True 24bit recording/32bit even?

[triphop]

It seems Audacity fakes 24 bit recording by padding zeros onto 16 bit. I don't envy some of you here that spent hours or months on a large collection.

To date, does any software (free or not) allow for true 24bit or higher input recording of vinyl? Protools? Adobe Audition, Soundforge, etc?

[SteveM]

It seems Audacity fakes 24 bit recording by padding zeros onto 16 bit. I don't envy some of you here that spent hours or months on a large collection.

Only in Windows. OSX & Linux versions of Audacity work fine. There are customized Windows versions out there somewhere that record in true 24-bit.

[triphop]

Ah! Good To Know... I love Linux, but currently my laptop runs it and I don't think the line in is part of a good card. Perhaps A Live CD or second drive on the desktop is in order......

[mmarston]

You might find this thread interesting: viewtopic.php?f=52&t=45483. User rkay5 made a generous offer to share ASIO-enabled Audacity, which does record 24 bit resolution in 2
Windows...

I had dubbed a bunch of records before finding out about the 16 bit issue. I don't think I'll have the energy to re-do those, trying to get others done... I might eventually revisit a few... but I have a lot more to get done in the meantime.

Best,
Mike

[rkay5]

Hi,
If anyone wants ASIO-enabled Audacity just shoot me a PM with your email address and I will send you the software.

[Trackside]

Seeing as how the DR of Vinyl is captured within 16 bits if you can control your input levels you have to ask why 24bit is important.

[ttechnics]

Isn't the DR of vinyl somewhere around 11 bits, seeing as surface noise seems to be about 50db below the loudest passages...

[mmarston]

Isn't the DR of vinyl somewhere around 11 bits, seeing as surface noise seems to be about 50db below the loudest passages...

Most of that noise is below 400 Hz due to the RIAA curve. If you plot the spectrum of a quiet groove you'll see what I mean. A clean undamaged disk will often have noise levels below -90 dB from 1 kHz up. And much of the LF noise is periodic, relating to motor vibrations, bearing imperfections, power line leakage, etc. both in playback & at the cutting stage, as well as surface imperfections, chiefly lack of perfect flatness. I just did this with a long lead-in groove, by far the biggest I component see here is related to arm/stylus resonance at 8 Hz. Once this is filtered out, I see small spikes at 24 (probably 3rd harmonic) as well as predictable 60, 120, 180 Hz etc. power line noise. All these are below -60 dB... and a general "lump" of non-periodic stuff around -80 and lower. Nothing visible above 1k with a -90 dB floor to the measurement...

Best,
Mike

[Trackside]

Isn't the DR of vinyl somewhere around 11 bits, seeing as surface noise seems to be about 50db below the loudest passages...

Most of that noise is below 400 Hz due to the RIAA curve. If you plot the spectrum of a quiet groove you'll see what I mean. A clean undamaged disk will often have noise levels below -90 dB from 1 kHz up. And much of the LF noise is periodic, relating to motor vibrations, bearing imperfections, power line leakage, etc. both in playback & at the cutting stage, as well as surface imperfections, chiefly lack of perfect flatness. I just did this with a long lead-in groove, by far the biggest I component see here is related to arm/stylus resonance at 8 Hz. Once this is filtered out, I see small spikes at 24 (probably 3rd harmonic) as well as predictable 60, 120, 180 Hz etc. power line noise. All these are below -60 dB... and a general "lump" of non-periodic stuff around -80 and lower. Nothing visible above 1k with a -90 dB floor to the measurement...

Best,
Mike

In other words 16bit is enough

[Dimal]

24bit recording's main advantage is when post processing is being considered. This ensures that any digital processing artefacts are relegated to well below the noise floor of human hearing capability.

I know that in my own personal situation, I find it very difficult to hear anything that falls below a -60db noise floor....

Mal.

[Trackside]

24bit recording's main advantage is when post processing is being considered. This ensures that any digital processing artefacts are relegated to well below the noise floor of human hearing capability.

I know that in my own personal situation, I find it very difficult to hear anything that falls below a -60db noise floor....

Mal.

Also it means you can be a little less critical about setting your gain and leave plenty of headroom in the recording. While 16 bits is enough that doesn't mean 24bit is not more useful.

[moddude]

Hi, i
If anyone wants ASIO-enabled Audacity just shoot me a PM
Ywith your email address and I will send you the software.

Hi rkay5,

Im a newbie and just acquired a usb ext soundcard capable of 24/96k recording of vinyls. I have 300+ LPs i wano to archive properly. Would appreciate it if you can send me a copy of the
ASIO-enabled Audacity for me to try on my project.

Thanks in advance!

[Steerpike_jhb]

There is no A/D converter available to the general public that will give you 24 bits of resolution. Such things are the domain of laboratories. No pre-amps or cartridges or input stages can do better than around 20 bits. The last 4 bits are lost in noise and no matter what software or PC you use, you will never get any music information out of them.

As said, the extra 4 bits can be useful to give you wiggle room for setting recording levels.
32 bit recording is just preposterous.
The PROCESSING that goes on in the digital domain may well need 24 bits or more, to retain accuracy when doing floating-point arithmetic, but that is a very different situation compared to the recording resolution.

[David.D]

Hi,
If anyone wants ASIO-enabled Audacity just shoot me a PM with your email address and I will send you the software.

Robert, that is a very kind offer. My email addy is on the way.
As far as I'm concerned, the more bits the better
Dave