Release FastTracker 2 Clone

GCC 4.9.3.. This is so old that it most likely will not generate the same code as the farnox' port of the software (which is important so that I can make the fix). Anyway, this is getting pretty complicated and difficult at this point, so I think there's nothing reasonable for me to do.
Later GCC can then be buit easily from there. It's just, the glibc used on the Pandora doesn't like to modern gcc version, that's why I stick with 4.93 there.
The integer based ARM core doesn't as far as I can find, but the A8 includes a floating point chip giving you FDIV/VDIV instructions at the very least. Takes some 30 or so cycles apparently, so isn't blazing fast, but is probably faster than any iterative approach.

Edit: More or less ninja'ed by ptitSeb ;)
If you could record the 32-channel song playing with the new -DLERPMIX compile and upload it, it would be great.

Ok what a hell of week i internet connection goes mad for all the week and plus i had a week uber full at work...:mad:

Anyway here is the recorded song (directly from the Fasttracker clone on the pandora with the option in diskop menu -> record to .wav and is the release 132 with -DLERPMIX )

the song is called Homecoming.XM and is available here

For me is a bit metallic the song in some part...but maybe is my ears, and with previous version was a bit better...i need to know your opinions about it.
Sorry, i'm an unmusical person, but for me, the wave file is missing volume, seems to me to be flatter than the xm file. This can depends on player.

Thanks for try and reply, i was looking at the code and noted that the source of version 132 did not have a reference of LERPMIX in code...odd, maybe the author did some mistake with this version..i'll retry with a more recent release.
Nobody did have time to test it?
Sorry, I couldnt be bothered to create a dropbox account again.
But could you tell me what I could do, for example, use a classic FT2 and export the wav and use the pandora (with latest FT2 port) and export the wav, then visualize it side by side in Audacity where you can import both wavs as stereo tracks?
Something like this:
Maybe you could do it...but i was requesting something more simple....just listen (with earphones attached to Pandora) the song that i linked , first with the release 130 (the last on the repo) and then with the release 131 that i shared some messages ago.
This just to prove that isn't my only impressions (of my ears) that is sound a bit better with version 130.
Audio memory is about 10 seconds so you need to play both files at the same time and switch between them to hear the difference.
I doubt the Spectrum view in Audacity will be of much use with subtle differences like a slightly more metallic sound. It would have to be very obvious to see that in such a graph.

It could also be a slightly different EQ setting between versions of fasttracker. Is there an EQ in there? If not, maybe it's hard coded and can be looked up in the source?

You could check that my way too, which involves a white noise and a 32 band spectrum analizer with hold function. :)
Ok just uploaded the new PND of version 1.01.
This time i didn't used any Pandora options on compile, as it seems (to me) that is sound more sharp without Neon or -DLERPMIX...but maybe it's only my impressions.
I compiled with:
gcc -D__LINUX_ALSA__ src/rtmidi/*.cpp src/gfxdata/*.c src/*.c -lSDL2 -lpthread -lasound -liconv -lstdc++ -lm -Wshadow -Winit-self -Wall -Wno-missing-field-initializers -Wno-unused-result -Wno-strict-aliasing -Wextra -Wunused -Wunreachable-code -Wswitch-default -march=native -mtune=native -O3

please test it and report your impressions (but please use some earphone to test), if someone want to test i have build also the pnd using classic CortexA8+Neon options and another one with old mixing routines (using -DLERPMIX).
  • Love
Reactions: rSl
Thanks @Farox grabbing my Pandora now... (but I really need to first take out the garbage and go buy linner)
  • Haha
Reactions: rSl
Diff between FT2clone_101.pnd and FT2clone_b6.pnd with the song "Space Debris"
Method used:
  • Use pnd_run to explicitly select the pnd to run
  • Use the save WAV file inside each FT2 to save the mod to wave. (101 too a LOT more time to convert, but maybe it was because of the SD cards, I have one pnd on each)
  • Load both stereo wav files into Audacity
  • Select only one of the stereo tracks and use the effect "invert" (which flips the sound horizontally)
  • Listen to the result. If wav's are identical, there should be silence
  • Merge both tracks (mix and render to new track)
  • Show a spectrogram for visual inspection
  • post images to the forums
  • Go to the supermarket because you are hungry and did this before house duty stuff (hang clothes, wash dishes, etc).

help from:

So it bleeds a bit in the high and mid frequencies, but not that much in the bass frequencies.
The song sound normal but very faint and after normalizing it sounds like with a high-pass filter. And some sounds are dimmed.


Will give an update later after listening to the wav's one by one (with that suggested 15 second interval)
Last edited:
  • Like
Reactions: rSl
Thanks for the spectrogram...but i don't know if comparing a .mod converted to wav is the correct way to test, as the sound pass thru a conversion routine.
If you have an earphone try to listen the same song (preferably one with many voices active, like some XM modules as i think with mod files there's no noticeable differences) and compare the previous release with the current.
  • Like
Reactions: rSl
tried with the home visit.xm but can not detect differences to be honest. even used alt+f4 to switch fast