NymphCast Network Audio / Chromecast Replacement?

[The Computer Audiophile]

Anyone tried this yet?

 

Using Raspberry Pi Like a Chromecast? Open-Source NymphCast Project Makes it Happen
 

“Who needs to buy a Chromecast stick when you build your own, open-source alternative? Software engineer Maya Poschis turning Raspberry Pis into streaming platforms with her latest project known as NymphCast. This open-source Chromecast alternative lets you stream video and audio to your television using a Raspberry Pi.”

https://www.tomshardware.com/amp/news/using-raspberry-pi-like-a-chromecast-open-source-nymphcast-project-makes-it-happen

[The Computer Audiophile]

@vortecjr Any chance of this on the Rendu / Sonicorbiter platform?

[The Computer Audiophile]

@bluesman have you seen this?

[bluesman]

53 minutes ago, The Computer Audiophile said:

@bluesman have you seen this?

Yes indeed!  It's cool, but the name is a bit misleading - it does not use the CCast protocol or the Cast SDK, so you can't 'cast to or from it.  It's a media server that will retrieve source files stored on the network and output them to a renderer (which has to be connected directly because Wifi use is limited to your WLAN & it won't output BT audio). So far, all I know about output connections is that it uses HDMI - I downloaded it when I read about it, but I haven't had time to boot it up yet.

 

On the face of it, I haven't read anything that differentiates it positively from the many other media servers available for RPi and other SBCs.  I have to find out for myself (and for the rest of us) if it will output digital audio via USB and analog via the audio output.  If it's limited to HDMI as the current scant documentation suggests, it's not a good match for powered monitors (which are pictured in the articles about it).  So if it doesn't use USB, I assume it'll at least drive them with an analog signal - which is great for all of the mp3 fans on AS  

 

Nymphcast now says it's a beta release on Github (last time I looked, it was alpha).  It only runs apps written in AngelScript for the Nymphcast platform. I don't think it's going to be a serious player (pun intended) in this game, but it's cool enough to deserve wringing out. I hope to be pleasantly surprised, but (as we need to learn to do today) I hope the best while planning for the worst.....

[The Computer Audiophile]

5 minutes ago, bluesman said:

Yes indeed!  It's cool, but the name is a bit misleading - it does not use the CCast protocol or the Cast SDK, so you can't 'cast to or from it.  It's a media server that will retrieve source files stored on the network and output them to a renderer (which has to be connected directly because Wifi use is limited to your WLAN & it won't output BT audio). So far, all I know about output connections is that it uses HDMI - I downloaded it when I read about it, but I haven't had time to boot it up yet.

 

On the face of it, I haven't read anything that differentiates it positively from the many other media servers available for RPi and other SBCs.  I have to find out for myself (and for the rest of us) if it will output digital audio via USB and analog via the audio output.  If it's limited to HDMI as the current scant documentation suggests, it's not a good match for powered monitors (which are pictured in the articles about it).  So if it doesn't use USB, I assume it'll at least drive them with an analog signal - which is great for all of the mp3 fans on AS  

 

Nymphcast now says it's a beta release on Github (last time I looked, it was alpha).  It only runs apps written in AngelScript for the Nymphcast platform. I don't think it's going to be a serious player (pun intended) in this game, but it's cool enough to deserve wringing out. I hope to be pleasantly surprised, but (as we need to learn to do today) I hope the best while planning for the worst.....

Great info. Touting it as a Chromecast replacement is a bit much then. Seems like UPnP. 
 

Guess I have more research to do. 

[bluesman]

8 minutes ago, The Computer Audiophile said:

Great info. Touting it as a Chromecast replacement is a bit much then. Seems like UPnP. 
 

Guess I have more research to do. 

It’s a crude CC substitute in that it’s wireless to the network and is controlled remotely by networked devices.  If HDMI is the only digital out (or worse, the only out), it’s not going to be useful to many audiophiles. It looks like a long run for a short slide but I’ll put it through it’s paces.

[bluesman]

As promised, I spent a bit of today's "social distancing time" exploring NymphCast.  My executive summary is simple:

1. It's a lot more work than it's worth, and I can't imagine why any audiophile would go to this much trouble to build a marginal solution to problems no one has.
2. The system architecture is strange to say the least.  Although the software you build on the Pi is called the server, the developer also calls it the receiver and it seems to be the renderer / player as its output is wired to the input of the device that will output your audio and video.
3. You have to build a "player" on another device to use this.  Per the developer, the player (which, I assume is the controller) "can be built on Linux/BSD/MacOS with a current GCC toolchain, or MSYS2 on Windows with MinGW toolchain".  There's also apparently an Android APK.
4. I got as far as executing the script that builds the "server" on the Pi and decided that I'd rather open a bottle of wine and sit in the living room with my wife.

If you're an intrepid experimenter and just want to see how you do with it, don't bother downloading and setting up the "prebuilt image" from the developer. If your Pi isn't running the current Raspbian image or a very recent precursor, just download the latest stock NOOBS or Raspbian image yourself.  The "prebuilt" image is just a September 2019 Debian Buster image with a few compromises that the developer must think make it "lighter", e.g. the embedded web browser is Dillo.  For those who never toiled in the trenches of early IT, here's the human incarnation of Dillo:

 

 

Once you waste time downloading, burning, and setting up the same OS image you already had on your Pi before (now without apps you actually use), you have another hour or two of work before you can even figure out if it will make music (and, if so, how good it is).  All I can say is that if I needed a device to wirelessly retrieve networked audio files and play them through powered speakers or a TV / HT system, I'd use Kodi, Emby, Plex, VLC, or another such proven solution.  Rooners can run Bridge on a Pi, and J River Media Center users can set up any DLNA software they like or run J River for ARM on the Pi to make it into a remote zone. 

 

I'd love to know what need the developer thinks she's satisfying with this, because I can't identify a single one.  It's always fun to play with new programming platforms and to make something work better in some way. And I applaud anyone who tries a whole new approach to an old standard and truly improves on it.  Maybe Nymphcast will be developed more fully into a package that keeps a Pi cooler, or speeds up the graphics, or improves SQ, or makes something else better.  But for now, I just don't see that in the cards.

[The Computer Audiophile]

5 minutes ago, bluesman said:

As promised, I spent a bit of today's "social distancing time" exploring NymphCast.  My executive summary is simple:

1. It's a lot more work than it's worth, and I can't imagine why any audiophile would go to this much trouble to build a marginal solution to problems no one has.
2. The system architecture is strange to say the least.  Although the software you build on the Pi is called the server, the developer also calls it the receiver and it seems to be the renderer / player as its output is wired to the input of the device that will output your audio and video.
3. You have to build a "player" on another device to use this.  Per the developer, the player (which, I assume is the controller) "can be built on Linux/BSD/MacOS with a current GCC toolchain, or MSYS2 on Windows with MinGW toolchain".  There's also apparently an Android APK.
4. I got as far as executing the script that builds the "server" on the Pi and decided that I'd rather open a bottle of wine and sit in the living room with my wife.

If you're an intrepid experimenter and just want to see how you do with it, don't bother downloading and setting up the "prebuilt image" from the developer. If your Pi isn't running the current Raspbian image or a very recent precursor, just download the latest stock NOOBS or Raspbian image yourself.  The "prebuilt" image is just a September 2019 Debian Buster image with a few compromises that the developer must think make it "lighter", e.g. the embedded web browser is Dillo.  For those who never toiled in the trenches of early IT, here's the human incarnation of Dillo:

 

 

Once you waste time downloading, burning, and setting up the same OS image you already had on your Pi before (now without apps you actually use), you have another hour or two of work before you can even figure out if it will make music (and, if so, how good it is).  All I can say is that if I needed a device to wirelessly retrieve networked audio files and play them through powered speakers or a TV / HT system, I'd use Kodi, Emby, Plex, VLC, or another such proven solution.  Rooners can run Bridge on a Pi, and J River Media Center users can set up any DLNA software they like or run J River for ARM on the Pi to make it into a remote zone. 

 

I'd love to know what need the developer thinks she's satisfying with this, because I can't identify a single one.  It's always fun to play with new programming platforms and to make something work better in some way. And I applaud anyone who tries a whole new approach to an old standard and truly improves on it.  Maybe Nymphcast will be developed more fully into a package that keeps a Pi cooler, or speeds up the graphics, or improves SQ, or makes something else better.  But for now, I just don't see that in the cards.

You're a star @bluesman!

[hackkitten]

On 3/18/2020 at 10:28 PM, bluesman said:

Yes indeed!  It's cool, but the name is a bit misleading - it does not use the CCast protocol or the Cast SDK, so you can't 'cast to or from it.  It's a media server that will retrieve source files stored on the network and output them to a renderer (which has to be connected directly because Wifi use is limited to your WLAN & it won't output BT audio). So far, all I know about output connections is that it uses HDMI - I downloaded it when I read about it, but I haven't had time to boot it up yet.

 

Hello, I'm the NymphCast developer :)

 

NymphCast will output to whatever audio/video output(s) have been configured. It's not bound to a specific AV port. As for its feature set, it's meant to eventually cover everything that ChromeCast can do, and more, even if it doesn't speak the same (proprietary, closed & encrypted) protocol. Right now it can be used for streaming media on the network, sending URLs to for streams which the server will then play back as CC does and use the (at the moment limited) NymphCast apps for services like SoundCloud.

 

On 3/18/2020 at 10:28 PM, bluesman said:

 

On the face of it, I haven't read anything that differentiates it positively from the many other media servers available for RPi and other SBCs.  I have to find out for myself (and for the rest of us) if it will output digital audio via USB and analog via the audio output.  If it's limited to HDMI as the current scant documentation suggests, it's not a good match for powered monitors (which are pictured in the articles about it).  So if it doesn't use USB, I assume it'll at least drive them with an analog signal - which is great for all of the mp3 fans on AS  

 

As stated NymphCast is not limited to HDMI output. I am aware that the documentation doesn't cover every single use case yet, but using it as a ChromeCast Audio replacement is definitely one of the primary targets.

 

On 3/18/2020 at 10:28 PM, bluesman said:

 

Nymphcast now says it's a beta release on Github (last time I looked, it was alpha).  It only runs apps written in AngelScript for the Nymphcast platform. I don't think it's going to be a serious player (pun intended) in this game, but it's cool enough to deserve wringing out. I hope to be pleasantly surprised, but (as we need to learn to do today) I hope the best while planning for the worst.....

 

NymphCast is at the moment in alpha testing prior to the v0.1 release later this year.

[hackkitten]

9 hours ago, bluesman said:

 

I'd love to know what need the developer thinks she's satisfying with this, because I can't identify a single one.  It's always fun to play with new programming platforms and to make something work better in some way. And I applaud anyone who tries a whole new approach to an old standard and truly improves on it.  Maybe Nymphcast will be developed more fully into a package that keeps a Pi cooler, or speeds up the graphics, or improves SQ, or makes something else better.  But for now, I just don't see that in the cards.

I'm not sure why you were expecting a smooth end-user experience considering that I only spent the equivalent of a month of full-time work on creating the basic ChromeCast experience from scratch. 99% of the work goes into implementing and testing the core functionality, trying to install or use the project at this early stage is bound to be a rocky experience, as it is for me and those who currently are testing the project.

 

The stated goal of the project is still correct: it will have all the features that ChromeCast has, and possibly more. As for putting it in equal footing as other projects, neither DLNA nor MiraCast are elegant or well-supported protocols. Both suffer from countless compatibility issues and are complicated, heavy-weight protocols. The main reason why I started NymphCast was to have something that 'just worked', after suffering through the humility of ChromeCast and Airplay for years and not finding any suitable alternative.

 

This is primarily a project to satisfy my own needs and to learn from as a developer. Also known as a hobby project. Anything else that happens is a bonus.

[Rexp]

What i would to see is a protocol similar to bluetooth that allows you to play 16/44 1411kbs on your phone wirelessly to a receiver. 

[bluesman]

8 hours ago, hackkitten said:

Hello, I'm the NymphCast developer

Hi!!  I'm thrilled that you're here and equally thrilled with the quality and content of your comments.  This is a great opportunity for an open discussion of many of the issues affecting what you and so many other solution developers do.  Let's explore how they affect you and the community of stakeholders drawn to your efforts, in the hope that we can be helpful to both you and the audiophiles who anxiously await the answers to their prayers

 

First, as I said in my discussion above, "it's always fun to play with new programming platforms and to make something work better in some way. And I applaud anyone who tries a whole new approach to an old standard and truly improves on it".  I've been there countless times since learning Fortran on my university's open source IBM 1620.  I've been a software designer, creator and developer since 1965 and have remained current. For example, I "attended" Macromedia University and got certified as a ColdFusion developer about 20 years ago, and have learned and actively use the Cs, J, JS, VB, Ruby, etc. I've written a few dozen medium scale (100k to 250k lines of code) healthcare software systems in current use and a few larger works (500k+ lines) also actively running in major academic medical centers today. I personally installed and supported at least one each of these in 13 university hospitals around the country.  I'm also a lifelong audiophile and professional musician with a pile o' Pis and other SBCs on which I've tried many novel things (some of which worked...)

 

The big disconnect between you and the audiophile community in this case seems to me to be the unreasonable and unrealistic impressions given by third party reports of your work (e.g. the piece on Tom's by Ash Puckett a few days ago that started this thread).  The web forum discussions that have sprung up (e.g. at opensourceforU and kitguru) also make it sound like it's ready for the kind of beta use that means "it works OK but there are bugs and glitches you can help us fix".  These were clearly neither written by you nor based on information provided by you, although your own descriptive web page is a bit on the overly optimistic side. Your opening  statement that "NymphCast is a software solution which turns your choice of Linux-capable hardware into an audio and video source for a television or powered speakers. It enables the streaming of audio and video over the network from a wide range of client devices, as well as the streaming of internet media to a NymphCast server, controlled by a client device" does give one (me!) the impression that it's sufficiently developed to be usable today by audiophiles with little or no specific computer knowledge or experience.

 

Sadly, such distortions of what you're doing and where you are with it do you a disservice by creating unrealistic expectations like that "smooth end-user experience" you mention.  Those who wrote that stuff about Nymphcast (and, therefore, those who read it) are either ignorant of or chose to ignore the simple fact that you "...only spent the equivalent of a month of full-time work on creating the basic ChromeCast experience from scratch".  Knowing where you really are in this project puts it in an entirely new light, i.e. you're working your buns off to create a small, economical device that will source network files wirelessly to drive an endpoint, and (in the immortal words of the Carpenters...) you've only just begun. 

 

I think you'd do well to consider changing definitive statements like "Nymphcast is a software solution..." to "Nymphcast is being developed as a software solution..." etc.  Although if you are successful it will someday do so, it does not now enable most Audiophile Style participants to stream audio and video over their networks.  In point of fact, very few will get beyond looking at the setup executable and throwing in the towel. At least for the present, they are much better directed to the solutions I suggest above. I'd also suggest dropping the Buster-lite clone and spec'ing a standard Raspbian instance instead.  I strongly doubt that swapping Dillo for Chromium is really necessary for smooth operation - and if it is, something else is wrong.

 

I would also caution you to be careful in describing your work on a web page that looks to all the world like it's offering a workable system (bugs and all).  I'm very happy that you have the opportunity to explain your project here, but those who learn of it from third party sources and go to it expecting something they can use today will be disappointed.  This will reduce their interest in looking at it in the future, even though it might then do what they thought it would do now.

 

Unless I missed it, there are other issues you might start addressing in your communications with the audiophile community, e.g. which file & output formats you plan to handle.  When ready for prime time, will a Nymphcast server on a Pi do everything (audibly) that Roon Bridge does? How do you expect it to compare in sound quality to JRiver Media Center driving a CCA’s optical output, and why?  Can you compare your projected performance to a widely used standard like MPD controlled by Cantata?  What's the library management capability of what you're calling the player (and I'm assuming is the control point)?  You'd also do well to respond quickly and accurately to every new thread about Nymphcast on any site.  The best way to keep people interested is to stop the dissemination of drivel and let them know that you're honest and well equipped to bring your project to fruition.  

 

I hope everyone on AS will chime in and offer you constructive input, so you can both learn and make Nymphcast as good as it can be.  We'd all love to be able to drop a download onto a card, fire up a Pi, and stream high quality music effortlessly to endpoints through our homes.  We'd all like to see our brightest and best pave the way for the next run of innovators, and you appear to have what it takes to be a leader!

 

Best regards -

 

David

[Cebolla]

5 hours ago, Rexp said:

What i would to see is a protocol similar to bluetooth that allows you to play 16/44 1411kbs on your phone wirelessly to a receiver. 

 

You require an iPhone - it's called AirPlay.

[Cebolla]

11 hours ago, hackkitten said:

As stated NymphCast is not limited to HDMI output. I am aware that the documentation doesn't cover every single use case yet, but using it as a ChromeCast Audio replacement is definitely one of the primary targets..

 

I'd suggest pencilling in gapless playback support as a must for the CCA replacement.

[Rexp]

6 hours ago, Cebolla said:

 

You require an iPhone - it's called AirPlay.

Sounds sh#t

[hatchum]

Nice! Looks like SoundWire on steroids.

 

Audio casting is already trivial between Linux systems thanks to Pulseaudio's networking capabilities, but both audio and video on multiple OSes is not a trivial feat and really hope the developer figures it out. It would be huge. Would be great if this could somehow overcome Android's audio resampling quirk to allow truly hassle-free, cheap and bit perfect USB audio over WiFi solutions. Of course, the feasibility of such a system would be very much network speed/bandwidth dependent, especially for video casting. It might be necessary to develop/implement some a low latency data compression scheme for it to work in most local networks. Hopefully, lossless for audio.