ghost4man, how much more testing and measuring do I need to do? The answer is - as long as it takes. Even if it takes years, you will find me patiently working my way through it

 

Davewantsmoore said it best - Paul Spencer has decades worth of knowledge which I don't, and even Paul has taken a year to design a crossover for his PSE-144. I also expressed my frustrations to Stuart of SGR, and he told me that he took about six months to get the crossover on his speakers correct. Stuart said, "the crossover is the heart of your speaker - it takes an experienced speaker designer a long time to get a crossover correct, let alone someone who doesn't do it for a living".  I have also read elsewhere that Mike Lenehan has made similar comments. Oh yes, Leigh and Leon from Kyron Audio took months to design their crossover, and they are using DSP, like me. Unlike me, all these people know what they are doing. 

 

What I can say is that the system is sounding better than I have ever heard it. The bass problem that has plagued me for years is gone. My friends think that it now has a coherence that it never did have, and the bass does not sound separate to the mid and high end as it did before. In less than a year, with the help of Acourate and the Acourate community (to whom I owe a massive debt), I have already designed a crossover which is superior to the standard passive crossover that came with my speaker. 

 

For sure, making a move like this has brought its own share of new problems for me to solve. I do not yet have a full grasp of what is possible with Acourate, and I know of people in the Acourate community who have been slowly chipping away at problems for years. But for me, what is important is that I have a way to objectively identify and solve those issues. I do not have to go on an equipment merry go round, trying this interconnect cable or that power cord or different valves in order to solve an issue which a new purchase may not solve anyway. 

 

Whilst I know of other people who use Acourate, I do not know a single person in Melbourne who uses Acourate as a crossover replacement. This, and my lack of knowledge and experience makes it more difficult, and I am probably struggling more than some others would.

 

In my favour, I am an experienced audiophile, and I know what music is supposed to sound like. I can hear that something is wrong. But correlating what I hear with what the microphone hears and what is displayed on my screen is another matter entirely. Sometimes I hear things that the microphone apparently doesn't hear - and then I wonder if the problem is with my ears, or whether I am looking at the wrong measurement, or made the wrong diagnosis, and so on. Having someone else to listen to my system and suggest a diagnosis or a measurement would help immensely - but I don't have any such friends. 

 

One last thing. I don't silver gild anything that I do. You won't find me going around telling people that my DAC is the best in the world, or that I am some kind of guru who knows what's best for everyone. I am on a journey like everyone else, and I make mistakes. I report my mistakes here - not because I want to scare people, or befuddle them with my superior knowledge - but because I want others who know more than me to come on this thread and tell me where to look. 

 

FYI I am in no way offended by your question. I would like to thank you for asking it! 

2 hours ago, davewantsmoore said:

Even with the "fix everything" buttons, that a system like Accourate, Dirac, or DEQX provides for you  (ie. linearise drivers, and apply XOs) ..... there is a still a lot of considerations if you want a good result.

 

 

Yeah, nail on the head again. If you want, Acourate has automated macros that greatly simplify operation. But the true power of Acourate comes from doing things manually. It gives you full control of anything you want to do, including things that I do not understand. I have used DEQX before, but even DEQX doesn't give you as much control as Acourate does. 

3 hours ago, Keith_W said:

Having someone else to listen to my system and suggest a diagnosis or a measurement would help immensely

 

 

If you want to "identify and remove errors" ....  then this isn't something which is difficult if not impossible by ear.

 

Take measurements, with a view to understanding what problems they show .... is the only way.

 

 

 

Listening is only good for things which are subjective assessments.

 

For better or for worse, I approach things differently. If I don't like the sound, I don't care what the measurements say. 

 

BTW, to answer an earlier question I had asked before we got sidetracked (re: the phase of the horns), more careful listening told me what was going on. In one configuration, the soundstage is wider, but instruments are poorly localized and soloists seem to be stretched out. In the other configuration, the soundstage is much narrower but individual performers are more precisely located. It turns out that the latter configuration was how I have had it all along, and I need to give my friend a bollocking when I next see him. 

On 11/12/2016 at 3:48 PM, joz said:

@ghost4man

i hope you understand Paul has been doing it much longer than Keith.

So one would expect him to nail it in a setup. After all he has done the same 

 configuration before to a standard where he would be happy.

Keith on the other hand is a hobbyist that has taken onboard something which he has never used before.

So I think it wouldn't be fair to the two lump the two into the same catorgory.

Im not sure how much experience you've had with measuring and setup?

But from what I have seen with any new measurements and setups there maybe more than a couple of options.

Not everyone settles for the first set of measured parameters.

They tend to be only a starting point.

@joz

 

I agree wholeheartedly.  It is certainly not the expectation that Keith would get it right in the first afternoon. I am not implying that.

BUT Keith has well and truly conducted a whole lot of testing which would suggest that he has well and truly done this above and beyond.

I know that some people dont feel comfortable in saying what I have said but a few private messages from people on the side watching what

is happening are understanding of my comments. I appreciate that these people may not represent the consensus here but it is worth noting.
 

10 hours ago, Keith_W said:

For better or for worse, I approach things differently. If I don't like the sound, I don't care what the measurements say. 

 

If you say this....  then I fear you might have fundamentally misunderstood me.

 

a)   The measurements cannot tell you what you like  (I completely agree with you)

b)   What you like cannot tell you what the design errors are in your speaker.

 

 

So....  you don't like the sound.   What is the cause?  Is it just that you like a "different flavour"? .... or do you have a design error in your speaker?

 

You will never know without investigating ..... and this question is of upmost importance for getting good sound.

10 hours ago, Keith_W said:

more careful listening told me what was going on. In one configuration, the soundstage is wider, but instruments are poorly localized and soloists seem to be stretched out. In the other configuration, the soundstage is much narrower but individual performers are more precisely located. It turns out that the latter configuration was how I have had it all along, and I need to give my friend a bollocking when I next see him. 

 

You've qualitatively described A and B.

 

This hasn't exactly said "what is going on" ....  Sure, your friend may need a bollocking, but if he has special ears and all that, then he may be hearing something that is not quite right .....   the only point is, that you don't know without quantitative data the difference between errors and flavour.

 

Errors can be addressed for an improvement in SQ .... Flavour (like it or loathe it) is your opinion, and never "wrong".

 

 

1 hour ago, ghost4man said:

BUT Keith has well and truly conducted a whole lot of testing which would suggest that he has well and truly done this above and beyond.

 

I know where you're coming from, and I agree that the tool is probably hasn't expedited the process  (but would I say "change"? Probably not).

 

... but who's to say "a whole lot of testing" is the right testing, or the data is interpreted the right way.   Speaker design is hard, it's easy to waste mammoth amounts of time and get nowhere great.

 

Hmmm... that seemed like me kicking the ladder out from under Keith.   Heheh.   Hopefully I am understood.  I don't mean to imply he is a fool with bad sounding speakers.

I have spent the past 18 months rebuilding an engine.

Could have (maybe should have?) bought a new one, but where is the fun in that?

Sometimes the journey is as important as the destination, the more elusive the sweeter the victory

This has gone quiet. 

 

What's happening? I am having withdrawals from reading this thread haha

What's happened is that I am happy with the way it sounds, and I haven't bothered tweaking it any further For sure there are things that I can do to massage more performance out of it, but I can't be bothered. It sounds great, I am happy. That's all there is to it, really. 

Incidentally, Bob Katz (yes, the legendary mastering engineer Bob Katz) recently posted a step-by-step tutorial on time alignment using sine wave convolution here. You have to register with Yahoo Groups in order to view the document. It looks as if this sine wave convolution method may be more precise than the current method I am using, which is: 

 

- delay each driver by a known amount (this spreads out the drivers in time so that it is easier to look for the pulse peak. If you don't do this, then all the impulses will arrive almost at once or with very little delay between them. It makes the peaks harder to find) 

- take a sweep. Note the position of each individual pulse peak. 

- Subtract the known amount from the pulse peak position. This gives you the delay. 

 

With sine wave convolution: 

 

- create a temporary crossover with no midrange. Subwoofer and tweeter only. 

- create another temporary crossover with no sub. 

- take a sweep. The result should look like this: 

 

 

- create a sine wave at the subwoofer XO frequency and convolve it into both curves. You will get this: 

 

 

- take note of the difference between the sub sine wave (teal) and the tweeter (black). Rotate the timing of the sub so that it matches the tweeter, and you get this: 

 

 

- go back to your real crossover, put in the settings, save, sit back and enjoy. 

What happened to..."I am happy with the way it sounds"....

I am  That's why I haven't bothered doing it! 

Uli also posted an automated time alignment method. It can be found in the acourate wiki.

 

I haven't read all the pages of the thread, but did you utilise the option of getting help from Uli via teamviewer?

 

I have used acourate to improve biamped vaf i66 significantly in my room.

Now I am using it to setup a PSE144 and i agree using acourate for the xovers is much more complicated.

Ah, you mean this? Interesting, I haven't seen that one before. It certainly looks easier than the sine wave convolution method, but it doesn't look as if it can be used to align a subwoofer. Have you tried it? 

 

No, I haven't tried chatting with Uli via Teamviewer. 

Haven't tried it yet. Not sure of the sub, but one option is to add a lot of gain to the sub to help the time alignment process - it makes the impulse response peak bigger. Of course the gain isn't used for playback.

 

Uli offers the free service via teamviewer, where he will run you through the processing of your system. If you are now happy with your system, it's rather moot, but it would be a shortcut to identify problems from the measurements and their treatments.

18 hours ago, Keith_W said:

What's happened is that I am happy with the way it sounds, and I haven't bothered tweaking it any further For sure there are things that I can do to massage more performance out of it, but I can't be bothered. It sounds great, I am happy. That's all there is to it, really. 

No way! True audiophiles are never happy with the sound

43 minutes ago, Mivera Audio said:

No way! True audiophiles are never happy with the sound

 

True, ...but eventually you get sick of playing and just want to tune into the music 100%.

 

Well, that is how I am anyway.  When my friends come visit lately and ask what I have done, after saying, "listening to the music", I am almost outcast!  Thier cries fall on deaf ears, I will install a few new toys when I am ready, for now they are resting in their boxes safely stashed away awaitingh for the im[pulse that will happen one day out of the blue.  Correct, there is just no common sense in that kind of thinking, but that is how it is here.

 

You keep tuning in to your Bach Keith.  Enjoy!

Nice to see you here, Mike. As you can see from my first post, the inspiration for this system came from your post in WBF. It was a real eye opener. 

Thanks. I'm glad to see a year later you finally have a system your happy with. We are much closer to our system becoming reality as well. Unfortunately everything takes forever with audio. There's going to be another free alternative you can try as well. It's called Daphile. Much better GUI than Hqplayer, and free. Has a convolution engine as well. It currently only does PCM-DSD upsampling but DSD-DSD is right around the corner. I'm liking the sound I'm hearing better than Hqplayer. I have a client using it as well and agrees. I read about your dislike for Roon so this might be just what your looking for. 100% headless as well. Just burn onto a USB stick and boot from USB. Then simply enter the IP address of the server into any browser and you have the GUI. Only problem for you right now is the Ravenna compatibility. But I'm working on that. I already have a working Linux Ravenna VSC for our system that's NADAC compatible. 

Thanks. I did have a look at DAphile but it's a Linux install. Merging doesn't have an official Linux driver as far as I can see, so it makes it a non-starter for now. I will be interested to try it as a USB boot once you have the Ravenna driver up and running. 

1 minute ago, Keith_W said:

Thanks. I did have a look at DAphile but it's a Linux install. Merging doesn't have an official Linux driver as far as I can see, so it makes it a non-starter for now. I will be interested to try it as a USB boot once you have the Ravenna driver up and running. 

Keith,

 

It might pay to dual boot with Mac and Linux.

I have on occasion run Linux and then used virtual box.

 

I would never go back to Windows. Our main desktop is a Mac which is very similar I feel to Arch Linux.

 

Thanks, but I won't consider using Apple. 

Edited December 29, 20169 yr by Keith_W

Just now, Keith_W said:

Thanks. I did have a look at DAphile but it's a Linux install. Merging doesn't have an official Linux driver as far as I can see, so it makes it a non-starter for now. I will be interested to try it as a USB boot once you have the Ravenna driver up and running. 

There's so many versions of Linux that Merging will never have one for every version. We want to have both the option of Daphile and Roon so this means working with Kimmo the Daphile guy to connect his GUI up to our VSC/DSP engine. I forgot to mention the current Daphile is also only 2 channel.