More digital clarity...

Good find, Jon. I think it probably puts to rest a bunch of arguments.

WOW.

THE 2ND ONE IN PARTICULAR.

ok 


cassette tape  =  6-9 bits deep at best.


the very best open reel analog tape recordings, with advanced noise reduction  =  13 bits.

batsbrew


so what the heck are all the 'audiophiles' really hearing?

That there has always been my argument. I knew for a fact that even a 2 inch tape machine spinning at 30ips wouldn't be 16 bit. I didn't know it was roughly 13 bit but have always wondered what it would be. (which is good to know!) Yet we get all these guys trying to sell us on 32 bit float points and excessive dithering conversations due to "math not being right". If my ears don't hear an issue that bothers me when comparing something, there IS no issue worth talking about in my realm. My rule of thumb is, if I can't hear a difference while listening 3 times to something while being super focused, it's nothing worth changing or discussing.
 
Dithering: I do it because I'm told I'm supposed to. When I've not used any dithering in a project and have just set it for 16/44 within the 24/48 realm during an export....nothing has ever sounded wrong or out of place to me. Have any of you heard any weird anomalies?
 
Dithering selections: I've not heard anything super blatantly obvious, however, on certain material, the type of dithering *can* make a difference. I dither using something completely different than most people...and it works for everything, so I don't stress about it at all. :)
 
It all turns out to be more hype for us to be sold on in my opinion.
 
-Danny

oh yea....

i always use the waves IDR dither when i render my files to 16 bit....
it does SOMETIMES sound better with the noise shaping.

jamesyoyo


Good find, Jon. I think it probably puts to rest a bunch of arguments.

Nah, folk will just try to create new arguments...
 
 
It was our own DrewFX that put the stair-stepping myth to bed for me quite awhile ago now, but seeing it shown here really brought home the idea that you can't plot a wave of a certain bandwidth without getting the resultant equation exactly right at any point in time.
 
Karyn also corrected me fairly recently in an excellent post she made in a thread about modelled DSP, when I said digital was cold and stark and she said no, think of it as being 'pure'.  A distinction that made a huge difference to me in how I approach the idea of adding 'coloured' signals to a mix.  It will save me a fortune in expensive 'modelled' plug-ins as I've really moved away from those things these days, and look to add colour as a seperate process in itself not as a by product of using an ecclectic mix of these kinds of things.  I regain control of when, where what kind and how much character I introduce as a result without thinking I have to add the latest marketing snake oil to make magic.
 
Which shows again there ARE some bright, clued up people around this place it's just a case of sifting the wheat from the chaff.

Jonbouy

jamesyoyo


Good find, Jon. I think it probably puts to rest a bunch of arguments.

Nah, folk will just try to create new arguments...
 
 
It was our own DrewFX that put the stair-stepping myth to bed for me quite awhile ago now, but seeing it shown here really brought home the idea that you can't plot a wave of a certain bandwidth without getting the resultant equation exactly right at any point in time.
 
Karyn also corrected me fairly recently in an excellent post she made in a thread about modelled DSP, when I said digital was cold and stark and she said no, think of it as being 'pure'.  A distinction that made a huge difference to me in how I approach the idea of adding 'coloured' signals to a mix.  It will save me a fortune in expensive 'modelled' plug-ins as I've really moved away from those things these days, and look to add colour as a seperate process in itself not as a by product of using an ecclectic mix of these kinds of things.  I regain control of when, where what kind and how much character I introduce as a result without thinking I have to add the latest marketing snake oil to make magic.
 
Which shows again there ARE some bright, clued up people around this place it's just a case of sifting the wheat from the chaff.

Well said Jon. I believe Drew and Karyn also set me straight on this as well as a few conversations with bitflipper. I really love those people because even though they bring the attention of science into things, they know how to explain things correctly to where even a dope like me can understand them. :)
 
On the "cold and stark" comment...this is common here with my older students that come from the days of tape. What THEY didn't understand was, because digital is so pure, it can also be unforgiving which means...a sound that might work in the analog realm with tape saturating or compressing, will not be the same in the digital realm. A sound that might be a bit harsh or brittle where analog can automatically curb this, won't be the case in digital.
 
So if anything, we just print our sounds a bit differently. An eq can warm something up if need be and with the additional clarity we get using digital, a good eq that low passes just the right way...and you can get similar results along the lines of tape minus the real saturation.
 
You know, I too was staying away from the modeled analog souding pieces, but the UAD stuff has changed my mind. I think a hybrid of the two (digtal and analog) makes for an awesome team. The fact that UAD has a good handle on "saturation" as opposed to what the others "claim" to be saturation (which is really a form of ugly distortion or a kind of bad clipping sound) is what makes up the real difference. The UAD stuff is so powerful and transparent these days, I wish it was a bit more affordable to people so they could have them in their tool box. They may not be for everyone, but when the right amount of "stuff" is used when using their plugs, it can really make a difference and take a little 'edge' off of the digital sound in a GOOD way. :) Good find on those vids...thanks for sharing brother.
 
-Danny

batsbrew


some of the anomalies of analog recording, is the very thing a lot of folks like the sound of.

that said, modern digital (meaning, 2013) gives you more clarity....
which mean, if you don't like treble being so 'clear', maybe you oughta engage that linear phase eq and smooth it right on down.......!

LOL

Exactly,
 
Once you've identified the elements you like about using steam powered technology you can remove all the good stuff from digital and make it sound as quaintly bad as old gear.
 
Add about 50 db of gain to a dithered silent track and you've got your tape hiss, put a big bump somewhere between 40-120kHz and slowly filter out anything above 6kHz and you've got some pretty cool tape-emulation going on straight-away.  Set up a limiter to give you a false ceiling level, get some extreme compression going starting 6db below the limiters peak to make like you are spanning the red line on a tape vU meter and you're done.  Instant ReVox...
 
Add some subtle wavering pitch shift to emulate wow and flutter to taste.

mike_mccue








The two places shown above that say "704"....


They are supposed to say "740".














Just saying.

Just saying what though?

mike_mccue








The two places shown above that say "704"....


They are supposed to say "740".


Just saying.

For those who don't want to bother checking the math themselves:

704*10 = 7040

480*11 = 5280

7040/4 = 1760

5280/3 = 1760


IOW, if we want to use 740 instead of 704, either the picture aspect ratio isn't 4:3 or the pixel aspect ratio isn't 10:11. 

[EDIT: I hope that further clarifies things.]

Danny Danzi


So if anything, we just print our sounds a bit differently. An eq can warm something up if need be and with the additional clarity we get using digital, a good eq that low passes just the right way...and you can get similar results along the lines of tape minus the real saturation.
 
You know, I too was staying away from the modeled analog souding pieces, but the UAD stuff has changed my mind. I think a hybrid of the two (digtal and analog) makes for an awesome team. The fact that UAD has a good handle on "saturation" as opposed to what the others "claim" to be saturation (which is really a form of ugly distortion or a kind of bad clipping sound) is what makes up the real difference. The UAD stuff is so powerful and transparent these days, I wish it was a bit more affordable to people so they could have them in their tool box. They may not be for everyone, but when the right amount of "stuff" is used when using their plugs, it can really make a difference and take a little 'edge' off of the digital sound in a GOOD way. :) Good find on those vids...thanks for sharing brother.
 
-Danny

Ah but I wouldn't be doing without my UAD stuff, that is part of my character adding arsenal.  What I mean is I'm not on a quest to find that latest bit of modelled gear.  I use a fair bit of Bootsy's freeware stuff for that as well as he has a really good handle on 'stateful' algorithms which means how they behave is dependent on what the model is doing at any given moment, or what 'state' it is currently in.  The easiest way I can think of describing what 'stateful' means is for example how an amp sounds different when it has warmed up rather than when it is cold.  So the two simples states there are warm and cold, it gets more fun when the behaviour of an amp changes when it becomes more under heavy load, so rather than just having a linear equation between low and high volume the actual calculations change depending on what the model is doing.  Check out his Levelling Amp, (ThrillseekerLA) to hear it working in action.
 
Another thing I've tried with some success when I want to add saturation and harmonics to sounds with strong fundamentals is to put a driven guitar amp model on a send and just roll in (-40-30db) of signal back into the clean signal, you'll want a reasonably clean but driven amp sound, but it sounds great if you've got thin strings or pads or want a vintage keyboard sound to have a bit of tube like character to it.

Learn something everyday:


http://en.wikipedia.org/wiki/DVD-Video

I've never seen anything but 720x480 (and the older 640x480) but it turns out there is a definition for 704x480 as well.

Non square pixels sure are fun. :-)


best regards,
mike

I like green lollipops the best.

It seems like he is describing a world that has passed us by.

Now a days we shoot in digital, we transfer in digital, we edit in digital, we playback in digital, and our display screens are built around clusters of pixels.

It use to be "line", screens and apertures with lots of A to D and D to A transfers.

Now a days thinking about pixels seems entirely appropriate.


I wish I had not made the initial casual and erroneous comment... :-)  Now I am having to actually think.



all the best,
mike

Put the lime in the coconut.

I think of bit depth on the Z axis. It defines the color space and light value dynamics rather than spatial resolution

I've seen plenty of vectors in video edit/composit/motion graphics applications but I have never seen vector math used in any digital video distribution formats other than "flash" and "silverlight".

Maybe Blue Ray authoring has some?



best regards,
mike