Complaint thread about speaker measurements

[amirm]

He doesn't seem to understand that, as I've demonstrated for him above, even if you measure in the nearfield, even in a large room away from boundaries, the room still afects the response. From what I'm gathering he measured his setup in the 'nearfield', EQ'ed it flat and believed that was the 'anechoic' response.

That seems to be his confusion among others. Somehow he thinks near-field means anechoic which it remotely is not. There are reflections, room modes, etc. in near-field just as well as far field.

 

[Deleted member 16543]

What? Speakers put each side of my monitor and listened at workstation seat is not "near-field?" What is it then?

Well, I'm just sayin that depending on the room dimensions they might be less near field than you think.
I would be truly surprised if a measured flat response (not gated like in spinorama measurements, but as measured according to psychoacoustics like Acourate does) actually sounded good to you.

 

[TimVG]

I didn't reply because it simply isn't true. There's no true intrinsic good value in the flatness of on axis response.
In fact, anechoic flatness goes out the window the instant that you equalize the true speakers response to an ideal PIR target.

1) It's most certainly true. I suspect you fail to realise a nearfield measurement is not an anechoic measurement. I suspect you lack some very basic knowledge of how to interpret measurements.

2) PIR or Estimated in-room response is an estimate based on the measured spinorama curves, it is not a target.

 

[Deleted member 16543]

That seems to be his confusion among others. Somehow he thinks near-field means anechoic which it remotely is not. There are reflections, room modes, etc. in near-field just as well as far field.

Really? We've been all writing a lot in the past hour or so (maybe too much), so it might have gone unnoticed that I specifically wrote that near field is an 'attempt' at anechoic listening.
There's a lot of gradations between far field and anechoic, I further noted. Gradations where the predicted in room response is less and less indicative of the perceived sound, the more you venture into the near field and getting away from the far field.

Funny enough, I just replied to you that you may be surprised that your near field speakers measured response can deviate quite a bit from the on axis response, depending on the room dimensions (and stuff in it).

Is there any way you can continue this conversation without implying I am an idiot? I thought we were past that. Consider this strike 2.

 

[TimVG]

Consider this strike 2.

You must be a special kind of stupid.

 

[amirm]

Well, I'm just sayin that depending on the room dimensions they might be less near field than you think.

Than I think? Sorry, you need to go and learn the terminology that you are throwing around. Here is the definition of Far Field from Klippel's ALMA presentation (which you can find in a number of other places):

Room dimension has nothing to do with it. It is all a function of wavelength, diameter of the source and listening distance.

In no way or shape are you in far field of a speaker on your desk. Not even close.

Here is Dr. Toole on the same topic:

With so little knowledge of the field and terminology, you proceeded to damn all the measurements and declared that you don't read them. Shame because if you did read them, then you would learn what you need to know.

 

[Deleted member 16543]

1) It's most certainly true. I suspect you fail to realise a nearfield measurement is not an anechoic measurement. I suspect you lack some very basic knowledge of how to interpret measurements.

2) PIR or Estimated in-room response is an estimate based on the measured spinorama curves, it is not a target.

1) See above
2) It's the target used when equalizing speakers that are less than ideal, or when the room is less ideal than the room used to calculate the predicter in room response. In any case, if it isn't an exact target, it comes very close to being one (for far field).

 

[Deleted member 16543]

You must be a special kind of stupid.

No. Not at all.. Also, I'm done with you.

 

[amirm]

Really? We've been all writing a lot in the past hour or so (maybe too much), so it might have gone unnoticed that I specifically wrote that near field is an 'attempt' at anechoic listening.

Pardon my French but what foolishness. You can't remotely approach anechoic listening. But sure, we can humor you. Post a measurement of a flat on-axis monitor in near-field and show us how your in-room measurements match it 1:1. Or even close to it.

For now, you think sound waves in low frequencies don't wrap behind the speaker, hit the wall and come back and combine with the direct sound? You don't think it hits the floor or the desk and combine with the direct sound? You don't think it hits the ceiling and combines with direct sound? How about the side walls?

Stunning lack of knowledge here and belief in myths about sound reproduction....

 

[TimVG]

It's the target used when equalizing speakers that are less than ideal

No, since it is an estimate partially based on the actual on-axis response of the loudspeaker in question! The rest is gathered from the early reflections and sound power curves.

or when the room is less ideal than the room used to calculate the predicter in room response.

Room used to calculate the predicted in-room response? What the **** are you on about? It's clear now you have absolutely no idea what you're talking about. For someone who doesn't want to be taken as an idiot you say a lot of stupid **** (and I'll refrain from commenting further, and I apologize for the outburst).

 

[Deleted member 16543]

Than I think? Sorry, you need to go and learn the terminology that you are throwing around. Here is the definition of Far Field from Klippel's ALMA presentation (which you can find in a number of other places):

View attachment 101889

Room dimension has nothing to do with it. It is all a function of wavelength, diameter of the source and listening distance.

In no way or shape are you in far field of a speaker on your desk. Not even close.

Here is Dr. Toole on the same topic:
View attachment 101891

With so little knowledge of the field and terminology, you proceeded to damn all the measurements and declared that you don't read them. Shame because if you did read them, then you would learn what you need to know.

There has been a misunderstanding.. What I meant is that even if you placed the speakers next to your monitor, depending on the dimensions of your room, you might be at or close to the point where the reflected sound power is the same as the direct wave power. It doesn't take much for that to happen.
I have no idea what distance you listen to your speakers at when you are set up for near field.
Nor what dimensions your room is.
It's not the same type of 'near field' we are talking about. The one I'm talking about has a lot to do with the room dimensions.

If we took the time to understand what the others are saying, without putting words in their mouth, this conversation could go on a lot more smoothly.

 

[richard12511]

I want a system which works properly without excessive dicking about.
I have generally got a system which gives me great musical enjoyment and have refined it by trial and error over the last 50 years. I haven't felt a need to change anything for 20 years or so.
If GLM isn't good enough, nor is anything else they are doing at Genelec, from my pov, so your opinion may rule them out. I'll go and listen for myself.
I don't intend to have a computer running at all once I have got the system running.
If I need one for a system then that system is a no-no for me.
I hate and detest dicking about with hifi when I could be listening to music
I could do with re-measuring my system for films with Audyssey and have been putting it off for weeks because it doesn't sound bad enough for me to put up with the loathsome task of re-doing the measurements.
I keep looking at the kit and saying to myself "I'll do that tomorrow".

Yeah, I knew you weren't a big fan of messing about with calibration. That's why I would recommend paying someone a few hundred dollars(or pounds) to do it for you .

I wouldn't disregard GLM completely based on my comments. I'm looking at it from the perspective of someone who kinda enjoys spending 100+ hours calibrating and improving performance. In terms of squeezing every last ounce of performance out of your system, I think other systems may have the edge, but for someone looking to do it once and forget about it, I think GLM is great, and definitely a huge improvement over no calibration at all. One good thing you might like about GLM is it allows you to take a smaller number of measurements(as little as 1) before calibrating, whereas Dirac strongly encourages you to take(9 to 17). GLM also allows you to store the filters in the speakers themselves, whereas with Dirac you need an extra box(or pc).

Do you use Roon(or something similar)? With something like Audiolense, I think you only need the PC to generate the filters and create the convolution file. Once you have that, you can just give that file to Roon, and the PC can go away.

I don't like what GLM does from the times I've heard it. You mentioned the flat target curve. I'm quite sure that's the culprit - it's fully adjustable however, but it does defeat the point of having an automatic system. You'd be using this as a secondary setup, correct?

I wouldn't say it's fully adjustable. That's actually one of my main issues with it, since none of their target curves (even manual) happen to fit my in room response(which always results in either shelved up or down treble.

 

[TimVG]

I wouldn't say it's fully adjustable. That's actually one of my main issues with it, since none of their target curves (even manual) happen to fit my in room response(which always results in either shelved up or down treble.

Oh that's a shame, for some reason I assumed it was in fact fully adjustable.

 

[Deleted member 16543]

Pardon my French but what foolishness. You can't remotely approach anechoic listening. But sure, we can humor you. Post a measurement of a flat on-axis monitor in near-field and show us how your in-room measurements match it 1:1. Or even close to it.

For now, you think sound waves in low frequencies don't wrap behind the speaker, hit the wall and come back and combine with the direct sound? You don't think it hits the floor or the desk and combine with the direct sound? You don't think it hits the ceiling and combines with direct sound? How about the side walls?

Stunning lack of knowledge here and belief in myths about sound reproduction....

There's research that shows how a reflection -20 dB from the first time of arrival wave is inaudible.
Of course, that depends also on the delta t of the second time of arrival.
This is stuff that relates more to psychoacoustics than raw measurements.

Therefore, I consider meeting that -20 dB requirement as a very good approximation for an anechoic environment.

Of course, this gets more and more difficult to achieve the lower you go in frequency. For example, even anechoic rooms are anechoic only above a certain frequency.

Anyway, to satisfy you curiosity, and in no way meant to work as proof of anechoic conditions achieved in a real room (but maybe approached to a promising degree?), in a 13 by 13 ft room I was able to get the SPL at room walls to be about 15 dB lower than the SPL at the sweet spot.
Very unusual type of speakers and tight setup.. definitely nothing similar to the typical near field you buy off the shelf (except the super cube 5, maybe) so I would expect near field speakers (as in speakers you listen to closer than the distance at which the reflected sound becomes as powerful as the direct sound) to sit somewhere in between my particular application and that of far field (as defined according to room reflections).
The more a certain speakers/room/sweet spot configuration approaches the reflections at -20 dB mark, the less the predicted in room response is indicative of how those speakers sound.
.. And yes, only above a certain frequency. We all know about room modes, 1/4 vs 1/2 space, etc...

I am as far as it gets from believing audio myths. ANY audio MYTH, by the way. Including the necessity of a flat on axis response
It's kind of funny that you imply I am the exact opposite of the type of person I am, I must say.

 

[watchnerd]

That seems to be his confusion among others. Somehow he thinks near-field means anechoic which it remotely is not. There are reflections, room modes, etc. in near-field just as well as far field.

It would certainly be strange for my LYD 5 monitors to have a "wall" switch on the back if near-field meant anechoic.

It seems like the designers of (at least some) near-field monitor speakers commonly understand that near-field isn't anechoic.

Not to mention desk reflections that are accounted for in the design...

 

[Sancus]

Oh that's a shame, for some reason I assumed it was in fact fully adjustable.

The ability to auto-generate filters to a target curve is limited and there's not that many options for the target curve, it's true. It'd be nice if they would add more, and it's kind of weird that the "manual" curve has such limited configuration options, it's not manual at all. But I get the impression that GLM is designed to be quick and simple, with the idea that if you want better you will probably want to do your filters manually anyway. I also suspect low latency is an important consideration.

It is fully adjustable in the sense that the filters themselves can be individually added and changed after generation to whatever you want. In theory I think you could even use REW or anything to generate filters and then plug them in. I believe the limit(at least for 8351B/8341A and probably 8361A) is 16 notch + 2 high and 2 low shelf filters.

 

[BYRTT]

Spinorama graph pattern is standardized as a 2 meter distance and below animation is ask CAD software set Z-axis to minus 200/400/600/800/1000mm relative to Amir's 2 meter spindata for KH 80 monitor, dont know if model is really valid but think result looks okay..

In relation above animation and KH 80 is nearfield monitor the ideal smooth on axis make sense, as farfield monitor say 2 meters and up it looks make sense tune for ideal power response over on axis..

 

[thewas]

That being said..
Let's take this very measurement.
This means that these speakers are (almost) ideal to listen to in far field.
What happens if you get closer with your listening position? You start to lose bass. The response tilts up and the speakers sound more 'clinical', right?
So, if these were speakers meant to be listened to in near field (IF...) they would not sound that great in their intended application.
Most likely, since they are almost perfect in far field, I'm willing to bet they are far field speakers.. or very bright sounding near fields.
So this is exactly the case I wanted to make. Same measurement, especially when boiled down to the predicted in room response, can mean very different things is the sweet spot is in the near or far field.

Since you mentioned Toole a couple of times, what he says is that above transition frequency the perceived tonality is dominated mainly by the direct sound which thus always should be flat and equalizing loudspeakers measured at the LP to match a pedefined room curve of a good one does not guarantee anything in terms of sound quality.
Here more about it https://www.audiosciencereview.com/...ut-room-curve-targets-room-eq-and-more.10950/

 

[whazzup]

Consider this strike3, Tim .

Wait, so he gets a prize?

 

[TimVG]

I don't see why further patience was required. I won't bother to reply to the poster in question anymore. However if you look at the last couple of pages I tried my best to explain, along with actual evidence, why this person has the wrong idea about loudspeaker measurements. He refused to accept any of it, unwilling to work along with any of us to get to the bottom of his 'issue', persistent in his opinion. Furthermore, he proceeds to hand out "strikes" as if he were a judge in a courtroom. Sorry, that's the line for me.