Just curious as to why still analog??

[snow lizard]

How about this:

Question: What is comparator?

A comparator is a device for measuring variations. We've been using them for over 100 years. The most commonly thought of comparators are tools for making physical measurements of variations in part size in the manufacturing world. They're starting to become obsolete with the advent of digital coordinate measuring machinery, although the method of making comparisons is still secondary to the physical act of doing so in the first place.

An argument can be made that an analog tape recorder head is a form of comparator, as it will register changes in the electrical signal it receives.

So what's wrong with digital? What's missing there?

It probably depends on what type of digital audio system you're talking about here. Older professional A/D converters and many of the current crop of prosumer models such as my Delta 44 are supposedly pretty crappy in the midrange frequencies. Apparently the newer converters like Apogee and such don't have this limitation.

As the technology continues to improve, hopefully everyone will be able to access pristine quality A/D conversion, and life will be sweet. Until then, analog recording still has a valid basis for comparison IMWAO, and rightly so.

Other aspects aren't so much about what's missing, as much as what's different. Natural tape compression or saturation or whatever it's called does not exist in digital. I'm also inclined to suspect that digital has faster transient response leading to greater detail in the higher frequencies that might be annoying if misused. Digital is finally able to register what many condenser mics will do to a sound source. A lot of folks think that many condensers sound harsh, or hyped in the high frequencies. Since everyone and their brother is always looking for the latest and greatest Chineese condenser microphones, that's part of the reason why things sound the way they do.

Another thing is compression. Any broadcast format for recording in the last 40 years or so is capable of rendering a lot of dynamic range. Having 144 dB of range available on a digital system is great, but the recent trend in overcompression during mastering is annoying at best. This is not the fault of the technology itself, but rather the executive that told the mastering engineer to make the album louder than the one that had just come out that had already pushed the process to its maximum. A bit of colouration from using, say, a ribbon mic in place of a condenser for its typically unhyped high end characteristics and flattering, non-linear frequency response, and realistic (albeit slower than condenser mics) transient response coupled with a bit of compression during tracking can help to bring a digital system a bit closer to what we've come to expect with analog, and you don't even have to flatten all of the dynamics to do it.

Using a reference condenser with a transparent preamp recorded digitally is probably a good recipe for a boring and sterile recording. Throwing some kind of variable in the chain to introduce colouration, be it a mic, pre, compressor or tape machine can be a big help for many types of music if that's what you want.

There also seems to be a lot of folks that think that tracking digital with no headroom for overs or transients is a good thing. I don't get that one at all.

Beyond that, we know from the old Neve consoles that overtones or harmonics from beyond the range of human hearing can affect frequencies that we can hear. I think the trend in future digital audio technology should allow for that better than "CD Audio" digital. DVD and SACD formats are a step in the right direction in principle I think, but MP3's and cheap computer audio are probably countering it. There also has to be a realistic approach to standardization, or at least cross-compliancy if new technologies are to emerge. In the end, being able to accurately replicate any even order harmonics, overtones, saturation and typical frequency curve bumps should be able to make digital pretty close to analog. Saying it and doing it may be two totally different things at present, but the gap is narrowing.

I still think that classic analog recorders have something to offer the audio world, be it a comparison to the format that was around when audiophile standards were more popular, a more comfortable recording enviornment for those who choose to use it, a way of getting a desireable recorded sound for specific instruments or musical style, or simply as a reminder that at one time there were folks who took the best of what technology was available at the time and used it to try to elevate their craft.

At least that's what I suspect. I'm just guessing, really. (except for the part about the comparator) Flame away!


sl

 

[Seeker of Rock]

Digital photography is not the same theory at all. Digital photography is far less accurate than digital audio recording for a whole host of reasons.

But isn't it the same concept of mimicing, by means of artificial data, real-life entities with bits of information? I would imagine by mere size of CPUs in a desktop unit or hard disk recorder, as opposed to relatively smaller digital cameras, that sound is far more advanced than photography. All this talk of digital being more accurate. I buy that, but believe the statement falls short...digital is probably more accurate at that which it "hears". I would probably use the term "better defined" than accurate, as that which it does not "read" as a signal strong enough to record, it will not. Accurate would elude to digital being able to sense and record all sounds (slight movement of a leg or arm, slight background breathing during a guitar or drum track, and anything else that tape may pick up ever so lightly that is not strong enough to register as a valid sound signal in digital) and from what I have heard, I am not convinced that digital does this. Maybe this would be considered desirable by many not to have sounds like these in the tracks, even at such a low, minutely audible level. I would prefer them, though, at barely audible levels as I think it is part of the character, the human side of music and recording.

 

[regebro]

But isn't it the same concept of mimicing, by means of artificial data, real-life entities with bits of information?

Sure, but digitizing images is much harder than audio, becuase you need to make really huge CCDs to get to the point that you can compete with film.

 

[evm1024]

Sure, but digitizing images is much harder than audio, becuase you need to make really huge CCDs to get to the point that you can compete with film.

Large CCD have existed for 20 plus years. Tektronix produced a 2k by 2k CCD a long time ago. Granted it was expensive but then again it was a backside device intended for astronomy and was about 2" by 2" which is huge for a single die.

So you are correct if you say that (to pariphrase) large pixal counts are needed to bring the spatial resolution of CCD to near film grain resolution.

But to say that digitizing images is harder than audio is incorrect. All of the difficulty in a CCD camera lies in the making of the CCD and once you have the process down then the devices just roll off your fab line.

Using clocks the charge in each well (pixel) is moved out of the CCD and into a sample and hold. That charge is measured by A/D. Each of the three colors are measured individually. The A/D is typically only 8 bits which gives 24 bits of color.

Tak, there are complexities in dark frames and well to well bleeding. However you do not need to worrk nearly so much about getting 16 or 24 bits of A/D and all the problems that are in those low levels of resolution.

I would say that digitizing and image well is much easier than doing audio.

regards

 

[regebro]

Large CCD have existed for 20 plus years. Tektronix produced a 2k by 2k CCD a long time ago. Granted it was expensive but then again it was a backside device intended for astronomy and was about 2" by 2" which is huge for a single die.

??? 2 K by 2k is 4 megapixels. That's what I have in my camera. It aint anywhere near enough to compete with film.

But to say that digitizing images is harder than audio is incorrect. All of the difficulty in a CCD camera lies in the making of the CCD

Which is the hard part, just as in the hard part in digital audio is making A/D D/As.

Tak, there are complexities in dark frames and well to well bleeding. However you do not need to worrk nearly so much about getting 16 or 24 bits of A/D and all the problems that are in those low levels of resolution.

No, but you have millions of points to sample, instead of 2.

I would say that digitizing and image well is much easier than doing audio.

why am I not suprised....

Motvallskärring is a swedish word for you. It means "old lady that floats against the current". It's used of people who always do things in unusual ways, or take contrary positions, or just refuse to move out of the way, seemingly just to be pain in the ass.

So, instead of having one A/D of 24 bits, you have three (one for each color) or 8, and you need to take millios maybe billions of samples instead of one. and this is your argument for that image digitizing is easier that sound digitizing?

I'm not impressed.

 

[evm1024]

??? 2 K by 2k is 4 megapixels. That's what I have in my camera. It aint anywhere near enough to compete with film.


Which is the hard part, just as in the hard part in digital audio is making A/D D/As.

No, but you have millions of points to sample, instead of 2.

why am I not suprised....

Motvallskärring is a swedish word for you. It means "old lady that floats against the current". It's used of people who always do things in unusual ways, or take contrary positions, or just refuse to move out of the way, seemingly just to be pain in the ass.

So, instead of having one A/D of 24 bits, you have three (one for each color) or 8, and you need to take millios maybe billions of samples instead of one. and this is your argument for that image digitizing is easier that sound digitizing?

I'm not impressed.

You're funny I enjoy talking with you. It is interesting how you speak with authority on things that you have a surface level of understanding.

Yes the TK4096 was just 2 megapixals but you missed the point. It was first made in 1985 (or so) and was 2" square! Plus it was backsided. This gives it a large QE and a bandpass at 10% QE from 300 nM to 1030nM. Outstanding for 20 years ago! And it did compete with film and did much better than film in it's intended application. And it cost $125,000 each. The world was in awe of this device 20 years ago.

CCD camera only have one A/D. CDD are read serially thus you need only one.

Question: How many samples does it take to read a 8 megapixal color ccd?

Answer: 24 million. 8 million per color.


Question: how many 8 megapixal images can be sampled with a billion samples?

Answer: 41.666


Question: How many CCD cameras have you built?

Answer: 0

I'll grant you that you have some knowledge. Sometimes at a greater or lessor depth. Perhaps if your offered you opinions as opinions rather than fact and refrained from berating those who disagree with you things would go better for you.

So the simple question is: What do you base your opinion on? Have you built a CCD camers?

Regards

 

[regebro]

You're funny I enjoy talking with you. It is interesting how you speak with authority on things that you have a surface level of understanding.

Well, it may be a surface level of understanding, but it's evidently vastly deeper than yours.

Yes the TK4096 was just 2 megapixals but you missed the point. It was first made in 1985 (or so) and was 2" square! Plus it was backsided. This gives it a large QE and a bandpass at 10% QE from 300 nM to 1030nM. Outstanding for 20 years ago! And it did compete with film and did much better than film in it's intended application. And it cost $125,000 each. The world was in awe of this device 20 years ago.

Yes, I'm still missing the point. What is your point? That doing something that is the photographic equivalent of 8-bit audio sampling was hugely expensive 20 years ago? Is that your point?

CDD are read serially thus you need only one.

So?

Question: How many samples does it take to read a 8 megapixal color ccd?

Answer: 24 million. 8 million per color.


Question: how many 8 megapixal images can be sampled with a billion samples?

Answer: 41.666

Again: Your point is?

Do you have any arguments? I can't see them. All you offer are insults, pointless statements and an appaling lack of knowledge, mixed in with weird claims. You never back up any of your statements, each and every discussion we have had has ended with me disproving you, and you switching the subject. I'm not having more of these daft discussions with you. You have a huge back catalog of discussions here. Go through them, and either prove why i am wrong, or admit that you were wrong.

Then maybe we can start discussing again.

 

[evm1024]

Well, it may be a surface level of understanding, but it's evidently vastly deeper than yours.
Yes, I'm still missing the point. What is your point? That doing something that is the photographic equivalent of 8-bit audio sampling was hugely expensive 20 years ago? Is that your point?

So?

Again: Your point is?

Do you have any arguments? I can't see them. All you offer are insults, pointless statements and an appaling lack of knowledge, mixed in with weird claims. You never back up any of your statements, each and every discussion we have had has ended with me disproving you, and you switching the subject. I'm not having more of these daft discussions with you. You have a huge back catalog of discussions here. Go through them, and either prove why i am wrong, or admit that you were wrong.

Then maybe we can start discussing again.

LOL, You a gem.

You voice an opinion on the difficulty of imaging and yet you do not know how a CCD works. You tell us that it takes billions of samples to digitize an image and when I show you that your math is off by at least a factor of 40 you ignore that. You say that there are 3 A/D in a ccd camera and have no idea that ccd read out in serial not parrallel. And this all tells us that you have no idea beyond the product brochure that you read.

You want proof? Pray tell what proof do you need? Shall I find a pundit to tell you ? I am basing my opinion on my experience of designing and building a digital audio system and on designing and building 3 digital cameras.

Pray tell what do you base your opinion on?

Regards

 

[monty]

Can't we just get back to politics or religion? Something everyone knows little about other than their opinions...please?

 

[regebro]

You tell us that it takes billions of samples to digitize an image

I said million, maybe billions, and this was to get a picture of the quality of a large format film camera, yes. That statement is completetly correct, but I understand how I was unclear, and why you misunderstood me. My fault. Maybe you should have said that this was the point you were trying to make?

So, now we have cleared that up, is anything else unclear?

Pray tell what do you base your opinion on?

Opinion? Please, we haven't even gotten past your daft factual claims yet. We are miles from discussing opinions.

You have a huge back catalog of discussions here. Go through them, and either prove why i am wrong, or admit that you were wrong.

I'm waiting...

 

[evm1024]

Thanks for acknowledging your error

I said million, maybe billions, and this was to get a picture of the quality of a large format film camera, yes. That statement is completetly correct, but I understand how I was unclear, and why you misunderstood me. My fault. Maybe you should have said that this was the point you were trying to make?

So, now we have cleared that up, is anything else unclear?
Opinion? Please, we haven't even gotten past your daft factual claims yet. We are miles from discussing opinions.


I'm waiting...

Actually I consider us to be caught up. We both chose which questions to answer and which to side step.

Between the straw men, begging the issue, obfuscation, petty fogging and other such tactics I can see that there is no point in presenting logical arguments for your consideration.

Yes, just getting your addmission that you were unclear is enough for me. Of course it would have been nice to have you say that you do not have indepth understanding of CCD cameras but I know that that is too much to ask for.

take care

 

[regebro]

Actually I consider us to be caught up. We both chose which questions to answer and which to side step.

No, only you do that, but it's nice of you to admit it. Or course, matched together with your talk about serious discussions, that makes you a hypocrite.

You have a huge back catalog of discussions here. Go through them, and either prove why i am wrong, or admit that you were wrong.

Still waiting...

 

[Dr ZEE]

How about this, then:

Analog head sees an object as one to observe and accept.

Digital fan sees an object as many to register and govern.

********
Analog head wants apple juice, so grows the apple tree.
Digital fan wants apple juice, so mixing water, sugar, acids, minerals, vitamins, phytochemicals, and what ever else ....
********

Analog head is a farmer.
Digital fan is a food factory employee

*********
Some day life will be ohhhh so sweet

The funny part is:
When analog head sais to Digital Fan: "I Love my apple juice from apples, it is better than your food factory mix, digital fan replies: "It's because of your lack of understanding and misuse of our product, also our hi-end hi-price ($1000 per glass) mixes are very close to real juice if you know how to drink it properly. Our technology is growing and our product is getting more and more affortable. Some day in very close future everybody will have access to affortable pure apple juice, indistinguishable from real juice, and then Nobody will ever have to dig the dirt again" .

 

[regebro]

I don't think that type of parallells have any real use, especially when it's so competely wrong as this one, but it's cute.

But if you compare a recording of a string quartet with a synthesizer using string modeling, THEN you would be somewhere in the vicinity of "not completely incorrect"-street.

 

[Dr ZEE]

I don't think that type of parallells have any real use, especially when it's so competely wrong as this one, but it's cute.

regebro, the analogy is not about recording nor technology ... it's a silly way to illustrate the situation we are in.

The reason, that you don't think that type of parallells have any real use is because you are reluctant to think period.

I know that I am COMPLETELY WRONG WRONG WRONG WRONG.....

ohhhhhhhhh, brother

 

[SteveMac]

Listen to these engineers digital recordings and you will see that the "digital bashing" is based on equipment made long ago.

I actually heard a track from "Brother's In Arms" on the radio yesterday. Great song but the recording really did sound bad over the radio. It was one I hadn't heard before. I don't remember other songs I've heard from that album sounding so digitized. I think the FM was somewhat to blame.

 

[arjoll]

I actually heard a track from "Brother's In Arms" on the radio yesterday. Great song but the recording really did sound bad over the radio. It was one I hadn't heard before. I don't remember other songs I've heard from that album sounding so digitized. I think the FM was somewhat to blame.

Remember that FM broadcasters do particularly nasty things to the signal (trust me, I've set 'em up!). If you're listening to a local station you're listening to a playout system based on MPEG2 or MP3 compression - at 256kbps if you're lucky. Some might use uncomressed 44.1/16 but its rare. Play off CD is even rarer. If its satellite fed then you're probably listening to a 128kbps stream - possibly higher, but unlikely to be 256kbps.

After that it goes through the station's processing. If they're running a 'hot box' like an omnia.fm or Optimod 8500 then its all digital, they could be cheap and using a David-II or any one of a number of other analogue or digital boxes. The cleanest I've heard is the Aphex 2020. Its analogue but digital boxes like the Optimod are more common, and with a Compellor in the front end can sound quite good (for FM that is).

Commercial broadcast is all about getting the hottest, loudest, fattest sound in your market - you need to 'stand out' on the dial so that people don't just go past you. Jazz, fine music and other specialist stations are more likely to have clean, transparent processing than rock, CHR or other formats. If everyone is loud and horrible, being loud and clean helps, but not many bother with the clean bit.

The point? What you hear on FM is absolutely nothing like what is coming off the playout system, let alone what was on the original CD.

 

[SteveMac]

Remember that FM broadcasters do particularly nasty things to the signal (trust me, I've set 'em up!). If you're listening to a local station you're listening to a playout system based on MPEG2 or MP3 compression - at 256kbps if you're lucky. Some might use uncomressed 44.1/16 but its rare. Play off CD is even rarer. If its satellite fed then you're probably listening to a 128kbps stream - possibly higher, but unlikely to be 256kbps.

After that it goes through the station's processing. If they're running a 'hot box' like an omnia.fm or Optimod 8500 then its all digital, they could be cheap and using a David-II or any one of a number of other analogue or digital boxes. The cleanest I've heard is the Aphex 2020. Its analogue but digital boxes like the Optimod are more common, and with a Compellor in the front end can sound quite good (for FM that is).

Commercial broadcast is all about getting the hottest, loudest, fattest sound in your market - you need to 'stand out' on the dial so that people don't just go past you. Jazz, fine music and other specialist stations are more likely to have clean, transparent processing than rock, CHR or other formats. If everyone is loud and horrible, being loud and clean helps, but not many bother with the clean bit.

The point? What you hear on FM is absolutely nothing like what is coming off the playout system, let alone what was on the original CD.

I've noticed that radio is sounding worse these days. I remember it sounding alot better before. The station is only a couple towns over. Funny, if they want it to be louder because if you turn it up it all goes to hell.

 

[regebro]

This is what happens when cowards can't handle the truth:

You pussy assed whiner! You're a mental midget. You have an unrelenting need to argue. You always think you're right, but you're not. Fuck off!

Pathetic cowardly looser.

 

[cjacek]

How about this, then:

Analog head sees an object as one to observe and accept.

Digital fan sees an object as many to register and govern.

********
Analog head wants apple juice, so grows the apple tree.
Digital fan wants apple juice, so mixing water, sugar, acids, minerals, vitamins, phytochemicals, and what ever else ....
********

Analog head is a farmer.
Digital fan is a food factory employee

*********
Some day life will be ohhhh so sweet

The funny part is:
When analog head sais to Digital Fan: "I Love my apple juice from apples, it is better than your food factory mix, digital fan replies: "It's because of your lack of understanding and misuse of our product, also our hi-end hi-price ($1000 per glass) mixes are very close to real juice if you know how to drink it properly. Our technology is growing and our product is getting more and more affortable. Some day in very close future everybody will have access to affortable pure apple juice, indistinguishable from real juice, and then Nobody will ever have to dig the dirt again" .

You know, I pretty much agree with your analogy. Just to reiterate and add my own: Analog is like a tree that gets planted in natural, God made soil if you like, that grows fruit as nature intended ... whereas Digital is a man made factory which produces "engineered" food with the idea to make it look, taste and be as healthful as the "real thing".

I see "Analog" as nature intended and I'm very comforted by that.

~Daniel