Axiom graphs - Axiom Message Boards

I am curious about the graphs that is used with Axiom speakers

How are these graphs done? At what db are they done and is there smoothing used with these.

The reason for asking is while trying to compare Axiom to speaker brand X it was said the comparison can not be made since two different methods were used.

If there is no set standard what good does it do to try and compare different brands using the supplied graphs from different makers.

I would think, though I can't say for sure, that Axiom uses the same methods the NRC does as per the Soundstage Article.

When making comparisons of graphs made by different parties you want to note am I comparing the same suite of measurements? The scale that is being used, unsmoothed-smoothed, output level, measurement technique (anechoic, quasi-anechoic, in-room, ect). All these factors will have a dramatic effect on how one should interpret and compare loudspeaker measurements.

Originally Posted By: Wid

I am curious about the graphs that is used with Axiom speakers

How are these graphs done? At what db are they done and is there smoothing used with these.

Axiom used to use non-smoothed graphs. At some point they changed to what a lot of other manufacturers use. I suspect it was a marketing decision based on excessive questions about 'all those bumps'.

Quote:

The reason for asking is while trying to compare Axiom to speaker brand X it was said the comparison can not be made since two different methods were used.

I think the case is over stated. It really depends on how much smoothing is applied.

Quote:

If there is no set standard what good does it do to try and compare different brands using the supplied graphs from different makers.

That's what ears and your favourite music are for.

Lots of people claim a greater understanding of measurements and speaker design than they have. Gene's, um some persons comments on another forum, put that into perspective.

Hi Fred,

I just wanted to note that we do not smooth any of our graphs. I am not sure where that piece of misinformation started but it could be related to someone comparing original series graphs to v3 graphs; the v3s are much smoother. There has been some talk lately of changing our graphs to a 60 dB scale and including some off axis graphs and the sound power. We will probably start changing the graphs on the site later next month.

Rick, the fact that you even have to ask the question illustrates the problem: a reluctance to provide technical details(this is again apparent in the lack of meaningful information about the DSP for the LFR1100). Some points are evident from the graphs themselves; the SPL levels are now shown in the low 90s of decibels. This is about 4dB higher than numbers in the past, but despite repeated requests, no explanation for the change has been forthcoming. Since those numbers coincide with the new sensitivity ratings, they were taken with an input of 1 watt.

Well, the topic came from another thread where it was claimed Axiom used smoothing and different levels to test their speaker. Hence one could not compare the two speakers in question because the measurements where done so different.

Originally Posted By: BlueJay [/quote

I am assuming Wid is getting his measurements off the Axiom website compared to that of Ascends off their website.

You can't compare the two simply for the fact that Axiom is using a 120db scale with smoothing applied while Ascend uses 50db scale un-smoothed.

These two factors alone can make a graph look a lot "smoother" when you cut down on the resolution.

Ascend also shows a whole suite of measurements including the polar response. To objectively compare the two you would have to compare the same suite of measurements using the same scale and preferably without any smoothing applied.

Originally Posted By: Wid

Well, the topic came from another thread where it was claimed Axiom used smoothing and different levels to test their speaker. Hence one could not compare the two speakers in question because the measurements where done so different.

Originally Posted By: BlueJay

I am assuming Wid is getting his measurements off the Axiom website compared to that of Ascends off their website.

You can't compare the two simply for the fact that Axiom is using a 120db scale with smoothing applied while Ascend uses 50db scale un-smoothed.

These two factors alone can make a graph look a lot "smoother" when you cut down on the resolution.

Ascend also shows a whole suite of measurements including the polar response. To objectively compare the two you would have to compare the same suite of measurements using the same scale and preferably without any smoothing applied.

[/quote

Yes, they are done differently and not comparable as I stated. You can't compare a single measurement of one speaker to multiple measurements of another speaker. Also consider the measurement technique is different between the two. Axiom obviously uses an anechoic chamber to measure their speakers and those are done quasi-anechoic.

Originally Posted By: JohnK

Rick, the fact that you even have to ask the question illustrates the problem: a reluctance to provide technical details(this is again apparent in the lack of meaningful information about the DSP for the LFR1100). Some points are evident from the graphs themselves; the SPL levels are now shown in the low 90s of decibels. This is about 4dB higher than numbers in the past, but despite repeated requests, no explanation for the change has been forthcoming. Since those numbers coincide with the new sensitivity ratings, they were taken with an input of 1 watt.

Hi John,

I'm not sure what meaningful information has not already been covered regarding the DSP. The dual DSPs perform frequency correction for the front and rear sections, along with a few other functions. The corrections tailor the system response so that the two sections of the speaker work together correctly.

That's the reason for the question, to try and understand better.

As Ian stated, Axiom does not use smoothing which prompted part of Ricks question.

I got that part Randy. What I don't understand is is the difference in how the anechoic chamber compares to them being done quasi-anechoic and to what affect the different db levels affect the graph being compared.

The main difference between the two is the anechoic chamber is built without any echoes and is not room influenced at all. Quasi-anechoic measurements are not accurate below the midrange where the room starts to dominate with this technique. So the amount of accurate information you can gather is more limited with the quasi-anechoic measurement.

Neither technique can measure low frequencies accurately. This is why subwoofers are measured outside.

Different db levels show at what SPL the frequency sweep occurs at. If there are deviations in the frequency response at increasing sweep levels it is a sign the speaker is stressing aka compression.

Thanks Doc. So can it be said a speakers measurement done quasi will not be considered to be measured as accurate as one done in an anechoic chamber?

If you look at measurements done quasi-anechoic, they never show the speaker below 200hz. On the other hand an anechoic chamber will give you accurate information below that. For a full range floorstander, the anechoic chamber will not accurately assess the bass response though.

Originally Posted By: Ian

Hi Fred,

I just wanted to note that we do not smooth any of our graphs. I am not sure where that piece of misinformation started but it could be related to someone comparing original series graphs to v3 graphs; the v3s are much smoother. There has been some talk lately of changing our graphs to a 60 dB scale and including some off axis graphs and the sound power. We will probably start changing the graphs on the site later next month.

Thanks for the correction Ian. My apologies for misstating the case.

I was looking at the old NRC measurements and did not take into account the difference in scale.

I do need to clean the troll crap off my shoes though. I've been wading through bad threads again.