New cheap portable player in the market

[speachy]

> I would do what I can to fund it as well.  I shied away from reading that enormous thread...  I wonder if it would be easier now than it would have been back when that thread was started.  With Arduino, Raspberry Pi, and similar boards readily available.  I think someone would just have to take the time to design it, build a prototype...

The tl;dr of that (and all other "rockbox should build its own hardware") thread:

1) Nearly all of those SBCs are designed to be always-on, powered through USB, with no thought to running on a battery. Retrofitting that after the fact is pretty janky.
2) Building a prototype is relatively straightforward (that is to say, about the same effort now as it was ten years ago)
3) Designing and fabricating a custom PCB is easier/cheaper than ever.
4) Designing and tooling up to manufacture e case is where most of the fixed cost comes from, easily in the five digits.
5) 3D printing prototypes likely won't give us the fine detail necessary, and the per-unit costs will be astronomical for anything we'd hope to sell.
6) These fixed costs don't change if we want to build one unit or one million.
7) Supply chain realities mean that we'd need to commit to building every unit we ever hope to produce up front, as there'd be no guarantee we could build another batch even six months layer.  So that's a lot of part inventory that needs to be purchased in advance.
We have to produce something not only price-competitive with everything rockbox runs on today -- both new[-ish] units like the Eros Q/K players but also the huge inventory of secondhand ipods.
9) Going after the "audiophile" market is a pointless exercise, as they are always chasing the latest shiny thing and finding reasons we don't measure up to the competition (lacking features like tube amps and "oxygen-depleted gold-plated optical connectors to reduce jitter")

Not to say this is impossible, but to give you an idea about the kinds of numbers we're looking at, I'd say we'd need about $100,000 in up-front funding for design and tooling plus the cost of parts to build an initial 1000 unit run, which will probably be closer to $100,000, ie per-unit costs of about $100.  And actually selling those things would require even more up-front money for things like business liicenses, regulatory approval, and minor hassles like getting a license from MPEG/LA so we can decode AAC and other necessary codecs.  So did I say $100,000?  Make that $200,000 as lawyers aren't cheap.

the tl;dr of that:   Any hypothetical kickstarter would probably need to be something like $300/unit for a 1000-unit run, for something roughly equivalent to an ErosQ/K. (Real econmomies of scale start to kick in over 10,000 units, and that requires even more money up front)  Or folks can just buy an old ipod instead.

> Unfortunately, I don't have time for such fun.  If it were me, I would take the easiest route using readily available parts to get something that works.  It would probably be bulky and ugly.  That's why I thought it might be easier to just take a device that already exists and port Rockbox to it.

... And then there's what I didn't mention above, (10):  It's going to take time to write the software needed to make all of this work.  And all of this effort is the equivalent of several full-time jobs.

In an ideal world, the sanest approach would be to reach out to one of the Chinese ODMs and ask them for a quote to produce 10,000 units of an existing player design, conditional on having our logo on the case and a full set of schematics.  That would be something that would easily translate to a kickstarter.

[saratoga]

If designing and making an attractive and functional mp3 player case was easy there wouldn't have been so many awful, janky devices made over the years

PCB is easy enough though if you don't mind it not being a portable device. 

[saucell]

> I would do what I can to fund it as well.  I shied away from reading that enormous thread...  I wonder if it would be easier now than it would have been back when that thread was started.  With Arduino, Raspberry Pi, and similar boards readily available.  I think someone would just have to take the time to design it, build a prototype...

The tl;dr of that (and all other "rockbox should build its own hardware") thread:

1) Nearly all of those SBCs are designed to be always-on, powered through USB, with no thought to running on a battery. Retrofitting that after the fact is pretty janky.
2) Building a prototype is relatively straightforward (that is to say, about the same effort now as it was ten years ago)
3) Designing and fabricating a custom PCB is easier/cheaper than ever.
4) Designing and tooling up to manufacture e case is where most of the fixed cost comes from, easily in the five digits.
5) 3D printing prototypes likely won't give us the fine detail necessary, and the per-unit costs will be astronomical for anything we'd hope to sell.
6) These fixed costs don't change if we want to build one unit or one million.
7) Supply chain realities mean that we'd need to commit to building every unit we ever hope to produce up front, as there'd be no guarantee we could build another batch even six months layer.  So that's a lot of part inventory that needs to be purchased in advance.
We have to produce something not only price-competitive with everything rockbox runs on today -- both new[-ish] units like the Eros Q/K players but also the huge inventory of secondhand ipods.
9) Going after the "audiophile" market is a pointless exercise, as they are always chasing the latest shiny thing and finding reasons we don't measure up to the competition (lacking features like tube amps and "oxygen-depleted gold-plated optical connectors to reduce jitter")

Not to say this is impossible, but to give you an idea about the kinds of numbers we're looking at, I'd say we'd need about $100,000 in up-front funding for design and tooling plus the cost of parts to build an initial 1000 unit run, which will probably be closer to $100,000, ie per-unit costs of about $100.  And actually selling those things would require even more up-front money for things like business liicenses, regulatory approval, and minor hassles like getting a license from MPEG/LA so we can decode AAC and other necessary codecs.  So did I say $100,000?  Make that $200,000 as lawyers aren't cheap.

the tl;dr of that:   Any hypothetical kickstarter would probably need to be something like $300/unit for a 1000-unit run, for something roughly equivalent to an ErosQ/K. (Real econmomies of scale start to kick in over 10,000 units, and that requires even more money up front)  Or folks can just buy an old ipod instead.

> Unfortunately, I don't have time for such fun.  If it were me, I would take the easiest route using readily available parts to get something that works.  It would probably be bulky and ugly.  That's why I thought it might be easier to just take a device that already exists and port Rockbox to it.

... And then there's what I didn't mention above, (10):  It's going to take time to write the software needed to make all of this work.  And all of this effort is the equivalent of several full-time jobs.

In an ideal world, the sanest approach would be to reach out to one of the Chinese ODMs and ask them for a quote to produce 10,000 units of an existing player design, conditional on having our logo on the case and a full set of schematics.  That would be something that would easily translate to a kickstarter.

thanks for this write up! if only some chinese maker would just make a player we could rockbox and sell it until the end of time all our problems would be solved!

[llnx1]

This look interesting, Anbernic often releases device source code of their devices

Has anyone seen this k3s soc ? I searched on google but didn't find anything about k3s, only v3s

I believe this device must have something like cortex A7 and at least 32 / 64mb ram for run Game Boy Advance, what should be enough to run Rockbox port

This a video from this device

https://www.youtube.com/watch?v=YVvBxdhKOQY

[speachy]

I believe this device must have something like cortex A7 and at least 32 / 64mb ram for run Game Boy Advance, what should be enough to run Rockbox port

I'd be shocked if it didn't have the hardware specs to run rockbox quite well, but the lack of a dedicated 3.5mm headphone jack is concerning.  It's also not clear if there's expandable storage (they didn't show one side of the device at all, so there _could_ be an SD slot there...) but it looks like a really nice form factor, and the guessimated price point (~$40) is pretty nice too.

[llnx1]

I believe this device must have something like cortex A7 and at least 32 / 64mb ram for run Game Boy Advance, what should be enough to run Rockbox port

I'd be shocked if it didn't have the hardware specs to run rockbox quite well, but the lack of a dedicated 3.5mm headphone jack is concerning.  It's also not clear if there's expandable storage (they didn't show one side of the device at all, so there _could_ be an SD slot there...) but it looks like a really nice form factor, and the guessimated price point (~$40) is pretty nice too.

Apparently it will be a Funkey S clone by Anbernic which have a internal SD slot, a SOC is a v3s (not k3s) and bad side is a lack of p2 port (need a USB C > P2)

https://linux-sunxi.org/V3s

For 40~45 dollars and being easy to sell (Aliexpress) I believe it is a good option to develop a port, let's stay tuned when it launches

[saucell]

great find guys! looking forward to reading this thread as more info comes out!

[speachy]

Apparently it will be a Funkey S clone by Anbernic which have a internal SD slot, a SOC is a v3s (not k3s) and bad side is a lack of p2 port (need a USB C > P2)

If I could wave a wand and build a "just for rockbox" player, I'd pick the S3 as its SoC -- It's essentially the V3s with double the RAM (ie 128MB) but in a much smaller BGA package.

[osaka]

The remaining unknown is the Shenju Jointbees chip, but I'm pretty sure it's an Actions clone.  It has the same form factor, and their site does list a similar chip with only 216 Kb RAM

The firmware for the chip is definitely different from Actions. My player enters some kind of upgrade (boot ROM?) mode if VOL (down) is held on reset: enumerates as mass storage VID_301A&PID_2800 and seems to accept C0 00, C0 03..06 vendor-specific commands (only managed to get a CSW back yet); IsActionsFirmware (CC) stalls the endpoint instead.
The company itself seems closely tied to Zhuhai Smartlink Technology Co., Ltd. as they use their USB Vendor ID and website looks almost the same. There's front page excerpts for the datasheets in English in the News section but sadly no other new information. They boast OTA function for the TWS headphone chipsets, I'm currently trying to locate an app responsible to pull some firmware out maybe.