Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[lendall]

The three screenshots attached tell the whole story, except the Files and Space Available (lower left corner of the screen) differs every time I try to do this, even with the same playlist.  I am using a 64 GB exFAT SD card connected to a USB2 port on my computer.  I have set the drive path of this drive in Disk Management to G.  I have no trouble copying files onto this G drive  (i.e., the SD card) using Windows Explorer, which recognizes the exFAT SD card just like any other drive.

Also:  Doesn't MC create a cache of converted files on the local hard drive, to expedite future syncing of handhelds or virtual drives?  I want for MC to create this cache where I know there is sufficient space. How do I set this?  I do not find it in Sync Handheld Options.  (BTW, in the "advanced" Sync Handheld page in the Wiki the section on Conversion is labeled as "Out of date.")

Thanks for any help or assistance.  Have pent a bunch of time on this and am running out of ideas.  BTW, I am using the latest build of MC20, and out of frustrations even did a clean install, restoring all of my libraries and playlists.

[BryanC]

You need to use a semi-colon to separate the supported filetypes in your handheld sync settings.

You can designate the location of the conversion cache in the main options screen under 'File Location'.

[lendall]

Thanks to BrianC.  I had tried the format *.mp3;*.ogg;*.wma (such as is used in a lot of software I use), but it did not seem to work.  MC "accepted" ogg mp3 wma without any semicolons, so I assumed that was okay.

So, we need to back up a couple of steps.  The reason I did the clean install of the latest build of MC20 and then dumped my previous conversion cache was because analysis was creating some very unpleasant results when I used the DSP effects under Sync Handheld.  Therefore I reanalyzed these playlists using the latest build, and as I watched the reanalysis take place I would say that at least 85% of my files ended up with different R128 and other numbers -- some radically different.  So I said Great!  That must have been the problem.  Now all I have to do is to run Sync Handheld again on the reanalyzed files, and everything will be okay.  Wrong. 

Please take a look at the four screenshots below.  The first two show the Apply DSP settings I selected in the Sync Handheld menu.  Notice that the instructions for Volume Leveling suggest that the user may wish to use it in conjunction with Adaptive Volume to avoid tracks having low levels.  The next two screenshots show, on the top, an original file from my library; and on the bottom, the parallel track from the conversion cache.  I opened these in Audacity where red shows clipping.  Obviously the converted tracks from the cache show horendous clipping and are unlistenable.  Spot checks on other files in my (new) conversion cache show the same problem.  If anything it is worse, and more consistently worse, than it was before I reanalyzed my files. 

So:  Where do I go from here?  I guess I could simply skip the idea of applying DSP to tracks when I convert them, although I am not happy about that.  It would be interesting if other people were to try this experiment and see what results they get.  I actually brought this issue up before on another board, and another user did try this and got the same result:  massive clipping on the converted tracks. 

[BryanC]

You generally don't want to use adaptive volume with music tracks, I don't know where you read that. You'll lose dynamics. It's generally meant for movie watching (especially DTS audio).

Also, make sure that whatever software or hardware you are using for playback doesn't use Replaygain (or R128) tags. At this point you've applied volume leveling to the tracks but you also have the Replaygain tags still present. You either need to disable Replaygain on your playback device or stop the permanent volume leveling modifications.

[Arindelle]

+1 to what Bryan says. Most players use  replay gain, or maybe there are handhelds that use R128. Regardless, they are in the tags already. You have permanently normalized these tracks, not just the metadata. Why are you unhappy about not applying DSP to your tracks, btw?

If I were you I would reconvert them form your masters and leave out the DSP. Otherwise you are going to have to remove any leveling your player is doing only for these tracks. Also as Bryan said, adaptive leveling is for pretty specific reasons.

Unless you are doing something original from a DAW maybe?

[lendall]

Where did I read that?  Actually on the MC20 DSP screens.  

"Volume Leveling  Volume Leveling adjusts the playback volume to the same level for all files.  The industry standard R128 algorithm is used to analyze files and apply the correct volume during playback.  Intentional volume differences between tracks on the album are preserved.  The overall volume of a playlist may be reduced.  Enable 'Adaptive Volume' on the left to compensate for this."

"Adaptive Volume  Adaptive Volume adjusts playback volume to enhance the listening experience for various situations.  Peak Level Normalize   Boosts low volume content, while preserving dynamic range.  Uses the peak level from Audio Analysis when available.  Works intelligently with Volume Leveling.  Night Mode  Dynamically adjusts volume to provide a fuller listening experience under lower volume conditions.  Loud or quiet parts of a movie or song will sound more similar."

Note that the above talks about "tracks on the album" and "songs," and does not mention DTS.  

Therefore all I was doing was following instructions.  The playback software I am using is MC20.  I am using an MAudio "Audiophile" sound card -- it's probably at least eight years old.  I do not recall anything about R128 in my sound card's settings, but of course I will check.  However in any event what I am talking about are waveforms of audio files which no other software has processed other than JRiver MC20.  I open these files in Audacity, but I do this all the time, and in this case only because I want to observe the waveforms and see if the clipping that I hear is actually present, which it is.  Audacity does not change anything on its own.  

So honestly, I am extremely unhappy at this point.  "Industry standard R128 volume leveling" was touted (by the JRiver developers) as the big new audio feature in MC20 -- I believe it may have even started in MC19.  I come along and say, "Hey, it doesn't work.  I've spent hours trying to make it work, and whatever I do, it causes massive clipping, and no one takes any interest.  I wish that someone on the JRiver staff would take twenty minutes, go through the steps I have gone through several times, and let me know what results they get; specifically, if they get the clipping I get.  If JRiver wants to see my actual audio files I would be happy to provide them.  I could put them on Dropbox or something.

[Arindelle]

Hey Lendall

Maybe you misunderstood something ... I think so.

Running audio analysis through JRiver "tags" the file with the metadata that is interpreted during playback. If you set volume leveling to on (and/or adaptive volume) in DSP studio for the zone you are playing in, the volume will be adjusted on playback. The audio part of the  files themselves is in no way altered, its just part of the tag.

So let's say you want to convert a flac file to an mp3 for playback on an apple or android phone. Their generic apps interprete replay gain "tags" and adjust the volume accordingly. However, when you chose to apply the dsp to the files themselves, which it appears you have by your audacity screenshots, this permanently "normalizes" (alters) the audio portion of the files. There are specific reasons to do this, but for the common man that has a modern phone or tablet usually you won't want to. A purist will never want to.

When you are playing the same track in flac, and you have leveling set in dsp studio, the MC playback engine makes the temporary adjustments based on the list of songs playing and sends them to your Maudio dac so they play at the same level.  When you are listening to an album, you would normally not have any volume levelling on in dsp studio and then it would be sent to your DAC with the playback volume exactly at the recorded level.

Now you should not have any clipping on the original tracks at all on your M-Audio DAC unless you converted the originals themselves  -- that indeed could be pretty tragic and hope that is not the case for you. But from your post's title, this is not issue ...  It is not surprising that you have clipping, especially on classical music (or the opposite maybe for rock music, very low volume levels) on your portable device though as you have chosen to "apply" the dsp as part of the conversion itself.  Without doing this the leveling information would have just stayed in the tags, and those tags would be interpreted by the handhelds player. R128 was touted rightfully so IMHO, but is primarily used in the JRiver playback engine  -- maybe there are handheld player that use this but I don't know of any off the top of my head. Some people might pemanently "apply" DSPs to boost the gain to make their player louder, or mess with the eq., but basically you do not want this checked for conversions.

For info, if you are not aware how this works, after you import music in your library, you should then run audio analysis or you can set it in auto -import to do this automatically. Once done all the "touted" benefits will be written to the tags (not the audio portion, just the metadata). On local playback you will select the features (volume leveling and/or adaptive volume) via dsp studio or set up a separate zone for when you want to use these features.

Then when you convert from say flac to mp3 or whatever, the tag info will go along with it and you will no longer have this issue.

In the mean time looking at your screenshots, I think the only thing to do is re-convert these files as they have indeed been permanently altered.

I hope this helps. Even though this must be frustrating to you, at least in the future you will really enjoy the music with dynamics intact and no clipping.

[lendall]

Thanks to Arindelle for taking the time to write this detailed explanation.  Strange as it seem, I really do believe I understand what is going on (that the analysis stats do not change the original sound files, that they are only added as tags for use if and when I choose to use the DSP features during playback, etc.).  95% of my listening is to 20th/21st century "modernist" classical music, which has an immense dynamic range, both among and within tracks.  My primary purpose in creating a cache of music via MC's Sync Handheld function is, needless to say, to convert tracks to a lossy format that will take up less space, whether it be on an SD card or a portable USB drive.  My choice of lossy format at present is OGG-Normal.  The original tracks in my library are in a variety of formats: flac, wav, mp3, higher resolution ogg.  But for me there is also a secondary purpose in creating this cache.  It enables me to apply a certain amount of volume leveling in advance, so that I can then take these files around and listen to them on my office computer or on any number of portable devices using a variety of playback apps.  That is why MC's facility for applying DSP Volume Leveling during the conversion process appealed to me.  Of course you can argue that if I were a purist this is not something that I would want to do, and I certainly respect that argument.  But there is very little point in being purist if the result is that the music you are trying to listen to is simply too soft to hear on the subway, or in your work environment -- or too loud to hear without bothering other people.  (And you would be amazed -- or perhaps not -- at how low a tolerance most people have for Boulez and Schoenberg, coincidentally the composers of the tracks for which I posted the clipping screenshots.)

From my perspective we always come back to the same place:  I used a feature in MC following MC's instructions (as quoted in my last post), and the result was that I got unlistenable tracks, unlistenable because the level was raised so high as to cause serious clipping.  When I run a straight format conversion, either using Media Center or Audacity or any other software, I never get this, which leads me to believe that it has nothing to do with my sound card or any other component on my system.  So I see this as a bug in the MC20 software, and as a bug it is something that JRiver should address.  It seems to me that JRiver has two alternatives: either to fix the bug, or to remove the "Apply DSP Settings" option from Sync Handheld.  I am at a loss as to why Volume Leveling and Adaptive Volume should work during playback (and yes, I have set import and ripping to analyze all tracks automatically -- and the tracks I posted were re-analyzed immediately prior to running Sync Handheld), but not during format conversion.  Again, this is looking like a bug, and one that is serious enough to be addressed in some fashion.