• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

DEQX Premate 8 digital active crossover / DSP

Mixed phase. I think the DSP Nexus has something like 1024 FIR taps per channel, which is not enough. So they also have min phase IIR's as well. The DEQX has a much more powerful processor, so it has 32768 FIR taps per channel in linear phase.



Older units were 96kHz. The sample rate does not matter unless you are concerned about latency.
Thanks again for the rapid replies :)
I am more concerned about high frequency resolution. In my view the higher the resolution in both the time domain (sample rate) and amplitude (bit) the less approximation in the dsp
I agree, they are not publishing enough information to convince me to part with my money. I will probably be seeing the DEQX guys next month at a hi-fi show. I can ask them a few more questions then.
 
Thanks again for the rapid replies :)
I am more concerned about high frequency resolution. In my view the higher the resolution in both the time domain (sample rate) and amplitude (bit) the less approximation in the dsp

Contrary to popular belief, neither higher sampling rate or greater bit depth give you more resolution.

Higher sampling rate means that higher frequencies can be sampled. For 96kHz, the Nyquist frequency would be 48kHz. Anything between regular 20Hz - 20kHz is exactly the same resolution. "Hi-res" audio is a marketing term that does not reflect reality.

When it comes to DSP, choosing a higher sampling rate actually LOWERS your resolution. The size of your bin depends on the number of taps you have and your sampling rate. So for e.g., if you have 32768 taps and you use 48kHz sampling rate, the size of each bin is (48000/32768) = 1.464Hz. If you use a 96kHz sampling rate, your bin is 2.93Hz wide. If you want to increase the sampling rate AND maintain the same bin size, you need more taps. And getting lots of taps requires lots of computing power which the DEQX does not have.

Greater bit depth means more dynamic range. For e.g., 16 bit audio has a dynamic range of 96dB, and 24 bit audio has a DR of 144dB. This means that if you play your 16 bit file at 100dB, the softest sound it can reproduce is 4dB. Just FYI your room's noise floor is typically 35-45dB and this is WAAAY below your room's noise floor.

The audible difference between 16 bit and 24 bit audio, or for that matter a sampling rate of 48kHz vs. 192kHz is so miniscule that I will say it is inaudible and challenge anyone who says otherwise to pass a blind test. And even if you do pick it, the difference is swamped by DSP. What is all important is how well you take measurements, and how well you design your filter. THAT is what makes a HUGE difference. I guarantee that anybody with functioning ears can hear the difference between a poor filter and a good one.
 
AFAIK the DEQX is the only hardware DSP unit I am aware of that has linear phase FIR filters. This alone is a massive advantage.

Surely there are some more other hardware options out there, but the FIR taps available is usually limited — current expensive “professional” digital processors appear to max out at 4-16k. Using mixed phase designs does not necessarily produce massively inferior or worse results — in fact, it is actually more practical.
 
Contrary to popular belief, neither higher sampling rate or greater bit depth give you more resolution.

Higher sampling rate means that higher frequencies can be sampled. For 96kHz, the Nyquist frequency would be 48kHz. Anything between regular 20Hz - 20kHz is exactly the same resolution. "Hi-res" audio is a marketing term that does not reflect reality.

When it comes to DSP, choosing a higher sampling rate actually LOWERS your resolution. The size of your bin depends on the number of taps you have and your sampling rate. So for e.g., if you have 32768 taps and you use 48kHz sampling rate, the size of each bin is (48000/32768) = 1.464Hz. If you use a 96kHz sampling rate, your bin is 2.93Hz wide. If you want to increase the sampling rate AND maintain the same bin size, you need more taps. And getting lots of taps requires lots of computing power which the DEQX does not have.

Greater bit depth means more dynamic range. For e.g., 16 bit audio has a dynamic range of 96dB, and 24 bit audio has a DR of 144dB. This means that if you play your 16 bit file at 100dB, the softest sound it can reproduce is 4dB. Just FYI your room's noise floor is typically 35-45dB and this is WAAAY below your room's noise floor.

The audible difference between 16 bit and 24 bit audio, or for that matter a sampling rate of 48kHz vs. 192kHz is so miniscule that I will say it is inaudible and challenge anyone who says otherwise to pass a blind test. And even if you do pick it, the difference is swamped by DSP. What is all important is how well you take measurements, and how well you design your filter. THAT is what makes a HUGE difference. I guarantee that anybody with functioning ears can hear the difference between a poor filter and a good one.
Regarding sample rates.. I simply do not agree. -This is why:

Let’s take a 44.1khz Sample rate. Given that a soundwave consists of a zero, top, zero, bottom. We can neglect the zeros, so a 22,05 kHz wave can be perfectly described by a sawtooth wave. Right?

But what about 18kHZ or any arbitrary frequency above lets say 1000Hz ? there’s not enough small points in time for a 44.1kHz system to “render” a 18kHz. These points in time is fixed so a 18kHz wave top and bottom sometimes coincide with the fixed places in time, most times not..

The same is true for lower frequencies, the chance for the system to “get it right” is just higher.

I believe that this is why cd treble always has the same “color” or flavor if you will, compared to analog systems and perhaps lag some air and fluidity.

Regarding dsp and digital crossovers, there will be a time component alteration when applying eq and slopes, and these amplitude values will have to be allocated to a certain timeslot rounded off to the nearest one. The higher the time resolution AKA sample rate, the less “rounding off” or approximation, the less distortion..
 
Regarding sample rates.. I simply do not agree. -This is why:
There is really nothing for you to disagree. Read up on the Nyquist–Shannon sampling theorem. It is mathematically proven.

The requirement of band limiting is central to sampled (i.e. digitized) signals. The signal can only contain frequency components below 1/2 the sampling frequency (the Nyquist frequency, = Fs/2) for sampled signals to work. The theorem says, given a sequence of digitized samples, there can be only one unique band-limited continuous time signal that will give you those samples. Therefore the signal can be mathematically perfectly reconstructed (and can be practically reconstructed to have errors far below human audibility, currently better than -120 dB for audio signals).

The sequence of digitized samples will NOT uniquely describe signals that are not band-limited to Nyquist. Imagine if you take a sample every minute. There can be a lot different signals that will give you the same samples. However, following the theorem, if you limit the signal to not having any frequency content with rate higher than 1/2 cycle per minute (e.g. with low pass filtering), there can be only 1 such signal.
Let’s take a 44.1khz Sample rate. Given that a soundwave consists of a zero, top, zero, bottom. We can neglect the zeros, so a 22,05 kHz wave can be perfectly described by a sawtooth wave. Right?
The sawtooth (or more correctly, triangular wave) that you envisioned is not bandwidth limited to Fs/2. It is just one of the infinite number of possible non-band-limited waveforms that will give/fit those samples.
 
A triangular wave is similar to a square wave, in that it is a sine wave + added odd order harmonics all the way up to infinity. The 3rd harmonic of a 10kHz is 30kHz. I assure you that the harmonics are beyond the audible range of hearing. Not to mention, your speakers are band limited and will be incapable of reproducing the said triangular wave.

In the meantime, do heed my warning about needlessly increasing the sample rate - you will need more taps. The most powerful DSP programs like Acourate and Audiolense are capable of 262,144 taps per channel, but the DEQX can only manage 32,768. The higher you push your sampling rate, the bigger your bins will become.

1725638151207.png


This is a simulation to show what happens if you decimate a subwoofer crossover down to various tap counts with a 48kHz sampling rate. If you double your sampling rate to 96kHz, it will look like the "16384" curve. Quadruple it to 192kHz, and it will look like the "8182" curve.

In short: do not worry about inaudible things like reproducing high order harmonics > 20kHz. Instead, worry about taking proper measurements and making good filters.

I will defer to @ernestcarl about mixed phase filters. I have less experience with those than he has, and I respect his expertise.
 
This is a simulation to show what happens if you decimate a subwoofer crossover down to various tap counts with a 48kHz sampling rate.

Doesn't look like the results from basic subwoofer xo filters (e.g. 24Hz HPF and 100Hz LPF). "Mixed phase" to me is just combining majority minimum phase EQ entries with a one or few backwards-time correcting phase EQ -- combined filters in a single FIR file more often than not is mixed phase rather than pure "linear phase". Many professional speaker manufacturers use FIR filters e.g. EAW, Fulcrum Acoustic, Meyer Sound, Neumann, Genelec. Fulcrum Acoustic, for example, mostly use 384 and 768 tap FIR filters at 48kHz and 96kHz sampling rate respectively. If I remember correctly, DSP correction for the separate the HF compression drivers, arrayed horn design in this speaker only uses +100 taps or so: https://www.fulcrum-acoustic.com/product/ahs-digitally-configurable-coaxial-horn/
 
I received my Beta DEQX Pre 8 in the in the very first batch shipped in February 2024. I expected to muddle through the programming of the Pre 8 following the instructions in the DEQX Pre 8 Preliminary Users Guide. I got nowhere. Everything I tried failed. I hated telling my audiophile friends week after week that I had had zero success getting the Pre 8 up and running.
Full disclosure is I am a little brain damaged from two TBI in the line of duty. I began to suspect that I was no longer smart enough to handle a component like the Pre 8. Then a couple of months ago I engaged DEQXpert, Larry Owens, to program my Pre 8 for me. After five sessions of three to four hours each with only limited success Larry told me that my Pre 8 obviously had a hardware flaw. He asked me to ship it to him in Colorado where he will either repair it or ship it back to DEQX in Australia for repairs.
I am so relieved. I am seventy-five years old, and I had come to believe that I was no longer smart enough to use a component like the Pre 8. Thank God for hardware flaws.
 
I received my Beta DEQX Pre 8 in the in the very first batch shipped in February 2024. I expected to muddle through the programming of the Pre 8 following the instructions in the DEQX Pre 8 Preliminary Users Guide. I got nowhere. Everything I tried failed. I hated telling my audiophile friends week after week that I had had zero success getting the Pre 8 up and running.
Full disclosure is I am a little brain damaged from two TBI in the line of duty. I began to suspect that I was no longer smart enough to handle a component like the Pre 8. Then a couple of months ago I engaged DEQXpert, Larry Owens, to program my Pre 8 for me. After five sessions of three to four hours each with only limited success Larry told me that my Pre 8 obviously had a hardware flaw. He asked me to ship it to him in Colorado where he will either repair it or ship it back to DEQX in Australia for repairs.
I am so relieved. I am seventy-five years old, and I had come to believe that I was no longer smart enough to use a component like the Pre 8. Thank God for hardware flaws.

Thats a horrible story. Sorry you had to go through thst - i hope deqx gets it running for you without fuss
 
Then a couple of months ago I engaged DEQXpert, Larry Owens, to program my Pre 8 for me. After five sessions of three to four hours each with only limited success Larry told me that my Pre 8 obviously had a hardware flaw. .
This is the most stunning aspect to the story.

It took an expert 15 - 20 hours to determine it had a fault?

I have an HDP-4 and know the software is complex to implement (not wave an iphone around and job done) but even if the new version of the software is more complex to configure, that long to proclaim "it's dead Jim"?

Peter
 
hello Don
I can well imagine the uncertainty.
I can reassure you that when everything is in order again you can look forward to great sound.
I've been running a first draft for a month now and I'm very impressed.

Like to hear from your experence when its playing .

Greatings Richard
 
Please remember what is written into the "Getting started quide".

Please keep in mind that this is beta software and there are aspects of the UI that are far from being in a refined state. They were quickly put together to allow testing of the fundamental operation of the unit and although the cosmetic refinement of the UI is in process, we’ll be using the basic and temporary ones to test the operation of the device.

You may see obvious graphical glitches in what’s being displayed on the web pages, and you may find others that we’re not aware of as yet, as the web UI is tested on different devices, browsers and operating systems. Feel free to screenshot and raise tickets in our customer service desk for issues that you see and think we may not have picked up already as this may help us. In most cases, we’re aware of them, and most of them are on the temporary screens that won’t be used in the final UI in any case, but your feedback is still welcome all the same.


--------------------------------------

I´ve had training session with DEQX Alan Langford. In the beta software there is several bugs that make the process of ´Identify Drivers´, ´Create A Speaker´ and ´Create A System´ really difficult. But when knowing the bugs and how to resolve the bugs the the endcome is great.

But I´m 100% sure we need to wait until the release version of the software is available. It should come in next 2-4 months.

BTW. Have you updated the latest firmware, V1.37? That is clearly better than the V1.36.
 
Guys, in a couple of weeks I will be seeing the DEQX guys in a Melbourne hi-fi show. Alan Langford and Kim Ryrie are there every year. If there are any questions you would like me to ask them on your behalf, post them here.
 
In brief, these are the differences. DEQX vs. Groundsound:

- 32k taps FIR filters vs. 180 biquad IIR's
- Linear phase vs. minimum phase
- Software <--- I have no idea how good the software is on the DEQX, I have yet to test drive it.

AFAIK the DEQX is the only hardware DSP unit I am aware of that has linear phase FIR filters. This alone is a massive advantage. If you want to learn more about why we should be using linear phase FIR filters, read this.

There are other differences, e.g. DAC's, power supply, and so on. All these other differences are extremely minor. I keep saying over and over again, the three biggest differences in DSP units are: the user, the software, and the hardware. In that order.
There simply is NO massive advantage of using linear phase digital filters in a practical multi-way loudspeaker environment. It is just a myth perpetuated by marketing people who usually have no engineering background. Even Siegfried Linkwitz and many other well respected luminaries in this field questions its benefits. Also what I find perplexing with this thread is how you constantly attack competing products and yet offer no insight into the hardware capabilities of this prospective product from DEQX and yet you yourself use a PC and a multi-channel DAC so why would you want to change it ? Isn't this what you can already do with an Okto DAC with little money down ?? Hello !!!
 
Let's dial the confrontational approach back a bit please . Not sure I read posts as attacking other products more elaborating on differences /explaining a personal choice . Thanks
 
Thank you @Jimbob54. I had him on my ignore list. I will provide some context: this guy manufactures a competing product for the DEQX. He runs his own forum with a "dark room" which is full of vile attacks against other industry members and some ASR members. He recently wised up and hid the subforum, but the Wayback Machine remembers. Perhaps he received another letter from a lawyer. It is a pity that he isn't as gracious as RME when discussing products of their competitors. Search his previous posts on this thread and you will see more examples of the same kind of reaction.

I am happy to have a discussion on the merits of linear phase with anybody who wants to chat in good faith, but not with this guy. He remains on my ignore list, and I will not engage with him. Anyway, back on topic.
 
  • Like
Reactions: OCA
Thank you @Jimbob54. I had him on my ignore list. I will provide some context: this guy manufactures a competing product for the DEQX. He runs his own forum with a "dark room" which is full of vile attacks against other industry members and some ASR members. He recently wised up and hid the subforum, but the Wayback Machine remembers. Perhaps he received another letter from a lawyer. It is a pity that he isn't as gracious as RME when discussing products of their competitors. Search his previous posts on this thread and you will see more examples of the same kind of reaction.

I am happy to have a discussion on the merits of linear phase with anybody who wants to chat in good faith, but not with this guy. He remains on my ignore list, and I will not engage with him. Anyway, back on topic.
Not true at all. It's still there for your benefit now. It's just that I didn't want my unsavory experiences developing a high-tech Audio product in Australia to cloud anything I do in the future :(

And what I said earlier in this thread about linear phase filters still stands unless you can refute it of course which is what you should be doing instead of shooting the messenger all of the time. Looks good on paper but in practice not so good :( The correct way of doing it is how Grimm and Kii audio (and maybe D&D too ?) have done it using a conventional IIR crossover with global phase correction applied usually implemented with an FIR filter ;)

 
Last edited:
All right, perhaps you could start by listing your objections to linear phase in your own words. I could quote my own authorities who have the opposite view to you, but that is an argument from authority (and a logical fallacy) and I won't do that.

In fact, the SNA show is on tomorrow. Alan Langford and Kim Ryrie from DEQX will be there (I checked). Maybe you could turn up and have a chat with them?
 
All right, perhaps you could start by listing your objections to linear phase in your own words. I could quote my own authorities who have the opposite view to you, but that is an argument from authority (and a logical fallacy) and I won't do that.

In fact, the SNA show is on tomorrow. Alan Langford and Kim Ryrie from DEQX will be there (I checked). Maybe you could turn up and have a chat with them?
Well that's exactly what you should be doing. Those references I cited present compelling arguments why you would not use this scheme to design a crossover rather than saying it's superior to anything else and they give valid reasons why. Quite simply two different drivers cannot guarantee proper impulse response cancellation off-axis no matter how good it is optimized on the main axis.

cheers
 
All right, the usual arguments against linear phase FIR vs. minimum phase IIR are the following:

1. High computation requirements vs. high CPU efficiency.
2. High latency vs. low latency.
3. Pre and post ringing vs. post ringing alone.
4. No phase distortion vs. phase distortion.

I am not going to discuss 1 and 2 to keep this post shorter. I generated these illustrations myself.

1728555705931.jpeg


Let's take a square wave and apply a linear phase low pass filter and high pass filter to it. If we look at the LPF and HPF, we see that both filters have pre and post ringing. However, when the LPF and HPF are summed, the pre and post ringing disappears. Thus, LP filters rely on proper summation of LPF and HPF, otherwise pre and post ringing will be exposed.

What would cause the LPF and HPF to sum improperly? Here are a few causes:

1. Asymmetric HPF/LPF slopes
2. Improper time alignment between drivers
3. Different directivity different drivers which means they sum on-axis but not necessarily off-axis. Linkwitz made this observation.
4. High Q aggressive phase corrections. Putzeys made this observation.

I should add that the mathematical causes for pre-ringing are now well understood and there are algorithms that can reduce pre-ringing to negligible levels. Since you bring that up, let's look at the on-axis step response of my speaker.

1728556807370.png


We can see a little bit of pre-ringing. Now let's go 30 deg off axis:

1728556855497.png


See much difference in pre-ringing? Even if there was, there is the phenomenon of pre-masking where any signal, provided it is low in amplitude and occurs less than 20ms before the main signal will be masked by the main signal.

Now let us look at what happens if we were to use a min phase filter. Let us take a square wave again and apply min phase LPF and HPF to it.

1728557069814.jpeg


Look at the output. The phase is completely distorted, and the post ringing does not go away. Now, you could argue that phase distortion is not audible, and post-ringing is masked by the main signal. Ohm's Acoustic Law and all that. 99.99% of speakers in the world use minimum phase instead of linear phase, and all of them distort phase in the same way. Quite simply we are used to hearing phase distortion like this.

Now I will wander into more controversial territory. I have the capability to generate both min phase and lin phase filters. I have directly compared them. There is no contest, linear phase sounds so much cleaner. The last time we met (maybe you have forgotten) you mentioned Bohdan Raczynski. You might be interested in what he has to say about linear phase. He cites a number of sources that show that there is increasing research that shows that phase distortion is audible.

So, do we choose linear phase, with negligible and inaudible pre-ringing? Or minimum phase with unavoidable phase distortion and twice as much post-ringing which can never be removed?

To wrap up - we should ask ourselves why should we choose FIR rather than IIR. FIR allows us to use min phase or lin phase. IIR forces us to use min phase only. We can get halfway there with mixed phase, but with very limited FIR taps. The ONLY valid reason to avoid using FIR is to avoid latency which is critical in AVR's so there are no lip sync issues. The other reason of course is so that you can use underpowered SHARC DSP chips which can't do high tap count FIR, put it into products like MiniDSP and sell it for cheap.

I am not letting DEQX off the hook here. I think it is incredibly overpriced for what it delivers. I will be meeting Kim and Alan tomorrow and I will have a chat with them.
 
Back
Top Bottom