Become a Site Supporter and Never see Ads again!

Author Topic: V2 levels  (Read 12848 times)

0 Members and 1 Guest are viewing this topic.

Offline Lil Kim Jong-Il

  • Trade Count: (6)
  • Needs to get out more...
  • *****
  • Posts: 6498
  • large Marge sent me
Re: V2 levels
« Reply #45 on: November 18, 2004, 10:15:35 AM »
Writing it down is a good idea.  I did that for a long time and I am usually pretty close on first setting.  But lately I've been hitting some different bands/venues and that has sort of thrown me.  A good example is the Landmark Mule show.  I was still set at the gain I used for the 930 shows, and then once they started, playing I knew I was going to have to drop the step gain. I did and my trim ended up at +4.  So I was off by just 1db.  It wouldn't be an issue if I wasn't always trying to use the full range but I like to get as hot a signal as I can.  I can't wait to go 24bit, then I'll be willing to run more conservatively.


Ben, are you from the midwest?  I met a taper Ben who runs B&Ks when I was on TOO tour in a few years back.  I think it was around Cleveland.
The first rule of amateur neurosurgery club is .... I forget.

Offline caymanreview

  • Trade Count: (22)
  • Needs to get out more...
  • *****
  • Posts: 9849
  • Gender: Male
    • shows ive taped  that are in circulation
Re: V2 levels
« Reply #46 on: November 18, 2004, 10:24:58 AM »
Ben is from illinois and runs 4022's

Offline BC

  • Trade Count: (1)
  • Needs to get out more...
  • *****
  • Posts: 2269
  • Gender: Male
  • Bongo Bongo
Re: V2 levels
« Reply #47 on: November 18, 2004, 10:33:05 AM »

Ben, are you from the midwest?  I met a taper Ben who runs B&Ks when I was on TOO tour in a few years back.  I think it was around Cleveland.


Yep, that was probably me, I taped TOO 11-29-02 Pittsburgh -> 12-2-02 Rosemont with my 4022's (I remember that was a week after getting my V3!) .
In: DPA4022>V3>Microtracker/D8

Out: Morrison ELAD>Adcom GFA555mkII>Martin Logan Aerius i

Offline Lil Kim Jong-Il

  • Trade Count: (6)
  • Needs to get out more...
  • *****
  • Posts: 6498
  • large Marge sent me
Re: V2 levels
« Reply #48 on: November 18, 2004, 10:42:51 AM »
Yeah, I think I walked in with you at the Cleveland tapers entrance and then saw you at some of the following shows.  It was freaking cold waiting to get into the Cleveland show.  I was running the AK50 and BK4023 rigs.

Cool.  Small world.
The first rule of amateur neurosurgery club is .... I forget.

Offline Todd R

  • Over/Under on next gear purchase: 2 months
  • Trade Count: (29)
  • Needs to get out more...
  • *****
  • Posts: 4901
  • Gender: Male
Re: V2 levels
« Reply #49 on: November 18, 2004, 12:36:15 PM »
It wouldn't be an issue if I wasn't always trying to use the full range but I like to get as hot a signal as I can.  I can't wait to go 24bit, then I'll be willing to run more conservatively.


If this is why you're running hot, I think you can drop back even now at 16 bits and run a little more conservatively.  Remember, you aren't really running at 16 bits, the V3 is running at 24 bits and then dithering down to 16 bits. Even if you aren't running hot enough to use all significant bits of the 24 bit signal (ie., peaks less than -6db), after dithering you will be using all 16 significant bits.  You'd have to run the V3 so low that you were only using 15 significant bits out of 24 in order for you not to be getting all 16 bits after dithering.  (Maybe you'd need 17 bits of info to accurately capture 16 bits via a dithering scheme, I'm not sure.  Regardless, you'll need more than a few bits less than the full 24 bits before you run into problems with not using the full 16 bits after dithering.)

So if you want to run hot so you can listen to it hot without normalizing, then keep running as hot a signal if you can.  But if you're worrying about getting the full range and maximizing your signal to noise ratio at 16 bits, then you can back off a bit and run more conservatively.  (Since you're dithering down from 24 bits, you'd have to be running at less than -24db peaks in order to not be getting the full 16 bits of info.) 

Of course, this is all in the theoretical.  From a practical standpoint, it might be that the ambient noise level of the venue -- given the noise in the PA system, the noise of the HVAC, the noise of the crowd, etc -- is such that you don't even need a full 16 bits of info to maximize your signal to noise ratio.  That is, given the high noise level of the system we're recording, we may well not even need 16 bits to capture it.
Mics: Microtech Gefell m20/m21 (nbob/pfa actives), Line Audio CM3, Church CA-11 cards
Preamp:  none <sniff>
Recorders:  Sound Devices MixPre-6, Sony PCM-M10, Zoom H4nPro

Offline dklein

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1184
  • Gender: Male
Re: V2 levels
« Reply #50 on: November 18, 2004, 01:19:53 PM »
I kinda got lost in there but I think you do need to go close to full scale even if you only want 16 bits.  When going from 24>16 it's the least significant bits that get dropped.  So if you're down 12 dB, you're at 22 bits on a 24 bit scale, or 14 bits on a 16 bit scale.  Maybe dithering can push you up a bit, but I don't think you can say hey, we've got 22 bits so there should be at least 16 available.  With 16 bit on this box, you're using bits 9 thru 24 and throwing away the first 8 (ignoring dither for the moment).

Did that make any sense?

Now if you took the 24 bit signal that was down 12 dB, transferred it, normalized it and THEN dithered it down, I'd think you'd have the full 16 bits of resolution.  Because you have pushed the data up to the 16 bit 'window' that is used (bits 9 thru 24).  But then again...that's what turning the gain up does.  The V3 can't do that for you because what would happen if you got a signal that hit full scale - where would it go.

So conceptually, the loudest point is fixed (0 dB, full scale, all bits on, whatever you want to call it) and resolution falls under that.  If you don't use the most significant bits, you don't really have the option of going further down for resolution unless you have the original 24 bit signal and you shift the whole thing.
KM 184 > V2 > R4
older recording gear: UA-5  / emagic A62 / laptop / JB3 / CSB / AD20 / Sharp MT-90 / Sony MDS-JE510
Playback: Pioneer DV-578 > Lucid DA 9624 >many funny little british boxes > Linn Isobarik PMS

Offline Todd R

  • Over/Under on next gear purchase: 2 months
  • Trade Count: (29)
  • Needs to get out more...
  • *****
  • Posts: 4901
  • Gender: Male
Re: V2 levels
« Reply #51 on: November 18, 2004, 01:54:05 PM »
I kinda got lost in there but I think you do need to go close to full scale even if you only want 16 bits.  When going from 24>16 it's the least significant bits that get dropped.  So if you're down 12 dB, you're at 22 bits on a 24 bit scale, or 14 bits on a 16 bit scale.  Maybe dithering can push you up a bit, but I don't think you can say hey, we've got 22 bits so there should be at least 16 available.  With 16 bit on this box, you're using bits 9 thru 24 and throwing away the first 8 (ignoring dither for the moment).

Ok, I guess we're even, 'cause I'm totally lost in what you're saying. ;)   I think you've got a fundemental misunderstanding of what dithering does (or perhaps me, but we're definitely not in agreement).  It is my understanding that when you say "I don't think you can say hey, we've got 22 bits so there should be at least 16 available" that you're mistaken.  My understanding of dithering is that is exactly what happens.  So you're right that if you're 12db down, you've only got 22 bits on a 24bit scale or 14 bits on a 16 bit scale.  But you'll have 16 bits on 16 bit scale that has been arrived at through dithering from 24 bits (or 22 bits in this case).  So, yes, the exact purpose of the dither scheme is to maintain 16 bits of true significant bits when the original signal is 24 bits (or 22 bits if we don't record hot enough to get 24 significant bits).  Even if you only have 18bits of info, you should be able to dither down to get 16 bits of true significant data.

The purpose of dither is essentially to add noise.  I'm going from a somewhat foggy memory here, but it is something like this. If you have 24 bits of info, the least significant 24th bit will vary between being a digital 1 or a 0.  Same thing with the 16th bit of a 16 bit word.  But the 16th significant bit of this 24 bit word will be set at either a 1 or a 0.  Simply lopping off 8 bits from the 24 bit word to get a 16 bit word will then leave this 16th bit at a given state of either ending in a 1 or in a 0.  This makes for a harsh sound and poor representation of the analog signal, since the least significant bit should vary between a 1 and a 0 and not be stuck at either one or the other.  The purpose of dither is to add just the right amount of noise to the signal so that this 16th bit (least significant bit) varies between 1 and 0 (and not to add so much noise that you affect the 14th or 15th bits).   Anything random will work, including just having enough noise in your electronic signal path of your recorder so that the 16th bit is random.  But the idea is to create a dither scheme that uses the information present in the (say 8 ) least significant bits that you're tossing aside to correctly add noise to the 16th bit of the now dithered word so that it will vary between 0 and 1.

It all depends on the exact specifics of the dither algorithm that is used, but done well, a good dither algorithm should be able to make good 16 bit signal (with 16 bits of significant info) from an 18 bit word (the Apogee AD500e did this), or a 22 bit word, or a 24 bit word.  So yes, it is my understanding that the dithering scheme does exactly what you think it doesn't do -- that is is builds a 16 bit word out of the significant data from a 24-bit word, even if that 24-bit word only has 22 bits of significant data.  It would depend on the specifics of the ANSR dither scheme, but it would be a pretty poor dither scheme if it didn't function to get 16 bits of signicant data out of 22 significant bits.

Edit:  stupid emoticons: the number 8 followed by a ) should not always = 8)
« Last Edit: November 18, 2004, 01:57:23 PM by Todd R »
Mics: Microtech Gefell m20/m21 (nbob/pfa actives), Line Audio CM3, Church CA-11 cards
Preamp:  none <sniff>
Recorders:  Sound Devices MixPre-6, Sony PCM-M10, Zoom H4nPro

Offline Lil Kim Jong-Il

  • Trade Count: (6)
  • Needs to get out more...
  • *****
  • Posts: 6498
  • large Marge sent me
Re: V2 levels
« Reply #52 on: November 18, 2004, 02:27:01 PM »
I don't believe the V3 realtime dithering also performs normalization.  Because the instantaneous value has no relation to past or future maximum values, you need the entire sample set to determine a normalization factor.  Therefore, the realtime dithered 16-value should be proportional to the original 24-bit value.

I don't think that dither operations just toss aside the least significant bits and then add noise based on the discarded value.  Properly converting between bit depths requires the entire range to be scaled such that a 6-bit value recorded at 24-bit resolution may become a 4-bit value at 16-bit resolution.  Noise is then added to counteract the additional quantization noise incurred during scaling.  As you mentioned, the value of the original signal or a brief window may be used to determine the appropriate value of the noise to be added.   If you just lop off the least 8 bits, quiet passages will flatline and no amount of introduced random noise will allow the original information to be resolved. 

Along these lines, I'm not really sure when we talk about "using all the bits" if that is an accurate statement.  I think "using the full range is more correct".  My understanding of LPCM is that the 16-bit quantized sample is assigned a value in the range from -32767 to +32767.  A sample set with a peak value of 16385 requires 16-bits to represent, but peak is just barely over half the supported maximum value.  The result is that there is a 50% degraded resolution of the softer passages.  So the gain should be adjusted to use more of the range so that low level information is represented in more detail.  Is this not correct?

No?  Maybe?


For me, it's not really about getting the best SNR.  I would expect the V3 analog section to have lower self noise than my mic (KM100).  I'm mostly just trying to maximize the resolution at playback.  I guess that could be considered an attempt to reduce overall quantization noise.  But even when my 24bit box arrives, I'll still be trying to use as much of the range as possible.

The first rule of amateur neurosurgery club is .... I forget.

Offline Todd R

  • Over/Under on next gear purchase: 2 months
  • Trade Count: (29)
  • Needs to get out more...
  • *****
  • Posts: 4901
  • Gender: Male
Re: V2 levels
« Reply #53 on: November 18, 2004, 02:48:22 PM »
I definitely agree that the V3 is not doing any normalization, since as you say that would be impossible.  I thought from your earlier posts you were most worried about getting the best SNR out of your tapes and given the effect of dithering a 24 bit signal down to 16 bits, going for the hottest signal on the V3 isn't that big of an issue.  As to the rest of what you're saying,  :-\ I'm going to have to give that some more thought.
Mics: Microtech Gefell m20/m21 (nbob/pfa actives), Line Audio CM3, Church CA-11 cards
Preamp:  none <sniff>
Recorders:  Sound Devices MixPre-6, Sony PCM-M10, Zoom H4nPro

Offline Lil Kim Jong-Il

  • Trade Count: (6)
  • Needs to get out more...
  • *****
  • Posts: 6498
  • large Marge sent me
Re: V2 levels
« Reply #54 on: November 18, 2004, 03:10:08 PM »
I definitely agree that the V3 is not doing any normalization, since as you say that would be impossible.  I thought from your earlier posts you were most worried about getting the best SNR out of your tapes and given the effect of dithering a 24 bit signal down to 16 bits, going for the hottest signal on the V3 isn't that big of an issue.  As to the rest of what you're saying,  :-\ I'm going to have to give that some more thought.

Cool.

I know that one could argue that the signal level produced by the lower 8-bits are so far below the values produced by a 18-bit to 20-bit signal that one can't really hear it.  I could buy that argument. 

I believe that the main problem with bit depth conversions using simple truncation is that the wave form will become discontinuous at nearly every 0 crossing and I think that is what ruins the sound of a recording when ANSR is not enabled on the V3 while feeding a DAT deck.  Maybe I'll play with the V3 if I get some time.  I never ran it without ANSR engaged.
The first rule of amateur neurosurgery club is .... I forget.

 

RSS | Mobile
Page created in 0.06 seconds with 38 queries.
© 2002-2024 Taperssection.com
Powered by SMF