Cubase SX 2 cuts high frequencies?

[SJ_Digriz]

gratz on a good post cookie. The EP article was BS when it was printed and is now ancient news.

Dithering always sounds better than truncation without dither.

I'd like to point out something in that quote up there. The statement is refering to 24 bit signals going to 16 bit. You don't need to dither if you aren't changing bit depths.

And, I have heard a zillion exceptions to that "Always".

[el davo]

don't loose one more night's sleep over the implementation of digital summing in cubase. put your hard earned cash in one (or more) analog summing device, e.g SPL Mixdream. Hardly overkill for the avg punter.

http://www.barryrudolph.com/mix/strictlysumming.html
http://www.spl-usa.com/

If you're a producer/writer and you can hear the difference, your mixes are likely to sound fkn fantastic already. Spend more time on the music instead.

[tibenmik]

I just want to jump in to ask a question re: Cookie's comments on limiting:

I recently used SX2 to complete an 'audio reel', a composite of about 15 to 20 songs/tracks in a 3 minute sound collage.

First off, each song/track began as a 44.1/32bit multi-track in SX2, where I determined the final mix, set the overall RMS at about -14dBFS, threw a -0.1dB limiter on the master outs, and exported a mixdown to 44.1/24bit.

These tracks were then imported to Tracktion v1 and arranged to fade into/out of each other. I then threw a HP filter at around 36Hz and a limiter at -0.1db on the master outs, and exported to 44.1/24bit.

Finally, I went back to SX2 again to take advantage of the UV22 dither algo. I simply imported the audio track made in Tracktion, kept the track and master fader at unity, threw the UV22 on the last post-fader insert, and exported to 44.1/16bit.

My concern is the use of limiting in each step. Am I correct in assuming that if I exported to 32bit in each step, then limiting is redundant as both hosts can deal with signals over 0dB due to the almost infinite dynamic range of 32bit? I'm considering re-exporting everything and only using the HP filter and limiter at the very last step, when rendering the Tracktion audio track to 16bit.

I appreciate anyone's thoughts on this.

_mt

[IIRs]

Yes, better to stick to 32-bit all the way, with only the one limiter at the end of the chain. That said, if you used 24-bit intermediate files, and the limiters were not applying much gain reduction (ie; just catching the odd peak) then you probably don't need to re-do it.

[tibenmik]

Thanks for the comment, PE. This makes me wonder: while a signal can not exceed 0dBFS in a 16bit file without audible distortion, how far past 0dBFS can a signal travel in 24 & 32 bit files before detrimental noise/distortion is audible?

_mt

[IIRs]

Thanks for the comment, PE. This makes me wonder: while a signal can not exceed 0dBFS in a 16bit file without audible distortion, how far past 0dBFS can a signal travel in 24 & 32 bit files before detrimental noise/distortion is audible?

_mt

24-bit will distort as badly as 16-bit if it clips. 24-bit provides better resolution for low level signals (allowing you to leave more headroom) but it doesn't actually increase the headroom itself. Its only 32-bit float that provides extra headroom (AFAIK more than you are ever likely to need.)

[tibenmik]

Sorted. Thanks again, PE.

_mt