The normalization and the dynamic compression works so far. After a few tweaks to the parameters, VLC can now also handle the files. I'm doing my first hot test right now. On average, it takes a good 30 seconds to normalize a normal series episode. Of course, you can also have this done directly during processing by an assistant or during transcoding.
It is in the nature of things that the sound has to be recoded during this operation. FFMpeg can read in pretty much anything, but there are limits when it comes to output. Because of this, the audio tracks only have a limited selection of codecs afterwards. Depending on the setting, an attempt is made to use the same codec (AC-3, DTS, ..) as for the input file. If this is not possible, the "fallback audio codec" set by the user is used. But you can also choose a fixed audio codec. You can choose from:
These codecs can also be selected as fallback. Bit rates, sampling rates and the number of channels are retained as far as possible. A little hint. If you enter Dolby TrueHD in the whitelist and thus always want to keep it as far as possible, you have to enter "MLP FBA" there. Because that is the Codec ID as it is issued by MediaInfo. There is also an odd thing about MP3. "Mpeg Audio" has to be entered in the whitelist.
With E-AC-3 and AC-3, both must be entered separately in the whitelist. E-AC3 is also retained when normalizing, but cannot be selected as a "fixed" or "fallback" codec. Because the difference to "normal" AC-3 is only marginal. The only noticeable effect is that the E-AC-3 can only run on a fraction of the devices on which the AC-3 works perfectly. More than 99.9% of all people can't hear the difference.
The normalization can be set by setting the peak value. This must be between 0 and 1. The best values are just under 1. Personally, I found 0.95 to be ideal, which is also the default value of FFMpeg. If you go even closer to 1, you can easily hear clipping in sensitive areas. Far below that, volume and dynamics are given away.
The compression can be set in 4 levels or left completely. (Level 0) Whereby the two strongest levels 3 and above all 4 are already very violent and can lead to strange results. Either "Easy (1)" or "Medium (2)" are recommended. In the case of medium, the dialogues are audibly increased without the volume difference to e.g. explosions being flattened too much. But everyone has to find out for themselves how it best suits them at home. That's why you can stop it or switch it off completely.
Normalization is available in the wizards and when transcoding videos. There is also a separate module to normalize videos. Nothing else is done here. So no optimizer functions, video is always copied, not changed, etc. This is primarily intended to quickly normalize already "finished" videos. If the videos still have to be processed anyway, you can better do that at the same time ...
I'm not quite sure whether it makes sense to include normalization for music, i.e. pure audio files. Volume differences within an album are desired, if they exist. There are no such dramatic differences in music as there are in videos. Although, the whole thing can be switched off completely at any time, so ...
Speaking of audio. I temporarily grayed out the metadata functions for music albums, i.e. deactivated them. After the last API change at TheAudioDB, this didn't work anymore anyway, because they hid the necessary function behind the paywall. The much more important enrichment with metadata and fanart for artists still works. Album information is available in the ID3 tags anyway. MP (or Kodi or Plex or, ...) can always read this information. Only the artist metadata cannot be saved in the Id3 tags.
TVDB is not back yet, it will take some time.
Let's see, maybe there will be a new version this week ...
I am currently uploading the new version 184.108.40.206. It's extremely slow right now (for whatever reason), so it can still take a while. But I wanted to use the time to write the article. As soon as the upload is finished, I change the download link and the version number in the first posting. Oh yes, I built in normalization for music too. If you don't like that, you can easily switch it off and it is switched off by default. For videos, normalization is available for all wizards (series, feature films and other videos) as well as for the video transcoding module. Here it is activated by default with 0.95 as peak level and medium dynamic range compression (level 2). If you don't like these values, you just have to set it differently once. There is also an extra module that only normalizes the audio tracks. This is intended as a quick solution for videos that are actually already finished ... The settings for this can be found in Transcoding Videos and, as usual, have the same effect on the wizards ...
Music normalization is available with the audio wizard and the audio transcoding module. The settings in the audio transcoding module are made here, as usual ...
After a long time there is a new version 220.127.116.11 from Media-Buddy. The download can be found where it is always found (see first post).
There are many improvements and fixes. Metadata grabbing is now much more stable and, above all, much faster. Other bugs, some of which have been around for a long time, have also been fixed. There will be support for TVDB, if at all, but only in the fall. TVDB wants to completely redo the API again. It's not worth adapting everything to the new API now (always assuming we get a free API key at all, which should actually work, since Media-Buddy is open source) and then starting all over again in a few months. We will tackle this when the new TVDB API is launched.
But that alone would not justify a jump from version 1.0 to version 1.1. There are also completely new functions. On the one hand, Media-Buddy now supports .nfo files in the "new" format for Kodi 19 and higher. In general, Kodi 19 also reads most of the "old" .nfo. But with a few downsides. Since the notation for external ID (i.e. IMDB-ID, TMDB-ID, TVDB-ID) has changed, unfortunately Kodi no longer reads this important data from the "old" .nfo. Media-Buddy now optionally supports both formats, old and new. In general, however, the new format is better because it is more flexible. MediaPortal, Emby and other tools should be switched to the new format as soon as possible, imho. You can now very quickly and easily integrate your own ID systems and/or other online sources, which are then read directly into the database. For example, you can use your own ID ranges for holiday films, which Kodi can evaluate and use, for example, for filtering.
Media-Buddy wouldn't be Media-Buddy if there wasn't a function for converting "old" to "new" .nfo.
This allows you to convert existing .nfo files into the new format in one go.
To round off the portfolio, Media-Buddy can now also edit the last media types that have not yet been dealt with, namely graphics (photos, etc.). Of course, Media-Buddy does not replace an image editing program such as Photoshop. But a lot of work that often occurs in everyday life can now be done in Media-Buddy and in the usual "Media-Buddy manner". As usual, you will find some "global" settings that (for performance reasons) are only read in when the program starts. Since you rarely have to change anything here, I can "get over it", I think.
Here you can specify the file types to be used. Separated according to "normal" graphic formats and camera-specific RAW formats. Many cameras offer the possibility to take pictures in JPG as well as in RAW. RAW files are actually not usable immediately, but have no losses whatsoever. They represent a direct image of the data read from the sensor. In the digital age, this is what comes closest to a "negative" of classic analog photography. Media-Buddy uses the OpenSource CLI tool ImageMagick for image processing, which can import various camera RAW formats. The most important ones (Canon, Nikon, Sony/Minolta, Olympus and Fuji) are preset. However, you can always add your own format and/or remove formats that are not required. You can enter any file format supported by ImageMagick here. A list of formats can be found here:
There is an option to move the originals to appropriate subfolders. This increases the overview if you have different edited versions of the same photo after editing. That's why the RAW and the "normal" formats are set and saved separately.
You can also have the images automatically rotated according to the orientation stored in the EXIF metadata. That's mostly true anyway. Then this option does nothing. But if an error has crept in, it can be fixed quickly and easily.
The third option is a toggle that lets you apply optimization for .jpg and .png files to the originals as well. Personally, I always leave my originals untouched and only work with copies. But if you like, you can do that here. More precise to optimize.
And of course you also have a window in which you can actually edit the photos:
Looks pretty familiar, doesn't it? At the top you select, as usual, the folder in which you have the photos to be edited. Now you can resize the photos. Attention is paid to the best possible quality. There would be faster ways to resize the images (most of the time), but this is the best way. Various preset sizes can be selected. Each size with a tick in front of it is then generated during the run. You can select none, one, several or all sizes. Behind the sizes you can select the file type that is to be created. In addition to the most common .jpg format, you can also select others such as .png or the particularly space-saving .webp format. The sizes are preset according to common screens, plus the size 800x600 pixels, which is particularly suitable for internet forums or social media platforms. In addition, you can also choose a size freely. The size must be specified in the widthxheight format. In the example above, it is 1280 pixels wide and 1024 pixels high. Other spellings will not work. The dimensions represent the maximum size in each case. An image retains its aspect ratio and is reduced accordingly so that it fits into the given dimensions. For example, if you have an image with a size of 1280x1024 pixels and select 720p, the image will be reduced to 960x720 pixels so that it can be displayed on an HDReady monitor (TV) without scaling. It is not stretched in width but keeps its proportions.
In the middle you can select two particularly good and important filters. "Normalize" optimizes the contrast and brightness of the photo. This almost always works very well. In rare cases, however, this makes the photo worse. That's why it's better not to touch the originals. So you have a second try later. Below that you can turn on the "UnsharpMask" filter. This will make the photo sharper. Of course, you don't actually gain more sharpness (i.e. more detailed information) with this. But the impression of sharpness is (significantly) improved, the photo looks "simply sharper". But too much of it is also not good, because ugly artefacts form at the edges. It is better to use the unsharp mask sparingly. Values of 2 or a maximum of 3 should actually always be enough.
And on the right there are two hooks, through which the converted images (if .jpg or .png) are reduced again without any loss. Two tiny CLI tools (jpegoptim.exe and optipng.exe) are used for this, which do nothing else than optimize and reduce the size of a .jpg or .png file. When downloading fanart for movies and series, these tools are applied automatically. You can save up to 20% of space without sacrificing. This means that these tools can also be used on files that already exist locally. If you don't want to make any other changes, set "Optimize originals" in the settings and deselect all filters and resizing options. Then the existing graphics are only shrunk. But you can also reduce the images that have just been converted, all in one go.
@Lehmden did I read your last post correctly that the new version does not support TVDB right now?
EDIT: Never mind, I was adding a new series and was using the old style TVDB-12345.txt file which was not working. It acted like it was working but in the end all it did was delete the TVDB-12345.txt file. I made a tvshow.nfo file and added some basic info in it and it worked.
Yes, TVDB has changed from a free service to a paid one a few years ago. So we needed to cancel TVDB support (as it simply didn't work any longer). In meantime it is possible for Open Source projects (like ours) to get a free API key. That's why we hopefully can add TVDB support again. But they are planing to change their API again in a few month. This means we need to do all the work again after the new API is active. So we will wait for the "new" API to re- implement TVDB again.
The method with "TVDB-12345.txt" still is working. But for this to work, the TVDB- ID needs to be available on TMDB. This is often but not always the case.. But you also can use a "TMDB-12345.txt" file with the TMDB ID too. This ID always exists on TMDB (of course). It always is part of the URL to the series. Here is an example: 'ht tps://www.themoviedb.org/tv/4588-er" is the link to the show "Emergency Room" (aka "ER"). The TMDB ID of this show is "4588" so you need a (empty) file named "TMDB-4588.txt" for "ER". This is valid for every series available on TMDB. I needed to add a space into the URL as the forum software always replaced those URL to a link, if they were recognized as an URL So instead of "ht tps" it has to be "https"
By the way, you also can use "TMDB-12345.mkv" instead of "tt1234567.mkv" for a movie video to be identified 100%. if you (or TMDB) don't know the IMDB-ID of a movie, this may help. Example: "ht tps://www.themoviedb.org/movie/19995-avatar" is the URL to the Movie "Avatar". The TMDB-ID is in this case 19995 and the name that should be given to the video file is "TMDB-19995.mkv" (or whatever file extension it has).
By the way, "Star Wars: Episode IV - A New Hope" has the TMDB-ID "11" so there is no fix number of digits for TMDB-ID...
Thanks for that explanation. I didn't know about all the different methods you could use for the ID file so that will be very useful. I sure hope TVDB gets their API sorted. It seems like all the constant changes have made their service less reliable.
There is a new version of Media-Buddy available for download, 18.104.22.168.
Bugs have been fixed and various external programs have been updated.
In addition, Media-Buddy can now automatically download trailers for feature films. They are packed with the film and can thus be used in various HTPC programs to create a "real" cinema feeling, in which one or two trailers are played before the actual film.
You only need to check the mark next to "trailer"...
However, most of the changes can be found in the image area.
On the one hand, you can now copy Exif metadata from one image to another with Media-Buddy. The Exif metadata contains data such as the camera used, the lens, the exposure settings, but also e.g. GPS coordinates and much more. Unfortunately, when editing the photos, this metadata is too often changed, damaged or even removed. With Media-Buddy you can now copy the Exif data from the unedited original photos to the edited "masterpieces".
All file types that can also be processed by Exiftool are supported. However, there are only a few file types that can fully contain all Exif files. In addition to the camera RAW file types, these are primarily .jpg, .tif and many of the native formats of image processing programs such as .psd (Adobe Photoshop) etc... .png also saves Exif metadata, but unfortunately only a very small part of the possible Data. Therefore I would not recommend .png as a format if you want to keep the Exif data. In addition to .jpg as a very heavily compressed but also very lossy format, you should primarily use .tif. Depending on the setting, .tif also compresses, although not nearly as much as .jpg. But the format is lossless. Especially with intermediate steps you should avoid .jpg as much as possible, because the losses add up every time you save and read in again. It is therefore essential to use .tif or .psd here, even if the files are much larger. Only when everything is really ready can you save the photo as a .jpg and thus save about 90% space with good quality.
As usual, you select the source and destination folder and Media-Buddy does the rest. The images must have the exact same file name so that the data can be copied. Only the file extension can be different. For example, you can copy the Exif data from .jpg files to .tif files. If the names do not match, the corresponding files are simply omitted.
In addition, the processing has been revised and expanded again. So you can now choose different color options color, black and white and sepia (the brownish "antique" look that old photos often have), and that in any combination.
Black & White:
And you can now choose two different variants when normalizing, "Nature" and "Studio". The easiest way to explain the differences is with examples. The "Nature" setting works very well for landscape shots and the like with many different colors and brightnesses. However, if you have photos with a monochrome background, as is often the case with studio. Photos is the case, it can very easily lead to color distortions.
First, the example from above is taken up again:
You can see immediately that the "Nature" variant has a color cast and the young lady's face has turned green. The "studio" variant is probably the best here.
As a counter-example, a recording from my Heligoland trip a few weeks ago. It has to be said that the photo was taken early in the morning under a gray sky and poor lighting conditions. It actually looked a lot like the original photo.
Here the "Studio" version is slightly better than the original, but the "Nature" version is by far the best.
The installation file can be found where it is always found (see first post). It is currently being uploaded. So it may take a moment until it's really online. And now have fun.