I know, it's a long text, but it's worth reading! Promised!
I have the idea to build a digital audio chain. I'd like to invite you to comment, discuss, or just like or dislike my idea. I'm interested in your opinion.
The digital audio chain should start at the sound source and end at the speaker. That sounds weird at first glance, but give me a chance to describe my idea in detail:
As a loudspeaker I plan to use active loudspeakers with a digital input. There are already many good active loudspeakers, such as from the manufacturers Nubert, Genelec or Canton. Of course there are many more good manufacturers. I think active speakers are very interesting. In contrast to conventional loudspeakers, active frequency cross-overs can be used in the active loudspeaker, which avoids phase shifts. Also, one amplifier per speaker can be realized in the box and perfectly tuned, which also has a positive effect on the sound.
The sound source can be e.g. a BluRay player or another device. Of course I can connect the sound source directly to the active box. But this has some disadvantages:
I want to realize the DSP with a Raspberry Pi (RPi). Because the RPi has no interfaces for sound output, I need a sound card. There is no sound card that has multiple switchable digital SPDIF inputs and at least 4 SPDIF outputs, therefore I plan to build such a soundcard myself. The sound card in conjunction with the RPi and a piece of software realizes exactly the characteristics described above.
What I love about this idea is the great flexibility. If I want, I can install e.g. Kodi or VolumIO on the RPi and then have a mediacenter. I can e.g. connect a blutooth dongle to stream music from my phone. Likewise, I could also implement multi-room PA with the system or integrate voice control. There are so many possibilities ...
Why make everything digital? Quite simply, it's cheaper and easier. When I realize everything digitally, I do not need expensive accessories or expensive cables to optimize the sound. I no longer have to decide if I want a good stereo or surround system. I can have both without restrictions. There are no loss of sound because the signal is transported lossless.
Well, enough written. Thank you for the reading! I'm looking forward for comments!
I have the idea to build a digital audio chain. I'd like to invite you to comment, discuss, or just like or dislike my idea. I'm interested in your opinion.
The digital audio chain should start at the sound source and end at the speaker. That sounds weird at first glance, but give me a chance to describe my idea in detail:
As a loudspeaker I plan to use active loudspeakers with a digital input. There are already many good active loudspeakers, such as from the manufacturers Nubert, Genelec or Canton. Of course there are many more good manufacturers. I think active speakers are very interesting. In contrast to conventional loudspeakers, active frequency cross-overs can be used in the active loudspeaker, which avoids phase shifts. Also, one amplifier per speaker can be realized in the box and perfectly tuned, which also has a positive effect on the sound.
The sound source can be e.g. a BluRay player or another device. Of course I can connect the sound source directly to the active box. But this has some disadvantages:
- Typically you can only connect a digital sound source to the active box.
- The active box can transmit surround signals such as decode and play DTS or Dolby Digital.
- Active speakers can only be used for stereo output.
- The DSP can switch between multiple digital sound sources.
- The device can decode digital surround signals, such as Dolby Digital or DTS.
- The DSP centrally controls the volume.
- The unit outputs the (surround) sound digitally. On the DSP you can then connect several active speakers via SPDIF. The first SPDIF port connects the channels FL and FR, the second the channels RL and RR, etc.
I want to realize the DSP with a Raspberry Pi (RPi). Because the RPi has no interfaces for sound output, I need a sound card. There is no sound card that has multiple switchable digital SPDIF inputs and at least 4 SPDIF outputs, therefore I plan to build such a soundcard myself. The sound card in conjunction with the RPi and a piece of software realizes exactly the characteristics described above.
What I love about this idea is the great flexibility. If I want, I can install e.g. Kodi or VolumIO on the RPi and then have a mediacenter. I can e.g. connect a blutooth dongle to stream music from my phone. Likewise, I could also implement multi-room PA with the system or integrate voice control. There are so many possibilities ...
Why make everything digital? Quite simply, it's cheaper and easier. When I realize everything digitally, I do not need expensive accessories or expensive cables to optimize the sound. I no longer have to decide if I want a good stereo or surround system. I can have both without restrictions. There are no loss of sound because the signal is transported lossless.
Well, enough written. Thank you for the reading! I'm looking forward for comments!
And you'll be routing Spdif cables all over the room (and in multiple rooms if you want multi-room)? Also, you will need to design a new card when Dolby Atmos becomes more mainstream and you need more channels?
Take a look at WISA and then traction it has been gaining - your new TV would be able to drive all your speakers and you can stream from any device using a WISA dongle.
Or, if you prefer something more mature, look at AES67, there are endpoint devices that cost less than $100 that you can connect to each speaker and stream to from any platform and you can scale as big as you like.
The RPI is a very nice toy but it is noisy (even for spdif a good stereo hat costs quite a bit) and the latest iteration is very power hungry. It also does not have the horsepower to do good room correction which most active speakers do not handle.
Take a look at WISA and then traction it has been gaining - your new TV would be able to drive all your speakers and you can stream from any device using a WISA dongle.
Or, if you prefer something more mature, look at AES67, there are endpoint devices that cost less than $100 that you can connect to each speaker and stream to from any platform and you can scale as big as you like.
The RPI is a very nice toy but it is noisy (even for spdif a good stereo hat costs quite a bit) and the latest iteration is very power hungry. It also does not have the horsepower to do good room correction which most active speakers do not handle.
Wealas, thanks for your respond and time. You have pointed out some very interesting aspects.
BTW, I am also interested in opinions on the "digital sound chain" itself.
At the first, I want to replace my current 5.1 system with a digital system. I don't want to install more speakers, because it is my living room. I guess, it is a common setup for many other living rooms, too. I want to wire it with SPDIF cables, that are thinner than conventional speaker cables. But you are right, there are other ways to transport the signal, like WiSA, you have proposed. I will take a look at it, because the technical specs are interesting: It supports 8 channels at max. 96kHz and 24bit. It has a lower lower latency and bigger bandwidth than Blutooth with SBC or aptX. I plan to realize the multi-room solution in a later step. Multi-room must be realized wireless like Sonos. Everything else ist not feasible. But till now multi-room is just a later stage of expansion.
Yes, WiSA will become an important standard. I will observe it. I intend to build a open source based system, that everyone can customize to his/her personal preferences. I think, it should be possible to integrate WiSA in the system.
Yes, I will need a design enhancement, if I need more channels. But currently, 5.1 is sufficient for me and I guess, max 7.1 for many others, too. But I will design the soundcard that way, that it is easy to add more channels. If there is a need for more channels, I'll do it.
I didn't know AES67 yet, but only AVB. It may be an alternative solution, but I have no experience with it.
Do you mean the fan, that cools down the RPi?
Yes, it may be. I have to test it out. But if the RPi is not powerful enough, fortunately, I can switch to a different hardware because of the usb interface. And room correction is a further cool idea!
BTW, I am also interested in opinions on the "digital sound chain" itself.
Rpi is noisy in a sense that the on-board power converters create a lot of interference that affects the i2s audio feed, but it won't matter as much if you use usb to your device.
I personally prefer many smaller devices compared to a central one. I want to send the same signal (any signal, maybe with the exception of Dolby types that Kodi can decode for me to PCM) to all speakers via a common network (wired or wireless) and have each speaker apply channel selection, room correction and active crossover and feed the signal to the amp. That way I can send the signal from any device on the network and have any device be the controller (for input and zone selection or volume). I won't need Bluetooth or dongles or other cables because all of my devices are already on my network.
The advantage of such approach is that I can have any number of speakers and I just have to configure them. Also, no single point of failure besides the network.
I personally prefer many smaller devices compared to a central one. I want to send the same signal (any signal, maybe with the exception of Dolby types that Kodi can decode for me to PCM) to all speakers via a common network (wired or wireless) and have each speaker apply channel selection, room correction and active crossover and feed the signal to the amp. That way I can send the signal from any device on the network and have any device be the controller (for input and zone selection or volume). I won't need Bluetooth or dongles or other cables because all of my devices are already on my network.
The advantage of such approach is that I can have any number of speakers and I just have to configure them. Also, no single point of failure besides the network.
Hi wealas,
Which Raspberry Pi are you talking about? There are lots of versions of the RPi and at least 3 different voltage supply schemes.
regards
Which Raspberry Pi are you talking about? There are lots of versions of the RPi and at least 3 different voltage supply schemes.
regards
Hey Greg, I don't think anybody has bothered to check the latest RPI versions, the one Dimdim tested here: The Raspberry Pi: Audio out through I2S | Dimdim's Blog seems to be the RPI 1 model B+.
However, I can't see them improving the situation without some serious price hike which has not happened in the later models. My very basic understanding is that suppressing high frequency interference on such a small board would require a serious number of PCB layers and additional filters which I wouldn't expect from a $30 board.
However, I can't see them improving the situation without some serious price hike which has not happened in the later models. My very basic understanding is that suppressing high frequency interference on such a small board would require a serious number of PCB layers and additional filters which I wouldn't expect from a $30 board.
I use JRiver Software in my PC as a digital preamp and crossover to feed Onkyo receiver via HDMI. This is stereo triamped signal utilizing 6 channels of the 7.1 receiver. I am able to modify each channel to whatever my heart desires. I can tailor the signals to the tweeters midrange and woofers while keeping the cost under control. PC is $650, JRiver is $28, used Onkyo $120, speakers are DIY at $400 total. I use Spotify but other uncompressed signals may be more desirable for younger ears. No CDs to buy or keep track of. Any album I have ever owned, heard of, read reviews of, lusted after is available. Really really cool.
Hello
Are you sure the speakers are an option now? Please read this article Digital speaker - Wikipedia and let me know, what your opinion. I am eager to learn, how the speakers overcome the problems (described in the article).
If these speakers are not true digital speakers, then I think better make a analog conversion somewhere in the signal chain.
Are you sure the speakers are an option now? Please read this article Digital speaker - Wikipedia and let me know, what your opinion. I am eager to learn, how the speakers overcome the problems (described in the article).
If these speakers are not true digital speakers, then I think better make a analog conversion somewhere in the signal chain.
I confirm with wealas, that the RPi should be the last model. The problem is, that I need USB 2 in the "high speed" mode with 480 Mbit to provide enough bandwidth for 8 channels. Unfortunally, USB is a very inefficient protocol and requires many cpu cycles to process. I have already test it successfully with a Rockchip RK3328 (rock64). This cpu has enough power. I hope, the RPi is also powerful and "clean" enough. Otherwise I choose a different board.
The RPi with $30 is really very cheap, although they produce the board in th UK. I guess, they can make this price, because they produce over 10,000 units per week.
@Atteju So far I have understood the wiki article you have linked, the described speaker type addresses small devices like smart phones, that are not suitable for hifi. The kind of speakers I referred is: Powered speakers - Wikipedia. I want to use powered speakers with a digital input. Some powered speakers are also digital amplified. I see many advantages in using powered speakers, particularly with a complete digital audio chain.
The RPi with $30 is really very cheap, although they produce the board in th UK. I guess, they can make this price, because they produce over 10,000 units per week.
@Atteju So far I have understood the wiki article you have linked, the described speaker type addresses small devices like smart phones, that are not suitable for hifi. The kind of speakers I referred is: Powered speakers - Wikipedia. I want to use powered speakers with a digital input. Some powered speakers are also digital amplified. I see many advantages in using powered speakers, particularly with a complete digital audio chain.
- Status
- Not open for further replies.