Why Analog?

[themaddog]

Firstly, I sometimes wonder why I occasionally find myself actually defending digital down on this little forum. What? There's no need to defend. Digital everything has had massive, worldwide acceptance. Hugely successful. Sure there are some still using analog in various areas, and that's fine. Vinyl record sales have apparently been on the increase but still they are a tiny minority.

No, I do it when someone seems to, intentionally or not, misrepresent the facts.

Funny that here we are on this Analog Only www forum which if not for digital wouldnt be here!

Here we are now on Analog Only exchanging photo, video and audio clips.
Compressed digital formats BTW like jpeg, mpeg. Apparently you/we judge that the loss of quality in both audio and video is an acceptable trade off compared to not sharing them at all. Good. So does most people. I use YouTube not because of its quality but its accessibility and the vast range of stuff available.

You maintain mp3's have been a "dramatic step back in terms of sound quality". Well in one sense yes but even with wav. too you could always record at lower bit and sample rates to cut file size, with a trade off in sound quality. Do we need hi fi stereo phone lines? Of course not. The phone service always traded off sound quality so we could have the service and at least talk to each other.

And it's not as if we started with some poor quality mp3 file and were hoping for an improved CD quality file down the track to supplant it. Our first consumer experience of digital audio was CD's. mp3's came along later as an efficient file size saving method.

Even so, it's dangerous to generalize about the audio quality of mp3's or similar audio compression codecs. There is no one standard of mp3 file compression, just as there's no one analog tape format, tape speed, track width. You cut your cloth according to your audio quality needs.

For much popular music, mp3's at a reasonably high bitrate are not a compromise in sound quality. Reason? Much popular music has narrow dynamic range and so little or no information at the deeper bit levels. So if you compress correctly you are only eliminating bits that are not even recorded on in the CD format or higher. They were blank all along, or at the least contained only noise that is irrelevent to the performance itself.

If you're a kid listening to an iPod on a noisy bus or train the last thing you need is a huge dynamic range anyway. All the quiet stuff that you might hear in a quiet living room is likely overwhelmed by the vehicle noise. You want the opposite. A narrower dynamic range to hear the quieter stuff above the vehicle noise while the louder sections of the track still dont blow your brains out.

The problem can come though when people take a standard CD with a big dynamic range, simply file compress it to a low bitrate and then wonder why the quiet sounds disappear. Of course they disappear because you told the codec to eliminate all sound below a certain threshold. It's just doing what it was told to do. Again, lack of understanding.

This in itself is not an analog vs digital issue. It's just a practical human hearing issue.

Any way, enuff from me for the moment.

Cheers Tim

I understand about this trade-off and tried to detail my feelings on that in my post. MP3's are great for what they are, but now they are the preferred "medium" (as an MP3 is really just a file that is stored on many unique kinds of physical medium). This is where I see certain drawbacks, and where I see music taking a few steps backwards in terms of audio quality. I know you write that most people can't hear the difference with the right kind of compression, but I think today, even with the best variable bit rate compression, the cymbals still sound like crap. As "mediumless" music becomes more or more mainstream, I believe recordings will become even more disposable, without the option of purchasing a CD for superior quality over an MP3.

On an optimistic note, the quality of MP3's, or whatever the next hip file format happens to be, should certainly improve with time.

My studio has both an analog recorder and a digital recorder. I like digital for what it is: cheap and quick. I do my demos on digital now, since I'm doing the whole one man band thing, and if there's a really good take that I wanted to keep, I could record to tape. But, when it comes to mixing, I'm much more comfortable and confident in getting things to sound as I want them to on my tape machine through my console. When it comes to mixing in the box, although the computer mixes everything consistently each time, I can't achieve the same quality I feel that I can on my mixer; I just don't have the patience. I haven't had the opportunity to work on a DAW that operates more like a tape machine (standalone multitrack hard drive like a RADAR or an Alessis Hard Drive ADAT), and who knows, maybe if I did, I'd be won over.

-MD

 

[SteveM]

Are you talking about legendary VS-880? I remember wishing one badly
Didn't it have some special "modes" of sort...
"Roland RDAC is a "lossless-compression" scheme."
...heh heh
Do you still have the machine? If so, keep it. VINTAGE!!!!!

Yep, well it's the 1880. It's still useful for certain things. The reverbs are prety good too if I use it as a return send. I'll probably keep it for a while.

 

[Dr ZEE]

Yep, well it's the 1880. It's still useful for certain things. The reverbs are prety good too if I use it as a return send. I'll probably keep it for a while.

Arghhhhhh, not so "legendary" then
Keep it forever. It's a cool applience. And you can't sell it for good anyways, maybe for about 300-400 bucks (???) ...
Is it SMPTE-slaveable????
You know what I did?... I refired my linked pair of AKAI DR-4d(s), installed new hard drive (Sea-Food type - Barracuda ) into the older one and installed SMPTE card into one of them. And I slaved DR4s to Teac A3340s , and also I SMPTE -ed Roland's 15 chanel "automation console" with VCA-unit into the "system" (here's the device: http://www.mzentertainment.com/studio_workshop_roland_cpr_800_vca_800_automated_mixing_system.html ).
So now I have three "legendary teac" tape tracks for important stuff (drums and bass get priority ), 4 pairs of digital(s) for not so important stuff... and I can mix all that crap on analog console and use Rolands' automated VCAs when I run out of hands to move things that needed to be moved during mixing.
I get a real kick out of watching , when hitting play of on a3340s, how everything else around obey and follow the orders from the Master

 

[SteveM]

Arghhhhhh, not so "legendary" then
Keep it forever. It's a cool applience. And you can't sell it for good anyways, maybe for about 300-400 bucks (???) ...
Is it SMPTE-slaveable????
You know what I did?... I refired my linked pair of AKAI DR-4d(s), installed new hard drive (Sea-Food type - Barracuda ) into the older one and installed SMPTE card into one of them. And I slaved DR4s to Teac A3340s , and also I SMPTE -ed Roland's 15 chanel "automation console" with VCA-unit into the "system" (here's the device: http://www.mzentertainment.com/studio_workshop_roland_cpr_800_vca_800_automated_mixing_system.html ).
So now I have three "legendary teac" tape tracks for important stuff (drums and bass get priority ), 4 pairs of digital(s) for not so important stuff... and I can mix all that crap on analog console and use Rolands' automated VCAs when I run out of hands to move things that needed to be moved during mixing.
I get a real kick out of watching , when hitting play of on a3340s, how everything else around obey and follow the orders from the Master

Yes, tape is the master. That's cool Mike.

Yes, it's got SMPTE. I've used like that quite a few times and it works very good. Once I had each machine going out a separate set of speakers, like quad and it sounded pretty great.

I lost my direct outs on it for some reason. Software.
It used to direct out 8 channels at a time. Which would be very useful if it worked. It'd be worth getting it fixed.

 

[Dr ZEE]

Steve, are you the original owner. If so you may be able to get software related problems fixed for no charge, you need to contact Roland (actually by phone, if it's possible), explain the situation and complain the hell and demand restitution or else!!!!.... heh heh heh. If you can't do this kind of talk, ask somebody who's good at it. I did so with my E-MU synth that was past the warranty period. I asked my wife to call them , got it fixed free that way. If you open this kind of gear, you'll not find much there that can be "fixed" diy style on the kitchen table ...

 

[SteveM]

I wouldn't even bother opening it.

 

[Tim Gillett]

One thing we haven't talked about is amplitude ceiling and how digital and analog mediums handle the "current" when it reaches that ceiling...herein lay vast differences between analog and digital...

Time for a nap.

OK lets talk about it.

Digital stays undistorted pretty well up to the 0 point and then just clips, much like an analog amp when it runs out of power rail voltage.

Analog tape more gradually enters distortion but eventually is just as badly distorted as digital. But analog tape overload can be a nice production effect if used wisely on some material.

But because digital has not the inherent noise of analog (any analog) recording means, you actually have much more headroom because you dont have to track nearly so hot to get above the added noise of analog.

Analog can only match this lack of noise with compander noise reduction circuitry which when well set up can be pretty transparent. But why resort to the complexities of NR when you have a recorder which doesnt even need it?

Besides, the common dbx NR pretty well negates the sometimes nice production effect of tape warmth because the non linear nature of the tape distortion (at higher levels) throws the dbx decoder into confusion.

If you want to use analog tape warmth you have to use either no NR (and live with the noise) or use Dolby which only tracks on the lower level stuff and leaves the high levels untouched.

Over to others.

Cheers Tim

 

[sweetbeats]

Maybe this will help to clarify my frustration with digital's "0 point": my primary instrument is percussion, and at least in my limited experience most percussive instruments have challenging transient peaks and analog (subjectively) handles them much more to my satisfaction than digital...unless I'm "printing" to digital using dynamics processing to try and manage those peaks (and if I were to shoot for average levels close to 0 the processing would have to be pretty aggressive which feels very limiting to me, no pun intended) I have to record well below 0 to avoid clipping and then I'm missing out on taking advantage of available bit depth and I'm closer to the noise floor of the rest of the system...tracking at 0 ~ +3 average levels compared to -10 is significant to me. I'm not trying to argue about what's better, I'm just bring an application issue to the table here that hasn't come up in this discussion that is a real-world issue for me in my neck of the woods.

 

[famous beagle]

But why resort to the complexities of NR when you have a recorder which doesnt even need it?

That's the thing. Plenty of people have given you plenty of reasons for this preference .... yet you just don't want to hear them I guess.

 

[evm1024]

SACD will save us all, but the point is....

Let's do take a look point by point.

OK lets talk about it.

Digital stays undistorted pretty well up to the 0 point and then just clips, much like an analog amp when it runs out of power rail voltage.

I notice that you minimize digital clipping "just clips" and liken a 0dBFS signal to analogs soft clipping. Digital's hard clipping and in many cases unpredictable results at 0dBFS are "the same as in analog". NOT!

Please show us the waveforms and spectra of a +1dBFS digital system and an analog amp driven 1 dB past the supply rails limits.

Analog tape more gradually enters distortion but eventually is just as badly distorted as digital. But analog tape overload can be a nice production effect if used wisely on some material.

Some really key words here. "Gradually enters distortion" for analog. So as you overdrive analog it eventually distorts as BAD as digital does as soon as it hits 0dBFS. Ouch, can you say harsh. We need spectra. All distortions are not created equal as well. Not saying distortion is a good thing (Hah, fuzz box) but some distortion grates like a chalk board. Question for reader Digital or analog is like a chalk board? Oh, but let's skip over that.

Oh, one more thing overlooked. Digital has the least distortion at just under full scale. All bits are used to represent the signal. As the signal level get lower and lower, digital has fewer bits to work with and the quality of the reconstruction filters come more and more into play. Maximum distortion at minimum signal.

Analog tape on the other hand has minimum distortion at minimum signal. Read that again. It is only at high levels analog tape begins to distort significantly and because analog does not run out of bits at low levels we are not always driven to record at the maximum level. We get to pick our level without fear.

Simple math... at full scale digital CD have 16 bits. For every 6.02 dB down we lose a bit. This is why I would like to see (as a minimum) 24 bit floating point for a new standard. Then perhaps digital can live up to its potential Dan Lavry not withstanding.

But because digital has not the inherent noise of analog (any analog) recording means, you actually have much more headroom because you dont have to track nearly so hot to get above the added noise of analog.

Let's ignore that digital systems are analog systems with all the noise that goes with it up to the ADC and after the DAC....

This is so misleading and not just for Tim. This may be one of the major reasons for CD's harshness. A rethinking is required. Digital has no headroom. Digital is a top down recording scheme. Max signal, minimum distortion, most bits used. Back off a few bits, oh say 3, to get some headroom and you get -18 dBFS for "0 VU". Now you are playing back mapped into 13 bits - count on the reconstruction filters to predict how the analog output of your DAC should be. 2^13=8192

The point being is that you have to track hot in digital to get the full potential. And everyone does. 96 dB signal to noise ratio is math not reality.

Plus, when tracking hot in digital you can still exceed 0dBFS even when all your "numbers" are less that full scale. Reconstruction filters have overshoot as a requirement. Take a look here: http://www.cadenzarecording.com/papers/Digitaldistortion.pdf

Analog can only match this lack of noise with compander noise reduction circuitry which when well set up can be pretty transparent. But why resort to the complexities of NR when you have a recorder which doesnt even need it?

Let's see complexity - companding or digital. Right Tim, digital is much less complex than a DBX compander.

Test - take a complex source (music) record it with the peaks 30 dB under the max (0dBFS and 0VU) on a digital and tape system. Now play it back with that 30 dB made up in the amp gain.

Besides, the common dbx NR pretty well negates the sometimes nice production effect of tape warmth because the non linear nature of the tape distortion (at higher levels) throws the dbx decoder into confusion.


If you want to use analog tape warmth you have to use either no NR (and live with the noise) or use Dolby which only tracks on the lower level stuff and leaves the high levels untouched.

I see that you understand some of the tradeoffs that are available to tape. But why cast this as a negative... unless you are biased?

Over to others.

Cheers Tim

So Tim, I've always wondered whats your agenda? What are you trying to prove? Why post and what results do you want to see?

Regards, Ethan

PS back to the question I asked. Have you thought about my comment "humans know what to do". Any thoughts on that?

 

[Dr ZEE]

....

(with)....digital..you actually have much more headroom ...Cheers Tim

oh boy,
it's just getting better and better ...

 

[Dr ZEE]

waveforms of.... "an analog amp driven 1 dB past the supply rails limits."
huh, Ethan, I've got some of those somewhere in my "archive", including videos. Those a beautiful

also, I don't like using term "supply rail" in that respect, but that is a detail that somewhat can be ignored, I guess...

 

[evm1024]

waveforms of.... "an analog amp driven 1 dB past the supply rails limits."
huh, Ethan, I've got some of those somewhere in my "archive", including videos. Those a beautiful

also, I don't like using term "supply rail" in that respect, but that is a detail that somewhat can be ignored, I guess...

I completely agree with you here. I just kept it simple. Some analog devices clip at the supply rails and others at some lesser voltage. Whole analog systems clip somewhere less and soft clip at that.

Like I say I was not trying to be exhaustively precise in my use of words and rather left it up to the thoughtful reader to understand the limits of those words.

Regards, Ethan

Now if only the horse would drink....

 

[Dr ZEE]

"past the supply rails limits" is much better than "analog amp ... runs out of power rail voltage" heh heh...
Don't mean to pick on words, at all, it's just sometimes drives me nutS, especially when a statement is a full of passion and being presented as a ray of enlightenment pointed toward the darkest corners of The House of "Ignorance"
************
Choose your words carefully!, that's what True Kings do

 

[SteveM]

I would have preferred: "Can't they all just get digital and leave all the good stuff for me?"

They've abandoned the tape decks. Now if we could just get them all to move on to digital mixing consoles.
We would be very wealthy people.

 

[Dr ZEE]

Water came, but the horse is still thirsty.

...if only the horse would drink....

Ethan? Have YOU "created" that (specific) phrase or is it borrowed?
I can't tell you how much I dig that one
Googled it and it took me straight back over here ...heh heh

 

[Dr ZEE]

They've abandoned the tape decks....

Yeah, I know what you mean, but it's not 100% so, not even close.
Well, it's not a serious part of "discussion" though ...(did I just say "discussion" ??? )

....wealthy people...

Hey, Hey! Choose your words carefully, Spartan!

 

[SteveM]

Yeah, I know what you mean, but it's not 100% so, not even close.
Well, it's not a serious part of "discussion" though ...(did I just say "discussion" ??? )


Hey, Hey! Choose your words carefully, Spartan!

Well Mike, if it were 100% WE wouldn't have them either!

 

[evm1024]

Borrowed

Ethan? Have YOU "created" that (specific) phrase or is it borrowed?
I can't tell you how much I dig that one
Googled it and it took me straight back over here ...heh heh

The original phrase is quite old of course.
http://www.phrases.org.uk/meanings/you-can-lead-a-horse-to-water.html

I guess I did originate this specific phrase "if only the horse would drink".


Regards, Ethan

 

[Dr ZEE]

if it were 100% WE wouldn't have them either!

Huh! My math sucks. ...LOL
Good one!