Is it Really Worth It?

[kemmler3D]

Is this the way audio DSP works? I can never correct both the frequency and the time domain responses?

You can definitely do both, in fact @OCA recently posted a long video on how to do this to a high standard... it shows up in this thread: https://www.audiosciencereview.com/...bout-the-eq-method-showed-in-this-video.55115

To do both AFAIK you need to use FIR convolution but people do it all the time.

 

[KEM]

You can definitely do both, in fact @OCA recently posted a long video on how to do this to a high standard... it shows up in this thread: https://www.audiosciencereview.com/...bout-the-eq-method-showed-in-this-video.55115

To do both AFAIK you need to use FIR convolution but people do it all the time.

Thanks for the advice. I've skimmed throughout the video.
Do any of the miniDSP products have both time and frequency correction?
Does anyone?

 

[staticV3]

Thanks for the advice. I've skimmed throughout the video.
Do any of the miniDSP products have both time and frequency correction?
Does anyone?

Yes and Yes.

FIR correction is not that hard to come by.

 

[kemmler3D]

Thanks for the advice. I've skimmed throughout the video.
Do any of the miniDSP products have both time and frequency correction?
Does anyone?

Basically anything that supports FIR supports both. Convolution works on both at once, they can both be done with the same filter. The limiting factor is the number of taps which corresponds to frequency resolution.

 

[terryforsythe]

OP here.

I used the EQ inside REW to equalize the data I showed in the first post. I did find an approach that would flatten the FR, but the time domain was untouched.
Also, I had REW create a minimum phase version of the same data. It did a great job on all time factors, but didn't change the FR.

Is this the way audio DSP works? I can never correct both the frequency and the time domain responses? If this is the case, then audio DSP is not for me.

I generated .wav files for FIR convolution using REW following a tutorial I found on YouTube. I applied minimum phase to the generated traces, and they definitely changed my frequency response. The only issue is that the results were slightly off from my target curve. I need to go back and play with it some more to try to get results that are closer.

I also used REW to generate biquad filters. The results were OK, but I still added additional PEQ filters myself to get it better.

Nonetheless, configuring DSP takes time to learn, and sometimes it takes a bit of trial and error until you get it right. That may not be your cup of tea, and that's OK. An easier solution may be just to use Dirac Live or something equivalent.

 

[MaxwellsEq]

Is this the way audio DSP works? I can never correct both the frequency and the time domain responses? If this is the case, then audio DSP is not for me

It isn't the case, so you should not reject DSP on this basis.

In case you are unaware, frequency and time are transforms of eachother. Every device you use, even if it's perfect and flawless (analogue or digital) that changes frequency response, MUST have an impact in the time and hence phase domain. This is most logical justification for "hairshirt" HiFi that lacks tone controls.

In reality, there are no perfect and flawless frequency adjusting devices. In theory, digital EQ, performed in the mathematics domain, has fewer artefacts than an analogue equivalent.

 

[KEM]

It isn't the case, so you should not reject DSP on this basis.

In case you are unaware, frequency and time are transforms of eachother. Every device you use, even if it's perfect and flawless (analogue or digital) that changes frequency response, MUST have an impact in the time and hence phase domain. This is most logical justification for "hairshirt" HiFi that lacks tone controls.

In reality, there are no perfect and flawless frequency adjusting devices. In theory, digital EQ, performed in the mathematics domain, has fewer artefacts than an analogue equivalent.

Thanks for the input. I do understand transforms and convolutions. My minor was Math, especially those parts that pertain to Optical system.

Since I am very new to the dsp universe, I'm planning to obtain software that will determine the correct filters to apply. As I learn, I can begin to tweak.
Is there a product that will improve both FR and Phase (and impulse response).

 

[staticV3]

Is there a product that will improve both FR and Phase (and impulse response).

Lots.

Most prominently, probably Dirac Live.

 

[MaxwellsEq]

Thanks for the input. I do understand transforms and convolutions. My minor was Math, especially those parts that pertain to Optical system.

Since I am very new to the dsp universe, I'm planning to obtain software that will determine the correct filters to apply. As I learn, I can begin to tweak.
Is there a product that will improve both FR and Phase (and impulse response).

An excellent way to learn is with a computer, since you will have thousands of times more processing power than is needed for FIR with hundreds of thousands of taps, and there are many, brilliant pieces of software (some free) that allow you to safely experiment. It's a great way to make mistakes without committing yourself to hardware. Once you've identified what works for you, you can buy some hardware that looks more like normal HiFi and blends in

 

[Mikig]

I want to ask you all a question:

when you use PEQ, REW, DSP, Crossover, computer programs, subwoofers I see that some use them more than speakers, what do you expect?
Can someone explain to me "in a nutshell" what you hear differently from your speakers? What do you hear from your seat that you couldn't hear before?

because in the end that's what matters; an audible difference that can convince you to say... yes, this is the way!
because ultimately, all the efforts we make have a single purpose, the satisfaction of listening...

 

[MaxwellsEq]

I want to ask you all a question:

when you use PEQ, REW, DSP, Crossover, computer programs, subwoofers I see that some use them more than speakers, what do you expect?
Can someone explain to me "in a nutshell" what you hear differently from your speakers? What do you hear from your seat that you couldn't hear before?

because in the end that's what matters; an audible difference that can convince you to say... yes, this is the way!
because ultimately, all the efforts we make have a single purpose, the satisfaction of listening...

You hear your room+speakers more than you hear your room, especially at lower frequencies. Here's what an uncorrected room can look like:

So, you pay thousands for a DAC and amp which are +-0.01dB between 20Hz and 20kHz and then you are actually listening to the above room curve, not the amp, not the DAC, but the room. There can be 30dB differences across an octave. Even the weirdest DACs and strangest valve amplifiers (and amplifiers from the 1950s) measure MUCH better than this.

 

[KEM]

You hear your room+speakers more than you hear your room, especially at lower frequencies. Here's what an uncorrected room can look like:
View attachment 375255

So, you pay thousands for a DAC and amp which are +-0.01dB between 20Hz and 20kHz and then you are actually listening to the above room curve, not the amp, not the DAC, but the room. There can be 30dB differences across an octave. Even the weirdest DACs and strangest valve amplifiers (and amplifiers from the 1950s) measure MUCH better than this.

And so do systems where one has employed wall treatments (traps) and hardware equalizers (GEQ or PEQ) to reduce the variation.

 

[MaxwellsEq]

And so do systems where one has employed wall treatments (traps) and hardware equalizers (GEQ or PEQ) to reduce the variation.

What?

 

[KEM]

What?

What's your question?

 

[MAB]

Right now, I have a (mostly) all-analog system.

Only exception is the Oppo BDP-105 and Bryston BDA-3 for digital files and media. No streaming. I am very pleased with the sound.
My listening is divided about half and half between LPs and CD/SACD/Digital files with a bit of FM radio thrown in.
I have used REW to assist me in placing the speakers, the sub and sound traps (Real Traps products). For the final tweaks I gently used a graphic equalizer (dBx 1231).
I've managed to tame all but one room mode.
Frequency response is"flat" to about 2KHz where I let the FR drop faster than the suggested curve since I have aversion to high frequency sizzle.
Phase response is not minimum phase. There's many more 360 phase shifts than are predicted for a minimum phase system. She attached graph.

View attachment 374675

OK. Great. But now I'm wondering if the sound would be even more enjoyable than now using a DSP for room correction (keeping the traps) and for crossover. This would eliminate my trusty, great sounding ML#38s pre, dBx EQ, Sublime Acoustic K231 crossover (75Hz) and Bryston BDA-3. In place I would have the MiniDSP Flex Balanced or SHD. II guess I could just buy one and get the answer. But I'd like to hear about other journeys such as mine before I spend the money.

And is all that excess phase shift wreaking damage to my sound?

I am quite emotionally attached to all my components and need a significant amount of persuasion before I replace any component. I have tried a few other pre and power amps, but have always come running back because the sound didn't get better, got worse, or brought other unwelcome problems.

Don't be shy about being overly technical. I have an MS in Physics (Optics/Electronics) and have spent my entire career doing product development.

Thanks for your attention!

Rest of System:
1. Technics SL-1200G TT
2. Parasound phono pre
3. Bryston 2.5B^3 power amp
4. Revel f208 Floorstanding speakers
5. Hsu ULS-15 Sub.

Yes on the MiniDSP route.

I used to use various analog preamps (Sonic Frontiers, Bryston, Apt Holman, etc.) and a Bryston 10B electronic crossover for sub integration. Also Linkwitz active crossover and filter boards that I kit-bashed for multi-amped speaker projects.

I replaced the Linkwitz boards with a MiniDSP years back, it was awesome and liberating. I later replaced the 10B crossover and Sonic Frontiers preamp with a MiniDSP SHD in my main system. I appreciate your conundrum on your ML38s, hard to part with something like that. I sold the SF Line-3 preamp, never looked back. I'm sure someone is enjoying that dual-chassis beast, I am glad that it no longer worries me about when I need to replace an army of tubes.

I now listen to JBL M2 with MiniDSP Flex as preamp plus DSP filters and EQ. Or a Genelec system with sub integration using MiniDSP. I still have a couple of analog preamps in use, like the classy vintage system in the guest bedroom (with a WiiM for connectivity).

Regarding the sound, the MiniDSP gear is quieter than any of my analog gear. I have a turntable/phono preamp hooked up to the analog input, it sounds as good as a turntable can sound. The Sonic Frontiers preamp was about as transparent as any, and doesn't impart audible changes to the signal, I can't tell the difference. I still occasionally listen to tapes since I have old recordings I made, I use the Apt for that since it has tape loops. I am not worried about the sound in this case, if I was I might add PEQ using the processor loop. I love old preamps, but I also love old bicycles and push lawn mowers, but I don't plan on riding TdF or mowing the back forty with a museum piece!

My only regret is I want more DSP power, and I would like to see other companies compete with MiniDSP. It's likely that both of my wishes are somewhat coupled.

I would get a MiniDSP Flex Balanced, with Dirac if you want. Get a UMIK mic too. And install REW and VituxCAD.

 

[KEM]

Yes on the MiniDSP route.

I used to use various analog preamps (Sonic Frontiers, Bryston, Apt Holman, etc.) and a Bryston 10B electronic crossover for sub integration. Also Linkwitz active crossover and filter boards that I kit-bashed for multi-amped speaker projects.

I replaced the Linkwitz boards with a MiniDSP years back, it was awesome and liberating. I later replaced the 10B crossover and Sonic Frontiers preamp with a MiniDSP SHD in my main system. I appreciate your conundrum on your ML38s, hard to part with something like that. I sold the SF Line-3 preamp, never looked back. I'm sure someone is enjoying that dual-chassis beast, I am glad that it no longer worries me about when I need to replace an army of tubes.

I now listen to JBL M2 with MiniDSP Flex as preamp plus DSP filters and EQ. Or a Genelec system with sub integration using MiniDSP. I still have a couple of analog preamps in use, like the classy vintage system in the guest bedroom (with a WiiM for connectivity).

Regarding the sound, the MiniDSP gear is quieter than any of my analog gear. I have a turntable/phono preamp hooked up to the analog input, it sounds as good as a turntable can sound. The Sonic Frontiers preamp was about as transparent as any, and doesn't impart audible changes to the signal, I can't tell the difference. I still occasionally listen to tapes since I have old recordings I made, I use the Apt for that since it has tape loops. I am not worried about the sound in this case, if I was I might add PEQ using the processor loop. I love old preamps, but I also love old bicycles and push lawn mowers, but I don't plan on riding TdF or mowing the back forty with a museum piece!

My only regret is I want more DSP power, and I would like to see other companies compete with MiniDSP. It's likely that both of my wishes are somewhat coupled.

I would get a MiniDSP Flex Balanced, with Dirac if you want. Get a UMIK mic too. And install REW and VituxCAD.

Quite a good perspective. You seem to understand my issue.

Why did you choose the Flex balanced vs the SHD? I am leaning toward the SHD only because I think it has the features I need. But I'm not sure.

You mention your system is quieter. How does it sound? Many on this and on other fora who switched from all analog to digital complain about sterility of the sound. I will have to say that my CD/SACD player (Oppo BDP-105 and Bryston BDA-3) does different from my other sources. I don't seem to enjoy it as much. I can't put this difference into words, but it is truly real. Blandness comes the closest. Could be my digital components.

Is it possible that my other sources - tuner, reel to reel, and turntable will suffer the same fate if I go digital?

 

[MaxwellsEq]

What's your question?

Sorry, I didn't understand your reply. I should have been clearer!

Are you replying to my first statement: "So, you pay thousands for a DAC and amp which are +-0.01dB between 20Hz and 20kHz and then you are actually listening to the above room curve, not the amp, not the DAC, but the room. There can be 30dB differences across an octave. " with "And so do systems where one has employed wall treatments (traps) and hardware equalizers (GEQ or PEQ) to reduce the variation."

OR, are you replying to my second statement "Even the weirdest DACs and strangest valve amplifiers (and amplifiers from the 1950s) measure MUCH better than this". with "And so do systems where one has employed wall treatments (traps) and hardware equalizers (GEQ or PEQ) to reduce the variation."

If the latter, I agree, if the former, I'm confused!

 

[KEM]

Sorry, I didn't understand your reply. I should have been clearer!

Are you replying to my first statement: "So, you pay thousands for a DAC and amp which are +-0.01dB between 20Hz and 20kHz and then you are actually listening to the above room curve, not the amp, not the DAC, but the room. There can be 30dB differences across an octave. " with "And so do systems where one has employed wall treatments (traps) and hardware equalizers (GEQ or PEQ) to reduce the variation."

OR, are you replying to my second statement "Even the weirdest DACs and strangest valve amplifiers (and amplifiers from the 1950s) measure MUCH better than this". with "And so do systems where one has employed wall treatments (traps) and hardware equalizers (GEQ or PEQ) to reduce the variation."

If the latter, I agree, if the former, I'm confused!

My comment addressed both. But it was poorly worded. I have an all analog chain (except for CD player/DAC). I have used traps and GEQ to tame the room response. It does a pretty good job. But the time domain is a mess (was even messier without the traps and GEQ!).

I am concerned about the time domain. I would like to have a system with little to no excess phase (as defined by REW). My system now has tons of excess phase. One concern is whether this is something audible or just a "nice to have".

Even all the great comments I've received, it isn't clear to me whether Dirac or some other software will create filters that can provide reasonably flat frequency response and zero excess phase simultaneously. And sound that is indistinguishable from analog.

 

[terryforsythe]

Can someone explain to me "in a nutshell" what you hear differently from your speakers? What do you hear from your seat that you couldn't hear before?

Below are measurements for my KEF LS60 speakers at my listening position with and without DSP. The traces are 1/12 octave smoothed.

No DSP at listening my position:

Below is with DSP at my listening position - green line is my target curve to which I tuned:

 

[KEM]

Below are measurements for my KEF LS60 speakers at my listening position with and without DSP. The traces are 1/12 octave smoothed.

No DSP at listening my position:

View attachment 375395


Below is with DSP at my listening position - green line is my target curve to which I tuned:

View attachment 375396

Clearly you've flattened the FR curve. What's different about the sound? On paper (OK then, on screen) the result is excellent. Do you like it? Prefer it?