Mod Archive Forums
Music Production => MilkyTracker => Tracking => MilkyTracker Tracks & Songs => Topic started by: qnio on July 26, 2007, 06:18:21
-
hello.
i notice a huge decrease in my .wav sample's sound quality (they're PCM uncompressed, 44100khz, 16bits, mono samples, i.e. a kick or a snare) when creating patterns in milkytracker or rendering them to .wav
i'm using ASIO4All as driver, 35ms buffersize, interpolation & volume ramping are checked, 44100hz as mixer resolution and linear frequencies.
is it milkytracker or my configuration?
thanks for reading
-
can you post links of audio examples for check?
-
you can attach small zip files of clips to your forum post. max 512kb
click 'additional options' at the bottom of the window when replying
-
ok, here's 2 .wav files: one rendered in milkytracker and the other on ableton live.
the pattern, as you might easily recognize, is a simple amen break chopped up in audacity. the same samples are used in both "sequencers" (since milkytracker is... a tracker), with no extra effects. all slices are PCM uncompressed, 44100khz, 16bits, mono.
-
ok I dont have here decent speaker system for check.. and yes, I notice that the volume is different on both versions, but not the quality
Try to render the milkytracker wav file using higher amp setting (200% or 400%)...
Im not sure if ableton do a render applying a normalize.. or if it have an autogain feature while export..
Anyway there is no audio downgrade... its just a matter of volume... so you can try applying normalize with an external audio editor to the milky rendered wave and compare again with ableton output
-
now they're at the same volume (i set the volume to 4x when exporting on milkytracker), and i still hear a difference on the quality
here are the samples
-
The milky recording has less high end frequencies thus sounds flatter and less brighter than the Albeton one. So to confirm, something is going on here.
-
The milky recording has less high end frequencies thus sounds flatter and less brighter than the Albeton one.
exactly.
i picked ableton just as a kind of "generic" DAW, for comparison purposes.
the goal of this thread is, since i like milkytracker a lot, to verify (and correct) this, IMO, flaw; thus making the software better.
thanks for reading. i'll be checking this topic
-
Did you try without interpolation? Interpolation is a mathematical operation which acts as a low-pass filter. It could (on least cases) cut freqs above Samplerate/4.
-
The Ableton Live sample is louder and exhibits more clipping than the Milkytracker sample. I suggest you compare with the original WAV instead of Ableton's rendered output. As well as the high frequency attenuation, Milkytracker is playing the sample slightly slower than Ableton - but once again we need to compare it to the original.
Also try disabling linear interpolation as jimee suggests.
-
MilkyTracker adjusts the relative note number and the finetune of the sample according to the sample rate of the WAV but it's only an approximation. So when milkytracker plays the sample at the given pitch there might be some accuracy difference for the sample rate resulting is slightly faster/slower playback thus causing the interpolation to miss certain frequencies in the original sample.
-
problem solved :)
here's the original kick sample, the same rendered through milkytracker WITHOUT interpolation and with interpolation.
the first two now sound the same.
thanks for the replies.
-
I had the same problem.
This thread helps :D