Alpair 7.3 notch filter

Status
Not open for further replies.
Hello lucky listeners!:worship:

finally i got my AP7.3 in a TQWT enclosure (Tuby from spectrumaudio.de).
I am completely impressed by the bass response and the room the speakers create with my restored Quad 303.

I have 3 questions:
- They sell it with a (selectable) RLC notch filter for 2-3kHz. I havent seen somebody using a correction for the AP7.3. Beside personal listening taste: Does someone experiment with filter for the AP7.3 ?

- After the first 50h and many comparisons i realized an increase between 10 an 14 kHz. On the one side this supports the extrem accurate response.
One the other hand this peak gives it a bit thin and metallic character when people are talking like in radio. People get always a very loud and metallic "S".
Does this effect decrease after a while? Did someone make the same experience?

Thank you guys...
HAL
 
frugal-phile™
Joined 2001
Paid Member
... in a TQWT enclosure (Tuby from spectrumaudio.de)...
One the other hand this peak gives it a bit thin and metallic character when people are talking like in radio.

I would suggest that the voice character is from the ill-matched enclosure, not the driver.

Also 50 hrs is not near enuff break-in time, i can't say for sure, i've never listened to a pair with less than 200 hr of initial break-in.

If you are measuring,make sure you also measure off-axis, on-axis measures are not very useful unless accompanied by off-axis measures.

dave
 
The enclosure and the voices are not the problem. I only refer to the peaks in high frequencies >10kHz which you can see in the official frequency diagram.
I will let them run and listen to it after 100h+.
It seems to me that the resonances are getting better.

Hello Hal,
Please read this thread:

http://www.diyaudio.com/forums/mark...nfo-flattening-markaudio-driver-response.html

Check your source, amplification and cables, as the Alpair 7 Gen.3 is an ultra micro-resonance sensitive driver. The thin metallic sound isn't the down to the driver, it has a reputation for emitting a sweet wide response.

Thanks
Mark.
 
Last edited:
I have read a lot of Marks speakers, but as a german guy i didn't see the most important statement of Mark;)
Thank you very much. That was the point!

Instead of "listening" to what experienced developers have created I used the recommended notch filter which ruined the character of the 7-3s. Voices were pale without fundament and secondly the notch filter emphasized the rich high tones like a loudness effect.
Now everything is fine:):)
Voices are real and putting them a few degrees out of focus dampens the high frequencies without loosing the stage.
Thanx Mark. You made my day!
 
I have read a lot of Marks speakers, but as a german guy i didn't see the most important statement of Mark;)
Thank you very much. That was the point!

Instead of "listening" to what experienced developers have created, I used the recommended notch filter which ruined the character of the 7-3s. Voices were pale without fundament and secondly the notch filter emphasized the rich high tones like a loudness effect.
Now everything is fine:):)
Voices are real and putting them a few degrees out of focus dampens the high frequencies without loosing the stage.
Thanx Mark. You made my day!

Hello HAL, Guys,

Glad to help. Its important to note that most full-range drivers aren't designed to be controlled by passive filtering.

I have no details of the filter that was recommended to you. I have the feeling it comes from someone who doesn't understand the design and function of an Alpair driver. Feel free to email me its details and any of its website links: support@markaudio.com (thanks)

Enjoy your project!
Thanks
Mark.
 
Last edited:
Voices are real and putting them a few degrees out of focus dampens the high frequencies without loosing the stage.

This is also how I set up my A7.3s. I find they sound quite good on any axis, but do prefere them toed out just a few degrees. This is something that I actually quite like about full range drivers. The ability to tailor the top end by merely re-adjusting the position of the speakers.

I've got a set of rather expensive bookshelves built by a big, respected canadian brand at home here and I wish I could simply toe them out a few degrees to tame the rather hot tweeters.
 
Hello HAL,
Many thanks for you help. I've sent a reply.

For beginners or those new to using full range drivers, here's the advice I'm currently circulating to the EU. Hoping those users with good English will translate into other EU languages and post on European diyaudio forums, many thanks:

The important message for users of Markaudio drivers (and similar designs) is to operate them without any filters. Most users will find it easy to adjust the axis of the boxes to get the sound they like. Markaudio drivers have very wide dispersion patten, the sound is spread very wide. Central stereo image is also wide so positioning the boxes by a little degree off-axis is normal for many users.

For a few users who have special circumstances, some filtering is needed for their purpose. They may have some sound reflection in their room that makes the highs sound too bright, or have a bright sounding amplifier, or a particular preference for old style music that they like played with less high range. But it is still best to try off-axis setting first or adjusting amplifier if its a tube type before adding filters to the drivers. Filtering components can add unwanted changes to the signal (noises) feeding the drivers. Most low mass Full-Range drivers are very signal sensitive, filtering can negatively effect their performance. So if a filter is applied, its important to experiment with various high quality components to see what works best.

Help from experienced Diyer's is welcomed to correct the errors of some users who mis-read the rising frequency data of Full-range drivers. They don’t understand the design function of Markaudio and similar Full-Range driver designs and make the error of applying a filter before trying to operate the loudspeaker with an adjusted on-off axis position.


Thanks
Mark.
 
Last edited:
For the last few years I have been head over heels for my CSS FR125SR drivers (with Dave's magic dots on) in a 'spiral horn' floorstanding cabinet. As a purist music only system with my homemade dac and amp (Buffalo & Charlize with Paul Hynes regs and BG's galore) this has made the most entrancing sounds I could imagine in a smallish living room.

Recently moved house and also moved to wanting to start playing TV and movies through those speakers. Finding the CSS drivers less clear on dialogue than I'd like, I picked up a couple of Alpair 7.3's secondhand. In those same cabinets I definitely find them clearer than than the modded CSS drivers and can follow dialogue at lower volumes (useful as my 4-yr old sleeps above the living room). But I do sometimes find the slightly elevated high frequencies less pleasing than the flatter response of the FR125SR. I get that similar issue with "S" sounds sometimes during dialogue and occasionally music seems less natural/involving because of it. Because of the shape of the living room people often have to sit outside the sweet spot, so angling the speakers outward is a less than ideal solution.

Having said the above, this evening my wife and I watched the film Dazed and Confused through them and did find the soundtrack sounded superb. So I definitely want to keep these drivers around in some capacity. But in view of what Dave says above, can an ill-matched enclosure cause such a problem with those high frequencies? I took care with the original cabinet design to minimize reflection of sound waves back through the cone and have used bracing. The cabinets sounded wonderful with the CSS drivers. I have no filters on the Alpairs.

So speaking personally, I really do like the A7.3's but I do think I would like them a lot more if they had a flatter frequency response higher up. Just wanted to share that / fish for insight from the more knowledgeable guys here...
 
Does the HT rig include something along the lines of Audyssey et al? I believe those will include some perfunctory digital EQ. In my own case (A7.2s in the front row / Denon AVR1610 so equipped) I don't detect any consistent issues with overall tonal balance - although certainly wider variations in quality of program material than one would expect with strict adherence to legibility target in audio engineering.

Of course with extensive HF hearing deterioration and the oft accompanying 24/7/365 tinitus, I'm quite likely not hearing things that could be measurably "wrong" or obvious to younger/more acute ear/brains. But IIWII

For "flatter/ smoother" response up top, there's always the 10P - and at some point soon, perhaps an intermediate sized paper.

Mostly wishful thinking that last part - Mark's plans do often pivot in surprising driections.
 
Last edited:
Thanks for the reply Chris. I'm not sure I would call it a HT rig as such - only 2 channels! I don't have any EQ available at the moment, being hooked on the idea of simplicity and purity of approach.

I gotta say you hit the nail squarely with HF hearing loss & tinnitus, both of which I do have. But that makes it seem strange to me that the HF bothers me. My hearing is reduced above 10K so if it seems wrong to me, how is it to folks with better ears?

Still, after all that the Alpairs are remarkable drivers. If Mark ever brings out a generation 4 of the 7's consider this my vote to design out the HF peak. :)
 
Thanks for the reply Chris. I'm not sure I would call it a HT rig as such - only 2 channels! I don't have any EQ available at the moment, being hooked on the idea of simplicity and purity of approach.

I gotta say you hit the nail squarely with HF hearing loss & tinnitus, both of which I do have. But that makes it seem strange to me that the HF bothers me. My hearing is reduced above 10K so if it seems wrong to me, how is it to folks with better ears?

Still, after all that the Alpairs are remarkable drivers. If Mark ever brings out a generation 4 of the 7's consider this my vote to design out the HF peak. :)
If using PC as player (jRiver Media Center or Foobar2000) then free EQ is available at source, and the simplicity and purity of approach is kept.
Other possibility is trying "Inductance Compensator"/"Zobel" filter, not that it do EQ your signal, but it makes some HF dampening and linearize the group delay electric HF signal thru speaker cable. This effect can sometimes makes one think the HF is less harsh. I attach picture with RC values for 7.3 done by WinSpeakerz program. If trying and values not at hand just try something near to make listening tests. For tests short run the resistor wattage can be small, for high volume long run use higher wattage types.
 

Attachments

  • IC.PNG
    IC.PNG
    21.2 KB · Views: 265
Last edited:
If using PC as player (jRiver Media Center or Foobar2000) then free EQ is available at source, and the simplicity and purity of approach is kept.

Just be careful, the included EQ of PC players often have unusual results due to the way they're implemented. The foobar one in particular produces very steep cliff-edge steps between bands which causes some odd results. I use the 'Graphic Equaliser' plug-in (foo_dsp_xgeq) which is smooth between points and also has much finer divisions instead for that reason.

Digital clipping and distortion also can result from poor software or poor setting of software - Don't ever touch the equaliser in a cheap sound card's drivers!
 
.....I may try the Zobel I guess.

Thoughts on placement. First choice at driver terminals. Second choice at speaker box terminals with same cable type and length as used for internal connecting driver to box terminal. This would make same electrical z elastic for filter as for driver referenced to box terminal. Third choice across speaker box terminals. Fourth, never place at amp end of speaker cable.
In my setup with 10.2M i run both zobel/inductance and resonance compensator (RLC), it's not a night and day mod but for my setup it gives kind of precision and smoothness in sound stage. Good listening
 
Last edited:
Guys, that last post by BYRTT made a lightbulb flicker briefly above my head. It reminded me to try some different speaker cable, as the one I have in is a touch bright and was put in 'cos it seemed to sharpen up the treble on the FR125SR's. But I got a couple of flavours of nice old Audioquest stuff buried in the cables box upstairs. Both have a much richer, smoother sound with no brightness at all. Tomorrow I'll see if they can help out. This may not remove the issue completely but I suspect it might be part of the picture. P'raps with the Zobel as well I'll be a happy camper - thanks for your help BYRTT !

Thanks for your patience folks - I gotta confess I haven't dabbled much in audio shenanigans since my little boy was born and I seem to have got rusty! Will report back. :D
 
Last edited:
..........Will report back. :D
Thanks, feedback is good to learn from.
Regarding speaker cables, for some it can be religion and respect for that, if you open minded here's another choice often used here at Markaudio threads. Use CAT4/5/6 SOLID core network cable. Use one set of the four twisted pairs for the 7.3. It's cheap to try and HF is not dominant. Use it myself up to 5 meters. Power to FR is not more than that about 0.5mm core is able to carry, some though use more than one pair (parallel) from the network cable. The thin core heals the skin effect in a way as 20-20.000hz is arriving at end (one freq) just as bad as another freq (no audio band freq has full skin depth).
 
Last edited:
OK, I tried the Audioquest cable and there is a significant improvement. The sibilance is much less noticeable and musical instruments sound more natural, although I did find with a couple of rock albums that the high hats were coming across louder than they should and detracting from the overall listening experience.

Thanks for the tip about CAT cable - I've come across this idea before but not tried it. The Audioquest stuff I have in now consists of several separate solid cores of various thicknesses in very pure copper. It was quite pricey back in the day. Is the CAT cable likely to sound better?

I kind of came to the conclusion earlier that the speaker cabinet is not right for these drivers as well. It worked excellently with the CSS drivers and with another 4 inch driver too, so I was hoping the Alpair would just drop in. But the bass just is not quite right - this is not the driver's fault though, just an incompatible box. I could try the Alpairs in a 10l bass reflex box I have empty, just for comparison.

I'm off to the big city in a week (currently living in the middle of nowhere) so will visit the electronics shop and pick up the parts for the zobel and perhaps the CAT to try. I really appreciate the time you've taken to offer help to me BYRTT. The kindness of people like yourself who offer help and place their knowledge in the public domain is a very beautiful thing about this hobby.
 
Status
Not open for further replies.