I love music, but don’t know what I’m doing with adjusting DSP settings or how to get the most out of my setup. Does anyone have any tips beyond play with the knobs until it sounds right? FusionDSP has a great guide on how to change settings, but not anything about what I should be aiming for.
I’ve got the FusionDSP installed, and I’m running 3.429 Volumio.
Pi4 → cheap 24b USB DAC* → 3.5mm cable → Amazon Echo Link Amp → Big cabinet speakers from the 90’s (MTX 250W 40Hz-22kHz, 3-way and largest speaker is 15")
'* considering upgrading this, since it’s probably the weakest link. but afraid since I don’t know what I’m doing I’ll spend money and not notice the difference. Might just go with still cheap higher bit DAC.
If you’re using DSP in your configuration, there are 2 ways:
- Measure the audio response of your room and compensate/correct using DSP
- Move knobs until you’ll find a setting that you prefer
For the first there is a pretty cheap solution posted on this forum: Room correction using your iPhone and HouseCurve
A DAC rarely makes a night and day difference, unless the DAC used is not properly engeneered, but you’ll notice about that soon enough.
Personally, I would have a really hard time telling a high-end USB DAC apart from the small GPIO card that sits on my pi4
As a common rule, you can stick to this one :
- leave everything that is digital on it’s own ;
- improve the “analog” part of the system.
The most noticeable difference will probably be the speakers.
If you don’t want to change those, try a better amplifier. I have no idea what the Amazon Echo Link Amp is, but I suspect it is not super high end stuff
Apart from that, it is difficult to say what to improve. Why ? Because we may be able to tell “This device is transparent and so the output is exactly the input”…
But I honestly admit that I have nothing against a slight “coloration” of sound.
So listen to Wheaten : move the knobs until you get something you like
If you like it better, it is better.
Even if not close to source
Eq is not something so easy if you want a good result…
Search the internet and you’ll find plenty of great guide.
Basically, a @Wheaten wrote, you need to measure your room FR, generate filters and use them in DSP.
Easy with tools such as HouseCurve, or HiFiScan.
More technical with REW but very powerful!
This is just few example though!
- try to get left and right speaker frequerency response match each other (this is more important for stereo imaging than having “perfect response”)
- make sure both speakers are equally loud to listening position.
- Time align, only needed if left and right speakers are not at equal distance to listening spot
at this point the stereo imaging should be rather good.
- From now on we do what @Wheaten suggested (remember to EQ both speakers equally from this point onwards, to not cause phase issues and bad stereo imaging)
Graphic EQ is quite easy, but the result may vary.
You may have good results with one setting while listening to a specific track, and horrible results with other tracks when using the same setting.
Parametric EQ is much more powerful, but I honestly admit that I search for settings for my end device (headphones) because without it I would be totally unable to tell what I am doing… I suspect some stuff, but would have hard time to explain what it does exactly
BUT : as a last example to date, I recently bought a Denon AH-D7200 headphone. Out of the box : too heavy on bass and so mids and trebles where totally “eaten up” by the bass.
They were splendid when there was no bass (i.e. : playing a piano concerto). But when bass came in, mids and treble almost not to be heard.
So I took some Graphic EQ and lowered the bass a little => splendid headphone
I also found PEQ settings and Volumio allows to apply the directly from AutoEQ. Good result.
But as I said somewhere else, for now I do not use Volumio anymore because I ran into to many problems with release 3 Hope those will be fixed soon with the info other people and I provided to Volumio team
Thanks everyone. Here’s what I did:
- Bought the HouseCurve app, used it.
- Tested after implementing in FusionDSP. Adjusted a few values manually and nicely matched the ideal curve (at least within a reasonably range).
- Took a detour - replaced my cheap USB DAC with a USB–>optical device, to just use my amp’s DAC. This resulted in occasional missing bits of sound. I don’t know if this is a problem from the pi’s end, a bad optical cable, or what, but this wasn’t acceptable.
- Swapped back in my cheap DAC (thumb drive sized).
Trying different types of songs, it sounds great. I don’t hear much of a difference from before I changed settings (turning on and off the DSP), but now I’m much more confident I’m not missing anything in the listening experience.
If you want to optimize the sounds from the components you have, the advice so far, and what you’ve done are great. But a lot more can be achieved by changing components. For myself, I want the experience of listening to a private concert, front-row-center. But that’s a big change, and it depends on how much you really like music . . . and are able to hear it. When I was in grad school, I heard a fairly high-end system, and I made a commitment to myself to some day own one. About 10 years later, I was able to put one together. Many years later, I have put together a digital-based high-end system. Volumio is a critical foundation of it, but all of the components had to upgraded. If you’re interested, I could tell you more about it. But if you like what you have, stick with it.
Yes please. I’d love to hear how to go further, especially if any component could make a big change.
the biggest advantage you can make, is in the analog part.
- Speaker cables
But then I would advise to go to a local reseller and have them demonstrate it. Never go blind on other people verdicts or reviews. It’s your ears, your sound experience, that determine was sounds ok. No one else.
Certainly, the “speakers” are the most obvious place to start. This does get quite complicated, and the first thing you need to consider is the combination of how good your hearing is, and how important accuracy is. As Wheaten wrote, this is all about what you experience and what you want. If you want to test your hearing, you should do some A-B comparisons at a high-end stereo shop . . . but make sure you are listening to electrostatic drivers, either speakers or headphones. I have the Koss ESP/950 electrostatic headphones. They provide more detail and more accurately than any magnetic drivers (headphones or speakers) are capable of. The reason is the response time. Electrostatic drivers have the fastest response time of any audio drivers. I suggest you listen to the same music with different sample rates . . . and to a record on a high-end turntable. With the best stylus, a record can reproduce up to about 1MHz frequencies. The electronics are very important, so I hope the stereo shop has Spectral pre-amps and power amps. What you want to test your hearing is accuracy of the sound reproduction.
“Science” declared that no human can hear frequencies above 20KHz. That’s true for a clear tone, but it’s BS as relates to high frequency harmonics. I, for example, cannot hear a clear tone of even 10KHz, but I can hear missing high frequency harmonics a LOT higher. By comparing the same music with different sample rates, I discovered that (with my Koss headphones), I can easily hear the difference between samples rates of 384KHz and 768KHz. Most people in the industry will tell you that’s impossible, but most audiophiles can hear missing high frequency harmonics. (To me, it sounds like there is ceiling to the frequencies, or like there is a blanket between me and the music.) The “scientists” somehow ignore that. Anyway, you need to find out what YOU can hear, and what is important to you. This is getting long, so I’ll make another response.
If you find that you are fairly normal and can’t hear (or don’t care about) the high frequency harmonics, it will make your search for sound much simpler. If you find that you are an audiophile, and can hear and appreciate the high frequency harmonics, you have to go against the herd. That’s what I fought throughout 2022 to achieve high quality music from digital sources. Records can be the best, but only if they were pure-analog mastered. Digitally mastered records are . . . digital. I have about 800 records and a few hundred CD’s. I treasure my analog records. I found that I could digitize my records played on a Sota turntable (like on the Doctor House series) with a Spectral reference cartridge, through a Spectral pre-amp (flat response from 0Hz-1MHz) and an RME ADI-2 Pro converter. (And, by the way, if you ask RME, they will tell you that no one can hear the difference.) I use the DSD256 digital format. The files are a little smaller than PCM 768KHz/32bit, and this sounds more accurate to me than PCM 384KHz/32bit. If you come to the same conclusion with your hearing, you can find a limited selection of DSD256 music on NativeDSD.com.
If you don’t have a problem with headphones, you can get electrostatic headphones much cheaper than electrostatic speakers. In the 80’s, I had electrostatic speakers: Magnepan MG III’s. I lost those, but in 2022, I bought a Magnepan Mini Maggie speaker system. I highly recommend them. They rival the MG III’s at a fraction of the cost of floor-standing electrostatic speakers. The Koss electrostatic headphones currently sell for about $1000, if you can find them. The best electrostatic headphones are some of the Stax, but you have to choose from a line of headphones. I think the SR-009S model is what I would get, but again, do some research, and if you can, listen to them before you buy. I’m getting pressure to cut down on my responses here, so maybe we should take this up in email . . . unless there are others who would like to participate. The speakers are only the first, and without other appropriate components, great speakers can be disappointing. Let us know what you find about your hearing, and what you care about, e.g. accuracy, warmth, file size, convenience, etc.