First of all, apologies if this has been already discussed. I have read many forums quite a lot about subject “Raspberry as HIFI/High-End audio player”.
My first impressions were that it can be, but then as I read more, there seems to be problems and hiccups along the way (clock&pops, crappy usb driver, clock problems -> jitter). This gives me impression that its far from High-End.
It seems Raspberry has at least one audio related problem: broken USB drivers causes pops and clicks in audio.
Also there is discussions people asking whether this and that DAC is good for Raspberry Pi. Also I haven’t found any discussion where people had actually A/B listened raspberry hifi setup vs Commecrial audio High-End player. You you know some, please point out.
So my questions is, Is it possible to achieve High-End, bit perfect performance with decent price rasperry pi with DAC and power supply (decent price is imho under 150€)?
AFAIK (one of the) best setup is RaspberryPi + Hifiberry i2s DAC. Both having their own decent power supply. Is this setup correct? Or does best setup require isoncronous DAC so that any clock problems are nothing to worry about?
Thank you for answers. With this thread I don’t want to argue something against Raspberry or Volumio. I am merely trying to find out some answers.
Hi Scorpius, have a look on the http://www.hifiwigwam.com/forum.php forum - there you’ll find plenty of audiophiles reaching new highs in performance with the Pi - many of them using exotic hifi / valve amps and speakers - several using Volumio too. The forum has a specific computer based audio section, as well as DIY audio sections. If you can’t find it easily then just do a search for IQaudIO on the forum’s home page.
Certainly the Pi’s i2s signals aren’t brilliantly clean and the Pi has a noisy power supply but audio output can be made extremely clean as can be seen below… (2nd and 3rd harmonics of the prime signal are 100db down!)
A consequent raspberry pi setup can be true high end without any doubt. I use the hifiberry digi with a good extra dac, have separated psu for each pi and digi. The dac is battery powered with automated charging. The platform is volumio. This combination performs near to an Ayon CD2 tube CD player that was highly recommended by high end press and priced around 4000€.
Further I run signal through a self built tube amp and the final stage are fullrange horn speakers delivering highest efficiency. 10 Tube amp watts are enough to punch your soul out of the body :mrgreen:
What is really important: If someone gets music files from a CD collection- like I do- extracting data has to be done very carefully!
No 24 times speed action, please, because this method kills sound quality- 4 times is recommended as a maximum- don’t ask me why but the difference is clearly noticeable!
Use Exact Audio Copy or similar ambitioned programs for audio extraction of your CD’s and make settings for highest quality.
Also important in this case is a good quality CD ripper like some devices from plextor- unfortunately there is no current production of high end cd rippers anymore. One more thing: WAV performs slightly better than FLAC! Beat me, but I’ve tested it a thousand times.
Yes, it can !
I run the Raspberry Pi as a high-end network player connected directly with Ethernet cable to the network switch.
I have tested different USB solution on the Laptop and I changed to the Raspbi with I2S connection.
Sounds incredible good. The AKOUO Reference DAC hardware is also just the finest.
I do not need a driver because the my AKOUO DAC has its own controller for input and volume.
I think the Raspberry with Volumio is the best way, to get music from the storage.
I ripped my CDs, play some Flacs, Hires and also recorded some vinyl with a good ADC to 24/96 wav.
Volumio plays all the files !!
There is no problems with Noise. The Rasbpi is powered with a normal mobile phone supply and there is no isolator on the I2S port. I was really surprised of that little noise and the low THD and I get the best performance out of my DAC that is made for absolute High End. (see akouo.at/dac.html)
A good power supply with a backup battery will be the next step.
The test signal (2 kHz WAV with-1dBFs) was played over the network.
Thank for this info. I need to read those posts in suggested forum. It seems PiDAC is awesome DAC for my purpose. Now two important questions pops into my mind:
Can I use paspbmc as a audio player in this setup on is volumio absolutely necessary for best sound? AFAIK xbmc will do bit perfect (does it?)so it should be fine? I love xbmc (I have movie theater also) and its iPhone app is the best for controlling it.
What affordable but good quality powers you recommend? I’mm willing to pay up to 50€ for good power. Does PiDAC need its own power supply or is one enough for both Raspberry and PiDAC?
Thanks metalrob for this. I am as well very interested. Could you tell us more? What is a “good extra dac” in your eyes? I don’t want any further volume or mixing stages, just the digital-analog conversion for to get the signal into my amp.
Morover: How did you realise the charging process for the battery? I’ve also thought about using a usb-battery as they are quite cheap as backup power sources for phones, but I was wondering if I can leave the charger connected… would result in noise from the switching right?
Would be very interested tp learn more. I’ve started another thread out of a similar motivation some time ago, but it somehow faded away. sabre-es9023-why-does-sound-bad-t950.html
Hi,
i absolutely agree that a perfect power supply concept is a basic requirement for good sound representation. If the power concept is not proof it makes no sense to play around with different chipsets. That was one of the reasons I decided to order the AKOUO dac spiri already mentioned a few posts ago. In my opinion they implemented a perfect analog design and power supply concept without compromises. The other reason was the used Sabre 9018 chip which is said to be one of the best - but unfortunately not one of the cheapest But I wanted the best as it will probably be the last DAC I build up for the next xx years. I pre-ordered the DAC last December and had to wait some weeks - since then played around with my CD-player. But a few weeks ago I decided to order a Raspberry Pi and tried to connect it to the DAC. Unfortunately I am not so familiar with software development and so I was afraid to stumble across driver installation. But since the AKOUO DAC has it’s own controller which is responsible for initialisation it was not necessary to install any additional peace of software. I enabled the I2S support of VOLUMIO and that was all - wow absolutely perfect sound and no software hassle!!!
The next step will be to find a UPS for the Raspberry Pi so that it will not crash the filesystem when not shut down accordingly. But that is another thing
A perfect power supply IS NOT a requirement as important as you think.
The DAC rejects noise extremely well because it has an op-amp inside and this works at least 10x faster then the power supply noise: the op-amp can reject noise up to 10 MHz, but the noise you have after the power supply is almost inexistent with frequencies higher than 20-80 kHz. This means that for the op-amp in the DAC the power supply noise is like a turtle or snail: very slow and easy to compensate.
There was a website that was showing how clean different power supply are (because they are selling “super clean ones”), so I asked them to use the same tools they had to check how much ACTUALLY is the quality of the sound (distortion and such things) affected by the noise and ripple of the power supply.
The answer was: with the DAC we tested, very very low. They didn’t answer so clear and straight, since they selling PSU, but they nevertheless stated that the effect is negligible.
So don’t waste money, just buy a PSU with 2x the required current (to be sure it is able to cover spikes, but not more, or you have other issues) and you are fine. Spend the money in things that improve the sound FOR REAL. Like room insulation or damping of reflections, these improve the sound SO MUCH (just try!) that no PSU can ever offer such similar improvement.
I’m currently building up a new power supply- a box with two completely separated 5V rails.
After this is finished I will do some tests and then I will report the configuration of my setup.
Also I will see if lag-na’s theory is correct. Testing is believing
At the moment I’m using two 5V battery power banks for digi and raspi. But this akku packs are known
for producing noise through regulation circuits- therefore the efforts in developing a better PSU.
The DAC I use is a RME fireface, powered by a smaller 12 V Car battery. The fireface accepts 9 to 24V.
A mytek 192 khz DSD DAC for example will raise the sound bar higher.
But also the good DAC Cards discussed in the threads here deliver excellent sound quality if they are properly implemented.
For charging I use a C-Tek Charger that is constantly hanging on the battery. A timer starts charging process from 2 to 6 am automatically.
But I can also activate and deactivate charging per pressing button at any time. That’s comfortable.
In the near future I will test a low power consumption SSD- USB- connected to the pi.
Currently A self powered HDD delivers file data. But this devices are known for producing noise either
through motor drive regulation circuits. We will see…
My aim: Optimization of even small details should bring clearly noticeable improvements in performance as a whole effect.
I totally agree. Room effects are by far the most noticeable. If room and speakers do not harmonize with each other a million dollar sound equipment will fail with respect to sound performance.
But if someone has optimized this point, further improvements can be done- even if it doesn’t have that big effect.
In the end it should be fun what we do
Just to be sure about the testing (thanks for doing it!):
place a mic in the room, about 50-100 cm (2-3 feet) in front of one speaker
connect PSU #1 and play a 1 kHz constant tone (you need a wav file that lasts a couple of minutes) ONLY through the speaker in front of the mic
measure with tools (for example roomeqwizard.com, it is should be enough) the distortion of the sound recorded by the mic
without touching speakers or mic, replace power supply
repeat measurement.
At this point it doesn’t matter if the mic is perfect or not, what you are looking for is the difference between the two measurements. If there are none, you are done.
Please record the sound from the mic and take pictures of everything, so that we can also publish the results (with your name of course) in my website, if it is ok with you.
As PSU, I suggest to use at least a normal phone-chareger rated at 5V, 1.2-1.5A (they are switching power supplies), better test two different brands (if you have an Apple original charger and an adapter DO test it, it is a very good PSU), then you can test the 5V battery pack (that is always very clean) and then your new PSU.
I will be looking forward the results! please send me a private message if I miss the thread (better open a new thread).
[quote=“Lag-na”]
A perfect power supply IS NOT a requirement as important as you think.
All manufacturers of real audio equipment have been wrong so far ??
I thought, we are talking in this thread about a raspberry high end player and not about the next replacement for a laptop output,
with 50 cent DACs, rejecting power supply noise, but don’t playing music.
[quote="
At this point it doesn’t matter if the mic is perfect or not, what you are looking for is the difference between the two measurements. If there are none, you are done.[/quote]
If you measure a not perfect mic, you will get the measurements of a not perfect mic. One with a good psu and one with a bad psu. Then we can compare a not perfect mic with a not perfect mic.
What will we see ?
A not perfect Mic.
We are not listening sine waves through microphones.
I think, that measurements are important as a technical orientation and as an insurance, that nothing has been messed up
during the development of an electronic component.
But after my experience measurements doesn’t always correlate with the facts I hear. I know that discussing this with technocrats
doesn’t make much sense. But in the end- and I’m absolutely convinced of this fact- I can trust my ears.
It happened uncountable times, that I heard a clear difference in the sound characteristics of two devices that showed (nearly) identical test protocols.
Noone would say, that tube amps deliver better results in the lab than transistorized amps.
Nevertheless I would NEVER prefer a transistor amp to a well built tube amp in the matter of sound characteristics.
This is subjective.
“High end” does not necessarily equate to high fidelity.
If fact many high end products sound pleasing BECAUSE they colour the sound, valve amps being a good example.
I think it’s best to strive towards the highest fidelity you can achieve and then optionally look at colouring the sound if that suits your taste.
These days DSPs are sophisticated enough to add these tonal colours and that’s why recording artists and studios use them all the time.
Also you don’t hear facts, your ears/brain profoundly affect what you “hear”.
Something that sounds awesome may sound dull on the next listen. Mood, emotion, hunger, thirst and many other factor affect how pleasing something sounds.
WAV sounding better than FLAC? The only possible explanation is a malfunction.
The funniest feedback I heard was that when one dealer preferred the audio quality of FLAC media played from his NAS using WD hard disks over Hitachi hard disks!
I Agree, but when I’m talking of High End, I mean highest fidelity.
No coloured sound, absolute neutrality and clear. And I can here this in every listening session.
If some audio device sounds not neutral, you don’t get the best reproduction of the music, made by the musicians and studio technician to which we have no influence. When the reproduction is very neutral and you can count on it, you can accept the sound like it is.
If the sound is coloured, warm and smooth, then your mood will be a sound factor and when you are hungry, don’t listen.
FLAC is a ‘lossless compression’. Can this be ?
Maybe WAV is the only way to get real audio performance, without special codec. I like 48kHz/24bit WAV and many good records are made with this format. I mean uncompressed PCM Data, because WAV is only the container, and can be also a compressed audio stream.
I too like neutral sound the most. Then every record won’t sound the same.
Lossless compression is like a ZIP-file. If you zip a word document and then unzip it, all the letters are still there.
It works like this:
original file: 10000000000000001111111111
compressed file: 1+150+101
The shortest line takes the least storage space on your hard drive.
Hardware requirements:
The first line requires more storage, system memory and network bandwith.
The second line takes a little bit of CPU to put together.
Compare that do destructive compression (mp3) which uses both the above techniques, as well as removing information that is the least useful to the listener in order to save space. It will remove subtle details. The higher compression, the more details will be removed. This is only used when processing media (music, video, images). A word document will never be compressed destructively because it will render text useless.
But I wanted the best as it will probably be the last DAC I build up for the next xx years. I pre-ordered the DAC last December and had to wait some weeks - since then played around with my CD-player. But a few weeks ago I decided to order a Raspberry Pi and tried to connect it to the DAC. Unfortunately I am not so familiar with software development and so I was afraid to stumble across driver installation. But since the AKOUO DAC has it’s own controller which is responsible for initialisation it was not necessary to install any additional peace of software. I enabled the I2S support of VOLUMIO and that was all - wow absolutely perfect sound and no software hassle!!!
