You have probably covered the topic some other place -
but could you please elaborate on how you control volume in your setup.
I would also like to know if you have tried doing your stuff using Vista and especially from Media center.
IMHO, The Media Center front end is a big plus when you have your familly using the system also.
I like your soundcard, but I think I will go for class D amps to save some enrgy 🙂
Thanks
but could you please elaborate on how you control volume in your setup.
I would also like to know if you have tried doing your stuff using Vista and especially from Media center.
IMHO, The Media Center front end is a big plus when you have your familly using the system also.
I like your soundcard, but I think I will go for class D amps to save some enrgy 🙂
Thanks
m0tion said:
Good point motion. It did occur to me too.
If you notice I swapped out the larger RAAL for the baby one, the difference is a ribbon length of 14cm vs. 7cm, The little RAAL also now uses the foam deflectors as seen in the larger RAAL's. Alex promises that this now makes the design comparable to a 1" dome disperation in the vertical and a fair bit better in the horizontal.
The stand for the center could also be made a couple more inch higher to further bring it inline with the main drivers.
Center channel placement is very nearly always compromised but it might just work. From experience the large RAAL in the LGT's is more than acceptable off-axis when the deflectors are used. There's still plenty of HF content even stood up from a couple meters away, the little RAAL should be better.
cph a decent switching amp will use as much as a linear a/b when sat in idle, the efficiency is greater when running, it's a bit of a red herring.
terry j said:
Hi Terry,
Yes that's right.
In simplistic terms; If you imagine the sample rate of the correction filter as its bandwidth, the resolution is determined by the bit precision and the number of samples(taps) within the filter is the length.
More taps offer more flexible and accurate correction.
Yes for a filter with 65536 samples then your looking at a delay of around 1.5 seconds using Convolver set to 16 partitions. Partitions split the convolving process up into multiple parallel tasks which are reconstituted at the end of each cycle. This greatly lowers the convolution processing time (down from 1500ms to about 10ms) but the filter length cannot be reduced without affecting the level of correction and accuracy so this is where the primary delay occurs.
This is no problem for audio only tasks such as music playback but forget it for movies or any realtime application. In these cases you need minimum phase filtering and again this is possible on the PC so you don't have to start being selective about what your audio system can or cannot do because of latency.
Don't be afraid of the computer! It should fear you 😀
I'm writing a loudspeaker DSP guide right now that will really break down the steps needed to achieve the things I've been talking about throughout the latter part of this thread and much more besides. It should make it accessible to virtually all interested parties regardless of technical know how.
That PCXO guide is pretty much outdated and was lacking a lot of information that would make it understandable to the casual non PC geek 🙂 There are newer, simpler and better methods now.
In my case the computer does everything - its the music server, its the gaming machine, its the movie playback device and a its the loudspeaker DSP and crossover. But the methods used for the latter part of that equation are entirely flexible enough to take external audio and process that.
Aside from this you can use programs such as AutoIT to create scripts that accomplish anything you'd do with a mouse and keyboard within the windows environment. So on windows bootup you could load Console set it to activate and load a particular crossover preset and that would be it. To the end user this would appear as a box you turn on, wait a couple of minutes for it to initialise and then from there you play music or whatever.
Such is the power of this setup that you can even add an IR with remote as I've done. This then allows you to setup more AutoIT scripts that tie in with the IR software hotkeys. You can then very simply switch between crossover and correction setups as the push of a remote button ie. low latency minimum phase for movies, quality linear phase for music. The limits are endless here.
So a streamlined and silent PC box is possible that would operate just as the DEQX or any other standalone hardward would. Of course when you wish to tinker under the hood you'd need to rig up a keyboard, mouse and screen but once set you can forget and still have all the convenience of a modern standard alone device. Its amazing how far it has come even in the couple of years I've been playing with all this.
cph2000 said:
Hi cph2000
For the volume control I'm attenuating at the source(the PC) for now. But Russ White is designing a flexible remote preamp that is configurable and expandable to allow multiple inputs and outputs. Its an excellent solution for active loudspeakers and/or home theater.
Re: Surrounds and centre channels
Grumpy_Git said:Shin
If you really wanted to do it properly you would just build a few more LGTs and use an acoustically transparent Projector screen!
Now theres a system!
Nick.
tinitus said:
I've seen these before.
I would say they'd be excellent for inbedded solution or where the user doesn't need maximum flexibility or quality.
The feature set is comparable to the DCX2496 ie. min phase filtering with PEQ and software configurable. The added bonus is the EQ automation.
Downsides are fixed DAC's, limited flexibility and limited power. Certainly a great answer to some situations but its not comparable to the PC solution.
ShinOBIWAN said:
I agree with your main message, but I think more of the good stuff can be had in the DVD Video department.
I have done tests with 5.1 setup and linear filters on my 5 year old laptop. With a modest delay on the picture through AC3 filter - and 32 k tap filter length it seems to work pretty well. But who knows? Maybe my laptop is blessed with a very slow graphics processor .😀
There is a correlation between number of correction paths and processing speed, so it may be harder to get full lipsync + full stability on systems that are both multichannel and active. A vital Duo processor should be able to handle the task pretty well though. In any case a no-compromize solution may be trivial to implement pretty soon if it isn't already. All that is required is cpu power and video delay control with a little headroom.
A minimum phase solution is of course a good fallback strategy if the linear phase route fails. And it can be made to sound very good too.
ShinObiwan do you use console as vst plugin controller?
If so, can you explain me one thing. When i play music from some program and set the output to ASIO for example, i hear it. Problem is that when i turn up console and regardless of connecting and disconnecting inputs and outputs in console and maybe some plugins, i always hear the same sound. How can i break direct connection between program that plays music and drivers, so that sound passes through console apllication?
Thanks
If so, can you explain me one thing. When i play music from some program and set the output to ASIO for example, i hear it. Problem is that when i turn up console and regardless of connecting and disconnecting inputs and outputs in console and maybe some plugins, i always hear the same sound. How can i break direct connection between program that plays music and drivers, so that sound passes through console apllication?
Thanks
Mx said:
Sounds like an issue that needs to be solved in the soundcard mixer/drivers.
Can you mute the playback channel within the soundcard ie. the one which is assigned and used by the audio playback software? Depending on the soundcard it might be a routing issue where you need to turn off any 'monitoring' that might be active on that channel.
What soundcard do you use?
I use creative live 5.1 for now (planning to buy emu or m-audio), and Kx drivers with it, which have a lot of routing options, but i never used ASIO with them. I have used crossovers, notch, peak filters and even wrote some plugins for myself inside the Kx drivers. But they have their limits, so i want to check this VST aproach.
I will soon be continuing my project, 2 way speakers with 1m tall ribbons, which i have designed (unique magnetic structure, field accuracy +/-0.01T ). Ribbon (i made one for now) sounds great up from 250Hz, but i must use 4 order crossover, and planning to do all the Media center and PC XO in one box.
You are right, there is probably some monitoring going on, problem is i can't find that in Kx router.
Best regards,
Mx
I will soon be continuing my project, 2 way speakers with 1m tall ribbons, which i have designed (unique magnetic structure, field accuracy +/-0.01T ). Ribbon (i made one for now) sounds great up from 250Hz, but i must use 4 order crossover, and planning to do all the Media center and PC XO in one box.
You are right, there is probably some monitoring going on, problem is i can't find that in Kx router.
Best regards,
Mx
Why is it low order crossover slopes are more musical than higher order ones?
It seems like a balancing act to me. I perceive a scale where at one end you have technical precision in some aspects of the sound such as the effects produced by stereo ie. imaging etc. and then on the other end you have musicality, dynamics and tonality amongst others. Roughly correlating to these are, respectively, steep filtering and shallow filtering.
I have 4 setup's(from shallow to steepest):
2nd order Linkwitz Riley
4th order Linkwitz Riley
8th order Linkwitz Riley
1st order Neville/Thiele
All are 3.5way with crossover points at 200, 600, 2500hz.
Each of these uses exactly the same measurement procedure for driver linearisation (ie. mic 40cm on axis) and all are room corrected at the listening position according to the same target curve and indeed the FR of each are very similar. They're essentially identical in all but crossover slopes and any resultant changes that illicits.
I've setup console to allow instant switching of the differing types and its interesting to compare the differences which are not subtle. Its a great test bed for such an investigation because of the ability to closely match on axis performance of each config and then quickly switch between them during playback allows to almost exclusively here the effects of just the crossover slope change.
Maybe I'm maturing as a listener and becoming less wowed by hifi special effects but I 've got to say the 2nd order sound wonderfully involving, dense and large.
A quick example would be plucked strings on a guitar, with the steep slopes you can really hear every single little detail - the initial pluck then the following transient and decay. Its impressive to hear but it sounds hollow in comparison to the the low order which loses the absolute detail but adds a natural dynamic richness and makes it more humane.
Which is more correct? I think one of my fundamental philosophies on how I should do audio just shifted.
It seems like a balancing act to me. I perceive a scale where at one end you have technical precision in some aspects of the sound such as the effects produced by stereo ie. imaging etc. and then on the other end you have musicality, dynamics and tonality amongst others. Roughly correlating to these are, respectively, steep filtering and shallow filtering.
I have 4 setup's(from shallow to steepest):
2nd order Linkwitz Riley
4th order Linkwitz Riley
8th order Linkwitz Riley
1st order Neville/Thiele
All are 3.5way with crossover points at 200, 600, 2500hz.
Each of these uses exactly the same measurement procedure for driver linearisation (ie. mic 40cm on axis) and all are room corrected at the listening position according to the same target curve and indeed the FR of each are very similar. They're essentially identical in all but crossover slopes and any resultant changes that illicits.
I've setup console to allow instant switching of the differing types and its interesting to compare the differences which are not subtle. Its a great test bed for such an investigation because of the ability to closely match on axis performance of each config and then quickly switch between them during playback allows to almost exclusively here the effects of just the crossover slope change.
Maybe I'm maturing as a listener and becoming less wowed by hifi special effects but I 've got to say the 2nd order sound wonderfully involving, dense and large.
A quick example would be plucked strings on a guitar, with the steep slopes you can really hear every single little detail - the initial pluck then the following transient and decay. Its impressive to hear but it sounds hollow in comparison to the the low order which loses the absolute detail but adds a natural dynamic richness and makes it more humane.
Which is more correct? I think one of my fundamental philosophies on how I should do audio just shifted.
ShinOBIWAN said:
Hi Shin,
My ears like B3 for the tweet-mid, and I used LR8 for the mid-bass to put as much human voice as possible on the midrange. There is some support for B3 on MTM and WMTMW in the literature, but a very close M-T physical spacing is also mentioned as important for B3.
I use descrete PASS-JFET diff-amps in my active crossover, and B3 uses just one stage, so this was a second plus at these sensitive frequencies.
ShinOBIWAN said:
The second order experience is more correct. Little details are considered necessary evils of the plucking techniques, i.e. noises. Acoustic instrument builders have developed guitars and violins and pianos etc. along with musicians during the centuries, for certain performing space formats. For example a Stradivari is loud for concert halls and an Amati is softer but more colorful for chambers. They were all heard from a distance. Beauty of tone and richness of palette for certain musical and performing criteria to be met, was their goal. Close up mics with the advent of multi track recording emphasize detail -falsely- over tone and body and space. They reveal noises -not meant to be picked up by audiences- that many of as try to emphasize in the replay chain thinking that more is better. Less is more.
I have also noticed the same thing. A steeper crossover does give a more detailed sound but less "musical" The beautiful of digital crossover is that you can change it any time with just a few key strokes! So we have the best of all worlds!
?????
I do not understand all this argumentation.
A high XO order allows a better separation of the drivers. It also suppresses undesired effects of drivers e.g. cone breakups (assumed that the breakup takes place in the stopband range).
A disadvantage is more oscillation with the step response of the lowpass filter where some people claim to be sensitive for.
So we have arguments for and against higher XO orders.
But to use the term "correct" is IMHO not correct. If the noise of music instruments is recorded then the playback system should reproduce the noise to be correct. Otherwise the playback system generates its own sound which may sound good or "musical" but is not correct anyway.
At least I expect to hear what is recorded. And if the recording is not good then I search for a better one. This can also be a recording which does not use close-up miking then.
Of course you can also treat such a recording by any method to get a pleasing sound, you may even apply artificial reverb, harmonizer or other stuff like harmonic tube distortions. Then it may sound pretty well for you. But then it is not correct.
BTW my preferred crossovers today are higher order Bessel crossovers which do not oscillate in the step response. One driver at the XO frequency I use has 10th order whereas the other driver is always of second order (automatically given by the subtractive approach in the XO design). It is possible to define if the steep slope is then on the right side (lowpass, bandpass) or at the left side (highpass, bandpass).
I was expecting this, that is why I wrote 'correct experience' and not correct replay. Its a matter of preference. Want a lab system sounding correct replaying the vast amount of musically wrong recordings? Great for research. Want a 'musical system' helping enjoy better most of your recordings? Great listening system. With emu cards and a PC it is reduced down to just an option. No sweat.
Best
Best
🙂
No complain.
My only problem is then to talk about quality of playback systems as my ears certainly experience music in a different way than your ears. There is no neutral criteria anymore in this case. Also statistics of pleased and unpleased audiophiles will not help.
- Status
- Not open for further replies.