Best way to convert 5.1 FLAC to stereo?

[ac16161]

Just came across some 5.1 FLAC files I'd like to play on my Linn streamer. I've done this successfully using the dBpoweramp converter tool, I selected WAV as the output type and asked for 2 channels. There are no other options there, but it seemed to work. Is this all I need to do, or should I be using one of the DSP tools for further options? Curious as to how the down-mixing is being done.

Thanks.

[Spoon]

You can control how the downmixing is done with the DSP effect 'Channel Mapper' or 'Channel Count'

[ac16161]

You can control how the downmixing is done with the DSP effect 'Channel Mapper' or 'Channel Count'

Thanks, I'd taken a brief look at those. In the absence of using the DSP's and just setting the number of channels to 2 (eg. this is an option when selecting WAV as an output type), how does the converter operate? Is there any mixing or are channels simply discarded?

[Spoon]

It mixes using best practice, however note you are squeezing 5 channels of audio into 2, there is the potential for clipping, so best to first use the DSP effect 'Bit depth' set to 32 bit floating point, then 2 channel, then volume normalize - peak to peak, then bit depth to 16 bit (with dither).

[ac16161]

Thanks. Although my main hi-fi amp is away for an upgrade, even through my PC system, adding the steps you suggested seems to create a file that sounds much more dynamic and open compared to direct-to-2-channel conversion. In practice, provided these DSP steps are loaded in order, can this be done in a single operation without having to complete four separate file conversions? I tried this with all DSP steps loaded at once and the final output FLAC file was not exactly the same size as when I proceeded with one conversion step at a time, though very close.

[Spoon]

Yes they should be done in 1 conversion, with 4 DSP effects in order.

[dexalama]

Yes they should be done in 1 conversion, with 4 DSP effects in order.

Thank you.

[ac16161]

It mixes using best practice, however note you are squeezing 5 channels of audio into 2, there is the potential for clipping, so best to first use the DSP effect 'Bit depth' set to 32 bit floating point, then 2 channel, then volume normalize - peak to peak, then bit depth to 16 bit (with dither).

Can I just check, does this method for converting 5.1 to stereo guarantee consistent volume across tracks? I was just listening to a concert I had backed up from a blu-ray source. I had split the overall file into individual files per song. On one track change there was a noticeable increase in volume. Might the volume normalize step be causing a problem here, since conversion is being done on a per-track basis rather than across the entire collection of tracks?

[ac16161]

ok, answered my own question - I removed the volume normalise DSP and the abrupt volume change on track change was gone. Is there anything else I should consider for optimising conversion?