If Audition dithers when it burns, then you don't need to dither before hand. I'm not sure if it does but I would image it does.

Dithering is the process of adding low level random noise. It is separate from the conversion process even though they are usually done at the same time. So it is possible to convert without dithering and dither without converting (which is what it sounds like you did).

I seriously doubt that dither is being applied during this process. It's more likely that your audio is being truncated. The reason for this leap of logic is simply that dither requires shaping, and there is no way for any application to determine how your dither noise should be shaped during bit depth conversion.

Also, I use several applications for audio, and none of them dither without your say-so, they only truncate. You should definitely be appplying dither in this case.

thanks guys. yeah, i looked into it, and no, i don't believe it's automatically dithering for me.....however, wouldn't i notice some pretty audible distortion if i have not dithered???
i've obviously been skipping this step for quite some time, lol, and haven't really been noticing any unplanned distortion....hmmm? idk.

although i know somewhat what dithering does (i don't know it fully. i just know it adds low level noise to get a fuller bandwidth coverage at lower bits...right?), would it also help with getting louder levels with the end product? -that might have been a dumb question....

jordy wrote: wouldn't i notice some pretty audible distortion if i have not dithered???

Not really. When dithering from 32 bit or 24 bit down to 16 only the least significant bit is in question. This corespondent to a -96dB amplitude change compared to full volume. If you don't dither, your audio is truncated (or it could be rounded)

A dithering algorithm will smartly alter the least significant bit to fake a higher bit depth. It's equivalent to a laser printer using very small dots of black and white to make shades of gray. If you truncate a photo, it would be just blotches of black and white. If you dither, you can see grays. Although with audio 16 bit allows 65536 shades of gray, and dithering allows you to fake it to make it seem like you have more.

would it also help with getting louder levels with the end product? -that might have been a dumb question....

Not a dumb question, but the answer is it does not.

thank you gecko!
that explains things alot better for me. i'll do some more research on it and start dithering my stuff.

hackenslash wrote: I seriously doubt that dither is being applied during this process. It's more likely that your audio is being truncated. The reason for this leap of logic is simply that dither requires shaping, and there is no way for any application to determine how your dither noise should be shaped during bit depth conversion.

Also, I use several applications for audio, and none of them dither without your say-so, they only truncate. You should definitely be appplying dither in this case.

Lots of programs I know of dither while burning. Dither does not require shaping either. There is usually a selection in the preferences to or not to dither. You can select the type of dither, shaped or unshaped. It's not rocket science. Shaped dither is random noise that is shaped. Looking in the manual should tell you if your program is capable of dithering during the burn.

I stand corrected.

I still highly doubt that Audition dithers automatically. Wavelab certainly doesn't.

From the Adobe website

"If you begin at 24 bit or higher, do not convert your material until the final stage of production"

"The preset dither depths and noise shaping curves (pdf=triangular, Noise Shape A) are usually adequate"

So, find the dither presets in the program and you should be good to go.

From Wavelabs website.

"We must mention the dithering options which include WaveLab's dithering algorithm and Apogee's highly-acclaimed UV22"

Just to extend the discussion a bit (if not hijack the thread) - Do any of you have strong opinions about the type/shape of the noise in the dithering process? I understand the theory behind the differences, but I frankly can't hear any obvious difference between the few options that I have.

Essentially, it moves the dither noise to different areas of the frequency spectrum. Since the noise itself is meant to be difficult to hear, it's hardly surprising that the differences are even more difficult to hear. If my understanding is correct, and it may well not be, it's so that you can hide the noise in different places, depending on the actual content of the recording.

Interesting question Mr. Bob.

I have seen the shape options but the default sounded good so I didn't mess with anything.

Next time I get a chance I will have a listen. Cheers for the fun mission.

I'm sure there will be smart dither in the future, but right now I only know of fixed dither shapes. flat dither is just that, low level broadband noise that is relatively flat across the spectrum. noise shaped dither tends to dip the noise at the ears most sensitive midrange and boosts it in the not so sensitive ranges. Some say it excites the content in these boosted frequency ranges, others say not. I feel it's not important as to what kind of dither you use, just as long as you use it. Keep an ear on it and check to make sure you aren't hearing any funky stuff if you're using exotic shaped dither's .

Thanks Michael. A great explanation, and clarified for me what I was only wandering around the edges of blindly before.

Incidentally, I can confirm that there is no auto dither in Wavelab, and it must be applied in the master section.

jordy wrote: ok, so i'm using Adobe Audition 3 for all of my recording, mixing, and my somewhat rough, improperly trained attempt at mastering. - but hey, i'm doing it all to learn from anyway.

I am not familiar myself with this program but it should give you a dithering plugin separately. I doubt it would dither as you convert...

after adding some limiters and stuff and i get my level fairly hot, i was wondering....should i dither?

Some limiters offer dithering and noise shaping options all in one plug. Does yours have it?

Edward Vinatea

hey thanks, Edward!
yes, some of the limiter plugins i've been using do have a dithering option. - there's just so many options on dithering it seems. -there's doesn't seem to be a "stock" dithering plugin with Adobe Audition, but there's options under the file and edit tabs at the top. i guess i'm just kinda uneducated in this area. the replies seem to have cleared it up for me. i just have to read up on how my program handles dithering. - it seems there may be sort of an auto- dither option that can peform right as i'm converting the 32bit file down to 16 during the burn process, however, i'm not quite sold on that idea seeing that some are saying each song may benifit from different dithering options- such as the shape of the noise and what frequency etc....

i haven't got to check though, if the program has indeed been dithering my previous material....if i load a song and look at the waveform, do you think i would be able to see the dither noise at the beginning or end of the song if magnified?

thanks for all the help guys

-jordan

jordy wrote: .if i load a song and look at the waveform, do you think i would be able to see the dither noise at the beginning or end of the song if magnified?

-jordan

You won't see anything. Dithering is a very subtle effect/process, somewhat arcane and thus you probably won't even hear a difference in the sound quality, certainly not a high amplitude.

Dither with a peak limiter when "bouncing" or mixing to 2 track stereo deck (for example a MasterLink ML9600) from high word length values to low, i.e. from 24 to 16.

So, always dither when changing word lengths. And, if you happen to save a lot of money to buy DAC/ADC pair of converters like the Lavry Golds, you would have world class dithering built right in. The price tag is about $15K

Edward Vinatea
Mastering Engineer

All converters have "dither" built in.

You may or may not be able to see the signs of dither depending on what kind of tools you have. If you have spectrafoo, they have many tools that you can use to visually see it. But you are more likely able to hear it on the fades of songs if it was truncated. it will sound "grainy" as it fades out.

okee dokee. thanks michael!

Michael Fossenkemper wrote:
You may or may not be able to see the signs of dither depending on what kind of tools you have. If you have spectrafoo, they have many tools that you can use to visually see it.

Have you ever "seen" this dither process? Do you have like a screenshot of it that you've made in the past to support this claim or is this something someone else told you about. I am really interested to know. Thanks.

Edward

sure, it looks like low level noise at -96db or so. If you get something like spectrafoo, you can also see all the different shapes of dither you have too. Sorry, no screen shots. I don't want to be known as the dither dude when I die.

Michael Fossenkemper wrote: sure, it looks like low level noise at -96db or so. If you get something like spectrafoo, you can also see all the different shapes of dither you have too. Sorry, no screen shots. I don't want to be known as the dither dude when I die.

Alright then thanks. But, then again, if you are viewing the screen in real time, one could argue that this is noise floor, not dither. And, if it's the graphic display of a program like spectrafoo, couldn't one mistake it with another type of noise source, like for instance, circuitry oscillation?

Assuming all you input is the sound of your mixer, is this something you can easily differentiate?

I am also assuming that these audio tests are conducted at very high volume levels (in order to hear the grainy sounding fade) which can be impractical for certain types of music and probably irrelevant to a generation of youth that are already half deaf ...:o

Edward

Well, that all depends on how good you are at identifying what you are looking at and what tools you have at your disposal. It's not "easy" but with ears, tools, knowledge, experience you can make an educated guess.

I'm not sure why listening to a fade out would be impractical on certain types of music. I also don't master for the lowest common denominator so what is irrelevant is if some of the users can or can't hear it. I can.

Michael Fossenkemper wrote:

I'm not sure why listening to a fade out would be impractical on certain types of music.

Oh, it's cool. I was referring to music like Club music, where the tracks are usually pumping from top to end and fading is delegated to the DJ (with his mixer in real time). I used to be a club DJ so I ought to know... :wink:

Edward Vinatea
Mastering Engineer

I see, so club music doesn't have sounds that decay? I'll keep that in mind. thanks.

*caution wasted ranting*

You used to be a DJ so you know what?
The art of playback is for chumps. I'm saying it here, not for the first time. Unless you are affecting with some sort of rhythm and feel the playback then all you are is a jumped up fuel inefficient Winamp.

I bet you all tracks end with an audio zero crossing. That's the pinnacle of fading, the zenith. The apex. The top, ending on an audio zero crossing is the final fade. So yeah. Got knowledge? Can you enlighten me, can you call me out. I'm so fixing for an argument I should call my ex and call her fat.

However I think I have more chance of winning this one so this one I shall choose.

Michael Fossenkemper wrote: I see, so club music doesn't have sounds that decay?

Of course there are sounds that decay. There are also beat break downs where all you hear is just a synth note. The point is that IMO, it is not possible to hear the effect of dithering in this type of music. I respect your opinion when you say that you can hear dithering. I just think I do when there is a fade and I am cranking up the volume of the master bus.

Edward Vinatea
Mastering Engineer

Ok, so don't use dither if you can't hear the difference. No skin off my back.

This argument (which was probably over before my attempt to comment on it) is really an apples and oranges deal. Dither, by definition, is about the last bit of signal. There is no club I've been in where the noise floor is such that the last bit (two, three, seven bits?) is audible.

I would really like to go to a club where that was possible. I realize that I would probably have to wear a tie and even still be turned away at the door. But I have no idea where such a club exists

Clubs are for people in the stone age. I use a bat, kind of the same principle as a club but mass produced and generally of a higher structural integrity than something I could fashion myself using a Garage Band saw.

For the record, Louisville Slugger. The junior sized one, it's better for indoor use.

Michael Fossenkemper wrote: Ok, so don't use dither if you can't hear the difference. No skin off my back.

Again, dithering is a useful technique that removes quantization errors and part of the linearity errors in A/D converters. The job can be handled by most quality AD/DA converters. ---

Edward Vinatea
Mastering Engineer

And what would be another useful technique?

Michael Fossenkemper wrote: And what would be another useful technique?

With dithering? I guess noise shaping --- :roll:

Edward

That's dither. Now that we have that straight, can we move on to more important conversations like "How do I make my songs loud"?

They're only important cause your knowing the answer keeps you in a job.

But these days everyone knows you just crack Waves L3 and choose the preset that makes it sound the worst.
Sorry, loudest.

"They're only important cause your knowing the answer keeps you in a job."

Many people who do Mr. Fossenkemper's job do not know the answer.

I would guess that what keeps anyone in a job is debts. Otherwise you'd be outta there.

Also, dithering is the final stage anyways, so if you've made puss it won't matter how you dither, it will still be puss.

"dither dude when I die"

You just can't make this stuff up ;)

one way to get a feel for the different dither algorithms is to repeatedly burn a file at 8 bit depth with different dithers.