For those that remember, in early copies of Windows (amongst other systems far better), there were sounds in the system folders or in Media which resided quietly amongst the alert sounds, etc. that one could play to test if their sound card(s) were working correctly. When testing hardware, I couldn’t get a proper test of the audio on some models as Haiku does not include sample music or sounds that test the range of the connected speakers. Usually, I transfer a track to test play, but it would be quite nice to have said files baked into the image.
Without going on any further, why not add a few such files to Haiku? Perhaps even a public-domain classical piece may be included for this purpose. It would literally take nothing to code, and would just mean tossing an extra 3-12 MB into the build. Then again, this is, of course, just an idea that I thought of posting… thank you for reading it if nothing else.
P.S. (If this belongs in the idea box instead, my apologies).
Our beloved devs seem to have this strict requirement that the whole distro must fit on a CD-R. Remember those? So any suggestion that is going to take up precious megabytes will not likely be accepted.
OTOH it would be fairly trivial to create a hpkg with appropriate sound files and make it downloadable in HaikuDepot. Don’t look at me: I’m not an A/V kinda guy and I’m half tone-deaf anyway. But that would be the way to go.
Actually, with package management, the size of the ISOs are now dramatically smaller (a current nightly ISO is around 280MB). We could even include sample videos and images, in addition to sample audio.
With WebPositive allowing replay of Youtube videos, and music from Grooveshark, Jamendo and a few similar services, the need for a bundled audio file to test sound is much lower.
Still needed is a set of default system sounds (we are still looking for someone who will design a sound theme that matches the Haiku look and feel - not as easy as it sounds).
We actually have demonstration packages with a set of audio and video files, pictures and miscelaneous data (e-mails, person files, etc). There is a reason for not putting those into a package, however. The goal for these is to demonstrate Haiku in a more "real-life" situation at demo shows and in other occasions. So, they are usually extracted to the home directory as user owned files would be. The "standard" demonstration run involves making changes to them (for example to show how a live query can be updated realtime, and how a MediaPlayer playlist can be based on such a query).
I'm not sure what the size of the release profile currently is. The nightlies are smaller, but we include some more software in the stable releases. While the package compression will certainly free some space, I don't know what the exact status is. And depending on the available space, there should be a discussion and vote about what to include/remove. My view on this is that it is important to have a good balance of a clean system (so as few packages as possible) and an useable one (so the main apps already installed). If we bundle a single song with Haiku, we have to pick it carefuly, as it can become strongly associated with the OS (as canyon.mid did with Windows).
From the point of view of testing audio setup the most useful things are typically
Speaker test, testing each speaker separately. This can have voice prompts, but obviously in a fully internationalised OS that's a lot of extra work. So it can also work to display visually which speaker is tested and play sound only through one at a time. This test verifies that all the speakers are connected, and correctly positioned. For headphones it checks they're on correctly.
Playback test. The test should begin with an audio prompt to demonstrate that the speakers work and are switched on (again it can be a voice prompt but that means extra localisation work) then the microphone is recorded for a brief period, and then the result is played back through the speakers. This gives confidence that the microphone works, and some indication of whether levels on the microphone are correct, without causing the user to wrestle with feedback
Both of these can be done without pre-generated sample sounds, although it can just as well work to include sample sounds for these tests.
With WebPositive allowing replay of Youtube videos, and music from Grooveshark, Jamendo and a few similar services, the need for a bundled audio file to test sound is much lower.
Still needed is a set of default system sounds (we are still looking for someone who will design a sound theme that matches the Haiku look and feel - not as easy as it sounds).
[/quote]
Hi
In my opinion, assumming the user has an internet connection set is not the most elegant way to go :).
+1 to the original idea.