Topic: Unexpected 24bit

[dgale]

Hopefully some Sound Forge (v.6.0) savy folks out there can explain something unexpected on one of my recent recordings...here's a recap of what happened:

I recorded a show as follows:
Sbd>R500 @16/48>RCA analog out>Oade dig mod UA5 @24/48>VX Pocket>SF 6.0

and everything worked fine...basically I was new to the UA5 and a bit distrustful that all would work out, so I put my R500 in front of it to make sure at least a 16/48 master got made.  Anyway, all worked out fine and the next night I was at a different venue and didn't have the time or space to set up the UA5 as well, so I ran:
Sbd>R500 @16/48>coax out>VX Pocket>SF 6.0

Now, here's where it got funky - when I opened a new record window in SF for the first set, I noticed it was set up by default for 24/48 as that is what I had run the night before.  So I changed the settings to 16/48 and ran the first set without problems.  For the second set, I started it up again but this time failed to notice it again had defaulted to 24/48 and for unknown reasons SF still locked in on the 16/48 signal even though it was set for 24/48...I know in the past w/SF, if I try to send it a signal with a different sampling rate then it is set for, it gives me an error message and I have to change the settings to the correct value before I can record, but I had not dealt with this in regards to bit depth before - I just assume it would do the same and give me an error message but apparently not (??).  Anyway, the recording ran at 24/48 with an incoming signal of 16/48 and it sounds fine when I listen to it.

So my questions are:
1) Why would SF lock in on a signal that is the wrong bit depth
2) What is the status of my recording - what did it do to covert the 16bit to 24?  Is this recording flawed in some way?

Not a huge deal as I have the 16/48 DAT back-up but was perplexed by this when I went to edit the file and noticed it was 24/48...should I toss the 24/48 file and transfer the DAT?  Does SF somehow convert the bit depth on the fly? 

Thanks for any info ya'll can provide.

[dklein]

A different sampling rate would cause a problem.  A different bit depth wouldn't.  I don't know SF but if you do an analysis with Wavelab or CEP you can find out the actual bit depth of the signal - I'm guessing that you never used the remaining 8 bits - they just get zero padded.  You should be able to truncate back to 16 without loss (again I don't know SF).

You can post a few second clip if you want me to look at it - flac it (not mp3).

[kindms]

I have done this samething when i used to use SF. Basically you have 8 bits of empty data on the 24bit recording fed w/ a 16/48 signal.

I cant remember SF not taking a digi signal regardless of what I fed it. I dont remeber it puking when the record rates didnt match. but that might just be my memory

[dgale]

So what happens if I dither this recording to 16 bit?  Is anything worthwhile lost or just the 8 empty bits?  Is this a simple fix or am I best off to just toss the WAV file and transfer my DAT?

[dklein]

So what happens if I dither this recording to 16 bit?  Is anything worthwhile lost or just the 8 empty bits?  Is this a simple fix or am I best off to just toss the WAV file and transfer my DAT?

You don't need to dither it because there is nothing in those additional 8 bits.  Dithering will only add noise.  Just truncate (save as 16 bit, no dither).

[trajhip2000]

I've had that happen to me too, altho more often the other way, sending a 24-bit in but only recording 16...

dklein is right, just open the 24-bit file in SF and then save it as a 16-bit file (no dither). when I was playing around with SF and 24-bit recordings a while back I came up with a way to test whether I was truncating or just removing padded zeros. I'd have to sit down in front of my SF-equipped PC to figure it out again, but I think it involved something like saving the 24-bit as a 16-bit file and then looking at the min/max binary values in the file.

Steve

[dklein]

As far as determining what you've got.  In Wavelab, turn on the 'bit meter' and watch it while you play music - it will either go to 16 or 24.  In CEP (Audition now), highlight a segment and go Analyze, Statistics and it will tell you the actual bit depth.