A Raspberry Pi as a music server

[threni]

I didn't know about this. Would it block ads on an Android TV YouTube app?

I don't think so. It's basically acting as a local DNS server with a blacklist.

 

[somebodyelse]

Thanks! Looks nice and cheaper than Hifiberry. On a second thought though, a USB connected converter frees up the the GPIO which I might actually try to use for other purposes. I wanted to try output trigger 12V signal to automatically power on the amplifier. So many options!

Most of the GPIO is still available on the unpopulated 40 pin header. The docs will tell you which pins are in use, and you can do what you want with the rest.

 

[Raitsa]

I didn't know about this. Would it block ads on an Android TV YouTube app?

Use this

GitHub - yuliskov/SmartTube: SmartTube - an advanced player for set-top boxes and tvs running Android OS

SmartTube - an advanced player for set-top boxes and tvs running Android OS - yuliskov/SmartTube

github.com

 

[glorious_iguana]

Use this

GitHub - yuliskov/SmartTube: SmartTube - an advanced player for set-top boxes and tvs running Android OS

SmartTube - an advanced player for set-top boxes and tvs running Android OS - yuliskov/SmartTube

github.com

I'm using it already, it's great, but the only downside you cannot cast videos from smartphone to the TV via this app. I was wondering if it's possible to remove ads from the official app.

 

[glorious_iguana]

I also came across the line of products by IanCanada centred around the hats that reduce time jitter such as FIFOPI Q2. Is jitter a really significant problem to worry about when using USB transports/DACs? Has someone been able to measure the effects? There is a whole thread on DIYAudio forum discussing I2S FIFO modules, but I don't feel like going through it all. Ian Canada also has a ReceiverPI hat (SPDIF input) TransportPi (SPDIF output). I'm trying to figure out whether this is snake oil or a worthy investment.

 

[tw99]

I also came across the line of products by IanCanada centred around the hats that reduce time jitter such as FIFOPI Q2. Is jitter a really significant problem to worry about when using USB transports/DACs? Has someone been able to measure the effects? There is a whole thread on DIYAudio forum discussing I2S FIFO modules, but I don't feel like going through it all. Ian Canada also has a ReceiverPI hat (SPDIF input) TransportPi (SPDIF output). I'm trying to figure out whether this is snake oil or a worthy investment.

IanCanada has a lot of products that are solving problems that don't exist. It's great for people who like to build stuff that looks complicated, but in reality none of it is needed. USB out from a Pi4 into a good DAC will be as good as you can get.

 

[charleski]

IanCanada has a lot of products that are solving problems that don't exist. It's great for people who like to build stuff that looks complicated, but in reality none of it is needed. USB out from a Pi4 into a good DAC will be as good as you can get.

Archimago did some measurements of jitter from a Pi 3's USB several years ago. I imagine that the output from the Pi4 will be the same or better.

MEASUREMENTS: Raspberry Pi 3 as USB Audio Streamer (with recommended CRAAP config & TIDAL/MQA arrives)

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

Of course its the DAC that really counts, but implementing a proper digital interface is not nearly as hard as some want to pretend.

 

[benanders]

Archimago did some measurements of jitter from a Pi 3's USB several years ago. I imagine that the output from the Pi4 will be the same or better.

MEASUREMENTS: Raspberry Pi 3 as USB Audio Streamer (with recommended CRAAP config & TIDAL/MQA arrives)

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

Of course its the DAC that really counts, but implementing a proper digital interface is not nearly as hard as some want to pretend.

Yes, good link. And another, in case one desired to condense all into one tiny chassis:

MEASUREMENTS: Raspberry Pi 3 & HiFiBerry DAC+ Pro (Audio Streaming)

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

The only reason I didn’t go the DAC-hat route was because my very large (by today’s standards) DAC is still perfectly fine, only its [obsolete] integrated transport x mobile app needed replacing.

USB implementation is purportedly inferior on the Pi 3 vs. 4, but this has not been tested for audibility/functionality in music files; the Pi 3 will consume less power in operation than the 4, a nice detail for any 24/7 device.

 

[valve_integrator]

IanCanada has a lot of products that are solving problems that don't exist. It's great for people who like to build stuff that looks complicated, but in reality none of it is needed. USB out from a Pi4 into a good DAC will be as good as you can get.

sorry that's total nonsense.
I think you never used one of Ian's DACs so you write stuff you know nothing about.

I have been using them on a CM4 for over a year. The DAC controller allows you to use a variety of output filtering, and the output is balanced XLR via a nice Opamp board which is upgradeable. (plug in 8 pin opamps).

The optional transformer output board using transformers from Biskek I have to admit to not trying yet (although I made up the board ready). You can use top quality OPT for it from others eg jensen or it's designed for Lundahl LL1544.

From what I am told the OPT version has the kind of sound you would expect from a typical NEVE console and already on mine is studio quality with opamps.

What you are proposing from a USB DAC doesn't even come close, and from what we can see, doesn't measure as well either.

 

[tw99]

sorry that's total nonsense.
I think you never used one of Ian's DACs so you write stuff you know nothing about.

I have been using them on a CM4 for over a year. The DAC controller allows you to use a variety of output filtering, and the output is balanced XLR via a nice Opamp board which is upgradeable. (plug in 8 pin opamps).

The optional transformer output board using transformers from Biskek I have to admit to not trying yet (although I made up the board ready). You can use top quality OPT for it from others eg jensen or it's designed for Lundahl LL1544.

From what I am told the OPT version has the kind of sound you would expect from a typical NEVE console and already on mine is studio quality with opamps.

What you are proposing from a USB DAC doesn't even come close, and from what we can see, doesn't measure as well either.

Seems like you're exactly the target audience, someone who likes to build complicated stuff. Lots of technical claims are made, but generally addressing issues that don't matter.

Amir has measured many USB DACs with state of the art performance, and that performance is available from a Pi4 USB out...

 

[Jack Harrison]

sorry that's total nonsense.
I think you never used one of Ian's DACs so you write stuff you know nothing about.

I have been using them on a CM4 for over a year. The DAC controller allows you to use a variety of output filtering, and the output is balanced XLR via a nice Opamp board which is upgradeable. (plug in 8 pin opamps).

The optional transformer output board using transformers from Biskek I have to admit to not trying yet (although I made up the board ready). You can use top quality OPT for it from others eg jensen or it's designed for Lundahl LL1544.

From what I am told the OPT version has the kind of sound you would expect from a typical NEVE console and already on mine is studio quality with opamps.

What you are proposing from a USB DAC doesn't even come close, and from what we can see, doesn't measure as well either.

Can I repeat my earlier question? How exactly are you using the CM4? Is this the ‘CM4’ which is a cut down RPi4 board?

 

[Jack Harrison]

sorry that's total nonsense.
I think you never used one of Ian's DACs so you write stuff you know nothing about.

I have been using them on a CM4 for over a year. The DAC controller allows you to use a variety of output filtering, and the output is balanced XLR via a nice Opamp board which is upgradeable. (plug in 8 pin opamps).

The optional transformer output board using transformers from Biskek I have to admit to not trying yet (although I made up the board ready). You can use top quality OPT for it from others eg jensen or it's designed for Lundahl LL1544.

From what I am told the OPT version has the kind of sound you would expect from a typical NEVE console and already on mine is studio quality with opamps.

What you are proposing from a USB DAC doesn't even come close, and from what we can see, doesn't measure as well either.

Actually you raise so many questions:
Why would any one want a DAC that gives the kind of sound you would expect from a NEVE console?
Do you know that very many usb DACs have xlr outputs (one of mine does, ain’t nothing fancy just performs perfectly).
Are you aware that Amir has tested a huge number of usb DACs that are measurably transparent to the human ear?
What do you mean by studio quality?
Do you think that a usb DAC with a way beyond the limits of human hearing measurable performance is not as good as a professional DAC which uses the same chip and has an similar if not exactly the same measurable performance?
Do you realise the subject of this thread is using an RPi as a music streamer?
Do you realise that Audio Science Review is basically an objectivist board which means opinions are expected to have at least a little bit of support from measurements and data and woo words don’t count for a lot?

 

[benanders]

I don’t think there’s need to broadly dismiss a device like the Ian DAC based on absence of supporting evidence for its superiority in one or more features (namely here, jitter).

Regarding that Ian device vs. Pi USB-to-DAC and jitter - the original query - there’s been anecdotal concern about the Pi 3 due to implementation of USB; as aforementioned, Pi 4 uses a different implementation. There’s been no properly controlled study demonstrating the Pi 3 USB implementation to be an audible issue for audio file transport.

That’s to say no (TMK) evidence for a real world (= audible) issue with jitter based on controlled studies for any of the mentioned devices.
So use whichever strikes your fancy and best matches the sum that’s burning hole in your pocket.

Edit: fixed typo that confused context

 

[Jack Harrison]

I don’t think there’s need to broadly dismiss a device like the Ian DAC based on absence of supporting evidence for its superiority in one or more features (namely here, jitter).

Regarding that Ian device vs. Pi USB-to-DAC and jitter - the original query - there’s been anecdotal concern about the Pi 3 due to implementation of USB; as aforementioned, Pi 4 uses a different implementation. There’s been no properly controlled study demonstrating the Pi 3 USB implementation to be an audible issue for audio file transport.

That’s to say no (TMK) evidence for a real world (= audible) issue with jitter based on controlled studies for any of the mentioned devices.
So use whichever strikes your fancy and best matches the sum that’s burning hole in your pocket.

Edit: fixed typo that confused context

I do agree broadly especially that people’s choices how to spend their money should be completely free. I frequent quite a few audio boards and what I find refreshing about this board is that anecdotal evidence is challenged by the actual evidence and often demonstrated to be almost entirely false. I have three options for my own RPi 3b. I can use my Allo Boss hat, I also have a spdif out hat, and I can take the usb out into a usb DAC. And without measuring I can’t distinguish any issues with the usb out and that’s backed up by those who have actually measured it. But the poster wasn’t criticising the usb on the 3b but the usb on the 4b which measures as well as any usb could do I think which is to say perfectly well. I also asked a series of totally ignored questions sincerely because what the person extolling the Ian DACs was claiming seems to me to be complete nonsensical gobbledegook - thinking that a DACs sound would be good if t was like that of a NEVE console, for instance and that the Ian Dac is ‘studio quality’ when presumably no one else’s is. But I expressed myself by way of questions because I’m not closed to learning something new, I can be wrong sometimes if not frequently, and also that the implementation of a cut down RPi4 which is what the CM4 is I believe might for some actual reason be better than an actual RPi4 intrigues me especially as actual RPi4s are as rare as priceless unobtainium at the moment.

 

[benanders]

I don’t think there’s need to broadly dismiss a device like the Ian DAC based on absence of supporting evidence for its superiority in one or more features (namely here, jitter).

Regarding that Ian device vs. Pi USB-to-DAC and jitter - the original query - there’s been anecdotal concern about the Pi 3 due to implementation of USB; as aforementioned, Pi 4 uses a different implementation. There’s been no properly controlled study demonstrating the Pi 3 USB implementation to be an audible issue for audio file transport.

That’s to say no (TMK) evidence for a real world (= audible) issue with jitter based on controlled studies for any of the mentioned devices.
So use whichever strikes your fancy and best matches the sum that’s burning hole in your pocket.

Edit: fixed typo that confused context

I probably should’ve used the “ @ “ in my last post to avoid confusion of who / which comments I addressed.
When there’s broad dismissal of a product like @tw99 suggested of the Ian DAC based on a simple question (jitter) like @glorious_iguana asked, it will tend to garner reactions like @valve_integrator produced. It’s a predictable pattern on fora everywhere.

I do agree broadly especially that people’s choices how to spend their money should be completely free. I frequent quite a few audio boards and what I find refreshing about this board is that anecdotal evidence is challenged by the actual evidence and often demonstrated to be almost entirely false.

Unfortunately the process isn’t that straightforward at ASR either, just different from many other fora that don’t focus on certain measurements. But measurements are data, not studies - not experiments - and depending on how they’re interpreted for answering questions, they can be more or less helpful. Making and proclaiming the assumption of a given DAC being a purchase for people who like to make things more complicated when all that was mentioned was jitter, is not an objective statement. Could save some folks their money, though.

I have three options for my own RPi 3b. I can use my Allo Boss hat, I also have a spdif out hat, and I can take the usb out into a usb DAC. And without measuring I can’t distinguish any issues with the usb out and that’s backed up by those who have actually measured it. But the poster wasn’t criticising the usb on the 3b but the usb on the 4b which measures as well as any usb could do I think which is to say perfectly well. I also asked a series of totally ignored questions sincerely because what the person extolling the Ian DACs was claiming seems to me to be complete nonsensical gobbledegook - thinking that a DACs sound would be good if t was like that of a NEVE console, for instance and that the Ian Dac is ‘studio quality’ when presumably no one else’s is. But I expressed myself by way of questions because I’m not closed to learning something new, I can be wrong sometimes if not frequently, and also that the implementation of a cut down RPi4 which is what the CM4 is I believe might for some actual reason be better than an actual RPi4 intrigues me especially as actual RPi4s are as rare as priceless unobtainium at the moment.

I agree with this, mostly.
That Allo Boss hat - seems it’s a pricey piece of kit, too. I recall a thread somewhere on ASR that shredded a guy for touting it anecdotally. Perhaps it would be chalked up to the Ian DAC category and put you right with @valve_integrator for even daring to possess it.
My point of course being, though I don’t know you at all @Jack Harrison , your typing and queries don’t strike me as being from someone who desires making things more complicated, nor someone who tends to tout a product arbitrarily. Good on ya for that.

I do not see reason to change from a Pi 3b to a Pi 4 because while the USB out is “better” by design on the 4, it will also be hotter and consume more electricity, albeit very marginally. Will there be an audible improvement with the 4? Show me a study that suggests it and maybe I’ll reconsider.
Pi 4’s are still very easy to get, but as with anything having a kink in the current production supply chain, the price has gone up quite a bit since only second-hand sales from product hoarders are an option now. Pi 4’s are only expensive now if compared to when they were available directly from the manufacturer. For what they can do as music transports with freeware, they still seem (IMO) to be within reason. Additionally, for those who (like me) use SATA SSD’s directly connected to the Pi for file storage, those are now far cheaper than 1-2 years ago; the money saved on a 4-8 TB SSD is more than enough to make up for the difference in cost of acquiring a new Pi board at present. Silver lining FWIW.

 

[valve_integrator]

Can I repeat my earlier question? How exactly are you using the CM4? Is this the ‘CM4’ which is a cut down RPi4 board?

CM4 is not a cut down anything.
It removes many of the Rpi3-4 bottle necks with PCI-e & e-MMC v SD. (e-MMC already 4x faster than SD).

BTW - As a result it handles audio much better and I am told *on "circle" NO FIFOS.

 

[benanders]

CM4 is not a cut down anything.
It removes many of the Rpi3-4 bottle necks with PCI-e & e-MMC v SD. (e-MMC already 4x faster than SD).

BTW - As a result it handles audio much better and I am told *on "circle" NO FIFOS.

Multiple claims and no explanations or specific examples applied to real world system applications - this is the sort of thing that some ARS fellas really jump on and go nuts over - hence the reaction (and/or lack thereof) to some of your initial statements.

I am not following your rationale for the importance of speed - Raspberry Pi Zero has been demonstrated to easily handle audio files… it’s certain OS’s and/or server apps that can run into trouble with highly limited system speed.

I recall the educated pontifications on Mojo Audio’s site actually suggesting the opposite: excess computer power is suggested to be a constraint for server performance in multiple ways.

The traditional concept of a server is appropriate speed, not high speed. And low power consumption. The latter usually (but not always) is incongruous with high speed systems. Obviously in the present kit comparisons we are talking about largely irrelevant differences in some of those specs, which also challenges the utility in going with the bigger/stronger/faster option if it cannot be demonstrated to perform better. And without either signal tests or pretty darned elaborate listening tests performed in direct comparison, comments become pure opinion, whether founded or not.

I could be missing something - am I mistaken? Please don’t mistake my queries for antagonism: they are not (whether the feel that way or not ).

 

[valve_integrator]

The traditional concept of a server is appropriate speed, not high speed. And low power consumption. The latter usually (but not always) is incongruous with high speed systems.

I could be missing something - am I mistaken?

yes missing the fact I quite frequently upload files to the file server which is also a router.

Despite the obvious ethernet overhead, it's clearly silly to equate low speed with compat with large files.
Quite commonly I upload a whole DVD to a CM4 so that it can be played back.

If you start sending 10 20 - 50Gb in files at a shot onto a server, you will welcome the extra write speed on NVRAM / SSD or e-MMC compared with those crappy SD cards - originally designed for digital cameras and USB readers.

No way would I go back to RPi3b or 4 after having had my hunches confirmed with a 4Gb mem CM4 compared with the 2GB version.

Nuff said??
Write speed 21mb compared with 389....

 

[JonP]

CM4 is not a cut down anything.
It removes many of the Rpi3-4 bottle necks with PCI-e & e-MMC v SD. (e-MMC already 4x faster than SD).

BTW - As a result it handles audio much better and I am told *on "circle" NO FIFOS.

Ehhh... you could say you're both right.

The Computer module versions of the RPi's are smaller format boards, with less features/ports, to get that reduced size. So, you could say "cut down".

And, the CM4 seems to have a few improved capabilities from a stock RPi4, as well, like that higher speed disk interface. So, you could say not a "cut down" device...

I'd also have to add that there's likely no impact on audio handling, transfer or playing, since a RPi 4 (or 3, or 1) are more than capable to handle that job.

Not going to touch the FIFO comment, for now. ;-)

 

[JonP]

I have three options for my own RPi 3b. I can use my Allo Boss hat, I also have a spdif out hat, and I can take the usb out into a usb DAC.

Not to be snarky, but you forgot one! ;-)
The HDMI is another audio out. Assuming we're talking digital outs that is. We won't mention the 5th one...

Remember... friends don't let friends use the RPi audio jack...

(actually, I think even that got improved, somewhat, from the truly horrible one back in the RPi 1 and 2 era)