I wanted to share a small improvement in preparing files for the DCC Studio to record on tape using a DCC175 with a computer link. Basically I took the latest mp2enc by Jac Goudsmit and adapted it to Foobar in a few steps.
The problem I encountered first is that mp2enc does not work directly with Foobar2000 as a command line encoder. I guess the problem is that Foobar does not wait for the encoder to finish working. So I created a wrapper batch file with just a single line:
These four parameters are passed to the batch file and eventually to the encoder:
%1 - input file
%2 - output file
%3 - track title
%4 - artist
Unfortunately, Foobar cannot run .bat files directly, so I had to create an exe file using a standard bat to exe converter. Now what’s left is to add a new custom converter in Foobar2000. Click to see a screenshot.
The encoder file name should point to the exe file generated from a .bat. The extension is trk, and the parameters passed on to an encoder are the following:
%s %d “%title%” “%artist%”
The rest of the settings are up to you, I decided to give the resulting trk files short names (track numbers). I just tested the output with DCC Studio and it is fully working, with artist and track titles displayed properly. With this tweak I can generate DCC files from any music I have, regardless of the format, in one click.
My knowledge of Foobar is quite limited, and there could be a more elegant solution, but here it is.
It’s great to see that others are interested in doing stuff like this.
From the looks of it, you probably wouldn’t need a batch file compiler if it would be possible to simply specify an input file on the mp2enc.exe command line. I’ll try to find some time for that soon.
By the way, it would be great if you would put your screenshot in the forum instead of imgur, so that it doesn’t get lost over time.
One additional observation. MP2ENC properly processes 24bit WAV files and preserves this extra data while encoding audio on DCC. I tested both DCC2WAV and MP2ENC, and also recorded RMAA test signal on a DCC951 deck via SPDIF. The graphs from RMAA below represent the output captured via the analog out of a DCC951 (unfortunately I cannot capture the output digitally)
DCC2WAV processes only 16 bit WAV files, and this is clearly seen when measuring dynamic range - the noise level is higher. MP2ENC and DCC951’s own recording indicate 18bit processing. But there is more. When testing the frequency range using a multitone signal (the hardest case for any lossy compression) I see that MP2ENC produces a different result, compared to DCC951’s own recording.
While DCC951 itself cuts off the high frequencies at 18.5Khz, the frequency response of MP2ENC extends beyond 21Khz. Same result (21Khz cutoff) is shown when using DCC2WAV. Apparently, using computer encoding offers somewhat (most likely not audible) better quality, or at least a more recent compression algorithm, compared to recording on a DCC951.
mp2enc uses 32 bit floating point internally so your observations are correct: It handles 24 bit sample files correctly and (probably because it uses a more modern acoustic model) generates less noise and better sound.
I just uploaded a new release of mp2enc.exe which takes a new command line parameter “-i” to specify an input file. That way you probably don’t have to use a compiled batch file anymore.
