Master clips after SRC

[wishtheend]

Working on mastering some mixes that were tracked @ 48khz/24-bit and I'm running into a weird problem where even though in the DAW that master peaks at -0.3db, once I render out to 44.1khz there are peaks above. If I render the mix in 48khz (24-bit or 16-bit) then it's fine, but any bit of SRC and the mix clips.

Has anyone run into this or have any tips on how to deal with it? I would prefer not to SRC before the mastering chain, but I'm thinking that might be my only option since there would still be plenty of headroom.

 

[timislegend]

first: what daw are you using?

second: is there any processing on the master? (eq/limiting etc).

 

[wishtheend]

Running Sonar atm. For the mastering chain I have an eq hipass @ 30hz, a comp doing 2db GR, then a combi of saturation and limiting to get the volume up with the limiter capped off at -0.3db.

I've narrowed it down to the sample rate conversion which I'm trying to do POST mastering chain but prior to dithering. If I keep the mix at 48khz, the export never goes beyond the -0.3db but the moment I go to SRC it causes these weird overs. I exported a 24-bit version without dithering to see if I only SRC/dither in another session file but same issue. Tried Sound Forge as well to rule out if it's Sonar but same deal.

I'm going to test doing SRC first prior to the mastering chain, and then just run a 44.1khz/24bit session to see if that fixes it but I'm just really curious what the cause is.

 

[Taurean]

Running Sonar atm. For the mastering chain I have an eq hipass @ 30hz, a comp doing 2db GR, then a combi of saturation and limiting to get the volume up with the limiter capped off at -0.3db.

I've narrowed it down to the sample rate conversion which I'm trying to do POST mastering chain but prior to dithering. If I keep the mix at 48khz, the export never goes beyond the -0.3db but the moment I go to SRC it causes these weird overs. I exported a 24-bit version without dithering to see if I only SRC/dither in another session file but same issue. Tried Sound Forge as well to rule out if it's Sonar but same deal.

I'm going to test doing SRC first prior to the mastering chain, and then just run a 44.1khz/24bit session to see if that fixes it but I'm just really curious what the cause is.

If I'm understanding your chain, the SRC should be in front of your brickwall limiter to catch intersample overs from the SRC process itself. That's why you are seeing overs.

 

[timislegend]

I'm going to test doing SRC first prior to the mastering chain, and then just run a 44.1khz/24bit session to see if that fixes it but I'm just really curious what the cause is.

for the record: this is how i always do it.

izotope rx (advanced) has the best software based src i have ever used. it smooths out any rippling or time-smearing artifacts in the passband.

i would recommend it.