measures using REW.

@andreasmaaan , this what @Blumlein 88 said is the point of this thred. I was hoping to get as many measurements of different speakers as possible so we can get an idea about what distortion they produce no matter what size and cost they are. The point was also to raise awareness among folks on this forum about the role of the speakers, their frequency response and distortion and how important it is to measure and adjust them as their impact on SQ is second to none.

 

[andreasmaaan]

Yeh fair enough @Krunok and @Blumlein 88. I guess I was seeing a trend of ppl talking about comparing different speaker distortion measurements taken in different rooms under different conditions and then trying to draw something from that. I think it might be an interesting thing for ppl to do for themselves to rule out any major distortion issues in their systems, but beyond that I’m sceptical

 

[Liberal Democracy]

@andreasmaaan , this what @Blumlein 88 said is the point of this thred. I was hoping to get as many measurements of different speakers as possible so we can get an idea about what distortion they produce no matter what size and cost they are. The point was also to raise awareness among folks on this forum about the role of the speakers, their frequency response and distortion and how important it is to measure and adjust them as their impact on SQ is second to none.

I'm confused so it's not an idiots guide to REW..

 

[Krunok]

Yeh fair enough @Krunok and @Blumlein 88. I guess I was seeing a trend of ppl talking about comparing different speaker distortion measurements taken in different rooms under different conditions and then trying to draw something from that. I think it might be an interesting thing for ppl to do for themselves to rule out any major distortion issues in their systems, but beyond that I’m sceptical

You don't think measurements made at home with REW are precise enough for the speakers to be compared distortion wise?

 

[andreasmaaan]

You don't think measurements made at home with REW are precise enough for the speakers to be compared distortion wise?

Serious question? no, I mean for example when your measurements vary that greatly from different locations in the same room with the same mic if the same speakers.. well you see what I mean?

 

[Krunok]

I'm confused so it's not an idiots guide to REW..

Well, no - I don't have skill nor time to do that. I do however believe that REW and rePhase are great tools (not to mention they are FREE!), with which you can enhance SQ of your system with the help of cheap mic and some basic skills.

Room EQ can really make your speakers sound better and with distortion measurement we could establish which speakers are worth the money and which are not. Pretty much same thing as @amirm is doing with DACs (and recently with amps), but we can't really expact him to start buying speakers and measure them for us, can we? We also can't expect him to do room EQ with our speakers for us, so this is something we should learn to do ourselves, right?

 

[RayDunzl]

I certainly have no idea what can I expect in terms of tolerance between my left and right speaker.

"Pretty close" would be the best I'd hope for, closer if you move each speaker to the exact same position in the room. If measured in situ the room will make different adjustments to different frequencies.

Left and right peakers at their positions - DSP room correction included, measurement at the listening position.

SPL

Distortion sweep - THD

But that THD is (I think) dominated by ambient/measurement noise (left speaker)

 

[Krunok]

Serious question? no, I mean for example when your measurements vary that greatly from different locations in the same room with the same mic if the same speakers.. well you see what I mean?

Sure I do. But I only started to play with this. It will surely take time to learn how to properly do distortion measurement as was the case with frequency response measurement. I'm not saying it is possible to do that in home environment with cheap equipment, but I will certainly try and see where it will lead to.

 

[Krunok]

"Pretty close" would be the best I'd hope for, closer if you move each speaker to the exact same position in the room. If measured in situ the room will make different adjustments to different frequencies.

Left and right peakers at their positions - DSP room correction included, measurement at the listening position.

SPL

View attachment 19638

Distortion sweep - THD

View attachment 19639

But that THD is (I think) dominated by ambient/measurement noise

View attachment 19640

Can you please try to measure distortion wth RTA and see how it correlates with this sweep distortion measurement?
You can pick a few frequencies, say the same I used (42, 100, 420, 1000, 4000) so you can compare it with the sweep distorion results. I have a feeling you will get different and much better results..

Btw, regading frequency response, I also believe you will get much more meaningful results if you use RTA and move mic within some space around the place where your head is at LP.

 

[RayDunzl]

I have a feeling you will get different and much better results..

Yes, same as you, a bit lower.

At what I think is an equivalent SPL, the harmonics of the electrostats are still burid in the noise floor.

The littlest JBLs (in this case the JBL 305p MKII), on the other hand, are not so discreet with their song...

 

[Krunok]

Yes, same as you, a bit lower.

Frankly, I believe sweep sucks for the very reason you mentioned. I'm inclined to use it ONLY to make room EQ for higher frequenies and even then you have to apply time window to remove refelctions. In all other scenarios I strongly believe RTA is preferable, both for frequency response and for distortion.

 

[Liberal Democracy]

Well, no - I don't have skill nor time to do that. I do however believe that REW and rePhase are great tools (not to mention they are FREE!), with which you can enhance SQ of your system with the help of cheap mic and some basic skills.

Room EQ can really make your speakers sound better and with distortion measurement we could establish which speakers are worth the money and which are not. Pretty much same thing as @amirm is doing with DACs (and recently with amps), but we can't really expact him to start buying speakers and measure them for us, can we? We also can't expect him to do room EQ with our speakers for us, so this is something we should learn to do ourselves, right?

I was just pulling your leg. As I understand it if you want to have semi accurate measurements for your speakers as speakers you need to go outside and do it there and even then I guess you've got the ground and a few issues to contend with.

Still maybe we could do some of that for smaller speakers.

 

[Krunok]

But that THD is (I think) dominated by ambient/measurement noise (left speaker)

When average SPL is 75-80dB that may be the case in average home environment but I still think THD of the speakers plays a role that significantly affects the SQ.

 

[Krunok]

I was just pulling your leg. As I understand it if you want to have semi accurate measurements for your speakers as speakers you need to go outside and do it there and even then I guess you've got the ground and a few issues to contend with.

I actually find it relevant to measure speakers in the very environment I'm listeing to them. That way you see which figures actually matter and whcih don't.

Still maybe we could do some of that for smaller speakers.

I see no reason to do that for home speakers of all sizes.