Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

Total Members Voted: 0

[Matt_G]

For those users here that are on OSX & that currently use Pro Tools, do me (& yourself!) a favour & download Reaper & post your thoughts about it here. The trial is unlimited & is fully functional http://reaper.fm

Matt

[Matt_G]

zmix wrote on Tue, 17 June 2008 01:20

Can you please provide some more information or possibly a link to qualifty this claim?

Here is some more links for you...

   http://www.gearslutz.com/board/music-computers/154664-wavebu rner-effing-up-my-tracks.html

   http://www.gearslutz.com/board/mastering-forum/154669-wavebu rner-problems.html

   http://www.gearslutz.com/board/mastering-forum/195415-more-w aveburner-tests.html

Be patient & makes sure you read through the threads & do the tests yourself before deciding that they are invalid.

Quote:

Gearslutz? Are you pulling my leg? The first relevant quote I found in that thread was this: Quote: Even 1.5.2 will only burn bit accurate discs from 16 bit source files with no processing. __________________ Obviously this was posted by someone who does not understand the following basic facts: 1:  You cannot burn a red book audio CD at a greater bit depth than 16, so any file with a greater bit depth than 16 bits will be truncated and dithered.   This process will not allow the original higher bit depth file to be extracted from the stored 16 bit file. So, yes, NOT bit accurate, but hardly unique to Waveburner.     2:  Any signal processing, even a level change will alter the original data.... Again, EDIT: Further idiocy in that *gearslutz* thread has resulted in the unsubstanciated claim that Waveburner truncates to 15 bits...   If you truncate and  - dither - a 24 bit file you will NOT get a null when compared to a 24 bit file or even a 24 bit file truncated to 16 bits.   Dither needs to toggle the LSB to be effective, therefor only the upper 15 bits will null against the original 24 bit file or the truncated to 16 bit file. Verdict? NOT A BUG

Respectfully you don't know what you're talking about. Please try this test on your copy of Waveburner (any OSX version has this same problem even the latest version 1.5.2). This bug does not exist in the OS9 version of WaveBurner so it's only in the OSX version, the following test validates all these claims...

1. Take a 16bit file convert it to 24bits with any application other than WaveBurner. Yes this will add an extra 8bits of padded zeros, stay with me...

2. Add this newly created 24bit file into WaveBurner & without doing any fades, volume changes, processing or adding any dither options (dither turned off in WB prefs)  bounce this region to hard disk & select 16bit as the word length.

3. Take this newly created 16bit bounce & line it up with the original 16bit file in the DAW of your choice (Pro Tools, Logic etc.) make sure they are lined up to sync accurately & flip the phase on one of the files & listen to the result (check the peak meters to see the results as well).

In all my tests (these were extensive!) Any 24bit to 16bit processing, bouncing or burning in WaveBurner (without dither!) resulted in a cancellation of only -90.3 against the original 16bit file. The same tests done with other applications such as WaveEditor, SampleManager & Pro Tools resulted in complete cancellation down to infinity.

Ok so it may not be dropping a bit but, the results are conclusive that WaveBurner is not bit for bit accurate when handling 24bit files & converting them to 16bit.

Do the tests as we all have done & you will see for yourself.

Matt

[tom eaton]

Hi Matt-

Can you think of a time you would want to go from 24 to 16 bit without dithering?  I can't.

-tom

[cerberus]

i just  delivered my first job that used my new system : with cubase 4 and izotope rx.  
i think it could be my best work ever, but that is subjective... i did however
feel more in control than ever, so if it is sent back for a recall, it's my
fault, not the soft....  the macpro is blazing, and stable like a rock.

one thing i have to say about getting a new system.  i upgraded to
a much finer a/d/a and i've unwittingly been bent toward some
kind of audiophilia.... for example:  i said to a client,
"would you like some of the pops and clicks to
be fixed?" and he kinda looked like he
wasn't quite sure what i meant!

and i observed that on my old system, those aren't
so noticible, of course. and likewise on the
client's krk monitors.

have any of you ever been "changed" because of
getting new gear?  how do you deal with this
problem in general?  maybe say:"is it ok to
ignore the pops and clicks i hear?" rather
than: "would you like them fixed..." ?

jeff dinces

[cass anawaty]

I had to send screen shots of spectral analysis to convince one guy they were there.

"Must have happened during the ftp transfer", he said.  

I always tell them, because I had one guy who had installed some firewire drivers for his AD/DA 16s, and they did something to the calibration that resulted in barely audible clicks, pops,and distortions.  He was very grateful I let him know.

[zmix]

tom eaton wrote on Mon, 16 June 2008 13:36

Hi Matt- Can you think of a time you would want to go from 24 to 16 bit without dithering?  I can't. -tom

[masterhse]

tom eaton wrote on Mon, 16 June 2008 13:36

Hi Matt- Can you think of a time you would want to go from 24 to 16 bit without dithering?  I can't. -tom

Well, the example that he gave would actually be one, though I don't know why someone would import as 24 bit to knowingly burn at 16 bit unless there was some sort of software limitation.

Nevertheless it should work reliably this way. For it not to work points to a potential underlying issue that reduces confidence in its use.

[jfrigo]

zmix wrote on Mon, 16 June 2008 11:20

Gearslutz? Are you pulling my leg?

A bunch of the same regulars over there as here as on Glenn's board. I don't think everybody loses their minds just because they hit "send" on a different board.

Do the tests yourself and report back on what you find. Your point about dither not nulling is certainly worth mentioning, but is it any less valid when you post it over there?

It also seems that there were more complaints than just this point. The 24-16 truncation was the main issue as I read it. Things like cache errors seemed less significant. It was unclear to me if Jerry and Matt said the 16 bit "rounding error" as it was called happened without dither or not. Jerry or Matt? Clarification?

[jdg]

i did the wavebummer test with dither (from WB)
and there was a munged bit.. it nulled to one bit, then i normalized that one bit and it was just distortion.  

i will re-do it.. as it was pretty damning in my mind, and i ran away at that point.

give me a few, and i will come up with how i did the test and repost.. its slightly different.

[zmix]

John,

I did some tests, and I suspect that the residual is due to the remainders in the floating point audio engine.  I have been studying the dither requirements of fixed and float point math and it's not insignificant. Perhaps Bruno could comment here on that...

In good news, I truncated to 16 bits in Logic with dither and unlike the undithered Waveburner tests, it nulled completely, with only the dither audible (once I added 48dB or so of monitor gain).

[jdg]

so, here is my dithered from 24bit in wavebummer, then burned to CD and re-imported
nulled against the same 24->16bit tone dithered in peak (both with POW-r 1)

i see and hear dither.. and i see and here some 1k tone (there is a harmonic too?!?! wtf)

[Fifthcircle]

masterhse wrote on Sun, 15 June 2008 06:22

Doesn't the new version of Sequoia have issues with dithering properly? Also how is the SRC?

There were some issues with dithering at version 10.0x.  They have been fixed with the newly released 10.1.

As for the SRC...  There were improvements made at Version 10 (if I remember correctly), but I still use R8brain for my SRC as I think it is a bit better.

--Ben

[Matt_G]

tom eaton wrote on Tue, 17 June 2008 03:36

Hi Matt- Can you think of a time you would want to go from 24 to 16 bit without dithering?  I can't. -tom

Yes when you want to null test a program's integrity, it's the only way you can do it reliably. Using dither with Waveburner doesn't solve it's problem.

Matt

[Matt_G]

jfrigo wrote on Tue, 17 June 2008 09:02

It was unclear to me if Jerry and Matt said the 16 bit "rounding error" as it was called happened without dither or not. Jerry or Matt? Clarification?

All the tests were performed without dither for the sake of getting an accurate null test without dither residue left over.

zmix wrote on Tue, 17 June 2008 11:24

]I did some tests, and I suspect that the residual is due to the remainders in the floating point audio engine. I have been studying the dither requirements of fixed and float point math and it's not insignificant. Perhaps Bruno could comment here on that...

When this same test using the same test file was performed with various 'fixed' & 'float' based hosts they nulled down to infinity. So you can't blame it on remainders in the floating-point audio engine... whatever that means!?

Quote:

In good news, I truncated to 16 bits in Logic with dither and unlike the undithered Waveburner tests, it nulled completely, with only the dither audible (once I added 48dB or so of monitor gain).

Which again disproves your floating point theory about WB. Just face it as we've all had to, WaveBurner is flawed when converting from 24 bit files to 16 bit. For me this makes it useless as I prefer to do all my fades for convenience sake at 24bit before doing the final dithering to 16bit. This is the best way to work with clients wanting to do fades & cross fades etc at the final CD sequence.

I've mentioned the bug to Apple & they may fix it eventually, but in the meantime I've moved on to something more reliable.

Matt

[TotalSonic]

cerberus wrote on Mon, 16 June 2008 14:16

i just  delivered my first job that used my new system : with cubase 4 and izotope rx.   i think it could be my best work ever, but that is subjective... i did however feel more in control than ever, so if it is sent back for a recall, it's my fault, not the soft....  the macpro is blazing, and stable like a rock. one thing i have to say about getting a new system.  i upgraded to a much finer a/d/a and i've unwittingly been bent toward some kind of audiophilia.... for example:  i said to a client, "would you like some of the pops and clicks to be fixed?" and he kinda looked like he wasn't quite sure what i meant! and i observed that on my old system, those aren't so noticible, of course. and likewise on the client's krk monitors. have any of you ever been "changed" because of getting new gear?  how do you deal with this problem in general?  maybe say:"is it ok to ignore the pops and clicks i hear?" rather than: "would you like them fixed..." ? jeff dinces

I often hear ticks or pops that the client wasn't aware of as well - usually if you let them know the exact timing they occur at and ask them to listen through a pair of headphones with hyped-up high end response they'll be able to finally hear them as well.  

Best regards,
Steve Berson