19.0.20 (8/5/2013)
1. Changed: Memory playback gracefully handles cases where memory allocation fails.
2. Changed: Added 'Peak Level (Sample)' field to report the peak level found in an actual sample value (doesn't account for inter-peak overs like R128 TruePeak).
3. Changed: Added 'Volume Level (ReplayGain)' that will be filled from tags or an old MC library on import.
4. Changed: After analyzing a file, the 'Volume Level (ReplayGain)' field and tag will be filled with a ReplayGain v2 value (which is the R128 level with a +5dB adjustment).
5. Changed: When a file has a 'Peak Level (Sample)' value and no 'Peak Level (R128)' value, the sample peak will be used (with 1.5dB extra headroom added to make it more similar to R128 True Peak).
6. Changed: When a file has a 'Volume Level (ReplayGain)' value and no 'Volume Level (R128)' value, the ReplayGain value will be used with Volume Leveling (adjusted by -5dB to make it more similar to an R128 value).
7. Changed: Removed some unneeded COM initialization / uninitialization from Media Network server threads (might help with ole32.dll crashes during idle).
19.0.19 (8/2/2013)
1. Fixed: The RAW image decoding cache was not enabled for the last few builds.
2. Fixed: The 'Check for Updates' Action Window could use the wrong button width for the button (may also apply to other places where the button text changes dynamically).
3. Changed: Updated German language file (thanks Bytestar).
19.0.18 (8/1/2013)
1. Fixed: A few 3d visualizations were not working properly on Windows 7 and Windows 8.
2. Changed: Removed 3d text visualization (it's no longer possible to create a 3d mesh from text using the D3DX helper).
19.0.17 (7/31/2013)
1. Changed: When getting thumbnails JRWorker will test-load the image embedded in mp4 files, and if not valid, will fall back to capturing an image from the video.
2. Changed: JPEG tag padding is stored as an APP9 segment instead of an APP14 segment because some Android devices would not properly handle an APP14 block that contained empty data.
3. Fixed: The option to start playback of the current playlist would not work when MC was configured to start when Windows loaded.
4. Internal: Switched most of the math used by hardware accelerated drawing away from D3DX to JRiver code (please report any new issues with Theater View, etc.).
19.0.16 (7/30/2013)
1. Changed: JPEG tagging would fail to read IPTC tags from newer versions of Photoshop.
2. Changed: The tag dump for JPEG files in the Tag Action Window shows the full file structure (list of JPEG segments, with a little information about each segment).
3. Changed: JPEG tagging understands APP15 segments filled with 0xFF bytes and considers them padding that can be removed on tag save (not sure what program writes them).
4. Changed: JPEG tagging always uses a consistent segment order: JFIF (if present), EXIF, XMP, IPTC, MJMD, Padding, other App blocks (if present), image data.
5. Changed: Removing tags on a JPEG removes the XMP, IPTC, MJMD, and Padding segments (but leaves the EXIF block).
6. Fixed: Importing mp4 video files with embedded tags did not get additional tags (for example Aspect Ratio) that were not already in the file.
19.0.15 (7/29/2013)
1. Changed: Hooked up new R128 volume leveling for handhelds, DLNA, and file conversion.
2. Changed: Red October uses LAV filters version 0.58.2 (thanks nevcairiel).
3. Changed: Revised how presets are saved and loaded in DSP Studio > Equalizer to hopefully be more clear.
4. Fixed: The tag dump in the Action Window would be blank when showing for the first file.
19.0.14 (7/26/2013)
1. Faster: True Peak analyzer when using SSE on 44.1 and 48 kHz content is about 20% faster.
2. Changed: Ripping and burning now use the updated R128 audio analysis.
3. Changed: When volume leveling is enabled but applying no change, it will still show in Audio Path as +0.0dB.
4. Changed: When using album gain for R128, the loudest track in a cluster will define the gain instead of the average (since quiet tracks may call for a gain unachievable by loud tracks due to their peak level).
5. Changed: When saving tags, the R128 volume is saved into the existing Replay Gain field (using a +5.0dB shift, as per Replay Gain v2 / David Robinson).
6. Changed: When saving tags, the R128 peak level is saved into the standard 'Peak Level' field (converted from decibels to a percent and stored like '0.944').
7. Faster: ListCombine(...) expression is faster.
8. Fixed: JTV video conversion lip-sync problem.
19.0.13 (7/25/2013)
1. Changed: Volume Leveling will not use an album gain for videos except for when the Media Sub Type is set to Music Video.
2. Fixed: Direct Sound playback would not work with devices that didn't support 24-bit audio.
3. Fixed: Video width and height fields could be filled incorrectly.
4. Fixed: When converting from some jtv television recordings, the converted files would have incorrect aspect ratio. Make sure you correct width and height problem first (see 2. above) before retry conversion.
5. Changed: Video width and height values in the database are applied before playing jtv files so videos will not be played at wrong aspect ratio briefly at the beginning.
6. Changed: Switched from htbackdrops.com to htbackdrops.org (their updated site) as a source for some Theater View online media backdrops.
7. Fixed: Audio Analysis was running a core audio engine thread at a higher thread priority than intended, causing possible playback or system performance issues while analyzing.
8. Changed: DST decompression (used by DSD, CPU expensive) automatically adjusts the thread priorities used for decoding based on the caller's priority (so DSD background analysis, etc. has less impact on system responsiveness).
9. Changed: Improvements to how memory playback manages CPU load when decoding demanding sources like DST compressed DSD.
10. Fixed: The ISO media type detector was not getting some ISO files correct.
11. NEW: Video aspect ratio is imported during thumbnailing.
12. NEW: Theater View fades down the user interface elements after 90 seconds of inactivity to prevent burn in (optional in Options > Theater View > Behavior).
13. NEW: Volume Leveling accounts for volume changes from JRSS downmixing, so that mixing different channel count content will provide a more consistent volume level.
14. Fixed: Exiting Theater View while the 'News' view was loading could crash in some cases.
15. Faster: R128 True Peak analysis uses a new brickwall FIR lowpass when upsampling that's much faster (readings will be mostly the same but could vary by up to 0.2dB from the previous resampler).
16. Faster: R128 True Peak uses SSE instructions when possible to help performance.
19.0.12 (7/19/2013)
1. NEW: Real-time 2xDSD output added to DSP Studio > Output Format (requires ASIO connection to 2xDSD capable DAC) (real-time DSD encoding not available for video playback).
2. Changed: Renamed 'Peak Level' field to 'Peak Level (R128)' and changed the units in the field to 'dBTP' from 'dB' to better indicate the values are R128 True Peak values (sorry, but this will require files to be reanalyzed).
3. Fixed: Some BD ISO files were not being properly imported as movies.
4. Fixed: Selecting a Direct Sound output mode was not working properly.
5. Changed: Red October uses LAV filters version 0.58.1.
6. Changed: Red October HQ uses madVR version 0.86.9.
7. Fixed: While playing a currently recording television show in transport stream format MC would play the ts file instead of going to the tuner that was doing the recording.
8. Fixed: MC could hang upon stopping when playing a video over the network (Library server, or YouTube etc.).
9. Changed: Volume Leveling will not use an album gain for Podcasts.
19.0.11 (7/18/2013)
1. NEW: Real-time DSD output for any content is possible using DSP Studio > Output Format > Output Encoding.
2. Changed: Tuned the program startup loader thread so that background tools begin running more quickly.
3. Fixed: Audio analysis fields were not visible in video views.
4. Changed: When importing ISO files of DVD or Bluray discs, the program will automatically mark them as Video / Movie for a media type, and search for sidecar cover art.
5. Fixed: OGG import could crash.
6. Changed: 'Peak Level' values calculated by the audio analyzer are R128 True Peak values.
7. Changed: Volume Leveling and Adaptive Volume in Peak Level Normalize mode target -1.0dB instead of 0.0dB (as per R128 recommendations).
8. Changed: Updated German language file (thanks Bytestar).
9. Fixed: With memory playback enabled, seeking forward in a track after the track had been fully cached could advance to the next track instead of seeking.
10. Fixed: Seeks could be unnecessarily slow when using memory playback.
11. Changed: Peak Level analysis uses -144.0 dB as a minimum.
12. Fixed: In some cases the output plugin settings in Options > Audio were not being configured for the correct zone.
13. Changed: When starting in Media Server mode, the program more quickly gets servers and TV recording running.
14. Fixed: "jpeg" is now an allowable format for cover art.
15. Fixed: The television recording conflicts warning did not work properly when stacked channels were involved.
19.0.10 (7/17/2013)
1. Fixed: 'Analyze Audio' was not available in the right-click menu for video files.
2. Changed: The internal library version has been incremented. It will no longer be possible to load a library from this version with an older version.
3. Changed: Revised the data type of the 'Peak Level' field so that it uses decibels and contains per-channel peak levels.
4. Changed: Removed Replay Gain fields, since Replay Gain has been replaced by R128.
5. Changed: Bumped up PIX01 limits to 200 images and 6 videos per gallery. Maximum gallery size is 600M.
6. Fixed: Server serving live television programs to clients did not work with the new channel stack system.
7. Fixed: A few possible cases where background threads that showed user interface could hang or cause a focus shift.
8. Changed: Added additional logging to the search suggestion system to help debug performance issues on some machines.
9. Fixed: R128 Dynamic Range analysis was not handling gating properly in build 19.0.9, leading to incorrect readings in some cases.
10. Changed: When parsing strings into numbers, parsing math expressions, etc. both types of minus will be honored ('-' and '‒').
12. Changed: Improved how Adaptive Volume reports in Audio Path, including more detail and the current volume change.
10. Changed: DLNA added missing evented variable.
11. Fixed: DLNA renderer wasn't sending events on track change when used with a controller that supported SetNextAVTransportURI (Media Steersman).
12. NEW: Added FirstNotEmpty(...) expression function to output the first value that is not empty from a variable size argument list.
13. SDK: Added MCWS/v1/Playlists/Add function to the web service to allow playlists and smartlists to be added.
14. Changed: Switched the naming in the audio / video bitrate monitoring filters used during video playback to 'JRiver' from the older 'J. River'.
15. Fixed: MCWS/v1/Zones/Unlink was not working properly.
16. Changed: When searching in Options, the available choices are also searched for most options instead of only the selected option (not available for all options, since some lists are built on demand when they're opened).
