Standmount speakers for people who dislike any brightness?

Since you mentioned intermodulation distortion, I recalled this article:
Bi-wiring Speakers: An exploration of the benefits -

Disregard the explanations, but look at the charts near the bottom of the article

So, that is why I suggested that you experimenting (not living with it long term, but just experiments) with crossover low frequencies to your subwoofer as high as possible and let your subwoofer take over all the low frequency and relieving your speakers from low frequencies.

Also, try doing bi amp your speakers to see if that would reduce intermodulation distortion.

 

I don't think you were that confused. The orange line is the difference in frequency response between the tube amp's output and the Anthem A/V receiver's, as measured in front of the speaker. So if the Anthem is flat as it claims to be, and everything else stays the same (as it did), then the orange line is the tube amp's deviation from flat, as measured at the speaker output.

Interesting. Maybe I'll try the 4 Ohm outputs - I haven't moved the microphone yet so it would be a fair test.

 

Yea, I think most most modern speakers would probably match better using the 4 ohm tap. In fact, that's the way my mid 90s era Conrad Johnson amp comes wired by default, you actually have to solder connections to change it, much easier with multiple speaker outputs

 

can you use arta and do a FR measurement and measure your two amps with your speakers?
I dont understand how the orange line is the FR. the orange line is not the Frequency Response right?

 

His chart is kind of confusing to me. There is dark green, light green, and gray/white area. Why there is dark and light green?

The orange line is supposed to be the difference between one green area (not sure which green) vs the gray/white area.

If the captured output of the grey/white area is from the tube amp, I don't see any issues with spike in 80 Hz. Instead, I see the AVR not able to handle 80 hz.

 

back2vinyl is overlaying two responses. The light-green/grey is the response of the tube amp, and the dark-green is the Anthem. The light-green is a result of the grey overlaid on the dark-green.

The orange dots are a plot of the differences. The peak corresponds to the tube amp having more energy in the region around 70 Hz. The tube amp also has more energy from 300 Hz to 3 kHz, but the Anthem has more above 3 kHz.

back2vinyl's thesis is that the Anthem (as a solid-state amp) should have a flat-frequency response. Obviously, he is measuring an in-room response, and so the plot is far from flat (whether that is due to the room, mic placement, mic characteristics, etc. is irrelevant). What he contends is that swapping the amplifier should not result in gross frequency response differences when the measurement is repeated, assuming the new amplifier also has a nominally flat response.

He cannot say definitively that the Anthem is right and the tube amp is wrong, but it seems most likely. To resolve the issue he would need to drive the speakers with a known high-performance amplifier.

 

My solution was to buy a pair of Reference 3A mm de Capos. They really just do everything right and are not the slightest bit fatiguing in any sense.

 

Oh no! Sorry I missed that.
Too bright? How can this be?
If anything, the pair I've heard (100's?) needed more top end.

 

Have you ever done that? I'm not sure how you'd do it with an RTA unless you have some very fancy and expensive measuring equipment.

I have in the past improvised a way of measuring the FR at the speaker outputs using Voxengo CurveEQ and I might try that if I get a chance after the weekend, just to see if the Anthem really is flat.

Thanks Black Elk - said it better than I ever could!

 

yes, arta allow easy FR measurements. I do it very often.

 

How do you connect the amplifier outputs to your PC?

 

I'm with you: avoid metal twitters, silver cables (both major brightness issues) and if you can relook at your amplification, tubes can definitely help.

 

Lots of tube amps have wide bandwidth. Mine is rated -0.5db at 12hz and 50khz. My preamp (tube hybrid) has even more bandwidth.

I'm curious about your in room response graph. Was that close miked? Looking at that graph I would imagine it would feel like a nail being driven into my ear. Pleasant sounding in-room curves are usually downward sloping. Yours are upward sloping.

Try doing a measurement at your listening position and see which way the graph slopes.

 

I use a yamaha soundcard, use my dac to output the pink noise via arta program and use a mic (input in my yamaha soundcard) to capture the speaker output

 

Yes, I'm sorry - when doing that chart, I was only really interested in the orange line (the comparison between the two amps) but if you ever started looking at the in-room responses in the background, they would indeed look a bit freaky! This is because: for some reason which I don't understand, Voxengo by default adds 4.5 dB per octave to the slope of the raw frequency response. This doesn't affect the comparison (the orange line). But it can sometimes make the slope itself look a bit weird.

Here is the same chart again but this time with the in-room responses adjusted to "normal" (no slope). The orange line (the comparison) of course stays the same.



As you can see, both in-room responses now slope downwards, as you'd expect with most setups. This is with one speaker only under test (front left) and with the condenser microphone only about a foot away from the speaker, pointing at the tweeter.

 

But I've already done that - see the chart immediately above! All you're suggesting is using ARTA instead of Voxengo CurveEQ but changing the software won't change the result. What we really want is a chart showing the amplifier's output WITHOUT any loudspeaker or room effects and I'm going to see if I can achieve that in the next day or two - I have done it in the past but it's a bit fiddly.

 

@Back 2 Vinyl that chart makes more sense. Looks like the MRX is better where you want it, down 3db from the Valve Zone in the 1k-2k region. Where the Valve Zone is giving you more upper bass, 60hz - 120hz, which should negate the 3dbs it's also adding from 1k-2k.

How much smoothing is happening to those charts?

When you turn on ARC in the MRX can you tweak the high frequency to be more to your liking?

 

ok
I dont understand what is happening in your setup. your -6db from 3khz to 20khz.
do you have some EQ?
I recommend arta cause its a bit more clean looking, and its free but voxego is okay. something is seriously weird in your setup, the downward curve is really exagerated. it supposed to be 1db per octave. in your setup it goes from flat at 1khz to -6db at 3khz. this should not sound bright, rather dull

 

I looked at measurements of the speakers at
ATC SCM 11 loudspeaker Measurements .

One interesting comment:

Though there is a rather demanding combination of 5.5 ohms magnitude and –42° electrical phase angle at 3.3kHz, this shouldn't overtax well-designed amplifiers.

I wonder if that could be the cause of your issue.

Can you try bi amp your front speakers? Tweeter with your class A amp and woofer with your AVR? See if that helps at all?

 

Another question, does the software correct for the mike's calibration file? Do you have a calibration file?

 

I will come back to some of those points in a moment but first, I've found a way to measure the frequency response of the amplifier at its speaker outputs while under load from the speakers. This is a direct, electrical connection from the speaker terminals into a DAC and excludes the loudspeakers and the room effects so it is just the pure, actual frequency response of the amplifier measured at the speaker terminals under load.

This is the frequency response of the Boyuu A10 tube amp using the 4 Ohm taps (not the 8 Ohm, as previously). The orange line shows the real, actual frequency response of the amp's output at the 4 Ohm taps, using the left-hand dB scale. In the background, the grey chart shows the FR of the pink noise test track I used and the green chart shows the FR of the amp's output - same as the orange line but using a different scale, this time the right-hand dB scale.




Now, here is the real, actual frequency response of the Anthem MRX 520, done the same way. The orange line is the FR of the amp's output measured at the speaker terminals under load, grey is the pink noise test track and green is the Anthem's output using the right-hand scale. Please note I forgot to turn off bass management, so the Anthem's bass output is cut off!




As you can see, ignoring the bass end of things where I forgot to turn off the bass management, the Anthem lives up to its claims of being flat through the most important part of the frequency range. In fact, far from being bright sounding, it gently rolls off the top end above 10 kHz and is 3 dB down by 20 KHz - not that I can hear anywhere near that frequency.

The new FR chart for the tube amp has a much smaller bass spike than we saw in my first attempt. That's because I forgot to turn off the Anthem's bass management for the first attempt, which distorted the comparison. Also, with this latest attempt, I used the 4 Ohms taps instead of the 8 Ohms taps and found this gave a better response. Even so, the pattern remains much the same.

 

can you explain how did you went from two very different FR?

yes, amps normally measure flat. doesnt mean they cannot sound bright though even though they measure flat.

 

Hi Khyl,

1. It's 1/6 octave smoothing.

2. No, unfortunately the ARC system doesn't allow you to customise the target response curve. I think Dirac does but not ARC.

3. I have a calibration file for the UMIK-1 but this particular microphone happens to be so close to flat that it doesn't worry me if I don't use it and I didn't have an easy way of using it here.

 

Fully explained in my last paragraph if you read to the end! Happy to answer any questions.

 

As you can see, both in-room responses now slope downwards, as you'd expect with most setups. This is with one speaker only under test (front left) and with the condenser microphone only about a foot away from the speaker, pointing at the tweeter.

you should measure 1 meter away from the speakers, one foot is not enough for the drivers to fully integrate.

in nearfield, at 1m, you should expect to have something quite flat in the treble

also, can you post measurements with the mic placed at the listening position. thanks