Simple answer: you have to be very fussy with digital, I repeat, very fussy. Lack of attention to detail will give you loss of fine detail structure. Every time.
Frank
> I am going towards measuring harmonics of analog
> designs with a digital ADC/DAC system!
It never ceases to amaze me the number of people
who use ' digital ' scopes ....
Anyone can see the steps on the traces, but yet
seem to be ambivalent to the ability of the steps
to hide sensed glitches ..............
> designs with a digital ADC/DAC system!
It never ceases to amaze me the number of people
who use ' digital ' scopes ....
Anyone can see the steps on the traces, but yet
seem to be ambivalent to the ability of the steps
to hide sensed glitches ..............
Exactly. What's being dealt with in a CD player, say, are always analogue behaviours, with some parts of that having a digital "meaning". Analogue can always adversely affect, interact, interfere with other analogue, and thinking of what you're dealing with, with that understanding has to be of prime importance if one wants "good" CD sound.
Frank
Wow. Thats great for you. you'll love it as long as you dont let yourself be bored.
I tell people anywhere that today, more than ever before, you must get a home paid off or you will never be able to retire... not as long as you have to make monthly house payments. And, making that happen is becoming harder to do. But, you've made it and are able to retire.
When do you make the big jump off and free fall? We'll have a party for you over here in beautiful, sunny downtown Cool, California!! BBQ and a glass of wine on that day. -Dick
No mortgages here not that I like to rub it into anyones face. Not my style, I can't quite figure your angle on this.
Is it the digitizer or the display steps you are seeing?
If you know how to use the tool to see what you need to see it matters little if its digital or analog. Sometimes digital is way better for seeing "glitches". If you don't know what to look for sometimes neither digital nor analog (nor analog distortion analyzer nor DSP) will help.
I have a couple of other FFT units -- one digital scope made by Tektronix -- will do up to 300MHz FFT - 1-2 GigaB/s clock ...no steps seen. But I want more bits for audio.
Its just thrown in there for any one who isnt close to retirement to think about.
So? When is your last day?
It looks easy from "digital" point of view, until you start looking at gain and offset calibration issues, reference voltage stability and noises, and similar analog issues...
Then you have (for whatever reason) a bad implementation. This has nothing to do with number of bits, which ONLY sets noise floor. That's been discussed to death here, and there's plenty of links to demonstrations of this (most notably at Werner Ogiers's site).
Unfortunately you are comparing apples and bananas. When you are talking 105 db. peak levels that is read on a standardized sound level meter. If you wanted to accurately record this you would want at least 10 db. of headroom.
Now when you talk about noise level that is a broadband measurement. As the ear can has the ability to analyze sound in critical bands, there is the ability to pick out signals 30 db. below the broadband noise level.
In speech systems it is considered acceptable when the signal to noise ratio is ZERO db.
Ed, you're comparing apples to pencil sharpeners. We're not talking about recording, we're talking about a playback medium, where the peaks can absolutely go right up to zero dB with no penalty (in fact, that's good mastering practice). That drops the noise floor 90+ dB, quieter than Randy's room or the concert hall, or the studio, or the mike preamps.
This has also been discussed to death in this very thread.
This has also been discussed to death in this very thread.
Yes SY,
This is old stuff. When the standards were set for digital audio the idea was that it would be as good as or better than the existing analog methods.
In a typical listening space the noise level will be NC35 or better. Although some argue that when you play back a recording at a lower level, due to Fletcher-Munson type of issues, you change the timbre.
But even with music peaks reproduced to read 85db. the player must reproduce the equivalent of 95db.
The noise level usually is mostly low frequency energy which is why the NC ratings count it less. But there still is a drop in mid and high frequency energy. So tones as low as -5db. may still be heard.
The result is that a true 16 bit recording should be fine for a typical room on much material, but it is not able to cover the entire range of human hearing.
P.S. I prefer a sharp pencil!
Last edited:
As long as the other noise sources dominate, the noise floor of the playback medium is not terribly important. And we're not talking about the range of human hearing (which are group maxima, not the same person), we're talking about the noise floors of studios, concert halls, playback rooms, and mike preamps, which dominate competent recording practice. And these noises are uncorrelated with the playback medium noise floor, so they sum as powers, increasing the influence of the dominant source.
So we're back to meter reading and test signals as opposed to playback of musical program. Maybe that's where we are differing, our definition of the goal of a playback medium.
So we're back to meter reading and test signals as opposed to playback of musical program. Maybe that's where we are differing, our definition of the goal of a playback medium.
No, it is not! Good mastering practice is to leave a couple of tenths of dB headroom. When using oversampling low pass filters on playback, under certain circumstances, inter-sample peaks might be higher than 0dBFS.
www.tcelectronic.com/media/level_paper_aes109.pdf
Last edited:
I would like to make a point about various people's opinions as to what is 'good enough' for highest quality audio reproduction.
Back when many here were just teenagers (1968) I used to debate the same thing at the Ampex Audio Department.
You see, my colleagues (and my boss) could NOT understand why I wanted to continually improve the product. For example, the performance spec for 15ips audio bandwidth might be up to 15KHz, no more. They would let the last 10KHz flap in the wind, one way or another. Hi fi standards at the time went from 20-20K, so even DYNA was more careful than Ampex.
One of my assignments was to look at various reproduce heads for their internal quality, or Q. I found out a number of things that showed that we were WAY back in development and were both adding several dB of UNNECESSARY NOISE at every playback of a tape, as well as making the frequency response almost impossible to compensate for. When I commented on this, I got all kinds of chaff: Don't rock the boat, it's to expensive to change, etc, etc. Same as here, sometimes.
Well, that was 1968, I gave my report on head noise, etc and left the Ampex Audio Department for Research. 5 years later, they were BRAGGING about the improvements in their head designs, exactly as I recommended. You see, the pushy, wild eyed dreamer, might, just might, be able to give a company a commercial 'edge' if only they are taken seriously at the time. Their suggestions at improvements will certainly come sooner or later, perhaps from a competing manufacturer.
Another story almost 15 years later:
Dave Wilson and I knew a guy, then, who worked at HP, and was very interested in audio. This guy bought one of the Sony F1 based digital recording systems based on video tape. This guy was elated! He was satisfied! He said the digital recording and playback was PERFECT! And he insisted he was right, to everyone else.
The rest of us we trying these things out and then trying to make them better.
Well, as time went on, there was some announcement from SONY that they had IMPROVED the digital recording process. At this point, Dave Wilson teased the HP guy about digital already being 'perfect', so how could they improve it?. His response was: 'Better than perfect' I don't know if this guy still likes audio in the same way today, but he found HIS answer back before CD was even available. The rest of us tried to make the F1 sound better through modification. People sent them to me for modding, and I did the best that I could with them. Where are they today?
Back when many here were just teenagers (1968) I used to debate the same thing at the Ampex Audio Department.
You see, my colleagues (and my boss) could NOT understand why I wanted to continually improve the product. For example, the performance spec for 15ips audio bandwidth might be up to 15KHz, no more. They would let the last 10KHz flap in the wind, one way or another. Hi fi standards at the time went from 20-20K, so even DYNA was more careful than Ampex.
One of my assignments was to look at various reproduce heads for their internal quality, or Q. I found out a number of things that showed that we were WAY back in development and were both adding several dB of UNNECESSARY NOISE at every playback of a tape, as well as making the frequency response almost impossible to compensate for. When I commented on this, I got all kinds of chaff: Don't rock the boat, it's to expensive to change, etc, etc. Same as here, sometimes.
Well, that was 1968, I gave my report on head noise, etc and left the Ampex Audio Department for Research. 5 years later, they were BRAGGING about the improvements in their head designs, exactly as I recommended. You see, the pushy, wild eyed dreamer, might, just might, be able to give a company a commercial 'edge' if only they are taken seriously at the time. Their suggestions at improvements will certainly come sooner or later, perhaps from a competing manufacturer.
Another story almost 15 years later:
Dave Wilson and I knew a guy, then, who worked at HP, and was very interested in audio. This guy bought one of the Sony F1 based digital recording systems based on video tape. This guy was elated! He was satisfied! He said the digital recording and playback was PERFECT! And he insisted he was right, to everyone else.
The rest of us we trying these things out and then trying to make them better.
Well, as time went on, there was some announcement from SONY that they had IMPROVED the digital recording process. At this point, Dave Wilson teased the HP guy about digital already being 'perfect', so how could they improve it?. His response was: 'Better than perfect' I don't know if this guy still likes audio in the same way today, but he found HIS answer back before CD was even available. The rest of us tried to make the F1 sound better through modification. People sent them to me for modding, and I did the best that I could with them. Where are they today?
Last edited:
SY,
The noisiest thing in my house is the clock.
Now my experience is that for symphonic music 20db of headroom is good, but 30db is better. So 16 bits would be the limiting factor.
We both agree for a competent recording under most circumstances 16 bits will work.
However in practice, safety margins are a good idea. As the technology allows it 24/192 can under the right conditions be better. Even though there are no true converters for that level!
Now there is a reason why so many old records sound so good. Who keeps and shouts about old bad records?
By the same measure there are a lot of bad CDs. The first generation were only 9 bit linear.
This is all old stuff. But I do object to you including mic preamps in the 16 bit limit. They are way better than that.
One of the projects I may get to is a mic preamp that is quiet, has headroom and AES output. I need this for the dynamic range of sports announcers.
The noisiest thing in my house is the clock.
Now my experience is that for symphonic music 20db of headroom is good, but 30db is better. So 16 bits would be the limiting factor.
We both agree for a competent recording under most circumstances 16 bits will work.
However in practice, safety margins are a good idea. As the technology allows it 24/192 can under the right conditions be better. Even though there are no true converters for that level!
Now there is a reason why so many old records sound so good. Who keeps and shouts about old bad records?
By the same measure there are a lot of bad CDs. The first generation were only 9 bit linear.
This is all old stuff. But I do object to you including mic preamps in the 16 bit limit. They are way better than that.
One of the projects I may get to is a mic preamp that is quiet, has headroom and AES output. I need this for the dynamic range of sports announcers.
- Status
- Not open for further replies.