16 vs 24 bit for vintage recordings, plus another question

otrfan123

New member
Hello,
I have a ton of vintage material to transfer and restore from open reel tapes, mostly 1930's & 1940's radio broadcasts. I am running my open reel deck through an old HHB CDR-850 cdr recorder (into the analog RCA jacks and out of the digital SPDIF), into CEDAR Duo Declickle & Auto Dehiss units, then into a computer dedicated to recording with the sound card set to 16 bit and 44.1khz. There is no setting for bit or sample rate on the HHB CDR-850. My question is, am I recording in true 16 bit 44.1khz? I had assumed I was as all of the material appears as 16 bit 44.1khz on Adobe Audition while restoring, however it just occured to me that by running it through the HHB CDR-850 first, I may be transferring the signal to something else (a smaller sample rate), then onto the CEDAR and then finally a 16bit 44.1khz sample. I am just using the HHB as a analog to digital converter and there is no way to set the bit depth or sample rate on the HHB.

My second question is: is it necessary to record vintage audio in 24 bit 96khz for audio restoration? Most of this stuff has no signal above 9khz and very rarely does the finest material ever reach 15khz. Is 16 bit 44.1 khz suitable for restoration of this sort of material? Does it make sense to use 24 bit 44.1khz? Any suggestions? Anything above 16 bit 44.1 khz and the files get absolutely huge.

Thank you very much for your help and opinions,
George
 
44.1 is fine, but 24 bit might be useful. You may not have small s/n ratios going on those tapes, but convert it over with plenty of headroom, and you'll be fine.
 
First off -- I have to admit, I've never heard anyone using Cedar units worrying about disk space... Or using Adobe Audition. But I digress.

If the capture is "final" then capturing at 16/44.1 is fine. If there's even a slim chance that anything else might ever be done to it, capture in 24-bit.

But Cedar, man. You're using CEDAR... I'd almost feel like the Cedar folks would come out and repossess the hardware if I were capturing at less than 96kHz (and don't get me wrong - I'm a big fan of 44.1kHz).
 
Whilst I would always advocate 24bits (44.1kHz 48 if going vid, no point in anything else* IMHO) for "live" tracking so that, as has been said, you can keep a good headroom and low noise floor, it is not really that important for duplicating low dynamic range material.

Then there is the general rule that it is better to keep the same sample rate/word length throughout a process if you can to minimize conversions.

The 16/24 issue really only matters when we are talking cheap converters that do not have the 24bit option. In these cases, USB mics and mixers, the ANALOGUE side of the gear is often so cheap that it is not up to the DR of 16 bit conversion leave alone 24bits!

To give an example. The (nonetheless useful and cheap!) Behringer UCA 202 only manages a record/play dynamic range of some 83dB when in fact 16bits should give us over 90dB. Not that this matters diddly for tape and vinyl! BTW I would use 16bit for tape duping but 24bits for vinyl, recording well down at -20 even -22dBFS and giving plenty of room for the clicks and bangs.

Re file size. I would invest in a big network drive. I bought a 2TB NAS a couple of years ago for only a little over £100. Should plug into the back of your modem/router and be available to all your computers.

*96kHz MIGHT be needed for certain plugins or processes.

Dave. (You beat me to my bllx on the the qwerty John! CEDAR? Yes! caused me to pause!)
 
I think you'll be better off using 24/96k quality for this kind of work, even if the source material is 16/44. If you are really so concerned about space, you should buy a cheap external hard disk just for the project, I personally wouldn't use anything less than 24/88 WAV quality for this kind of work.

Ben.
 
Hello,
Thank you for the responses. I get the comments about the CEDAR, it's a long story and I was able to get these as part of a project for a steal - typically this sort of source material wouldn't deserve this processing. I use Adobe Audition spectral display mode to view the output to make sure there were no defects in the transfer before sending it along.

So using my HHB CDR-850 as a A to D converter doesn't effect the bit depth and sample rate, correct? Only the final capture rate matters, correct? I just want to make sure I'm not changing bit depth or sample rate before the final capture.

All of this material is mono so there is no mixing to be done, just potential future restoration. Until recently I assumed 16bit 44.1khz was fine for future restoration. So much of this stuff is a couple of generations removed from the original source material that it seemed like a waste to go 24bit 96khz. Everything I'm transferring was recorded before 1958 and there is rarely any signal above 9khz and nothing above 15khz.

Thanks,
George
 
Since you are using th HHB as the converter, you have no choice but to go 16/44.1. If you change the bit depth of the files later in the chain, you arent accomplishing anything other than adding eight more zeros at the end of the 16 bits that you captured.

Since everything after the HHB is a digital connection (Im assuming), you can't change the sample rate anyway, otherwise the clocks won't sync.

I do agree with your thinking. With really limited dynamic range, no real need for a large bit depth and with a really limited frequency response, not much need for high sample rates.

The only real argument in this situation for higher bit depth would be to push the processing rounding errors down farther. But even that depends on how quiet you really get these recordings.

But, if the HHB is your only ad converter, you have no choice but to be 16/44.1, so the discussion is moot.
 
"The only real argument in this situation for higher bit depth would be to push the processing rounding errors down farther. But even that depends on how quiet you really get these recordings."

Yay Jay! I have just had the thought...Even the very best FM receiver on the very best transmission would only deliver a noise level of 70dB below peak deviation and even the cheapest, nastiest 16 bit device is going to be at least 10dB better than that. Anything coming off MF transmissions is going to be far,far worse. So IMHO 16 bits at 44.1kHz is overkill! (there is the caveat that certain processes might need 96kHz to work)

Dave.
 
The HHB CDR-850 has excellent converters and is a bit different animal than plain 16/44.1. I've owned this model for many years. The 1-bit ADC's are described by HHB as being equivalent to 20-bit although there is no direct comparison to how standard 16-bit CD converters functioned at the time. They are also 48kHz. There is only a conversion to Red book 16/44.1 if the data is written to CDR. The signal passes through at 48kHz to the SPDIF outputs just as it came in. This unit was built for HHB by Pioneer and is based on the pioneer PDR-555RW, which is still an awesome consumer CD recorder compared to today's standards. There is also a Fostex unit, Model CR300, also built by Pioneer and internally identical to the HHB CDR-850.

I love this Burner and would not hesitate to use it for archival purposes, even stand-alone to the internal CDR as well as just the converters.

The sound card on your PC is doing the 16/44.1 conversion. But I don't know anything about the Cedar... whether it has an auto SRC or it's passing the sample rate as well.
 
Why would anyone design a cd burner with a fixed 48k sample rate? Conversion normally does more damage than just starting out at 44.1k to begin with.

If the connections are all digital, it must be the cedar that is doing the conversion. Stock computer sound cards will not do on the fly sample rate conversion on a digital signal.
 
Why would anyone design a cd burner with a fixed 48k sample rate? Conversion normally does more damage than just starting out at 44.1k to begin with.

If the connections are all digital, it must be the cedar that is doing the conversion. Stock computer sound cards will not do on the fly sample rate conversion on a digital signal.

According to the manual they didn't!

"24-bit A/D conversion with dither on recording
S/PDIF digital input and output (coaxial and optical
TOSlink), with full DAT track ID and PQ-code recognition/
transmission, and SCMS defeat
AES/EBU digital input and output
Switchable dither on digital inputs
Word clock input
Built-in sample-rate conversion (32-96kHz) on inputs"

Dave.
 
Thank you all for the responses. I read the manual and the only mention of the bit depth is "1 bit", which didn't make sense to me, so being a CD recorder I just assumed it was converting the analog signal at 16 bit. Now, if the HHB is converting at 24 bit, and my soundcard is capturing at 16 bit, do I have a problem with dithering? Would my sound card automatically perform dithering or should I set my sound card for 24 bit and use software to perform the dithering later. I'm surprised a CD 24 bit CD recorder wouldn't have an option for 16 bit conversion.

George
 
Why would anyone design a cd burner with a fixed 48k sample rate? Conversion normally does more damage than just starting out at 44.1k to begin with.

If the connections are all digital, it must be the cedar that is doing the conversion. Stock computer sound cards will not do on the fly sample rate conversion on a digital signal.

I agree generally speaking that down or up conversion should be avoided and I prefer resampling in most cases. But the HHB CDR850 sure does sound good, so I can't really take issue with why they did it that way. But it was designed to convert anything from 32kHz to 48kHz for standard CD 44.1. 32kHz was a standard for TV satellite data and 48kHz as most of us know is the standard DAT sample rate. The service manual is out there. You can google it and find it pretty easily. And the specs for the converters are all in there.

The next model, the CDR830, had even broader capabilities and the converters were 24-bit. Tell you the truth there is so much crap out there, mostly theory, and so much of it outdated, about how some ideal converter works that I don't think any of us really know it all. There's so much in the design that people don't even talk about because so few have a working knowledge of what goes on in between A/D conversion to writing to the physical CDR media and back to analog through D/A conversion. And if someone is just using the converters and not writing to the CDR its a whole different path from input to output depending on the design.

I don't get it all myself and I couldn't begin to design one of these things, but I do know so much of the abstract theory people throw around on these forums doesn't get to the brass tacks of what's actually going on in a specific physical device because there is no one way that every manufacturer builds converters.

So even a question like 16 vs 24-bit or 48 vs 96kHz sampling rate can't be easily answered. It just begs the question, "What specific product are you using?" Theoretically every DVD audio recorder should sound better than any CD recorder, but if you bought it at Wallmart for $99.96 does it really sound better than something like the HHB CDR-850, which was designed for professional studio use and listed for something like $1500.00 in 1999? Not likely.
 
24Bit is a must for vintage recording I believe. I have some LPs and 24Bit is far better than 16Bit in this case

Whilst I agree that 24bits is better for vinyl there is really no point in better than 16 bits for tape.

A few years ago I was asked to copy my daughter's 45 collection of punk to CD since she no longer has the means to play them (at 43 you would have thought she would have grown up?!!). I cleaned the discs,over 50 of them, as best I could but they still had horrendously loud clicks on them. To stop these driving the converters into overload I set my average level to about -24dBFS iirc and ran at 24 bits 44.1kHz.

Once I had a batch in the can I ran them through Sony Soundforge de-clicker (which you can "borrow" once, for a month) and then exported the result as 16bit .wavs for burning.

I still have the original 24bit files somewhere but then I am not hard up for storage but once done there is no merit in keeping vinyl files as 24bits if you ARE light on disc space.

Dave.
 
Back
Top