Still awaiting customer support since 15.01.2015

I already addressed this in the German “lounge” since I didn’t find a suitable place to talk about this.

I also contacted the “Head of Support” through the forum about this (20.02.2015), but I’ve yet to get an answer.


I filed a support inquiry on 15.01.2015, which got confirmed through mail at the very same day at around 4pm. I can’t recall everything I wrote in this mail anymore, but it had to do with bugs and severe meter offsets (mainly in the Analysis module in the batch processor, and the “Signal Meter” in the WAV Edit view) in WL8.0. Also isn’t the first time I sent this in.

Yes, WL8.5 is out, and I I think WL9 will hit by Musikmesse 2015. But v8.0 is still listed as “support”, so I can at least expect an answer to my question if this was fixed in a consecutive update.


Furthermore, I got the information that “answers could be delayed due to fairs”, but my inquiry was sent in 10 days PRIOR to Winter NAMM 2015, and the next audio related fair is almost 6 weeks away.

I can live with a delay of 3-5 work days. But no answer at all in over 1 month is unacceptable IMO.



I’d love to get a status update on this.
Thank you very much.

I already addressed this in the German “lounge”

Please reformulate here in english.

I I think WL9 will hit by Musikmesse 2015

This is not planned.

You’ve got to be… sigh

Okay… I can’t get everything together anymore, but I already addressed several times through the support mail, that the “digital meter” in the WAV view, can be up to 2dB off, if you set it up in a custom form.


Recently, I am heavily using the batch processor for loudness normalisation.
Not only does the built in limiter of the Normalization module not(!) limit to -1dBTP (but rather -0,7dBTP, measured with TB EBU Loudness), but the “Audio Analyzer” is also off by SEVERAL dB.

Especially if you use this chain in the processor:
→ (Multipass) Loudness Normalizer → (Monopass) Audio Analysis

The dBTP values are off by up to 5dBTP, the ILk values sometimes don’t match up either. If I run the files from the same batch, in the WAV window, through either the global analysis or a third party tool ilke TB EBU Loudness, I have higher readouts of random values (dBTP max, ILk, MLk, etc). This shouldn’t happen!

On top of that, I’d love to see “RMS” (average loudness, and maximum) with a selectable timeframe for measurement in ms to be dropped with the Audio Analysis as well.



I don’t have any hopes that this will be fixed in WL8.0, I am not on WL8.5 (as I am not happy with the bundled features, AAC surround support is also not available) - and WL9 comes out lord knows when. So I’m not upgrading.


I had a more detailed report in German, where I’ve yet to get an answer for (since 15.01.2015!). But if you want it reports in English, mention so in BIG BOLD LETTERS on your support page. Then I know that my time wasn’t wasted.

There is no know issue i this domain. If you think a measurement is wrong, please provide a sample file and tell us what values are wrong for that file, and why you think they are wrong (what is your refeence?).

Why doesn’t that answer surprise me?



I can’t provide the material, as this was chart music I had to adapt for a local theater a couple of weeks back.

The material in question was first declipped in Cubase, and was roughly around -20LUFS (with sometimes a DR of over 15!) to -13LUFS prior to the loudness normalization, depending on the production. Everything had to be pulled up/down to -16LUFS, so that the theater tech barely had to do any fader movements.

The Loudness Normalizer (Batch Processor especially) doesn’t limit to -1dBTP - period. Even if the tolerance is pulled down to 0,00dB.

And the Batch Audio Analysis module was always about 3-5dB off in terms of the dBTP readout after everything was normalized. (again, I checked this with EBU Loudness by ToneBoosters) The batch processor chain was: Loudness Normalizer → Audio Analyzer. The CSV file it dropped, showed offset values.


Regarding the “Meter offset” in the WAV editor - you probably already heard of my K-System v2 concept which is built upon the EBU R-128/ITU-R BS.1770-x specs (though only uses a custom reference level, and a different color coding system). But setting a custom reference level (i.e. -16LUFS, -14LUFS, etc) results in an up to 2dB offset on both the MLk and SLk meter compared to third party tools (Nugen Audio, ToneBoosters, etc).

You know of this report for over 1 year at this point. I also sent that in through mail and never(!) got an answer. As AE, that is RELIANT on exact metering tools, this is an absolute NO-NO.



You also know of the infamous VST3 Audio Forward Bug that you shrugged off on to the developers (same answer as well “not our fault, it’s a third party developer issue”).



I didn’t get any answer to my Wavelab support inquiry mails, and Steinberg is a German company. Sometimes I do have the impression, that things are ignored on purpose. And Wavelab is a Mastering tool where metering is essential - it has to be precise (remember how many Wavelab versions it took until the K-System meter was correct?! - same issue this time around with the EBU R-128 realtime meter). Yet issues are ignored or pushed on to the next paid version (my FR to implement RMS avg and RMS max readout for the Batch Analysis module for example - RMS is not important, but EBU R-128, DC offset and various peak types is?! - and let’s not talk about surround support and AAC rendering in the same realm - 2.0 is a dying format).

I was always confident and hoped “these issues will be fixed in the next version”, and silently continued to invest more money. I am not(!) supporting this anymore. Disappointing support. no demos, disimprovements with every consecutive version on terms of what worked before… This is frustrating, and I’m with WL since WL3(!), Cubase even longer (since mid 1990ies) - so I have my 20 years full of being a “Steinberg” member.




Is that enough information for you now?
Or will this be disregarded as any issue and tagged as “solved”?

How will we go from there?

And why didn’t I get ANY answer to my support inquiry via the support engine?!
I thought the Steinberg boards are no support section, hence the user account.

I can only repeat myself - this is frustrating.

You seem upset and am sorry about that, but what is wrong with my answer?.. A test case is needed if you think something is wrong.

Sure, I’ll drop you 5GB of edited data, copyrighted and refurbished chart material from the last 40-50 years, just to test the metering, analysis and Loudness Normalization tools - only to get an answer “we can’t reproduce what you report!”. It wouldn’t be the first time. Actually, it wouldn’t even be anything new to me at all.


I think I’ll sit down next week (as I’m about to head out for this weekend), with my copy of WL8.0.x and provide some videos for you people, do another “data drop” and “manual checks” with third party tools (read: waste several hours/days to prove once again that something is wrong, only to be shrugged off as “non-issue”). A TeamView session wouldn’t solve this in a short manner as confirmation of values takes time.

But in order to get this sorted - please RESPOND to my mail, so that I have a ticket number and a contact person - in either language (German, English - I don’t care anymore at this point, as long as I even get an answer - unlike the last couple of times, with both Cubase and Wavelab), then this can be easily forwarded to the dev team and not be dragged out at the boards (which is said to be NO CUSTOMER SUPPORT board!).

Where is the problem in doing that?! No repeats at the boards necessary.



That is an understatement as of late.

I’ll drop you 5GB of edited data, copyrighted and refurbished chart material from the last 40-50 years,

You don’t describe a problem related to the quantity of data. Therefore, why a single 3 min non-copyrighted file shouldn’t be enough to demonstrate a problem?
You quote an software as reference, but why do you think it should be right instead of WaveLab?

But in order to get this sorted - please RESPOND to my mail, so that I have a ticket number and a contact person

I am a tech guy and I only give answer in this forum.

provide some videos for you people

Please don’t loose time with this, because it can’t help. Rather provide audio data and describe why you observe wrong.

Won’t be happening, as “test tones” alone won’t cut it. Other than maybe testing if the Loudness Normalization module “really” limits like it should (I do have something for that to test ISP limiting). But this doesn’t clear the offset of the Audio Analysis (batch process).

And even then - I am still on 8.0.x, not 8.5. So in order to get a maintenance bugfix, I need to pay for the upgrade.



Because the software I use compared to Wavelab is by the creators and forerunners of the EBU R-128 specs in the first place? Thomas Lund (tc.electronic, though I use LM2 mostly at clients rigs, not mine) ringing a bell? Jon Schorah (Nugen Audio)? Jeroen Breebaart (ToneBoosters)?

Wavelab is maybe the forerunner in terms of CD authoring, but it isn’t the forerunner in terms of metering (anymore).



You’re not only “a” tech guy, you’re the main developer/creator of Wavelab, Philippe. Which is feeling even more like a slap in the face, considering all the bugs and reports I sent in via mail in the last couple of years.

If you only respond in the forum, then Steinberg (company, and the support division mainly!) wasted my time with this - plain and simple. If “tech support” is only done through this clunky web board, then “Wavelab” shouldn’t be on the official support web-from other than for license activation issues.

A short mail message with the info “we forwarded this, but Wavelab support is only done through the boards” would have saved me a lot of time, actually put me in the picture. And then this thread would have looked different as well.



Again - I won’t be able to provide “chart content” that I had to edit up, as this is copyrighted and therefore distribution outside of my clients would be illegal. This is where I encountered the issues, this is what I base my reports upon. Unless you find a different solution, you have to take my word for it (my findings). Hence the planned videos and CALC (OpenOffice) sheets. Or at least “screenshots” and the CALC sheets.


It’s your call.

WaveLab metering has already been heavily tested.
I have to repeat myself: if you think there is a special case that don’t match (and I would be surprised), looking at a video won’t help.

Just a guess with regard to your issues… AFAIK the loudness normalizer doesn’t behave like a peak normalizer. It is normalizing with regard to the loudness, not the peaks. Could this possibly be the source of the irregularities you are experiencing?

Your resistance to provide a “fair use” sample for PG to examine truly boggles my mind in this thread. What do you think PG is going to do with your item - use it as a profit generator?

If you really want t get to the bottom of this - (you continue to talk like you never want a resolution) - man up and send PG a sample so he can see what’s up?

There is nothing “illegal” about sending a professional courtesy sample for technical support. If your “client” truly has an issue with this - might be time for some new clients…

VP