The room is about 12 by 23 feet with the speakers on the narrow end. I am about 7 to 8 feet away depending on where I put the speakers. I think I have tried every reasonable position of the speakers on that wall up to about 3 feet out. It makes a difference in the bass to low mid range regions, but doesn't get rid of the harsh sound to my ears.
Now that I have some acoustic panels on the wall behind the speakers (new this week), it has changed FR at my listening position some and I feel a greater need to cut the tweeter level a bit more.
Previously, I had these speakers in an MTM box, and even switching from the wrong box (it did have the correct volume & port tuning) to one built to Zaph's specs did very little change to the overall harshness in my ears. Since I had the wrong box before, I always attributed the problem to that.
Can you separate the woofers and the tweeter and take some measurements of the tweeter with the crossover attached. Likewise for the woofers.
Regards
Mike
Regards
Mike
Yeah I guess I should do that. The tweeter isn't a problem but I don't know how best to measure woofers with minimal room effects though since I need to be far enough to measure both at once. Also not sure at what height and distance would be best for measuring them.
This is one of the parts of taking basic measurements that I don't understand very well: How to accurately measure 2 woofers... even more complicated in a cascaded 2.5 setup that require both to be connected in order for the crossover to be accurate. Even near-field response seemed strange when I tried yesterday which I thought might be because of cancellations from the other woofer.
Also, when you take the woofer measurements, do I take them at 1 meter? at the upper woofer height? at tweeter height? None of these positions is exactly the way they would sum at tweeter height from normal listening distance.
All PC based (FLAC, MP3, streaming) connected through optical connection to Yamaha RX-A3030 HT amp. I have tried an IPOD touch connected to analog input on the amp just to see if it could be a DAC issue. I don't believe it is. I did try a different HT amp but that one was so cheap and sounded quite bad. Hard to draw any conclusion from that. I had a DIY Honey Badger amp at one point and that had no affect on the harshness and only very minor subjective difference in other areas. I was using the Yamaha as a DAC/preamp with the honey badger.
I thought it might be something like that. You can't expect for it to play good if you feed it with bad quality files. Try to run one of audiochecker software to check your FLAC files are they real or not. It has to show 100% cdda. Sound of MP3 can be described as you described it.
If you listened music on loudspeakers that wasn't designed properly before ZRT, it might be that you got used to that sound and now you can hear how bad the recordings really are.
There are sh*tload of diyers in Canada so you could look if anyone close to you made an F5 or similar. It's not state of the art but it sets a standard how no bs home amplifier should sound these days. Invite him for a drink and a listen to your new speakers. Maybe your Yamaha is really good - but the ones i listened were not. Not compared to well executed Pass clone amp. For preamp you can use passive step attenuator for neutrality.
If you listened music on loudspeakers that wasn't designed properly before ZRT, it might be that you got used to that sound and now you can hear how bad the recordings really are.
There are sh*tload of diyers in Canada so you could look if anyone close to you made an F5 or similar. It's not state of the art but it sets a standard how no bs home amplifier should sound these days. Invite him for a drink and a listen to your new speakers. Maybe your Yamaha is really good - but the ones i listened were not. Not compared to well executed Pass clone amp. For preamp you can use passive step attenuator for neutrality.
Last edited:
Keep the mic at tweeter height on tweeter axis. 1 meter away and tweeter 1 meter from floor. Raise your speaker if you have to.
Set Omnimic window to 5 msec. This is accurate to about 500Hz. Below that, it will record your room reflections when you set to "blended". Don't worry about it being inaccurate. The purpose is to check the crossover region.
Measure the response of the tweeter. What you want is to see the nature of the high pass slope.
Measure the two woofers together with mic at same position (1 meter, tweeter axis). Again, it's to see the low-pass slope.
Disconnect the lower woofer.
Measure the response of the top woofer alone.
Your post illustrates the divide between the "musical" vs. "analytical" school of speaker tuning. Should I tune my speakers to sound good on 20% of my music collection but sounding bad on 80%? Or should I tune it to sound good on 80% while making my other 20% sounding less than it should be? Is there a perfect solution?
I've seen countless of commercial speaker kits on various online sites (which I won't reveal) that are tuned so flat as if they were meant for a grading excercise. I wouldn't go near them even with a 10 foot or 8 foot pole. To me, it's a lot harder to design speakers that reveal the beauty of music even if it may have to sacrify something else. I am not saying speakers should sound dark and veil. There should be a way to make them sound musical without being bright, edgy, or analytical. I think that is where an experienced designers make their money.
Is it too much asking for both ? Is it so hard to invest a little bit of effort to find good copies of music you enjoy so much ?
If only flat frequency response was the only thing that's needed for speaker to play good. There's whole set of measurements that can/must be done to describe the sound of just loudspeaker - and another one for speaker/room interaction. Even then you must listen to it. Mic and ear must both say Yes.
Bad masters and lousy remasters past decades made any audiophile seriously crippled if without an equalizer. Only worse thing is lousy remaster of rock/heavy metal in MP3 format. It's doomed to sound unacceptable to a person without some serious hearing loss ?
Last edited:
This might possibly be the issue, no matter what though, they dont sound right to me. There is a harshness that is present in just about all music. It's worse on more complex music probably because the problem frequency(ies) get hit more often or more prominently. This is not a music file problem for sure. This is across the board. Does not seem like it can be a power amp problem either since the Honey Badger power amp made no real change to the problem sound. Unlikely to be a DAC problem since the IPOD really seemed to have the same issue but just removed a bit of clarity from the sound. Did not try the ipod with the honey badger but if that were to solve it then there would be really something freaky going on in the Yamaha, since I tried eliminating the DAC and I tried eliminating the power amp section.
All the people that say how great these are... how many of you have listened to them? It is quite a different design from the 2-way ZRT and seems to have been done as an afterthought. Also, no phase tracking between mid and tweet? Is that not something you usually strive for? Not saying all of this is necessarily a very bad choice (i'm not qualified to say) but the whole thing is less pleasant to my ears than a handful of ~$2000 CAD speakers that I auditioned last month. (BTW the ZRTs cost about $2000 CAD so value wise it's a fair comparison)
I will play around with tweeter level some more when my resistors get here, and I will see how much that can help soothe my ears. After that, I may just move on to something else. I have the itch to build something anyway.
I look through a lot of Zaph designs and his writing and it appears to me that he favors the "neutrality/analytical" over "euphonic" sounding and I am not saying which one is right or wrong, but if a pair of speakers favor any of these two extremes are probably not a good solution. Of course overly euphonic is probably not good either. His designs may favor some very good recordings but there are a lot of so so recordings out there. You can't just design speakers to absolute neutral and in a way that is sort of an easy way out, because if people criticize you, you just going to point out "hey, mine is perfectly flat, it's your recordings that are the problem". Well 70% of my recordings are like that.
To me a good design has to achieve some compromises or appealing to something sensible.
I'm sticking by my original analysis here:
Zaph 2.5 way modelled with Zaph's own FRD and ZMA bass files, AFAIK. I've winged it on a slightly different tweeter.
I wouldn't give up on this expensive investment. I found a third order series MTM or TMM looked much better, below. Tweeter crosses around 2.2kHz on steeper slopes. Less 5kHz cone breakup from the basses. I haven't allowed for the 3 ohm tweeter, but that's not hard.
Zaph 2.5 way modelled with Zaph's own FRD and ZMA bass files, AFAIK. I've winged it on a slightly different tweeter.
I wouldn't give up on this expensive investment. I found a third order series MTM or TMM looked much better, below. Tweeter crosses around 2.2kHz on steeper slopes. Less 5kHz cone breakup from the basses. I haven't allowed for the 3 ohm tweeter, but that's not hard.
Attachments
This taken from his ZRT-2way design. If you look at the frequency response on his website, you can see the 1kHz - 2kHz is about from 1db to 2.5db higher vs. the 400hz - 800hz region and at the same level as the bass which is around 200hz to 100hz. There is also a slight peak at 3khz which usually means forward. From 4khz to 8khz, it either at the same level or just slightly higher as the voice region (which is about 400hz to 1khz) and this will make vocal sounds with excessive sibillance. I guess some recordings would sound good on this design, but most would sound overly bright. Zaph no doubt is a good designer, but I guess his preference is straight ruler flat.
Last edited:
This is his ZRT-2.5 which has similar frequency response to his 2way version. The region between 1Khz - 3kHz is either as high or even slight higher vs. the 100hz - 500hz. The 3kHz - 8khz is a little bit high as well which makes vocal sounds with excessive sibillance.
I'm sticking by my original analysis here:
Zaph 2.5 way modelled with Zaph's own FRD and ZMA bass files, AFAIK. I've winged it on a slightly different tweeter.
I wouldn't give up on this expensive investment. I found a third order series MTM or TMM looked much better, below. Tweeter crosses around 2.2kHz on steeper slopes. Less 5kHz cone breakup from the basses. I haven't allowed for the 3 ohm tweeter, but that's not hard.
Well I could use this as a learning exercise to try and do something using my own measurements. The main problem is not having speakers in my living room while I tinker. I do have other drivers that I could use to learn with though. And they aren't exactly low quality drivers either, it is possible for me to get something pleasant to my ears with those other (popular lower-end Seas) drivers and I can still keep the ZRTs together in the meantime. My real goal is to soon move into a 3 way to avoid the 6.5" to 1" compromise. I've known that I am not partial to this tradeoff from the beginning, yet somehow I keep building them.
Actually 6.5" + 1" is not a bad compromise if the xover is done right given some of the constraints. There are a lot of good 6.5" + 1" out there. Scan-speak drivers are as good a driver as there can be for a 6.5" + 1" design.
This taken from his ZRT-2way design. If you look at the frequency response on his website, you can see the 1kHz - 2kHz is about from 1db to 2.5db higher vs. the 400hz - 800hz region and at the same level as the bass which is around 200hz to 100hz. There is also a slight peak at 3khz which usually means forward. From 4khz to 8khz, it either at the same level or just slightly higher as the voice region (which is about 400hz to 1khz) and this will make vocal sounds with excessive sibillance. I guess some recordings would sound good on this design, but most would sound overly bright. Zaph no doubt is a good designer, but I guess his preference is straight ruler flat.
Yes. Some mild EQing has helped, and lowering the tweeter level a bit more may help more. I had built a different Zaph project before and that was bright too. Not harsh in the same way though... just bright and a little weak in the bass all the way to lower midrange probably mainly due to Vifa XG18 limitations. My descriptive words about the ZRT might not match how others would describe it. It could be a certain distortion either by the tweeter playing low or coloration in the treble on the woofers... I dunno.
Troels Gravessen also said that he felt these woofers were not a good choice for a 2-way so there must be something in them that didn't sound right to his ears and it could be that is what I hear too.
Pick your poison. Seems I am figuring out mine. If these are as good as it gets, then 6.5 to 1 is not for me.
I don't think frequency response has much to do with a good sounding speaker. And phase alignment can be a bit woolly without sounding bad.
Power response and impedance play a part in the sound too.
FWIW, below is the steeper circuit I was playing with. One of many ways to skin this cat. Using Scanspeak FRD and ZMA on the bass which I believe to be about right. The tweeter circuit needs more adjustment though, because your 3 ohm tweeter is going to work differently.
Impedance correction on a bass has a lot going for it, IMO. Damps the voicecoil.
Power response and impedance play a part in the sound too.
FWIW, below is the steeper circuit I was playing with. One of many ways to skin this cat. Using Scanspeak FRD and ZMA on the bass which I believe to be about right. The tweeter circuit needs more adjustment though, because your 3 ohm tweeter is going to work differently.
Impedance correction on a bass has a lot going for it, IMO. Damps the voicecoil.
Attachments
I am not saying this design as good as it get. I was saying the drivers are some of the best out there but need proper xover design. Zaph xover is probably not my cup of tea either.
- Status
- Not open for further replies.