@phofman thanks for the reply and links . do you think the "master clock issue" would explain the loss of bass? and would a "reclocking" board put it right ?
@ranshdow i only have 1 rpi4 and 1 DAC so cannot test them against anything else ! i have tried probably 10 OS's in total, all have the same issue.
@tjrep i will have to look into the clock chips in the dac, it seems. cheers
@jean-paul "It would be honest to add up all cost of various tested HATs and all stuff and then see what that money could have bought." i sort of agree with you but, you can upgrade an rpi set easier than a stand alone dac, i'd bet!
@ranshdow i only have 1 rpi4 and 1 DAC so cannot test them against anything else ! i have tried probably 10 OS's in total, all have the same issue.
@tjrep i will have to look into the clock chips in the dac, it seems. cheers
@jean-paul "It would be honest to add up all cost of various tested HATs and all stuff and then see what that money could have bought." i sort of agree with you but, you can upgrade an rpi set easier than a stand alone dac, i'd bet!
@ jean-paul Oh, I agree totally that the Pi is a cheap singleboard pc based on a telephone/tablet SOC, but nevertheless you can do amazing things with it.
If you take into account how the Pi has been designed to be used and accept that here ARE limitations, you can still build pretty good systems with it.
My setup sounds amazing and can compete against way more expensive streamers, but I don't exepect it to go one-on-one against a 2, 3 or 4000 Euro streamer.
If you take into account how the Pi has been designed to be used and accept that here ARE limitations, you can still build pretty good systems with it.
My setup sounds amazing and can compete against way more expensive streamers, but I don't exepect it to go one-on-one against a 2, 3 or 4000 Euro streamer.
Tjrep, it can go one-on-one with cheaper hardware. Yes RPi can sound good but at a cost ... and it will still be bunch of PCB's with cables everywhere 
Never mind, what floats your boat isn't it?
As a critical DIYer I subject all hardware to critical viewing, so also the devices I like. Nothing is perfect (thankfully!).
Never mind, what floats your boat isn't it?As a critical DIYer I subject all hardware to critical viewing, so also the devices I like. Nothing is perfect (thankfully!).
Last edited:
No I do not think so. IMO if the loos of bass really occurs (measured, really confirmed?), there is a problem in the analog part. Low-capacity coupling capacitors, power issue, etc. No increased jitter (if it actually occurred) can attenuate low frequencies in the digital signal. It would take a digital filter (DSP) to do so.
IMO the reclocking board is a technical nonsense. Instead of slaving the I2S transmitter with precise clock placed close to the DAC, the I2S is left at master mode and a large FIFO is placed between the master I2S and another master clock by the DAC. Eventually every FIFO will under/overflow, no two independent clocks run at exactly same frequency. Professional deployments use adaptive reclocking for merging two clock domains, not dumb FIFOs.
If the I2S transmitter allows to be clocked externally (which is not always the case, e.g. in an SPDIF receiver), the only technically sound solution is using the single master clock by the DAC. Such arrangement is equivalent to PCI/PCIe or USB-asynchronous soundcards - the I2S is clocked by DAC clock, and data to the input buffer of the I2S transmitter are transferred as needed to fill the buffer appropriately (via DMA or USB, both controlled by the I2S side monitoring the buffer level).
There are RPi DACs with master clocks. Most allow to be powered by a separate power supply input, not from the 5V pins of RPi.
As of the noise radiated by the RPi, being induced in the DAC circuits - I have not seen any real measurements. For the peace of mind one can always use a fully-enclosed metal heatsink (e.g. Cooling Aluminium Alloy Case Passive Metal Heatsink For RaspberryPi 4 B Computer | eBay + e.g. 2x20 Socket Riser Header for Raspberry Pi HATs and Bonnets | The Pi Hut), effectively shielding the digital board.
If one wanted to galvanically isolate the digital signals from RPi, low-cost signal isolators can be used (and some DAC boards use them, typically inexpensive chips made by SiliconLabs). But that is often for breaking a ground loop and keeping the analog circuits floating. An RPi powered by a class II power source, networked via ethernet (galvanically isolated by design) or WIFI, and not connected via USB to any class I device will not introduce a ground loop.
RPi is a standard source of digital signals, like any other SBC. There are ARM64 chips/boards which allow multiple I2S outputs (8ch), offer master clock inputs with configurable internal dividers for the bitclock (which may simplify the design if the DAC chip itself does not offer the feature). But most audio projects (moode, volumio) do not support them fully, some have only outdated linux kernels.
Last edited:
Can I suggest that you download and try one of the 'audiophile' OS's, such as Moode, Volumio, PiCorePlayer etc?
For the sake of 30 minutes or so, a download and an SD card, this should help to assess whether your issue is the hardware or the OS. I run all Pi variants with a selection of DAC Hats, none of which display the issues you describe.
I would suggest Moode as the most configurable, to get the most out of your Pi/DAC combination.
IMO there is nothing "audiophile" on those audio distributions, apart of having all the audio softwares preconfigured and having a simple-to-use configuration GUIs etc. The drivers are identical with raspbian (or better the raspberrypi official kernel linked in one of the posts above). Anyway software (apart of DSP) cannot cause attenuation of low frequencies in the audio data stream.
@ga77a - did you measure LF attenuation with a sweep signal or are you reporting a perceived effect?
Which is why I wrote 'audiophile'; perhaps I should have written 'audio focused'
There is something amiss with the OPs setup, I was simply offering an alternative approach to try to identify the cause/source.
Trying a different distribution is OK, but honestly I do not think this is a software issue. The setup needs some diagnostics - if the RPi4 model has enough RAM to run REW, then e.g. using some reasonable USB soundcard for capture and measuring frequency response of the DAC, as already suggested here. If the DAC is found OK, then going step by step to the speakers - measuring FR behind the preamp, behind the power amp (using some power resistors and voltage divider). The very standard diagnostics.
software settings and software, OS, network, cables...
...everything can have quite an impact on the audio performance.
...and there's an explanation for everything.
And guys, don't let you intimidate by those ignorant guys, who scream show me the measurements. These guys are clueless! And these guys never showed a relevant single measurement proving their point! "relevant" that's the key here!
Audio Precision (the No1 reference measurement gear manufacturer ) has clearly stated that their measurement equipment won't be able to reflect the real sound experience on real world music signals. ( I wrote an article on my blog and you also find a related Audio Precision presentation on Youtube.) Side note: The guy from Audio Science Reviews was in that very same AP presentation on RMAF...
...and guess what!?!? He keeps spreading the "what I can't measure, you can't hear" message.
Where does this leave us? We - who realized the importance of listening tests - simply need to accept there are ignorant people out there.
OK. We need the standard measurements to start with. No question.
But.
We also need to trust our ears! If we can't measure it, there's a pretty good chance we can hear it. That's the storyline!
Real music signals and real world chains are very complex.
I know. Next comes the double blind A/B crap. Look. The audible changes are
usually pretty obvious.
Anekdote:
While prepping (audiophiling ) my Gustard A18 yesterday, I was also looking (had to) into the "audiophile fuse" subject once more.
Of course a snakeoil subject for our old-school snake oil screamer audio engineers. I am an engineer myself btw. For those who'd be interested I'd recommend to watch and listen to this video. (Note: Instead of paying $150+ for a audiophile fuse I might try a silver wire. )
Anyhow.
Now. In science you'd approach a subject like this this way: "What the heck can make such an obvious difference? ...let's find out!"
The snakeoil-screaming audio experts over here, would simply ignore it, and keep pushing their crap message "show me the measurements". This attitude seriously slows down developments. That's the problem here!
*********************************
And yes. I do also experience greater bass slam ( and more ) on optimized audio operating systems. I did it 10 years back and still do it.
Still do it!?!? This or that tweak even works on a pretty advanced Gustard A18 that I just have in front of me. According to ASR: "All standard specs are beyond audible threshold". Amen.
Enjoy.
...everything can have quite an impact on the audio performance.
...and there's an explanation for everything.
And guys, don't let you intimidate by those ignorant guys, who scream show me the measurements. These guys are clueless! And these guys never showed a relevant single measurement proving their point! "relevant" that's the key here!
Audio Precision (the No1 reference measurement gear manufacturer ) has clearly stated that their measurement equipment won't be able to reflect the real sound experience on real world music signals. ( I wrote an article on my blog and you also find a related Audio Precision presentation on Youtube.) Side note: The guy from Audio Science Reviews was in that very same AP presentation on RMAF...
...and guess what!?!? He keeps spreading the "what I can't measure, you can't hear" message.
Where does this leave us? We - who realized the importance of listening tests - simply need to accept there are ignorant people out there.
OK. We need the standard measurements to start with. No question.
But.
We also need to trust our ears! If we can't measure it, there's a pretty good chance we can hear it. That's the storyline!
Real music signals and real world chains are very complex.
I know. Next comes the double blind A/B crap. Look. The audible changes are
usually pretty obvious.
Anekdote:
While prepping (audiophiling
) my Gustard A18 yesterday, I was also looking (had to) into the "audiophile fuse" subject once more. Of course a snakeoil subject for our old-school snake oil screamer audio engineers. I am an engineer myself btw. For those who'd be interested I'd recommend to watch and listen to this video. (Note: Instead of paying $150+ for a audiophile fuse I might try a silver wire.
)Anyhow.
Now. In science you'd approach a subject like this this way: "What the heck can make such an obvious difference? ...let's find out!"
The snakeoil-screaming audio experts over here, would simply ignore it, and keep pushing their crap message "show me the measurements". This attitude seriously slows down developments. That's the problem here!
*********************************
And yes. I do also experience greater bass slam ( and more ) on optimized audio operating systems. I did it 10 years back and still do it.
Still do it!?!? This or that tweak even works on a pretty advanced Gustard A18 that I just have in front of me. According to ASR: "All standard specs are beyond audible threshold". Amen.
Enjoy.
@phofman & @abraxalito yes, sub-bass attenuation is measured and confirmed. i noticed it straight away when i hooked up the rpi4, the only thing that changed was the introduction of the rpi4 , i have also downloaded a subwoofer test mp3.
@yatsushiro yes i have tried many OS'S including picoreplayer
hopefully tomorrow i will ba able to power the rpi4 from a battery, and see if it makes any difference.
@yatsushiro yes i have tried many OS'S including picoreplayer
hopefully tomorrow i will ba able to power the rpi4 from a battery, and see if it makes any difference.
... hooked to the IQAudio DAC. So the change is RPi4 + your DAC. Or did you connect your DAC to another RPi (2, 3) before this test to confirm the DAC was OK?
