normalization 360

S

smg

Member
I spent a lot of time mastering the current track I'm working on,tweaking the MBC settings to get the type of sound I had in mind...then when I normalized it the sound changed completely,all highs w/none of the mid/low range aspects I'd achieved w/the MBC settings....

Fortunately my set up allows for undoing things so I set things back before the normalization ...

I could rerecord the mastered mix,adding the same MBC settings used again as a way of increasing the overall vol (I happened to do this accidentally already and saw that's what happens)...but not sure if the vol increase overall will preserve the results done initially to the mix w/MBC.

Another approach would be normalizing the untreated mix then adding MBC...however the manual
recommends against this as it potentially causes distortion....

I'm wondering why the normalization came out the way it did and what the best approach would be....
 
Last edited:
1) Why are you using a multi-band compressor.
2) Why are your 'normalizing'?
3) Post a clip so others can hear (before MBC, after, and 'normalized')
 
Kind of the same questions as [MENTION=39487]mjbphotos[/MENTION] - especially "why ... normalizing?"

I never normalize. I put loudness and peak meters on the master bus, then use compression and limiting to get things where I want them, dither, and bounce (assuming I'm targeting a pseudo-mastered level, e.g., -14dB LUFS for a video or maybe SoundCloud upload).

I'll play with MBC on the master bus if I'm not happy about something and can't seem to make EQ do it for me. Sometimes it surprises me and actually does what I hope for, but I've got a lot to learn about it.
 
Appreciate the replies-
I posted a thread a few days ago on the Tascam users forum(Building up larger mix using DP008ex) that goes into some detail about the current approach I'm using..if you read it over the following will be more understandable-

Each time I make the stereo master track I record the individual source tracks mix at a lower volume so once it gets re-recorded w/the MBC it will be at normal vol to continue putting the mix together...

I use the MBC when mastering the SMT each time to set up the levels of the three bands/to get the individual parts to work together...i.e.if the kick sounds too boomy I lower the low range band levels,if I want it to have more punch I raise the mid band level ...of course all the other instruments are affected by whatever you do to each of the 3 bands so its a delicate balance and sometimes a compromise between getting one to sound like I want without messing up the overall sound....

Probably as I get better at things I'll have a better sense of levels/individual instrumental sounds working together during the initial tracking without encointering the type of masking problems I had to compensate for w/MBC ...and will use it more for enhancing the overall sound of the submix SMT each time instead of having to work w/the individual band settings to get things balanced.....only.been doing this for a couple months and this last track had a lot of instruments/sections etc...

My understanding of normalization is that it's used as a matter of course as the final step in the mastering process to bring the overall level up to maximum vol...so I appreciate getting a chance to hear the viewpoints/approach used to maximize the track volume from people using a different equipment set-up....

To return to the main thing I'm wondering about,any ideas about why the normalization result was a completely different
mix in terms of frequency range distribution instead of just preserving the mix I made and making it louder as I expected would happen would be a great help in understanding these things....

I appreciate the offer to review the track itself in its different stages by uploading it here and ,while my current set-up/approach isn't really geared towards doing that(i.e.I don't own a computer and would have to go over to the library to post the file....its so hot right now in Reno NV where I live that I'm just staying here w/the AC)....further on down the line I'll be looking forward to getting peoples' feedback on the tracks I'm making....

This leads to something else I'm looking at doing...over there at the library they have Audacity installed on the public computers and I took a look at the normalization functions in the Audacity manual...maybe I'll use that instead ...wouldn't mind any info about that from people who've used it- especially as regards getting the normalization to take place without changing the mix ......
 
If you wish to continue using the MBC and Normalization, here's a page out of my DP-03SD manual showing the MBC and SBC presets,
along with descriptions of music styles they're applied to.

I didn't see any presets in the 008 manual I have, but you could certainly use these as a guide to get you in the ball park.

RDqJzkW.jpg
 
Hey thx Spantini..glad to see you reply here as in my understanding the DP03SD is an earlier or later version of the DP008ex w the exact same capacities(let me know if I'm wrong but after downloading the DP03SD manual the other day and looking through it I think this is the case...)

I have the same page in the manual(pg 82)showing the same info...

One thing I noticesd yesterday when working w/the MBC parameter settings is that a given threshold/ratio/level setting used on one preset doesn't sound the same on another...i.e.parameters I used to setu p the Basic CD mastering preset gave a completely different effect on the Dance preset...
'm assuming this has to do w/the preset variables you can't adjust that must be different for each(i.e.frequency band divisions(crossover)/attacks release/knee)...which is one of the drawbacks of working w/both the Tascam mastering presets( as well as the COMP/EQ function on my ZOOM RT223 which just has adjustable settings +/-12 for "compressor sensitivity")in terms of applying things from online sites that give settings to use for specific effects....anyways I've figured out how things work to some extent and can.translate info given for more sophisticated compressor types to what I've been doing w/my set-up)

As I said before I'm happy to see you reply man as you're using the same kind of recording equipment as mine....and wonder if at some point you'd have a chance to check out the thread I referred to above on the Tascam Users forum and let me know what you think in terms of what I'm describing/asking about...
 
The way I read that DP-0008EX manual is that normalization is peak normalization. And, I misspoke/typed when I said "I never normalize," because, at least according to Wikipedia, what I'm doing by manipulating compression/gain/limiting to achieve -14dB LUFS (integrated), is actually loudness normalization (a "normal" mastering step). Live and learn.

Without seeing the audio files before/after each step, I'd say it's speculation to guess what is causing what you are hearing. Is it possible you've cranked the MBC so much that the DP-0008EX's normalization is reducing the overall track level?
 
Normalize should not affect dynamics or frequency response at all. It's just a gain change.

...unless...

1) It's happening before the MBC, in which case it should be obvious why it sounds different until you adjust everything for the new input levels, or

B) It's such a large adjustment that you're getting into Fletcher-Munson and psychoacoustic phenomena.

I'm leaning toward #1. If that's the case, bounce it once with the MBC in place and then normalize that new file without any further processing. If it's the other, you just have to do a better job of matching levels at your ears. Use the headphone volume knob.
 
ashcat_lt said:
Normalize should not affect dynamics or frequency response at all. It’s just a gain change.
...
Click to expand...
That’s what I thought, but that’s just peak normalization.

Again we are only guessing but I’m actually leaning toward the FM thing, though for the reason I posted i.e., normalization (either kind!) might be reducing overall level in this case.
 
keith.rogers said:
That’s what I thought, but that’s just peak normalization.
Click to expand...
Well no. Normalizing is any time you measure the signal and then add gain (or attenuation) to compensate whatever you measured to some standard. Whether you're analyzing the peaks or some average (RMS, LUFS...) doesn't change the fact that it's only gain. I think it's obvious that if you measured peaks, your standard should be peaks and likewise with an average. If you're doing anything else to make it meet your standard - compressing, clipping..., then it's not normalize, it's compression or clipping or whatever.
 
Keith/Ashcat-Thx for weighing in here-

Both possible causes suggested for what happened w/ normalization yesterday hypothesized the cause as being things that were actually not involved-

-the MBC I did was always done w/vol levels as the focus each time...made sure that the clipping indicators on the meters weren't showing when re- recording the mixdown w/the MBC added....

-although the normalization before MBC option is something I've considered (re-first thread post above)this was not done here....what was normalized was a normal vol range mixdown w/added MBC...

Something caused the result to have just highs...not sure what ...
 
You know that's exactly what I was talking about in the first post this AM-

Since I'm new to this and trying to develop a 'go-to"approach to things,submixes I think will work in the context of the larger mix sometimes don't and,unlike working w/a larger track capacity,I'm trying to put things together using a strategic approach to 8 tracks(re- thread on Tascam Users forum referred to above)where things are mixed down and copied to tracks to make new ones instead of being on separate tracks throughout the mix so they can be adjusted as needed...

This is something I plan to change when I can... upgrading my current equipment to the DP24SD...other thread posts here go into this somewhat....but as it is now I have a drum mix balanced in of itself that using MBC with sets up better in the mix....and as stated this AM I'm hoping to develop a better sense of EQ'ng and setting levels for other instruments so the overall balance will allow for everything to be clearly heard at the intended level for the mix concept...
For ex this last mix had a lot of parts including sustained pad and I wasn't able to predict how the drum track would sound once everything was added...

This is why the normalization result being focused on here was so pronounced in terms of a mix that had been mastered to offset some of the masking-type situations and level imbalances ...
 
Once you did your MBC adjustments to the final mix down......should we assume that you then rendered that mix down to another wav file with no effects.........and then normalized that final wav file? Your description of the final steps you took still has me a bit confused. (could be me of course) As Ashcat has pointed out.......normalizing any track should not affect anything except volume.....and if the track / wav form is already peaked prior to that........peak normalization would have very little effect. Certainly it would not have the drastic tonal effect you are hearing.

The only other thing I can think of would be if the wav form was SO SMALL that normalization would raise the volume SO MUCH as to affect HOW you hear what was already there. What do you listen to your mixes on? Poor headphones and poor speakers could be fooling you a bit.

Hey.....just my thoughts.
 
smg said:
Hey thx Spantini... As I said before I'm happy to see you reply man as you're using the same kind of recording equipment as mine....and wonder if at some point you'd have a chance to check out the thread I referred to above on the Tascam Users forum and let me know what you think in terms of what I'm describing/asking about...
Click to expand...

You're welcome, man. I have read your threads, but I can offer no more than what you've already discovered. I have always used the default
Mastering setup (which I don't know the name of because I never went looking).. it's probably the first preset on the chart : "Basic CD M".
That sounded good to me so I stuck with it, but now I see you're swimming in the deep end of this I may jump back in and try all of them
just for the hell of it.
 
ashcat_lt said:
Well no. Normalizing is any time you measure the signal and then add gain (or attenuation) to compensate whatever you measured to some standard. Whether you're analyzing the peaks or some average (RMS, LUFS...) doesn't change the fact that it's only gain. I think it's obvious that if you measured peaks, your standard should be peaks and likewise with an average. If you're doing anything else to make it meet your standard - compressing, clipping..., then it's not normalize, it's compression or clipping or whatever.
Click to expand...
(N.B. This is fairly OT.) We are probably splitting semantic hairs here, but, sure, in a world with no upper limits, you can normalize loudness to anything, but in the context of mastering, "normalizing" your tracks to a fixed loudness almost always involves some limiting and usually compression because (e.g.) if you've got a master bus loudness of -18dB (LUFS) with a dynamic range of -6dB, you can't deliver a music file [that doesn't clip] to a streaming loudness target of -14dB LUFS without doing something about that dynamic range. Maybe that's not normalization in strictest sense, but it's what I'm calling normalizing [stereo mixes'/tracks'] loudness to a consistent target, and that process is, from what I've just been reading, generally called normalizing.

I have NO IDEA what "normalization" in the OP's case is, and without files to measure, nobody does.

Edit: The example of -18dB LUFS and 6dB of dynamic range was a boneheaded math mistake because only 4dB of gain would meet the loudness target without causing any clipping. A better example might be if I was mastering that mix for a CD with a loudness target of -10dB LUFS on that mix and it had 10dB of dynamic range. Then 8dB of gain would push the peaks to +2dB, or well past clipping.
 
Last edited:
Mickster actually the way the DP008 works is that it normalizes the stereo master track itself prior to any conversion to a wav.file(although it can be converted prior to normalization as well)..the process as you described it is what I did...i.e.MBC to normalization w/out adding anything else...and the monitoring system used was at the same vol used to set things up before nornalization...so the vol level change expected was there along w/the unwanted mix change...

Spantini do you use the DP03SD to do the kind of thing I described over on the Tascam Users Forum thread?(i.e.making the stereo master track then cloning it back to 2 tracks in the multitrack mode to continue building up the mix?)
 
Last edited:
keith.rogers said:
..if you've got a master bus loudness of -18dB (LUFS) with a dynamic range of -6dB...
Click to expand...
...then your loudest peak hits -12dbFS, and you can easily turn it up 4db to get your LUFS to -14 and still have have 8db before anything clips. It will still be crushed to shit.
 
ashcat_lt said:
...then your loudest peak hits -12dbFS, and you can easily turn it up 4db to get your LUFS to -14 and still have have 8db before anything clips. It will still be crushed to shit.
Click to expand...
Haha. The numbers side of my brain wasn't working when I typed that on my phone, obviously. You know what I meant. Let's say it has 8dB of dynamic range, or you're starting at -15db LUFS and want to master for CD, so going to -10dB LUFS.

I'll go back and make an edit mention the error.

The point is, moving to the mastered loudness target would cause peaks be at 0dB or above, so I cannot just apply gain. i don't believe anyone "masters" using peak normalization, because that could leave the loudness of your tracks all over the place.
 
appreciate you guys going into detail here re-specifc #'s..although I don't have metering now to be able to implement this(just ones that monitor input/output/track levels that show -12 dB >O dB w/6 in center + w/out anything else specifically determinable...if I'm understanding how they work correctly) ,I'll definitely use these parameters once I have a bunch of tracks ready to transfer to wav. and look at them on Audacity...

Not sure but after reading everything today and reviewing things from yesterday to find the X factor that caused things to sound like they did,I think I might have listened to the normalized track w/the MBC settings used to make the pre-normalized master still on instead of being turned off...I'm going to deal w/this track again in a few days ( spent so much time on it I need a break,started on some new tracks) and then normalize it again + check to see if that was the problem....
 
Last edited:
You must log in or register to reply here.
Back
Top