I know for a fact that a large part of the Software Engineering team at Devialet is actively lurking around here, although prohibited by management from officially posting on this forum.
I strongly believe everybody has a very big misconception about the size of Devialet, and especially of the Software team.
They used to be a group of very dedicated people trying as hard as they could to deliver the very best product to their customers.
I also know that in the past, the Software Engineering team has repeatedly been in conflict with management regarding the lack of a dedicated Quality Assurance team, explaining how the mileage of different customers has varied drastically. So technically yes, for quite some time, customers have really been beta testers.
Despite all the comments from seasoned engineers about the absolute necessity of having a QA team discovering bugs before the customers did, especially given the Devialet price point, for quite a while, testing fell on the engineers themselves.
So testing essentially happened in perfect R&D environment conditions, done by the same engineers which were already overwhelmed by the scope of features, switching of webradio providers, and other marketing gimmicks (who remembers "Live on Phantom", live opera streaming on Phantom Premier?).
No surprise here that DOS1 had many holes to patch. Not to mention that DOS1 was entirely tied to the Phantom Premier and Dialog hardware in its design.
What people don't seem to see is that the very same DOS2 that now powers their Phantom Premier is the very same OS that was released on Phantom Reactor without any of the issues or complaints that people seem to report with this migration.
In terms of features, it's definitely bare bones, but the strategy is clear: adopt streaming standards and let people choose their player.
It seems perfectly OK for new Phantom Reactor customers, where the reception is much better than that of Spark and DOS1.
Moreover, this migration is not mandatory, so much so that you have to download a new app to perform it.
If users are happy with what DOS1 and Spark offers them, they're free to keep it!
However, it's true that a lot of users complained about the lack of usability of Spark, especially in here. Something one could explain by the continuous refusal from management to hire UX and UI designers, or even to follow the guidelines from the one-off freelancers on that matter.
Well aware of that, Devialet finally (after much lost time) decided to start from scratch with a basic application to configure and control the system, and let users stream their music their own way, hence focusing on audio rendering rather than content delivery.
More protocols are definitely to come with AirPlay 2, Roon RAAT and maybe Google Cast. So maybe people should just embrace that. Let's also note that in terms of HiRes formats, nothing beats UPnP which is readily available on Premier and Reactor.
What that means is that those issues (especially regarding stereo sync) will definitely be fixed with an upcoming release, that the foundation of DOS2 is way more solid than DOS1, that multi-room is definitely on the horizon, and one could even dream of mixed multi-channel setups with Premier and Reactor.
My only concern now as a owner of a pair of Gold Phantom Premier is that most of the senior developers have left the company because of all this repeated friction with their management. That could negatively impact the velocity with which this roadmap is executed.
I strongly believe everybody has a very big misconception about the size of Devialet, and especially of the Software team.
They used to be a group of very dedicated people trying as hard as they could to deliver the very best product to their customers.
I also know that in the past, the Software Engineering team has repeatedly been in conflict with management regarding the lack of a dedicated Quality Assurance team, explaining how the mileage of different customers has varied drastically. So technically yes, for quite some time, customers have really been beta testers.
Despite all the comments from seasoned engineers about the absolute necessity of having a QA team discovering bugs before the customers did, especially given the Devialet price point, for quite a while, testing fell on the engineers themselves.
So testing essentially happened in perfect R&D environment conditions, done by the same engineers which were already overwhelmed by the scope of features, switching of webradio providers, and other marketing gimmicks (who remembers "Live on Phantom", live opera streaming on Phantom Premier?).
No surprise here that DOS1 had many holes to patch. Not to mention that DOS1 was entirely tied to the Phantom Premier and Dialog hardware in its design.
What people don't seem to see is that the very same DOS2 that now powers their Phantom Premier is the very same OS that was released on Phantom Reactor without any of the issues or complaints that people seem to report with this migration.
In terms of features, it's definitely bare bones, but the strategy is clear: adopt streaming standards and let people choose their player.
It seems perfectly OK for new Phantom Reactor customers, where the reception is much better than that of Spark and DOS1.
Moreover, this migration is not mandatory, so much so that you have to download a new app to perform it.
If users are happy with what DOS1 and Spark offers them, they're free to keep it!
However, it's true that a lot of users complained about the lack of usability of Spark, especially in here. Something one could explain by the continuous refusal from management to hire UX and UI designers, or even to follow the guidelines from the one-off freelancers on that matter.
Well aware of that, Devialet finally (after much lost time) decided to start from scratch with a basic application to configure and control the system, and let users stream their music their own way, hence focusing on audio rendering rather than content delivery.
More protocols are definitely to come with AirPlay 2, Roon RAAT and maybe Google Cast. So maybe people should just embrace that. Let's also note that in terms of HiRes formats, nothing beats UPnP which is readily available on Premier and Reactor.
What that means is that those issues (especially regarding stereo sync) will definitely be fixed with an upcoming release, that the foundation of DOS2 is way more solid than DOS1, that multi-room is definitely on the horizon, and one could even dream of mixed multi-channel setups with Premier and Reactor.
My only concern now as a owner of a pair of Gold Phantom Premier is that most of the senior developers have left the company because of all this repeated friction with their management. That could negatively impact the velocity with which this roadmap is executed.
