How important is 24 bit?

  • Thread starter Thread starter sai
  • Start date Start date
S

sai

New member
I Understand the theory about the 16 bit and 24 bit business. I want somebody's input on how big a difference 24bit really does as opposed to 16.

Is it really worth considering a mixer (I'm considering one) and spending some extra bucks that writes at 24 bits?
 
The difference is 8 bits......... 24-16 = 8

The difference is also in having 16777216 values to represent your signal digitally instead of just 65536.

As well, the difference is similar to looking at a JPG using only display settings of 16-colors vs. a million colors.

Can YOU see the difference? ;)
 
I can get flopy disk boxes with 120.795.955 bites inside for just $2.5. I f someone is interested please email privately.
 
As well, the difference is similar to looking at a JPG using only display settings of 16-colors vs. a million colors.
Click to expand...

Actually it would be more like the difference between looking at a jpeg using 65536 colors and 16777216 colors. Semantics.. :)
 
Assuming you are using a decent mic and pre and know what you are doing the difference is very noticeable.
 
...and how many people in the mp3 clinic, can you therefore tell if they record at 24 bits or less than...?

I know my answer, but I'd be interested in others.

this may be a loaded question too, I suppose
 
Yo people of bits:

I can run 24 bit out the back of my 2816 into tape and the tape is wonderful.

I do that when I'm in a hurry.

But, I do record most stuff in 16 bit so I can burn the CDR.

Green Hornet:p :p :p :cool: :cool: :)
 
mixmkr said:
...and how many people in the mp3 clinic, can you therefore tell if they record at 24 bits or less than...?

I know my answer, but I'd be interested in others.

this may be a loaded question too, I suppose
Click to expand...

Half the people are still using casette. I'd bet very few are recording in 24bit and it usually shows.
 
mwarkentin said:


Actually it would be more like the difference between looking at a jpeg using 65536 colors and 16777216 colors. Semantics.. :)
Click to expand...
Well I had mentioned those numbers in the 2nd point I made -- I just didn't want to be repetitive! ;) :D
 
Yo Fly in the Knight:]

I'm not sure I understand your dithering question. But, the 2816 can record in 16 or 24 bit and 44 and 48.

44/16 is what I use most of the time. I can record in any combination of the above; however, the unit will not burn a CD in 48/24 mode.

Green Hornet:p :p :p :cool:
 
Allow me to add my 2 cents (and as I like to say, it's in an E*Trade account, so now it's worth even less ...):

* I've recorded into a Roland VS880 -- 18-bit converters, 24-bit mixing section, 16 bits actually stored on the hard drive

* I've recorded into an Echo Gina 20, which is a PCI card/break-out box (with converters in the box for reduced noise issues than if they were on the card) with, as the name implies, 20-bit converters.

* I've recorded into a Pro Tools Digi 001 as well as a Digi MBox, both 24-bit converters.

* I've bypassed the converters of all the above gear by going SPDIF thru the 24-bit converters in my Mindprint DI Port.

What I've found is, the farther down this list I went, the better the sound quality became. Unquestionably so. AND, it proved to me that the bit resolution at which data's stored is just as important as the bit resolution of the converter capturing the sound into the system -- the DI Port was a vast improvement over the VS880's stock converters, but the DI Port through the PC into a 24-bit session was much better. This was true even when the 16-bit sessions in the Roland were in their "uncompressed" data mode (meaning, actually 16 bits .. it's this weird thing Roland does in some recording modes but not the mode I tested, intentionally).

It also proved to me that the quality of converters of the same bit resolution can affect the sound, as can the pres/gain stage.

I typically record in 24/44.1, then use the dithering algorithms in my software (PT LE) when making a stereo master. It sounds better than not dithering and better than a 16-bit session, in my case.

I hope this is useful.
 
Yo, Bee-in-the-Bonnet - Sorry my Q was too succinct - I'm not very familiar with the Yamaha workstations, other than hearing very little negative feedback. What I should have asked, was why you weren't taking advantage of the 24 bit's extra headroom, but still at 44.1 to keep rate conversion out of the equation - then dithering down from 24/44.1 to 16/44.1 for your "scratch" CD's. I'm not sure if there is a time penalty for dithering, but there is usually one for sample rate conversions depending on processor speeds and architecture.

If there is no time difference between a realtime dub (analog) and burning a CDR from 24 bit down to 16, then I would have to assume that recording at 24/44.1 would be your best quality option on the 2816. Just my "iggorant savage" assumption, since I do everything thru a 20 bit converter into Samplitude, work and store at 32 bit float, and only dither back down for CDR - all at 44.1 until I find a proper interface between my DM-24 and DAW, at which time I'll be at 88.2/32float with 24 bit inputs. My new DAW, once it's optimized, should be capable of over 40 tracks at 32/96, so I see no reason not to go for it... Steve
 
Blue Bear Sound said:
As well, the difference is similar to looking at a JPG using only display settings of 16-colors vs. a million colors.

Can YOU see the difference? ;)
Click to expand...
Bruce makes a good analogy, but the point even goes beyond the difference in final resolutions between two files. Even if the final resolutions are the same, the path by which they got there is also very important

I have a bunch of friends who are graphic designers. They always do their productions in a higher resolution than what they ultimately need. Just like audio, all the processing they do to their images degrades them significantly. When they do everything in high resolution then dither down at the end, the final product is higher quality.

I did a quick experiment in Paint Shop Pro.

In the first case I took an 8 bit color image and did a few processing effects to it, rendering each successive generation also as an 8bit color image. In the second case I started with the exact same 8bit color image and did the exact same processing, but rendered each generation in 24 bit color, then finally dithered down to 8 bit color at the end.

Attached is the original image.


barefoot
 

Attachments

  • origional.webp
    origional.webp
    8.8 KB · Views: 111
Now here are the two different versions – 8 bit processed and 24 bit processed. Can you tell which is which?
 

Attachments

  • compare.webp
    compare.webp
    21.6 KB · Views: 123
24 bit is the 2nd one.
you can see (top left and bottom left) where the colors overlap...a LOT smoother look..
 
Yo Knight-flyer firefly:]

The 2816 has more headroom, in any configuration, than a convention of bald-headed men and women.

I have used 24 bit and run tape out of the back end of the 2816; great results; good results using 16 bit too.

I may need more inculcation bit by bit.

Thanks,
Green Hornet
 
The eye and the ear are two different devices, and the way they are calibrated is radically different.
 
Its true... but which is more sensitive?? In the physical realm... light and sound are the same thing... light is just faster.

xoxo
 
You must log in or register to reply here.
Back
Top