posted 02 September 2003 11:14            

I was just reading the Dolby documentation on mixing 5.1 surround sound, and their description of the LFE (subbass) channel using a lowpass filter at 80 Hz to separate the deep bass sound effects from the main stereo tracks. They record the LFE track 10 dB below its actual level, and leave the playback system to boost the audio back up. That gives them a lot more headroom.

This seems like such a good idea, taken one step further. Namely, when mixing CD stereo disks, send the audio channels through a low-shelf -10 dB filter before burning the CD, then have the playback system boost back up in a complementary fashion.

Doing this restoration in analog leaves the digital domain free to exploit a whole lot more headroom in the mix.

So my question to you all is, is this already routinely done by recording engineers for stereo CD's? Or do they actually burn the bass levels that are present in the source sound. Going through a low shelving filter after the mic pre's would save headroom in the mix tracks, and the monitoring system would compensate.

Cheers,

posted 02 September 2003 13:29            

With vinyl records that is exactly what happened. Ever heard of RIAA filters. Early DAT recorders had an emphasis option and a digital flag that went along with it. But it seemed that because digital recording was so superior (technically) and relatively hiss free in comparison to analogue tape and vinyl records, that most Dat users decided that emphasis wasn't necessary. Also there was no guarantee that all the manufactures would have identical emphasis filters and in those days every one believed that digital was perfect. So eventually no one used emphasis just to standardize things. Similar to the old BETA v VHS for home videos.
CDs where then just copies of those dats.

I think this is what happened, but if anyone knows different please say.

posted 02 September 2003 20:31            

Thanks for the replies. I understood the old RIAA and tape preemphasis were to overcome the problems of noise in the highest frequencies, not so much to gain headroom in the bass. Am I wrong about this?

Anyway, I see your points about the failure of reproduction systems to properly undo the prefiltering. I guess that is all the more reason to need 24 bit recording, to gain an extra 8 bits of headrooom plus noise immunity.

Past reading about the big-guns consoles, like the Neve's and others, indicate that they like to run at least 30 dB of headroom for transients (plus bass room). My dinky A&H 3300 has about 16 dB of headroom. But digital 16-bit formats typically use only 10 dB of headroom -- just as with the Capy. That never seemed like enough to me.

Cheers,

posted 03 September 2003 09:27            

What you should realize is that "noise at the high frequencies" and "not enough head room" are one of the same problem but just at different extremes. If you record a signal at high level you have very little head room but a good signal to noise ratio. If you record the signal at low level you get a poor signal to noise ratio which makes hiss (mainly at high frequencies), but a lot of head room. In most signals the low frequencies are of a greater amplitude and more likely to hit distortion. So attenuating the low frequencies relative to the high frequencies gives you ether, a better signal to noise ratio or, more head room, depending no which way you want to look at it.

does this make sense?

posted 03 September 2003 10:28            

In analogue mixers it was relatively easy to produce a large dynamic range between the noise floor and the point of distortion, but not so with analogue recording devices. One of the main jobs of the mixer was to get just the right level on tape. If you were to gave yourself 30 db head room on an analogue tape you would hardly hear to signal as it would be buried in hiss. The 10 db of head room on digital tape was like shear luxury.

Then they started making digital mixers and more care had to be taken at the point of entry into the mixer. Now 24 bit devices are common place and maybe we should give our self's more head room but until we all have 24 bit dat recorders and 24 bit CDs it is probably better to stick with 10 db head room .This way that we can record straight to these devices digitally without having to readjust the dynamic range.

But one day this might all change, I hope.

Pete.

posted 07 September 2003 10:18            

""I understood the old RIAA and tape preemphasis were to overcome the problems of noise in the highest frequencies, not so much to gain headroom in the bass. Am I wrong about this?""

No. But there's more to it. With vinyl, to be able to cut a laquer master with any decent length to the album side, you need to keep the pitch, (distance between groove centers) small. The cutting needles excursion would be far too wide, (for in-phase audio) or worse, far too deep, (for out-of-phase audio) in the bass freqencies. So, if you cut the master without the pre-emphasis then the high frequencies end up in the noise floor, and you'd only get a few minutes on the record.