Mic Problem & CPU

[Frost]

Hey everyone,
Just saw alchemy and loved it. Needs a little bit of getting used to but nonetheless very fun. I'm looking forward to see it develop.

One problem I encountered is the mic shapes. It does not work. I've readjusted the mic volume to the point that the speakers are booming with the "silence" sound and clicks from movement (noise reduction is off), and I shout but nothing seems to happen. Mic expand, also, does nothing.
Another problem with mic shapes is that, every time I choose it, the cpu goes to 100% usage and stays there, until I switch it back to another method, then it just falls to ~0%. Is this behavior normal with sound input routines or is it some java bug in my computer?

Thanks for the great job.

[NDean]

Hey,
I dont know alot about this but when I got it I also thought mic wasnt working, untill I discovered I had to put my cursor on the shape I want to alter. Plus, its effects seem subtle anyway, apart from expand.

CPU usage wise I havnt a clue!

[Karl DD]

Hi,

Firstly can you give some more info, i.e. bugs/topic73.html
Along with some info about your system, especially the type of audio card etc...
Mic stuff does depend on the system audio so this can potentially cause problems.
But give us some more info and we will do our best to see what the problem is.

Karl

[Frost]

Hi,
I'm running windows vista 32bit. Several jre (se and ee) versions are installed, but i think jre1.6.0 update 11 is the main one, though, is alchemy ran by java? it seems to run without invoking jvm.
Audio card is onboard, soundmax integrated digital hd. In my experience, soundmax mic in with vista is quite prone to error, so this might be caused by the card.

Another problem is, well not a problem but when auto record is selected, during the save cpu usage spikes, which causes the cursor/tablet to become unresponsive and jump around, creating straight lines. That may be reduced by, I don't know, keeping several pages in the memory until it is flushed to the disk less frequently? yes, pages may be lost if it crashes, but it seems quite stable, never crashed for me. Or better yet, save it as a background job in another thread, slooowly?
Also not a bug, but undo please? I know the "happy little mistakes" we make, and the process of experimentation, but some methods are very uncontrollable and it is not uncommon to have a huge blob of paint covering all of the work you've done in a flash. Since the shapes are so fluid, it may be impossible to get the "just right" look you had before. Anyway, just my opinion.

[Karl DD]

Hi,

Thanks for the info.
Alchemy is written in Java.

When you run MicShapes can you get the output from the console?
Should look something like this:

Selected Format: PCM_SIGNED 11025.0 Hz, 16 bit, mono, 2 bytes/frame, big-endian

You might need to download the debug version and take a screen grab of the console window that pops up:
http://al.chemy.org/files/Alchemy-006-Debug.zip

For the session PDF saving, Alchemy uses iText - which I think, but am not sure, runs stuff in a background thread. Each time the session file is saved it is cumulative, so each page is copied into the file and a new one added at the end. Having really long session files can cause things to slow down and I would be interested to know how big the PDF files are when you experience unresponsiveness?

Re, Undo - check out this thread: ideas/topic129.html

Hear from you soon.

Karl

[Frost]

uuh.. sorry, forgot that I asked this question. I will post the debug info asap.