Hey now!
Welcome
to rosemary.dyndns.org
2/5/7: Thanks to zep— for the insight: you don’t need to install another otheros.bld; check the ps3-sdlmame-readme for current details. BTW, sticks & front-ends are much better now than in earlier releases.
2/2/7: Also, you need to install a different otheros.bld. Here’s one to replace your existing one: otheros.bld
1/30/7: The ps3 “sixaxis” joypad works (albeit crazily sometimes; needs some improvement). Also, I hacked together a front-end (psmallphront). And I wrote some documentation, which I hope answers your basic questions:
1/27/7: Been getting mame running on sony playstation 3, which I view as a blu-ray player with brains. I’ve got video & sound working on a self-contained, bootable linux .iso that boots up & runs one game- robby the roto (video boots into 480p, btw). It needs sixaxis support, a front-end, etc. As I’m mid stream in development, sorry for not responding to emails (yeah, I suck). I’ll offer it as a torrent if there is sufficient interest, but the latest work-in-progress will keep appearing here: ps3-mame-cdr-test.iso.gz
11/2/6: Got a win32 GUI working for resample-1.8! Its a quick kludge for a compression ui tool called “multi-frontend” (as found here http://members.home.nl/w.speek/multi.htm)- along with committing some minor atrocities inside resample-1.8. New versions of resample binaries & source are posted @ the original links below. I am contacting the multi-frontend developer to see if he would add appropriate resample command line handling.
So now you can drag & drop files into the front-end window to enqueue for resample processing, much like flac w/ its frontend. Here’s how:
After you download & install multi-frontend, start it up. Then select the “Codec:” FAAC (other Codec settings may in fact work too, btw- but I know this one works). Into the “Options:” field (just below “Codec:”), enter the text:
-to 44100 –r
Next, click the “Locations” button in the right-hand button column. For the “FAAC” location, instead of faac.exe, specify the (new) resample.exe from where you put it.
Queue up some files (drag & drop, or click the Add button at the top of the right-hand button column). Then hit the “Go” button (bottom of that right-hand button column). Kick back and watch the dots, then notice the generation of one *-16-44100.wav for each file queued.
11/1/6: I’ve been hacking on softwares lately. I modified a cool open-source audio resampling command-line tool by Julius Smith at Stanford University’s Center for Computer Research in Music and Acoustics; thought it may be of public interest – “resample-1.8” – original found here:
http://www-ccrma.stanford.edu/~jos/resample/Free_Resampling_Software.html
Here is my modified "-r" win32 .exe zip
Here is a recent version of win32 cygwin1.dll (zipped for less bandwidth), which you may need.
Here is the source tarball (see source code notes below) modified "-r" source code .tar.gz containing my kludges.
Resample-1.8 converts a recording made at one sampling rate to another sample rate. Julius Smith’s site has a white-paper discussion of the mathematics, which will be of interest to those who wonder about an algorithm for this operation.
The added functionality of my modified version is to permit sample-size/bit-depth adjustment, specifically to map from 24-bit samples to 16-bit samples (& “get the endian of win32 cygwin compilation to be valid”, etc.).
The new “–r” semantics are: map (truncate) to a 16-bit sample size, force the output to be of “RIFF” (Microsoft pcm wave) format, and suppress the generation of a “clm” riff section (This clm section normally contains text which describes the resample operation. Flac complains unless the file is ‘canonical’ with respect to having no ‘extra riff’ sections, as this clm chunk is.).
Usage is as per:
resample.exe –t 44100 –r source.wav dest.wav
I use this to batch process recordings @ 24/192 from my sound-devices 722 down to 16/44.1 (which makes them usable for playback on our home slimdevices squeezebox network audio player, flacing, and getting closer to cd-format files). The results is music to our ears ;)
Your mileage may vary wrt this advice, but fwiw: if your recording levels are low, you should “normalize” or amplify the signal *before* processing. This is to ensure that you ‘move the significant audio data’ into the ‘higher-order bits’ of each sample. As long as you’ve got at least 16 bits of significant data at the peak-case sample, the end result should sound crispy.
I (or somebody else?) may put together a more graceful interface (based upon the open source “flac” user-interface) in days to come, make it a little more user friendly.
Note that I’ve only tested with 24-bit -> 16-bit redux. Also, you may wish to run shntool to obtain proper red-book audio cd-frame sized files (no clicks between tracks).
Source Code Notes: Resample-1.8 uses sndlib for audio file i/o. The 24->16 kludge is gross but effective. Interface changes through the call tree would be nicer (i.e. adding an argument to specify target sample bit depth, instead of the parallel *_24_to_16() functions I added). I was stumbling around somewhat randomly through the source code, so there plenty of kludges. Like I forced the LITTLE_ENDIAN constant to be true- if you’re compiling for another endian, you’ll want to fix this in the sndlib.h
I am contacting Julius about this functionality- please direct comments about this ‘-r’ version to phopely@seas.upenn.edu