It would be really nice to have presets for LUFS and LKFS normalization in all outputs at the same time. Actually it only works for the main output and only for LUFS.
thanks
LUFS and LKFS is essentially the same thing though. Same measurement standard can apply and it’s just target levels that are different.
I think it’d be a good idea btw. I can see how different mixes that need the same spec could be output simultaneously.
I tried to get a specific LKFS with the loudness adjustment in Nuendo but it never reaches what I need. I tried to reach integrated -27LKFS, but every time I try, I have to target a diferent amount .So I think the measurement is diferent between LKFS and LUFS.
I don’t think that’s correct. As far as I know LKFS = LUFS. Just different words.
Thank you Tumppi and Mattias for the input but it seems that there is a problem here. I usually use Nugen LMCorrect 2 to correct the loudness of my audios and always with success. I tried to use the loudness normalization with Nuendo …first with the same target as LKFS (-27) and than analyzed it with Nugen but there was a difference of 0,07 between LKFS and Nuendos LUFS. In other words: exported with -27 LUFS with Nuendo and Nugen analyzed it at -26,3. Then I exported the audios again using 27,7 in Nuendo and Nugen showed me -27 LKFS. But it doesn´t stop there. I tried to do that to another audio (Nuendo 27,7 LUFS) but it gave me another result…26,6 LKFS. In other words…I think its not reliable. I may be misunderstanding something.
By now however even “LFKS” in an updated software will measure with some sort of gating, and the results should be very similar.
If there is a difference then based on what Tumppi dug up Nuendo should be “more correct” if it uses “LUFS” since that should include gating, and only the older “LKFS” wouldn’t (i.e. your Nugen).
Either way, typically deliverables give us +/-1LU, so if you’re seeing anything below that you should be fine.
Thats it Mattias…“gating”. New specs from clients don’t want “gating” anymore. As I said at the beginning of the post it would be nice if we had some presets for Loudness and could export simultaneously to different outputs.
Thanks for all the help.
I don’t understand. Your clients want the original specs for loudness measurement? They specifically asked for that?
The “gating” is embedded in the algorithm that measures loudness.
These are the new Netflix specs. ITU R B.S. 1770 (not 2 or 3).
Ah, right. I remember now that I’ve seen that.
On the topic of nomenclature “LFKS” vs “LUFS” I’m not sure those words matter. It’s going to be a matter of being able to choose which version of 1770 is used for measuring and/or processing.
Anyway, while I think there are workarounds I agree with your feature request. It would be good if we could;
- select more than one source for processing/loudness-normalization
- select the specific measurement standard that precedes it…
+1
As a feature, obviously it would be nice feature to have different “scale” options as even tho Lufs is great not every channel sees the greatness and thus there are so many different formats…
I think this is one of those areas where there’s a good feature introduced into Nuendo and it’s a bit “left behind” after introduction. I haven’t used this feature for different reasons, one reason being that stems need to sum to the main mix (although I realize that there are probably situations where nobody notices a minor difference), but as I said I can see the use for it. More importantly it seems like it could be a relatively cheap and easy addition.
I could imagine just another drop-down window either in the export dialog or in project setup (or have it follow metering preferences set in Control Room) where we’d get to choose BS.1770-0/1/2/3 etc…
LKFS = LUFS = LU related to Full Scale, nothing more nothing less.
Just like dB is to dBFS.
Netflix ALSO uses gating! Just a different one…
They use dialog gating, the same gate that was used to measure the old dial norm using Dolby hardware ten years or so ago.
At this time AFAIK this new measurment is only available in Nugen VisLM and in the old Dolby LM100 and Dolby media meter (discontinued).
The algorithm is freely available from Dolby so it shouldn’t take long for it to get implemented in more meters.
I have VisLM and my very first quick tests showed that
I had already hit the spec on two episodes of previous R128 mixes. No change needed.
This does not mean it will be the same result for everyone, just that the difference is a bit less than you would expect.
I agree with Erik.
So Erik and Mattias, what you are saying is that,if I export my mixes with Nuendo at the specific spec Netflix is asking (-27LKFS–2TP ), they are NOT going to reject it ? I mean, the difference LMcorrect 2 gives me…about 0,4 to 0,6 LKFS is okay ? Or not ?
Thanks
No. But
Nuendo does not yet have the Dialog gating model that Netflix uses. Only VisLM does at this time.
You can compare it to dBTP and dBFS.
Both uses the exact same dB scale to measure and relate to the signal. However the measurment model is different and so is also the measured result.
To try to clarify further, the program I tested the VisLM Netflix measurment was mixed to standard EBU R128 @ -23LUFS. When I measured the mix with the updated Netflix algorithm using VisLM i was in spec.
This does not mean that any R128 mix will be. Just that with the type of mix I did for that show I wouldn’t have hade a problem to mix to the new spec either.
At this time, until others implement the Netflix model (should happen fairly soon though) you will have no choice but to buy and use VisLM.
I agree with Erik of course.
I have no idea how Netflix actually QCs the incoming mixes so I don’t know how picky they are and how far off you could be when mixing using a different algorithm. I have to say that I’ve never gotten a kickback from QC because of a loudness reading having read correctly but the mix having been out of spec, despite having used r128 from Steinberg as well as the built in Nuendo loudness meter, Dolby Media Meter and iZotope Insight (v1), all with everything from 1770-1 to -3.
I have however ended up in trouble a couple of times despite being fully compliant with specs where the specs weren’t detailed enough. So in general I prefer more detailed specs with a good QC department rather than looser ones with QC people that are random (sorry, I’m getting off topic now). Anyway, a while ago I moved to using separate loudness meters for stems as well as master. This way I can monitor the average of dialog as well as music & effects. A&E for example specifies m&e sitting about 10dB below dialog and total loudness to measure -24LFKS. So by using separate meters I can ensure that dialog is sufficiently above. They don’t seem picky about levels though, and Netflix might be.
I’d trust Erik’s advice 100%.
PS: Erik, do you know if Insight v2 measures using dialog gating? I know they incorporated a visual aid for dialog intelligibility, but I’m not sure if it’s just the visual separate meter or if they have a model that accounts for the correct type of gate in the final integrated measurement…
Nope, never used insight so I couldn’t tell.