Creating 16/44.1 audio file

[midnightsun]

I want to capture the stereo output from my analog console at 16/44.1 to compile a redbook CD. I want to use my Cranesong HEDD for A/D and send a 16/44.1 digital signal via SPDIF. Then capture the signal on my Macbook Pro in a PT11 session at 16/44.1 Sounds simple enough.

However, I am using a Tascam US-366 at my interface, i.e. SPDIF from HEDD to US-366 and USB out to the MBP. The specs for the US-366 state that it supports 16 and 24 bit but internally runs at 24 bit. Therefore, I presume that the TASCAM US-366 is taking my 16 bit signal and jacking it up to 24 bit and then Protools is truncating the 24 bit signal back down to 16 bit.

Any thoughts are greatly appreciated. It maybe that I have a misconception regarding bit depth. I certainly don't want to this much trouble if my bit rate gets changed back and forth.

[cwsand]

Originally Posted by midnightsun

I want to capture the stereo output from my analog console at 16/44.1 to compile a redbook CD. I want to use my Cranesong HEDD for A/D and send a 16/44.1 digital signal via SPDIF. Then capture the signal on my Macbook Pro in a PT11 session at 16/44.1 Sounds simple enough.

However, I am using a Tascam US-366 at my interface, i.e. SPDIF from HEDD to US-366 and USB out to the MBP. The specs for the US-366 state that it supports 16 and 24 bit but internally runs at 24 bit. Therefore, I presume that the TASCAM US-366 is taking my 16 bit signal and jacking it up to 24 bit and then Protools is truncating the 24 bit signal back down to 16 bit.

Any thoughts are greatly appreciated. It maybe that I have a misconception regarding bit depth. I certainly don't want to this much trouble if my bit rate gets changed back and forth.

Interesting question. Seems to me that taking a 16 bit signal and bringing it up to 24 bits is not really going to change the resolution of the outputted data stream, so by dropping it back down to 16 bit you shouldn't lose anything. The original audio was quantized at 16 bit, so raising it to 24 bit shouldn't change the quantization. It'd be nice if someone could confirm this.

[Eric Seaberg]

Yup, you're only adding 8-zeros to the end of the data word and they will 'fall' off when you go back to 16-bit. Actually, there's no need to dither, either.

[midnightsun]

thanks for the input.

[midnightsun]

Originally Posted by Eric Seaberg

Yup, you're only adding 8-zeros to the end of the data word and they will 'fall' off when you go back to 16-bit. Actually, there's no need to dither, either.

Eric, regarding no need to dither. Are you saying that if I go from 16 to to 24 bit and back to 16 bit there is no need to dither (presumably because the math in the truncation process yields no garbage)! Or are you saying that you just don't do dither at all (I have heard folks saying that dither really isn't necessary, but that has not been my practice).

A few years back I did many hours of A/B ing and found that the best sounding redbook CDs came from A/D to Cranesong HEDD > 16/44.1 > to CD burner (Tascam DVRA1000). Now my workflow has changed and I want to burn my demos in a separate DAW using Wavelab. I am after the same quality but want to ditch the DVRA1000. I have been doing some comparisons and am still undecided. I want to figure this out and move on. I am just not wanting to unwittingly add garbage by going back and forth with bit depth. I wonder if I would be better off sticking with 24 bit all the way through until burning the CD.

[Eric Seaberg]

Yes, you should dither if you're going from an ORIGINAL 24 bit file to 16 bit. However, in the scenario you're talking about, there isn't any data in the last 8 bits. You're taking a 16 bit file, inputing into a 24 bit recorder and adding nothing, then spitting it back out to 16 bit. If you're not editing it or making any changes to the file, there isn't any reason to dither.

However, if you imported the 16 bit file into a 24 bit ProTools session and edited to made changes to it, then you would want to dither on the way out. In this case, you've added new information to the last 8 bits of the data.

If you started with a 24 bit file, and edited it in ProTools (or your DVRA1K, which I have and LOVE), dithering should ALWAYS be done at the last possible moment and only ONCE.

Can your Cranesong not operate at 24 bit? If nothing else, even if you're only doing a CD now, you should make your master mix the HIGHEST resolution possible. You never know what you may do with it later, i.e. high sample rate/high resolution downloads.

[midnightsun]

Eric, thanks for the thoughtful response. You clearly answered the question for me. Yes, my Cranesong will do 24bit. However, what I am trying to accomplish is to make quick demos that sound good. Turns out that the Cranesong at 16/44.1 sounds better than most converters at 24/88.1 I would take my 16/44.1 from the Cranesong and feed it to the DVRA1000 to print a redbook CD and I was totally happy with the result. For higher resolution 88.1 or 96, I just bounce to open tracks in the PT session.

I recently picked up Wavelab 8 and have it on a second DAW. I have been feeding my final audio from PT out D/A and into analog EQ and Dynamics into the Cranesong and printing into Wavelab 8. I was erroneously thinking that I should print 16 bit Wavelab; thus, avoiding dither. However, I am making adjustments in Wavelab 8, so I am just going to print to Wavelab8 at 24 bit and dither as I convert to 16 bit on the way to my final CD print.