It is more complicated than that and trying to make it simpler will lead to misunderstanding.
FIR filters can be minimum phase and to all practical purposes have no appreciable latency. It then depends on how the convolver works as to whether it will add latency, but that is not then due to the filter but the software processing.
If the FIR filter is linear phase or mixed phase it will have some latency depending on the length of the filter and where it's impulse centre is located.
As krivium said above it is adding a time based correction that causes latency to be added. In order to process something that has already happened you delay the signal to allow it to be processed before being heard so you don't have to bring out the time machine.
Whether you can improve the sound by using an FIR filter is a difficult question. If you know what you are doing and why then I think the answer is yes. But on the flip side you can totally destroy the sound as well.
Using a linear phase crossover filter does have some potential benefit over a minimum phase one especially as it gets steeper. But as always the best result for a particular situation can't be decided based on rules you read on the internet
I did already understand the weightings but I don't think I was understanding the context of your reply. Did you mean that the true SPL could be up to 12dB higher than shown on the meter depending on the weighting used?
If you use IIR XO and a steep slope (causing high group delay)....and clean it up with FIR downstream isn't it essentially the same thing?
I couldn't figure out how to test the latency of my system but this is pretty ingenius...set my mic right next to my mouse with my first person shooter open....record a click...measure distance in between mouse click and gun shot lol
Last edited:
Jriver does have an ASIO input driver so you can use it to process audio in real time although I suspect you would need to test that properly with the latest version to see if it met your needs.
If you are already using DAW software then you should be able to set a convolution plugin either singly or in multichannel to process your mix straight out to your active system.
For playback of regular music files then Jriver is miles ahead, you could use both within the same system, just not at the same time.
I'm glad I am not the only one. I have everything R128 volume levelled so I end up somewhere around the -6dB from reference mark to take account of the dynamically challenged music in my collection
My background noise is pretty low in the 30dB range to 40dB range, not the inky blackness of a studio but not bad for home with no real effort being made to get there.
I haven't got the latency tested yet but from what I've read I'll get 1ms from it.
So I installed it...so this can basically do for me what Jriver does also? If so, thats pretty cool.
At one point in space, yes, over a wider area is less certain.
Is it simple, yes, is it easy to replicate with simulation, no. Modelling the frequency dependent time delay through the intended material is not so easy. I have had the same idea but shelved it.
It is not the same thing but whether that makes any practical difference to how you use it is hard to say. Equalizer APO sends the audio through the windows subsystem, Jriver can use ASIO or WASAPI and bypass that. If you use metaplugin in Jriver then there really isn't anything that you can't do within the limits of the number of audio channels you have available.
Whats the path to super cardiod for someone like us..or you...or we...or me...??
Do I need to pick up four 10" sealed back woofers, place them on the sides of my boxes and tune them with dsp?
Last edited:
Phase can be flattened with Finite Impulse Response filters, in exchange for latency. The duration of the FIR filter needs to be at least twice the period duration of the lowest frequency it’s describing.
1000ms/20=50ms, 50X2=100ms. That’s 100 ms at 20 Hz.
100ms of latency would make for an obvious echo or lag between a live source and the FIR output, but no problem for playback only.
Merlijn van Veen - The harmonics lead the fundamental
The above link has a video that allows you to audition between IIR with varying degrees of phase shift and FIR with flat phase.
Even using headphones with a flat phase response, I can not hear the difference between them.
That said, I could easily detect the comb filtered, messed up frequency response of the combined sound in the video linked in post #6471 at 5ms, while still not hearing an "echo" with that little delay.
I find the audibility of phase changes to be much more pronounced when it is done based on an in room measurement at low frequencies. If outdoors or anechoic I don't imagine I would be able to pick it at all.
Depending on the weighting and time constant (Slow, Fast or Impulse) used, the "true SPL" could be far more than 12dB higher than shown on the meter.
For instance, a 30 Hz continuous sine wave would read 40 dB less than actual on "A" scale.
Dropping a spoon on a tile floor might read only 70dB SPL "Slow", while on "impulse", could read 110dB or more.
Attachments
Why would you want a supercardioid pattern?
Putting that question aside, rear or side speakers with dsp delay and filtering can change the polar pattern, like Kii uses.
Fulcrum acoustics has a patent on passive cardioid too, there is useful information on their site to look at.
Attachments
At 2:00 he changes the xo and I can hear the phase shift after this....its subtle.
Its pretty bad test material...the music has all sorts of automated filtering going on lol...and the melodic notes rarely constant or short sustains, with the tempo....a very slow played piano with long notes in sustain would of been choice....00
ON second though maybe I'm just hearing the processing errors of the phase sweep? as it goes from one polar to the other?
Last edited:
I am just repeating things I've read that are likely incorrect...judging by the diagram you just provided (ty ty)... I don't know what type of cardioid pattern I'd be creating by placing one of these Beyma Speakers - Beyma 10MCF400Nd closed back carbon fiber midrange speaker with a neodymium magnet - Beyma 10MCF400Nd 400 watt AES 10" speaker for all midrange applications. Beyma 10MCF400Nd sealed midrange and other Beyma 10" speakers here. on each side of the mid box...or maybe between the the mid and horn even....those are pretty expensive though...I have enough amps...I wonder if it would be effective. What type of pattern do you get when you put the cancelling source on the sides?
I just learned you can use VCad to model it...but with something like the product I listed, used correctly...is that going in the "right" direction?
Last edited:
I like the "cardiod" options. Which is why I showed how to do this in my book both actively and passively, although I didn't discuss the use of side ports with damping. That is a good idea if all the gotca's have been worked out.
The Dutch 8c has left an impression on the Mastering Engineer community apparently....the only thing unique about it to me is the cardioid polar....The polar has been polished of course for sales but...non the less, it is apparent to me, that there is a rise in Sq to be had in every bit of deliberate directivity....so I kinda looked at it like, am I doing everything I can do....no....is active cardioid a possibility yes?....
The product I listed above, the Beyma, what you guys think of that mounted on the sides for the role? Maybe theres some better suited tactics for active cardioid?
The product I listed above, the Beyma, what you guys think of that mounted on the sides for the role? Maybe theres some better suited tactics for active cardioid?
You can get passive super cardioid by building the D&D but then opening up the back. Adjust the resistance medium, rockwool in my case. I thought it was going to be difficult but it was easy. The end result is not a product you can sell commercially or that would survive shipping but you can DIY it.
The 8C is a good all in one package for those that need or want to put the speaker near the front wall and don't want the speaker to be too big.
The midrange is passive cardioid and the polar looks to be quite consistent, but it is pretty wide from 1K down.
The back sub drivers operate below 100Hz and are designed to be used close to the wall behind them to leverage output and avoid the cancellation dip by putting it out of band for that driver.
They are not part of the cardioid arrangement.
You could look at the Genelec W371A Woofer box for some inspiration, wildly expensive but seems like a good idea.
The Ones | Designing the new W371 Adaptive Woofer System. - YouTube