Sorry for the long diatribe below, but I want to give the details. My 48/24 mixes are louder after dithering in Ozone to 44.1/16. Why?
 
Here's the basic issue: I make a mix, then I make a small M master to bring the level up using the ozone maximizer to -0.5,-0.5 on the master bus (according to SPAN) often after a compressor and a pulteq eq on a mix bus. I record in 48/24. I export the mix and the master with no dithering at 48/24. Cool. good to go. I'm using Ozone 5.
 
Now for my own amusement only, I want a CD and perhaps a mp3. so I need to dither to 44.1/ 16.
 
Then...and here's the issue:
 
My dithering alone in ozone produces a song with a peak anywhere from -0.4 to 0.0. Why?
 
Here are the specifics:
 
Bad Hair Day mix: -0.5,-0.5 with Ozone 5 on the master bus using the maximizer only.
 
1. Now I turn the Ozone dither on to Mbit+, high, 16, normal, DC filter on. No other changes and export audio: sonar set to stereo, 44.1/16 no dithering. Incidentally, for the maximizer, the setting was IRC II, -0.5. Then I bring up the song in a test project, the CD master is -0.2,-0.2. Why?
 
2. I go back to the mix and try IRC III (for some strange reason) and repeat as above. In the test project, the new audio at 44.1, 16 is -0.3,0.0. Why?
 
3. Then I take the original master at 48/24 and -0.5,-0.5 in a new project, insert Ozone with all modules off including the maximizer and dither to 44.1/16 and the new master is -0.4,-0.3.
 
I repeat with a different 2nd song mastered in Ozone to -0.5,-0.5. With dithering only in Ozone, the new CD mix is -0.4,-0.4.
 
OK Why?
 
I'm not going to be mastering my own music, so this isn't mission critical but I've never checked this before and who knows what I gave clients over the years?
 
Why does dithering in Ozone increase the peak levels a s much as 0.5dB?
 
I hope I have explained this well. I would have assumed all my CD masters would also be -0.5,-0.5.

The offshoot is, I may want to shoot for -1.0,-1.0 before uploading my mixes to YouTube, etc. rather than -0.5,-0.5.

Make sure you're enabling the "True Peak Limiting" option (under the threshold sliders in O7, don't know about O5). Measuring peaks is less precise at 44.1 than 48 KHz due to the greater time between samples. Turning this option on enables internal oversampling to get around that.
 
-1 dB is a better target for MP3/YouTube etc. If you find that setting the brickwall limit to -1 causes excessive compression, turn the Input slider down a little.

bitflipper
Make sure you're enabling the "True Peak Limiting" option (under the threshold sliders in O7, don't know about O5). Measuring peaks is less precise at 44.1 than 48 KHz due to the greater time between samples. Turning this option on enables internal oversampling to get around that.
 

Bitflipper, I was really hoping you would reply as this has me really perplexed.
 
In Ozone 5 there is no "True Peak Limiting" option. And it is not a limiting issue as in #3 above in the original post. A song at 48/24 is dithered only (Maximizer and all other modules are off) and the original song at -0.5,-0.5 becomes 44.1/16 at -0.4,-0.3. this was attempted with two songs with similar results. Thus it is a Ozone dithering issue. I haven't checked it using only Sonar dithering.
 
That being said, in Ozone 5, in the dithering module, there is Auto-blanking, Limit Peaks, and Harmonic Suppression check boxes. I had them unchecked.
 
Do you think Limit Peaks is your "True Peak Limiting".
 
The help file below gives this info under dithering:
 
Auto-blanking
Selecting this option instructs Ozone to completely mute dither output (i.e. dither noise) when the input signal is completely silent (0 bits of audio) for at least 0.7 seconds. 

Limit Peaks
Dither noise is random in nature and has a very low amplitude. However, after noise shaping, especially in aggressive dithering modes, the high-frequency dither noise is significantly amplified, and the overall dither signal can show spurious peaks (up to -60 dB FS for a 16-bit quantization). If such high peaks are undesirable, you can enable the Limit Peaks option to effectively suppress the spurious peaks in the noise-shaped dither.
Suppress Harmonics
If, for some reason, any dithering noise is undesirable, simple truncation remains the only option. Truncation results in harmonic quantization distortion that adds overtones to the signal and distorts the timbre. In this case you can enable Suppress Harmonics option to slightly alter the truncation rules, moving the harmonic quantization distortion away from overtones of audible frequencies. This option doesn't create any random dithering noise floor. Instead it works more like truncation, but with better tonal quality in the resulting signal. This option is applicable only in the modes without dithering noise and without aggressive noise shaping.
 
Thanks again Bitflipper, I sincerely appreciate your response!
 
And yes I agree, were I to make a 44.1/16 or mp3 for upload, I would make it at -1.0 to -1.5 as opposed to -0.3 or -0.5 for my 48/24 mixes which I do upload unless they are for iTunes or YouTube, etc.

In Ozone 5 and earlier, the option was called "Prevent inter-sample clipping". Same feature, different label.
 

^^ In Ozone 5 is the "Intersample Detection" button in the lower left of the Maximizer window.
 
Just FYI - iZotope has free pdf guides that are written very nicely and almost in a universal manner (can apply to any plugins with the same features). This page has them all listed. If you scroll down to the Ozone 7 section, there is one specific to Dithering (still the O5 version), and another for Mastering (updated for O7). The third that is useful is the Mixing one listed under Alloy2 at the top of that page.

Well, maybe I've gone daft, but why will this help?
 
Yes I keep the Intersample Detection unchecked. But it is not a part of dither but a part of the Maximizer module.
 
All I am doing is dithering a 48/24 project to 44.1/16 to make a CD using Ozone 5. The maximizer and all other modules are off. When I do so, the levels increase from -0.5,-0.5 (acquired by using the maximizer in a separate project) to a high as 0.0. That is not good and very counterproductive.
 
Oh well, I'll befuddle and fiddle about some more when I have some time.
 
When I finish my current project, I will check out the dither in Sonar, IK Multimedia and Pyro and see if they will dither to a CD audio at -0.5. If not, I really don't understand dithering.
 
Incidentally, Metteleus, I have all the amazing pdf guides from Izotope and have read them often for years now. They are great! However, I do not have the current ones for Ozone 7 as you suggest. Now I'll get those as well so thanks for the heads up :)
 
 

Higher sample rates result in more accurate analog peak detection in the limiter. "True peak" detection just means data is temporarily oversampled prior to hitting the meters and the peak detector in the Maximizer module. 
 
Digital to analog conversion can result in real peaks as much as 6 dB higher than your highest digital sample. Then, MP3 conversion can further add another 3 dB due to ringing in the filters.
 
Often, this isn't noticeable. After all, any 3 to 9 dB increase happens in the analog world, where it has no consequence on all but the crappiest of portable devices that don't have 9 dB of analog headroom.
 
However, if you subsequently import that track from a CD or an MP3 into your DAW, you're going to suddenly see those higher values. Had the "True Peak" option been enabled on the initial export, they might still be a little higher but they'd be closer to the limit you'd set for the maximizer.
 

1. When you resample, the samples end up in different places so some intersample peaks become "on sample".
 
2. Dither is adding noise, which increases the level.

Bitflipper, Drewfx1, those last two comments help.
 
So the word is, when I initially limit using ozone, even though I am staying at 48/24, check Intersample Detection and this will then be of benefit should I eventually dither to 44.1/16 regardless of whether I use sonar or IK Multimedia or Pyro or Ozone. If I limit with another plugin, I assume there is no such feature. I had thought Ozone was a better dithering plugin than others but maybe not...I'm starting to see less benefit to Ozone as the years go by. I'm glad I did not upgrade to 7.
 
And secondly, I knew dithering is adding noise, but I didn't think it could add up to a half of dB in a 3 minute song.
 
Like I say, I've never checked this over the last ten years so it comes as a big surprise.
 
Thanks so much for the explanations. I don't think I've ever seen a caution statement anywhere e.g. "Dithering may add a dB or so to your peak values even when a limiter is on as is often the case in Ozone!"