One massive advantage of tape over digital is that it will take peaks and compress them as the level gets larger. With digital, when you hit "0", you're all out of 0's and 1's and you get a very hard clip. That might be why digital sucks the life out of music.
So tape based technologies will accept overloads and do something a lot nicer with them than hard clipping. They actually use this as an effect in recording studios (listen to Motown tracks for this example of badly overloaded tape machines).
-Chris
So tape based technologies will accept overloads and do something a lot nicer with them than hard clipping. They actually use this as an effect in recording studios (listen to Motown tracks for this example of badly overloaded tape machines).
-Chris
Interesting theory about clipping, but most digital recordings aren't clipped. However, too much dynamics processing of any kind as a part of the loudness wars has been accused of sucking the life out of music. Maybe that's what you're thinking of?
The other thing is a lot of early digital did sound bad. The old Pro Tools Mix systems come to mind. But, modern hi-res professional digital can sound quite good, hard to understand how you could say it would suck the life out of music.
Last edited:
Hi Mark,
Nope. Clipping can be a problem in recording depending on the compressor they are using. If they aren't using a limiter then they can run into digital clipping. The professional studios likely wouldn't, but the smaller guys using "Pro Tools" or whatever is popular today certainly can. The character of the limiter / compressor is everything to a recording. With tape they could allow the natural tape overload to slightly compress the rest of the music after processing. But with digital, they have to be much more careful as there is no compressor effect. Just a nasty raw clip.
You tend to see the master mix instead of what they do with the bed tracks before mixing. Coming off real instruments, they have a huge dynamic range to deal with. To crush that down into the allowable dynamic range can be challenging - depending on what sound they are looking for from that instrument. Even the human voice can be a challenge to deal with. With tape, the occasional foray into the + levels aren't going to be audible. With digital, the limits are extremely audible. BTW, a Studer with Dolby SR has a wider dynamic range than the digital equipment had in the 90's.
Once the beds are down on tape, mixing is easier as there are no more retakes and the tracks are fairly well controlled and processed. That is what you see. No, you will not see any excursions into the 0 Vu or above with this material. To my ears, an AAD CD always sounded the best. The same holds true for an LP, the same code would look like this: AAA (not the car people).
-Chris
Nope. Clipping can be a problem in recording depending on the compressor they are using. If they aren't using a limiter then they can run into digital clipping. The professional studios likely wouldn't, but the smaller guys using "Pro Tools" or whatever is popular today certainly can. The character of the limiter / compressor is everything to a recording. With tape they could allow the natural tape overload to slightly compress the rest of the music after processing. But with digital, they have to be much more careful as there is no compressor effect. Just a nasty raw clip.
You tend to see the master mix instead of what they do with the bed tracks before mixing. Coming off real instruments, they have a huge dynamic range to deal with. To crush that down into the allowable dynamic range can be challenging - depending on what sound they are looking for from that instrument. Even the human voice can be a challenge to deal with. With tape, the occasional foray into the + levels aren't going to be audible. With digital, the limits are extremely audible. BTW, a Studer with Dolby SR has a wider dynamic range than the digital equipment had in the 90's.
Once the beds are down on tape, mixing is easier as there are no more retakes and the tracks are fairly well controlled and processed. That is what you see. No, you will not see any excursions into the 0 Vu or above with this material. To my ears, an AAD CD always sounded the best. The same holds true for an LP, the same code would look like this: AAA (not the car people).
-Chris
If you allow digital recording to clip - you are not using the tools correctly. You can't criticise a tool if you break it through misuse....
I heard this before from recording engineers, The band complains it's not loud enough. So the engineer compresses it to satisfy the complaint.
Hi Anatech, It used to be a problem back in the days of 16-bit recording. Nowadays with 24-bits there is no need to compress or limit before the A/D, and clipping is rarely an issue. With more than 100dB of dynamic range available there is no problem leaving adequate headroom. In addition recording programs light up red lights with one single clip, and count the exact number if any occur, so there aren't any sneaking by.
The 0dB digital level is -18dB from clipping in most 16 bit recorders. Even with a Vu meter (and a minimal experience) there is no reason why you never enter into saturation.
Anyway, who is using Vu meters nowadays ?
(On a magnetic tape, you can allow +10dB).
Once a mix is finished, you can process-it to limit some peaks and give some more gain in order to normalise the level to the max available.
In fact, it is just a different way to work.
it is strange to see complaints about the dynamic of digital, while we are at >100sB, and we were at ~70 with magnetic tapes.
Last edited:
Which links to the question. How little compression can you use on a drum kit and still get a usable recording. I need to fire up audacity on this Rickie Lee Jones* and see how high the peaks from the snare really are, but it needs the volume control racking up 9-12dB from a regular 'pop' recording to get the lyrics at a sensible level. It's a nice change from tracks where the vocalist cuts in and the VU meter stays pegged at -10.
* This is the one that Bob Cordell used to show you need 250w/ch to handle good recordings without clipping.
* This is the one that Bob Cordell used to show you need 250w/ch to handle good recordings without clipping.
I think everyone except you is using 0dB to mean 0dBFS, not sure which specs you follow.
Bob Ludwig for one!
Attachments
None. But, it depends on how good the drummer is. After that it depends on genre and what type of sound is desired.
And the big yellow VU meters around the outside. You asked a question, I answered. Had you asked 'Who only uses VU meters and nothing else' I wouldn't have said anything.
Hi Mark,
The last time I was inside a recording studio was in the 1990's. Therefore the issues I remember are of that time period, and possibly musicians trying to record themselves today. No matter if you are using 16 bit or 24 bit, or even 32 bit (I have heard of), it still takes some skill to avoid clipping in the digital realm.
Engineers that were used to tape had to relearn everything as far as how they worked for the digital recorders. Can you imagine what it was like for the guys at Motown to make that transition? That's if they ever did.
These days they use peak meters as a VU meter doesn't show what they need to see.
Hi gpauk,
16 bit recording was a problem in those days. Dynamic range was very much an issue, and the sound of a low level signal was very poor on the digital equipment compared to tape. Don't forget that with the lower levels you are only using 12 bits, maybe 14. Then the quality really did suffer. Analogue tape was far superior to digital in that situation. 24 bit really does change the game a bit, and 32 bit is better still. I don't know why it was 32 bit instead of 30 bit or some other count.
-Chris
The last time I was inside a recording studio was in the 1990's. Therefore the issues I remember are of that time period, and possibly musicians trying to record themselves today. No matter if you are using 16 bit or 24 bit, or even 32 bit (I have heard of), it still takes some skill to avoid clipping in the digital realm.
Engineers that were used to tape had to relearn everything as far as how they worked for the digital recorders. Can you imagine what it was like for the guys at Motown to make that transition? That's if they ever did.
These days they use peak meters as a VU meter doesn't show what they need to see.
Hi gpauk,
I can see your point, but I wasn't criticizing anything at any time here. What I was doing was mentioning some of the problems I had actually seen happening in a studio environment. In many ways, back then a new engineer taught on digital recording would have an advantage (technically) over a seasoned engineer. That's because the new tools were so completely different than the old ones.
16 bit recording was a problem in those days. Dynamic range was very much an issue, and the sound of a low level signal was very poor on the digital equipment compared to tape. Don't forget that with the lower levels you are only using 12 bits, maybe 14. Then the quality really did suffer. Analogue tape was far superior to digital in that situation. 24 bit really does change the game a bit, and 32 bit is better still. I don't know why it was 32 bit instead of 30 bit or some other count.
-Chris
Then there are those of us who do live events.
First off there really are no 24 bit accurate audio A/D converters.
Now even if we could get a 147 dB range converter it would be difficult to get the analog circuitry to match.
Now a normal speaking voice is around 60 dBa at one meter. (I see a lot of cites for 70, but I have never measured that.) Using a large diaphragm. Condenser microphone an announcer will deliver about 75 dBa into it. For reasonable amplification use the noise floor must be 20 dB or more below the ambient level in the stadium or about 35 dB. As the delivered announce level should run around 95 dB that means 60 dB of range below that is desirable.
Now when the announcer starts getting excited he can exceed 85 dBa. However he will also move closer to the microphone. At the 1 cm. distance that level becomes 125 dBa. Add the 60 dB of margin from normal use and for some reason the mixer inputs cannot handle 185 dB of dynamic range. In theory the input gain trim should be adjusted, but as no human operator is that fast...,
The other issue is that most modern microphone preamps run on 15 volt rails or about 10 VRMS maximum. From the noise floor of a 150 ohm source that limits one to about 150 dB of dynamic range. But of course reality sets in and you don't see that.
As in my opinion one can never escape all clipping the issue becomes how do you handle it. My approach involves dynamically changing the noise floor.
Of course those who don't recognize the problem make no attempt to deal with it.
First off there really are no 24 bit accurate audio A/D converters.
Now even if we could get a 147 dB range converter it would be difficult to get the analog circuitry to match.
Now a normal speaking voice is around 60 dBa at one meter. (I see a lot of cites for 70, but I have never measured that.) Using a large diaphragm. Condenser microphone an announcer will deliver about 75 dBa into it. For reasonable amplification use the noise floor must be 20 dB or more below the ambient level in the stadium or about 35 dB. As the delivered announce level should run around 95 dB that means 60 dB of range below that is desirable.
Now when the announcer starts getting excited he can exceed 85 dBa. However he will also move closer to the microphone. At the 1 cm. distance that level becomes 125 dBa. Add the 60 dB of margin from normal use and for some reason the mixer inputs cannot handle 185 dB of dynamic range. In theory the input gain trim should be adjusted, but as no human operator is that fast...,
The other issue is that most modern microphone preamps run on 15 volt rails or about 10 VRMS maximum. From the noise floor of a 150 ohm source that limits one to about 150 dB of dynamic range. But of course reality sets in and you don't see that.
As in my opinion one can never escape all clipping the issue becomes how do you handle it. My approach involves dynamically changing the noise floor.
Of course those who don't recognize the problem make no attempt to deal with it.
BTY 32 bits would cover the audio range from 0 dBa to completely removing the entire planet's atmosphere. Just think of the rail voltage you would need!
I can't tell you how many times this scenario happened to me with the same results initially. There is some element of "shock of the new" with the digital file.
Remember Paul Harvey and "The Rest Of The Story"? The rest of the story with this is when you listen to the tape weeks or years later. The tape usually wins.
Because we were talking about monitoring signal to avoid digital clipping .
Not about the various instruments you can use to show average, minimal, max, peak etc.. levels to get visual ideas about dynamic.
Looking for controversy?
It is like your question about this -18dB VF dBFs.
Do-you often worked with professional analog mixing desks and digital recorders (where do you put your 1,23 Volt reference ?), or various labs, TVs etc. to ask such a question ?
[edit] I had made a little research for you : http://www.homerecording.be/forum/attachments/f20/12946d1362385990-zb-reflevel-analogdigital.jpg
Last edited:
Hi Ed,
Yes, I know. I'm just repeating the jargon that they use. The 32 bit thing comes in as soon as they modify the signal in any way, even changing the volume can generate long digital words I'm told. Then at output time, it is truncated to whatever standard they are recording with.
Your experiences with the voice and microphones is repeated in studios all across the planet. At least the singers are a little better controlled as far as keeping the mic in front of them. Well, most of them anyhow. As pointed out earlier, drums are tough, as is piano. Most other instruments have better controlled output levels and can be recorded reasonably easily.
Your job is far more difficult than a recording engineer's would be. Your "recording" on a single take with no punch ins and no redos. Your performance is a single, uninterrupted take pretty well with no way to stop and fix something that isn't front and center. My guess is that over-easy compressors and limiters are your prime weapons. Intelligibility and not dynamic performance would be your priorities, am I right?
Best, Chris
Yes, I know. I'm just repeating the jargon that they use. The 32 bit thing comes in as soon as they modify the signal in any way, even changing the volume can generate long digital words I'm told. Then at output time, it is truncated to whatever standard they are recording with.
Your experiences with the voice and microphones is repeated in studios all across the planet. At least the singers are a little better controlled as far as keeping the mic in front of them. Well, most of them anyhow. As pointed out earlier, drums are tough, as is piano. Most other instruments have better controlled output levels and can be recorded reasonably easily.
Your job is far more difficult than a recording engineer's would be. Your "recording" on a single take with no punch ins and no redos. Your performance is a single, uninterrupted take pretty well with no way to stop and fix something that isn't front and center. My guess is that over-easy compressors and limiters are your prime weapons. Intelligibility and not dynamic performance would be your priorities, am I right?
Best, Chris
No, just poking a pin into inflated pomposity. YOU were not following the twists and turns of the thread, and you asked a non-specific question. You got an answer which you didn't like so you are trying to twist words around. Meh.
Your little research shows 0dBFS to be full scale. Please explain what you meant by "The 0dB digital level is -18dB from clipping in most 16 bit recorders." in context of the diagram you linked. That says that +4dBu is -18dBFS which is not what you said.
Try and be a little more accurate in what you post and these confusions will not appear.
- Status
- Not open for further replies.