@Basel-Shishani

Hi Basel.

I do all of my audio editing outside of MEP and VPX and have had no problems to date.

As far as I'm aware there is no problem with deleting .hdp/.h0 files but so far I have never bothered.

However I never use original files in a project in case of problems but always start a project using backup copies just in case. You can never have enough backups of original material.

Ray.

@Basel-Shishani

Hi,

Ditto what Ray mentioned.

Here is a link to a video that I sent to Magix showing the problem with Edit Wave Externally:

We discussed Edit Wave Externally in this thread and in this thread.

From what I recall about what happened with Audio & Music Lab, Edit Wave Externally always showed up when the right-clicking on an audio object or the audio part of a video object (VA on separate tracks). The program would do a mixdown, put the wave file on a track, usually track 5, and open the wave file for editing in AML; the original audio would disappear from the timeline but could always be brought back. Edit externally, exit and the updated wave file is now the one used in MEP. If you do an undo, it would disappear and the original audio brought back. Also, I seem to recall that looking at the files after one session where I exported the new wave file, MEP/AML somehow kept the previous version as a take. This is the way it should work, and we should be able to manage takes.

As I pointed out in my tutorial on audio, Edit Wave Externally is trying to open the HDP file instead of the wave. For Magix products that understand this, they use the hdp file to open the wave file. This needs to be corrected so that all audio editors work as the external editor.

Also in the tutorial, I show how to do a mixdown in MEP of one or more audio objects by using a range and/or muting tracks. Once done, there is a wave file on the timeline that can be opened in an external audio editor and modified. For now, to use an external editor that does not open the audio file from within MEP, you have to check for the location of the wave file, and open it in the external audio editor. Hopefully, this will get corrected some day.

Final note for now, if the waveform in MEP has not been changed after editing a wave file, delete the HDP and H0 files and reopen the project. The updated waveform will be created along with the HDP and H0 files.

John CB

@browj2, @CubeAce: Thanks for the tips, and the tutorials were helpful.

Now I was happily applying my newly learned techniques when I hit this glitch:

I'm modifying the audio file externally, just normalising and applying effects without changing the sequence, then when I relaunch MEP, the new modifications show up in the audio object. If I rename the original file, MEP prompts me for the new file name.

This was working fine until in one movie the new object was showing unsynced with the video segment, and seemed to be playing a bit slower. After some digging and head scratching, it turns out that the older file was recorded at 44k and the newer version was generated at 48k. [The project setting is 48k].

What MEP is doing here is that it's silently adding a 'play speed' effect to the modified audio object, which slows it down and hence the loss of sync with the video.

This is an image of the effect showing for the audio object: (I can add a video if the description here is not clear)

[I couldn't add this effect to the audio object manually because MEP requires a video object for the 'play speed' effect]

Now that I understand what's going on, I can fix the sampling rate and problem is solved. But I'm thinking this is a bug, or it could be some default behaviour that can be adjusted somehow. But the lack of user alert is still problematic. What's the expert say on this?!

@Basel-Shishani

Hi,

I'm not quite sure what you are doing. Was the audio wave object initially part of a video clip or recorded separately? If so, then I presume that is was sent to the audio editor and that is where a change occurred, like going from 48 to 44 Hz and then back to the timeline where MEP will adjust it for 48 Hz. Is this what you did?

To adjust the speed of an audio object, use Effects, Audio Effects, Timestretch/Resample.

John CB

@Basel-Shishani

Hi

. . . . just normalising and applying effects  . . . .

What effects are you adding in the external editor?

Normalisation can be done in MEP, as can many other audio effects and cleaning..

John EB

@johnebaker

I'm aware of the audio editing features in MEP, but I like to normalise on the command line using ffmpeg as I find it quicker:

> ffmpeg -i .\original.wav  -af loudnorm=I=-18:LRA=2:TP=-1.5:print_format=summary -ar 48000  original-mod.wav 

# input file original.wav turned out to be 44100 when I was assuming it was 48000. This command generates original-mod.wav as 48000 then I reload it into MEP for the same object that imported original.wav. 

After that I delete original.wav and re launch MEP, which prompts me to enter the missing file, and I choose original-mod.wav.

The processing is out of order because I forgot to process some older files before adding to MEP, but normally (now) the file is processed first then added to MEP.

[I could be wrong about the benefit of this workflow and might change it as I get more familiar with MEP, but this is how I was doing it].

I was assuming original.wav was always 48k sampling, but in one instance it was 44k. When I reloaded the modified version (original-mod.wav) into MEP, the sampling has now changed from 44k to 48k when the file got reloaded, and I got the symptoms I described, where the audio object was slowed down, and the 'Play speed' effect added automatically for the object as shown in the attached picture. But this was done without alerting me about the problem.

@Basel-Shishani

Hi Basel,

Please take note of my last reply where I note about using Timestretch.

For normalization, I would never do what you do. Normalization is almost instantaneous in MEP. Select one or more objects with audio and hit Alt+N. Everything selected gets normalized. But, the main reason to not do what you did is because you now don't know what happened to your file.

Example: I have 3 video clips taken of me walking with the dog in the forest. There is background noise of traffic way in the distance. In one clip, I'm not moving. In the second, I move and you can hear leaves crackling along with background noise. In the third, I call the dog. Normalize these. In 1, the background noise is now extremely loud; in 2, it's less but still very loud; in 3 the background noise is soft as my voice takes over and goes to 0db. Doing what you did, I would have a lot of trouble getting the sound to be at the same level for the background.

Normalizing in MEP, you get the adjusted level on each clip as the centre handle on the audio part, so you can see how much the volume was raised (or lowered) for each object. I can now look at setting for object 3 and adjust objects 1 and 2 to the same level. Of course, I may also want to reduce the level by a further 1 db.

Like John EB, I rarely use external audio editors, as almost everything that I need can be done in MEP/VPX using the built-in tools or VST plugins. I use SF Audio Cleaning Lab for spectral cleaning, but that is about it.

John CB

@browj2

Thanks. The thing is that EBU R128 is available only in VPX, that's why I resorted to ffmpeg/loudnorm. I get your point that external processing is not suitable for more demanding cases but in this case it was just a talking head, so I could get away with it. Maybe there's an EBR R128 VST that works with MEP out there, but I haven't checked.

@Basel-Shishani

Hi,

I knew I should have done the tutorial on loudness and LUFS. There are a couple of long discussions about this that I'll try to find.

I almost never use normalize to EBU R128. This standard is for overall loudness. YouTube, for example, uses -14 LUFS with 0db as the cut-off, as far as we know, and will reduce the overall volume to get -14 LUFS. EBU R128 uses -23 LUFS, IIRC, which is below what you'll get on the internet. This is an average value for the entire video and allows peaks up to 0 db. You should have none above -1 db.

In the discussions that I mentioned, I used a few songs from Magix on the timeline, plus some video clips with sound and some narration, I think. When you playback the songs, each one has a different volume. If you normalize each to 0 db, you'll still say that one song is louder than another. When the average is calculated, you'll find that one is at -7 LUFS, another at -11 LUFS, and another at -14 LUFS. The difference is in the dynamic range.

Get YouLean Loudness Meter 2, install it (VST plugin to use in MEP) and take a look at some tutorials about it. Here is a great tutorial on this:

This will allow you to visually see (and hear) what is happening at various points and overall. Also, use your ears.

Once you get an understanding of this, you'll find that you should make the sound adjustments within MEP, not individually. This is what mixing is all about. There is more, of course, and I really should do that tutorial.

John CB

@browj2 Thanks a lot. I'll check the plugin.

@Basel-Shishani

Hi

to add to @browj2's comment:-

I too use the Youlean Loudness Meter for ensuring that the volume levels in my projects are all the same perceived sound level, there is nothing worse then jumping from one movie to another and having you ears blasted due to audio level differences.

I also use the Melda Productions MDynamic EQ VST for compressing the audio - IMO it gives a cleaner audio then user the built in compressor. @browj2 do you still use this?

John EB