R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2] 3 4 ... 8   Go Down

Author Topic: NEW DITHER SHOOTOUT (#1) NOW POSTED!!  (Read 17058 times)

Andy Simpson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 714
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #15 on: September 21, 2005, 11:37:35 AM »

Ronny wrote on Tue, 20 September 2005 18:51

I didn't analyze these songs in keeping with the honesty of the test, but I did notice that each zip file had a packing ratio difference that ranged from 26% to 35% and am confused as to why this would be so, all of the examples unzipped were 5,207KB, but zipped ranged from 3,433KB to 3,945KB a difference of over 500KB. No doubt some folks will report differences between some of these files, but I'm wondering if the packing ratio on the zipping is going to be an issue.
.......


My guess is that some of the dither types form patterns that suit the (lossless) compression algorithm and some don't.

With this logic, I predict that the smallest file is the truncated one.

Andy
Logged

TotalSonic

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3728
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #16 on: September 21, 2005, 11:48:49 AM »

andy_simpson wrote on Wed, 21 September 2005 16:37

Ronny wrote on Tue, 20 September 2005 18:51

I didn't analyze these songs in keeping with the honesty of the test, but I did notice that each zip file had a packing ratio difference that ranged from 26% to 35% and am confused as to why this would be so, all of the examples unzipped were 5,207KB, but zipped ranged from 3,433KB to 3,945KB a difference of over 500KB. No doubt some folks will report differences between some of these files, but I'm wondering if the packing ratio on the zipping is going to be an issue.
.......


My guess is that some of the dither types form patterns that suit the (lossless) compression algorithm and some don't.

With this logic, I predict that the smallest file is the truncated one.

Andy


The packing ratio could also be showing as different for the individual files simply because of the order in which the files were packed into a single archive.

If people really think this is an issue I can make the individual test wav files available for direct download - but I really think this is y'all barking up the wrong tree.

Best regards,
Steve Berson

bblackwood

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7036
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #17 on: September 21, 2005, 12:14:43 PM »

Errr, yeah. Why  not just try listening to the files instead of worrying about how the data compressed?
Logged
Brad Blackwood
euphonic masters

Ged Leitch

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1057
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #18 on: September 21, 2005, 12:59:16 PM »

Yes indeed, lets not let this Dither thread go off on a silly tangent about audio changes in zip or stuffit files for goodness sake.
Steve has done a lot of work on this so it would be only good manners to adhere to the topic.
Cheers.
Logged
http://bitheadmastering.co.uk/

"...But I don't wanna be a pirate!"

Tomás Mulcahy

  • Full Member
  • ***
  • Offline Offline
  • Posts: 211
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #19 on: September 21, 2005, 01:24:46 PM »

Maybe I am stating the obvious here- the files are different sizes because they comprise different data, because different kinds of dither were used. This will not affect the test, because we are trying to hear the effect of different types of dither!

It's not "bad manners" to bring up the issue, it is interesting after all. But let's not get carried away, for the purpose of this test it is not an issue.

Great idea Steve, I hope I can hear differences, I have become very blas

Ronny

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2739
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #20 on: September 21, 2005, 02:43:48 PM »

bblackwood wrote on Wed, 21 September 2005 12:14

Errr, yeah. Why  not just try listening to the files instead of worrying about how the data compressed?


I'm not worried about it Brad, just brought it up because of the different packing ratios that I observed when unpacking. Why not participate in the test and post your results rather than question the validity of the question? I'd really like to see some more ears getting in on this. Some people may pick up things that I can't and if so I'll switch my parameters from real world listening levels to whatever it is that helps anyone else detect differences.

WRT to Tomas' comment. For one, dither added should not change packing ratios to a degree of a half a meg over 5 megs, so I suspect that packing ratios vary with or without dither and on the same file. Steve if you would pack the same file 3 or 4 times and get different ratios than we can all be not concerned with the packing issue. In any case, I feel that the lossless files are just fine to determine differences in dither relative to tone or dynamic sonics.  
Logged
------Ronny Morris - Digitak Mastering------
---------http://digitakmastering.com---------
----------Powered By Experience-------------
-------------Driven To Perfection---------------

bblackwood

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7036
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #21 on: September 21, 2005, 02:58:29 PM »

Ronny wrote on Wed, 21 September 2005 13:43

Why not participate in the test and post your results rather than question the validity of the question?

FWIW, I'll question any statement here that I deem necessary. You guys are totally disrupting the thread with a discussion that, while you may find it interesting, has nothing to do with mastering.

And as for my participation, I'll gladly do so when and if I have the time. There's a reason that WoMP III hasn't taken place yet and it's not due to my laziness...
Logged
Brad Blackwood
euphonic masters

Ronny

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2739
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #22 on: September 21, 2005, 03:42:04 PM »

bblackwood wrote on Wed, 21 September 2005 14:58

Ronny wrote on Wed, 21 September 2005 13:43

Why not participate in the test and post your results rather than question the validity of the question?

FWIW, I'll question any statement here that I deem necessary. You guys are totally disrupting the thread with a discussion that, while you may find it interesting, has nothing to do with mastering.

And as for my participation, I'll gladly do so when and if I have the time. There's a reason that WoMP III hasn't taken place yet and it's not due to my laziness...




Testing for dither has everything to do with mastering and making sure that test parameters are on an even keel is crucial to all tests. I spotted irregulaties and I feel it's appropriate to question why this is so. You have the right to question any statement as it's part of your job as moderator, but I felt that your participation in the test was more much more important than questioning why someone was questioning the test parameters. This is your forum and you run it the way that you want, but I felt that packing ratios being different "may be" pertinent to the validity of the test files and is no way off topic or disrupting this discussion. All I said was I wonder if this may impact the test.

Anyway, I just zipped file 01 3 times and got the same packing ratio and same exact file size  each time, so maybe the dither is causing the files to have different packing ratios after all. I've participated and given my results and I'll not question anymore test parameters, but the time to catch any parameter irregularities is now, not after the test is in full swing.  
Logged
------Ronny Morris - Digitak Mastering------
---------http://digitakmastering.com---------
----------Powered By Experience-------------
-------------Driven To Perfection---------------

TotalSonic

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3728
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #23 on: September 21, 2005, 04:06:45 PM »

One of the questions being proposed then is:
can using StuffIt to zip files change their data?

My answer is the following:
StuffIt (and similar programs such as WinZip, WinRAR, etc.) has been designed as a lossLESS compression format.  Very often it is used to compress executable (i.e. application or installer) files.  Unlike audio - if an executable is altered in even the tiniest way it can render the file completely and utterly unuseable - which is nearly always shown as a "corrupted file" error message when attempting to unzip the archive.  With a very very long track record of successful use I believe I can more than safely say that zipping and unzipping a digital data file is not altering its data whatsoever.  I will not entertain contrary beliefs without strong evidence very clearly presented and backed up - so if the doubters here have any - go ahead and present it on a seperate thread.  

The second question being proposed is:
Is it possible that these different files are packing slightly differently because they are processed with different dither algorithms?

My answer:
If each file zips with the exact same ratio when packed individually (and not in a group - which may be "stacking" the files) via Winzip or Stuffit or RAR - then the answer is yes - it is indeed possible.

SO: let's hear some more feedback on the sound of the files themselves!!  

AND: Where are the "golden ears" at???  Let's hear from some of those who have on this forum claimed that dither radically effects tone of areas that isn't disappearing to the noise floor.

Best regards,
Steve Berson  

Thorhallur

  • Newbie
  • *
  • Offline Offline
  • Posts: 11
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #24 on: September 21, 2005, 06:43:07 PM »

the compression thing seems relatively simple to me.  Just like andy said, the dither noise randomises the wav file making it harder to compress.

Just try zipping an audio file consisting of white noise, it won´t compress at all.
Try the same with a wave file consisting only of digital silence and the result will probably be like 99% compression.
(and for fun, try it with a sinewave too)

but I´ll stick to the topic from now on (or better yet, just stick to reading what you guys say Razz )
Logged

dcollins

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2815
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #25 on: September 21, 2005, 10:59:35 PM »

TotalSonic wrote on Wed, 21 September 2005 13:06



The second question being proposed is:
Is it possible that these different files are packing slightly differently because they are processed with different dither algorithms?



Not only is possible, it's the only explanation.

Google "run length encoding" or Huffmann encoding for more...

DC

William Boyle AKA Elfy

  • Full Member
  • ***
  • Offline Offline
  • Posts: 138
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #26 on: September 22, 2005, 02:59:36 PM »

High guys, just thought id add some info with a few more dithers used with the test file i downloaded.
I took an online dither test, called the great dither shootout.
Its probably not 100% accurate, but the results seemed  clearly in favour of the megabit max dither. 2nd was waves idr ( type2 ultra)
So i emailed the guys in charge to see where i could  obtain that exact dither from, and basically they said it is unavailable, but
izone izotope 3 plugin has incorporated it.
They did give me the email of the guy who wrote the algorithm so i emailed him and i asked him what he thought of the izone dither, and he said it is an updated version of his called the mbit+.
So anyway, this plug has quite a few variations of dither on it, so i tried them all on the test file.
I also tested about 15 variations of the new sony oxford limiters dither.
Now track 10, i believe has IDR, type 2 ultra,
but besides that it is clearly the most inaudible to me during the fade out.
So i took this track and compared it to all of these other dither combos i tried with the oxford and izone3. Plus few others.
A lot of them were average, a few were quite inaudible, but none really came close to this track 10 beauty.

Now these conclusions of mine were determined by cranking everything up, so it might be a bit, non realistic.

One thing i do have to say about this dither i believe to be the most inaudible is that at the very end of the fade out, the sound
changes to a truncation type sound without dither, if you get what i mean, or i could be using incorrect terminology.
But it gives a zipper type effect, and i would be tempted to fade this out, which might defeat the purpose of using this dither in the first place.

Also, If this dither that i believe to be the most inaudible turns out to be the waves idr type 2 ultra dither.
Waves does not recommend the use of ultra for cd mastering in there product manual. It says with such high noise shaping it is possible that problems could occur during playback on some systems with inferior d/a.
What do you guys make of this.

But anyway thats what i can make of these tests.
Logged

bobkatz

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2926
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #27 on: September 22, 2005, 06:41:39 PM »

elfy wrote on Thu, 22 September 2005 14:59



Now track 10, i believe has IDR, type 2 ultra,
but besides that it is clearly the most inaudible to me during the fade out.




Danger Will Robinson.

Quote:



One thing i do have to say about this dither i believe to be the most inaudible is that at the very end of the fade out, the sound
changes to a truncation type sound without dither, if you get what i mean, or i could be using incorrect terminology.
But it gives a zipper type effect, and i would be tempted to fade this out, which might defeat the purpose of using this dither in the first place.





Cautions against using Waves type 2 have not been stated strongly enough by Waves. In actuality it was invented to help when dithering at even lower wordlengths like 8 bits, where the hiss of the dither can become as annoying as the distortion without the dither.

You should not use type 2 on a 16 bit file as it will by definition add grunge. It is incomplete dither basically, not sufficient dither amplitude to fully dither the signal so you hear the effect of the grunge and then full cut out of sound that you would get if there were no dither at all, only you can get a little bit lower level resolution than with no dither at all.

Quote:



Also, If this dither that i believe to be the most inaudible turns out to be the waves idr type 2 ultra dither.
Waves does not recommend the use of ultra for cd mastering in there product manual.




It's the type 2 that they should be warning against.

Some have made warnings against using ultra noise shaping was that under certain circumstances some potential clicks or ticks migh occur if one were to reedit noise shaped material. Or reprocess (redither) it and the cumulative high frequency noise could become audible. It's been shown by many experienced engineers that the worries about the clicks are just that, worries but with no practical effects. The concerns about the cumulative high frequency noise are much more valid, but if you make a 16 bit dithered file you are not expecting it to be reprocessed anyway.

Apparently the noise shaped dither has no effect on codecs, either, to the best of my knowledge.

The issue with inferior dacs is that to reproduce the dynamic range that is "encoded" in high resolution dithered material the DAC has to have good low level resolution to begin with. For example, if you take a 24 bit file and dither it to 16 bits, you have to have a DAC which is linear to "nominally" 24 bits (let's say 20 bits) to properly reproduce the 16 bit result. But I am not certain what the audible consequences of using a poor DAC would be, it would be no worse than using a poor DAC with the original 24 bit file as far as I can see. The dither won't affect the poor DAC, just that you won't hear the low levels of the music that's mixed with the dither properly.

Anyway, choose whether or not you use noise shaped dither, ultra or otherwise by your ears, there are few other concerns.

I received an excellent 16 bit DAT for mastering, and in the end mastered it with POW-R 2 or 3 with excellent results. The master was several dB louder than the mix so that may have helped by "separating the dithers". But anyway, to my observations, noise shaped dither on top of previous flat 16 bit dither doesn't seem to hurt at all. Caution is advised... listen for artifacts. I checked and checked and checked and my master sounded warm and full and fat and punchy and clear and musical, and the client loved it. Maybe the dither had nothing to do with that  Smile

BK
Logged
There are two kinds of fools,
One says-this is old and therefore good.
The other says-this is new and therefore better."

No trees were killed in the sending of this message. However a large number of
electrons were terribly inconvenienced.

Ronny

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2739
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #28 on: September 22, 2005, 08:24:15 PM »




Don't tell me that you are going to give a lecture on dither and not take the test, Bob. Have we not had a million dither discussions without some valid tests under our belt? Isn't this a thread on identifying dither types and discussing sonic changes that each brand of dither may or may not impart to the audio signature? You should be able to identify each dither type and give us a review of how each one sounds on your Lipinsky's. Won't take 30 minutes.
Logged
------Ronny Morris - Digitak Mastering------
---------http://digitakmastering.com---------
----------Powered By Experience-------------
-------------Driven To Perfection---------------

TotalSonic

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3728
Re: NEW DITHER SHOOTOUT (#1) NOW POSTED!!
« Reply #29 on: September 22, 2005, 09:08:18 PM »

This is the reason why I included Waves IDR Type2 with Ultra ns to boot - it's supposedly a "no-no"
so if we go by BKs post - people should be able to identfity the example that was processed with it by its "grunge" - but so far that hasn't happened -

then again -
So far we've had over 25 posts on this thread - but only 3 actual evaluations!
c'mon people - step up and give us some feedback!

thanks in advance...

Best regards,
Steve Berson
Pages: 1 [2] 3 4 ... 8   Go Up
 

Site Hosted By Ashdown Technologies, Inc.

Page created in 0.063 seconds with 18 queries.