Print

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

[GreggP]

Just recently upgraded from MC15 to MC17, so not sure if the problem I'm about to describe is also with MC16.

I have a 5.1 system. My sound device uses an ASIO driver. I am running Windows 7.

All of my speakers are full range. However, I still like to copy the signals below 80 Hz from all speakers to the subwoofer. The L, R, C, RL, RR speakers are Era Design 5s and they can use a little help below 80 Hz.

So, when I don't play 5.1 sources music (2 channel stereo, 4 channel quadraphonic, or just 5 channel surround (a lot of surround classical music doesn't have a sub channel) I want frequencies below 80 Hz copied to the subwoofer. I don't want any other upmixing or downmixing. So, with stereo sources, I want 2.1; with quadraphonic (4 channel), I want 4.1 (it is very important that the music is NOT mixed to the center channel speaker, because that really messes up quad recordings); and I want the 5 channel surround to become 5.1.

This worked great with MC15. Using the DSP Studio, I have the following settings:
Output Format --
Sample rate: no resampling
Channels: 5.1 channels
Mixing: no mixing (unused channels silent)
Bitdepth: source bitdepth
Subwoofer: nothing checked
Room Correction --
All speakers, except subwoofer
Volume level: 0 dB
Distance: measured distance for each speaker
Bass Management: Crossover: 80 Hz / 24 dB/octave
                          Routing: Copy bass to subwoofer / 48 dB/octave
Polarity: nothing checked
Tools: nothing checked

These settings worked great for everything I play using MC15. I can play 24bit 96 Khz multichannel tracks from another library server or locally. There's a bit of buffering for the high resolution files at the start, but then they play fine without any more gaps or stuttering caused by buffering.

Now with MC17, using the same settings
Output Format settings --
Sample Rate input frequencies -- all outputs set to no change
Bitdepth: Source bitdepth
Channels: 5.1 channels
Subwoofer: nothing checked
Room Correction --
All speakers, except sub (BTW, nice UI improvement here)
Distance: measured distance for each speaker
Volume level: 0 dB
Bass Management: Crossover: 80 Hz / 24 dB/octave
                          Routing: Copy bass to subwoofer / 48 dB/octave (NOTE: the bass is NOT getting copied to the sub)
Polarity: nothing checked
Tools: nothing checked

Playing multichannel high resolution music with MC17 can result in a lot of buffering, and repeated gaps and stuttering from additional buffering when playing a track. This sometimes goes away if I restart MC17.

These are the only DSP Studio features I'm using. To summarize, playing all music with my 5.1 system works great with MC15. All sources, except 5.1, have the bass frequencies below 80 Hz copied to my subwoofer (stereo -> 2.1, quad -> 4.1, etc.). With MC17, when it plays all sources, except 5.1, it does NOT dopy bass frequencies below 80 Hz to the sub.

BTW, I tried JRSS v2.0 surround sound just to see if/how it would work. It sounds terrible with my multichannel high resolution music. Lots of weird distortion. This isn't a big deal, because I shouldn't need it to accomplish what I want here.

[Matt]

I think you're reporting three problems:
1) Buffering / performance issues
2) Room Correction not redirecting bass
3) JRSS distortion

Let's focus on #1 first.

What is your JRMark (Help > Benchmark)?

Does adjusting ASIO settings / buffering in Options > Audio > Output mode settings... help?

Are you playing from a slow source, like a NAS?  What if you try from the C:\ drive?

[GreggP]

I think you're reporting three problems:
1) Buffering / performance issues
2) Room Correction not redirecting bass
3) JRSS distortion

Let's focus on #1 first.

What is your JRMark (Help > Benchmark)?

Does adjusting ASIO settings / buffering in Options > Audio > Output mode settings... help?

Are you playing from a slow source, like a NAS?  What if you try from the C:\ drive?

Like I said in my post, the buffering/performance issue seems to get corrected after restarting MC17, but it would be great if we can find a fix so it never happens.

When I first noticed the problem, I was playing a track from Miles Davis - Bitches Brew. This is a quadraphonic recording, ripped from a DVD-A and converted to a 4-channel FLAC. The bitrate of this track is about 6000. While I type this message, I am playing the 1st track and it is performing poorly. Lots of buffering, etc. After I let it play for about 30 seconds, it seems to play without all the buffering, but it took a long time for it to settle down.

My benchmarks:
=== Running Benchmarks (please do not interrupt) ===

Running 'Math' benchmark...
    Single-threaded integer math... 4.477 seconds
    Single-threaded floating point math... 2.652 seconds
    Multi-threaded integer math... 2.278 seconds
    Multi-threaded mixed math... 1.448 seconds
Score: 1750

Running 'Image' benchmark...
    Image creation / destruction... 2.164 seconds
    Flood filling... 1.180 seconds
    Direct copying... 3.154 seconds
    Small renders... 3.152 seconds
    Bilinear rendering... 1.890 seconds
    Bicubic rendering... 1.725 seconds
Score: 1659

Running 'Database' benchmark...
    Create database... 0.826 seconds
    Populate database... 3.603 seconds
    Save database... 1.256 seconds
    Reload database... 0.195 seconds
    Search database... 2.358 seconds
    Sort database... 1.825 seconds
    Group database... 1.058 seconds
Score: 1934

JRMark (version 17.0.68): 1781

I tried adjusting the ASIO settings and it doesn't make a difference when I increase the buffering length. The exact same settings don't cause any unusual buffering with MC15.

All of my media is stored on a media server (an unRaid server). I have your basic gigabit network (Asus gigabit router, Netgear gigabit switches, etc.). I haven't had any performance problems over the network, including playing recorded HDTV, DVDs, Blu-rays and any music tracks using JRMC15, W7 Media Center, foobar2000, MPC Home Cinema, etc. I have an HDHomeRun and HDHormeRun Prime and am able to watch and record several HDTV programs simultaneously to different computers on the network. The drives on my unRaid server spin down after a period of inactivity, so they may have just spun up when I started playing the 1st track. However, I've never noticed this performance problem with MC15.

[GreggP]

BTW, once the buffering problem goes away, as it did after about 45 seconds, I can't easily reproduce the problem. I'll see if I can figure out how to reproduce this so it is easier to diagnose.

UPDATE: no luck reproducing this... I just upgraded to 17.0.71 and it seems to play flawlessly with no buffering problems. So I don't know if the upgrade fixed this problem or if it went away just as before when MC17 was restarted. Oh well, can we try tackling my 2nd issue -- bass not getting redirected to subwoofer?

Thanks!

[Matt]

It would be good to try playing from the C:\ drive, just to see if it makes a difference.

Also, turn off from memory playback if you're using it since the stress on the NIC could be a problem.

Finally, try disabling auto-import.  Perhaps the NAS doesn't like the program talking to it from multiple threads at once?

[GreggP]

I copied the album folder to my C: drive and it plays flawlessly. There is hardly any buffering when the track starts and it plays great for the rest of the track. Transitions to the next track with no noticeable buffering. Works great from C: drive.

Same album buffers a lot from the server. Again, this wasn't a problem with MC15. And apparently is still a problem with 17.0.71.

Next, I turned OFF 'Play files from memory instead of disk (not zone specific)'. Now it plays flawlessly. Just as it plays from the C: drive. Hardly any buffering at the beginning. Very responsive.

NOTE: play from memory is ON in MC15 and I don't have the buffering problem with MC15, although it does take a few seconds to buffer the track at the start.

I haven't tried turning off auto-import yet, because turning off playing from memory has fixed the problem.

Is there any advantage to playing from memory? I must've turned it on for some reason.

Thanks again for your help.

[Matt]

Next, I turned OFF 'Play files from memory instead of disk (not zone specific)'. Now it plays flawlessly. Just as it plays from the C: drive. Hardly any buffering at the beginning. Very responsive.

Problem #1 solved.  Good.

Is there any advantage to playing from memory?

The theory is that it causes less hard drive activity.  If you're playing from the network, it should not be used.  Many systems don't behave well when data is moved quickly on the network (NIC fights with soundcard, etc.).

Now for #2, is it still happening?  How do you know it's not moving bass?  The best approach is to play pink noise and watch Analyzer.

[GreggP]

Now for #2, is it still happening?  How do you know it's not moving bass?  The best approach is to play pink noise and watch Analyzer.

It still does not copy frequencies below 80 Hz to the subwoofer. I know this because I have an application that allows me to monitor each channel on my audio interface. I can adjust the levels on every channel and monitor all signals going in and out of the interface. When playing MC15, I can see the signal light up the meter on channel 4, the sub or LFE channel, when playing any audio source (2 -> 2.1, 4 -> 4.1, etc.).

When I play a 5.1 source with MC17, channel 4 shows a signal and the meter lights up.

When I play the same sources with MC17, there's no signal on channel 4 and its meters don't light up.

I've tried looking at the analyzer while playing my 4.0 source and if I'm interpreting it correctly, it seems to show 5 different channels playing -- that's if each colored line indicates a different channel. There are 2 orange-ish lines, 1 yellow, 1 green, and 1 blue. After setting crossover to 'No crossover (play all frequencies), the analyzer still shows all 5 colored lines.

When I play a 5.1 source the analyzer shows 7 different colored lines. My app shows a signal at all 6 channels, included channel 3 - center and channel 4 - subwoofer.

[Matt]

It still does not copy frequencies below 80 Hz to the subwoofer. I know this because I have an application that allows me to monitor each channel on my audio interface. I can adjust the levels on every channel and monitor all signals going in and out of the interface. When playing MC15, I can see the signal light up the meter on channel 4, the sub or LFE channel, when playing any audio source (2 -> 2.1, 4 -> 4.1, etc.).

When I play the same sources with MC17, there's no signal on channel 4 and its meters don't light up.

Does 'Room Correction' say 'Enabled and processing data' at the top?

I've tried looking at the analyzer while playing my 4.0 source and if I'm interpreting it correctly, it seems to show 5 different channels playing -- that's if each colored line indicates a different channel. There are 2 orange-ish lines, 1 yellow, 1 green, and 1 blue. After setting crossover to 'No crossover (play all frequencies), the analyzer still shows all 5 colored lines.

In Analyzer, one line (the top one) is the sum of all channels.

[GreggP]

Does 'Room Correction' say 'Enabled and processing data' at the top?

Yes

In Analyzer, one line (the top one) is the sum of all channels.

Okay, that explains why I see 5 lines for 4 channel and 7 lines for 5.1. Maybe the color of the top line (for the sum) should have it's own color, like white.

[GreggP]

Has any progress been made with fixing the Room Correction feature that is suppose to copy bass frequencies below the crossover setting to the subwoofer?

With each speaker set to 'copy to subwoofer' and crossover set to 80 Hz, I still don't get anything at the subwoofer.

[Matt]

Has any progress been made with fixing the Room Correction feature that is suppose to copy bass frequencies below the crossover setting to the subwoofer?

With each speaker set to 'copy to subwoofer' and crossover set to 80 Hz, I still don't get anything at the subwoofer.

JRSS and Room Correction work together to create the best possible subwoofer.  This means bass management is smart enough to not fold in redirected bass when JRSS already built an ideal subwoofer from 2 or 4 channel music.

However, you've disabled JRSS.  This is fine, but the work-together logic is still engaging causing this problem.

Look for a fix in a coming build.

[GreggP]

I haven't had to use JRSS with MC15 and the bass management worked exactly like I wanted.

The problem I have with JRSS is twofold:

1. When you play a quadraphonic recording (4-channel), it will upmix it to 5.1. A center channel added to the quad mix ruins the original mix. Just like adding a center channel to a stereo mix (actually it's worse because the relationship between left and surround-left, and right and surround-right is more critical).

2. When I tried JRSS with my 24bit/48,88.2,or96KHz multichannel recordings, I was getting a lot of awful distortion.

[Matt]

2. When I tried JRSS with my 24bit/48,88.2,or96KHz multichannel recordings, I was getting a lot of awful distortion.

This is not a common problem.

Can you reproduce it?  Could you provide a sample file?

Thanks.

[mojave]

1. When you play a quadraphonic recording (4-channel), it will upmix it to 5.1. A center channel added to the quad mix ruins the original mix. Just like adding a center channel to a stereo mix (actually it's worse because the relationship between left and surround-left, and right and surround-right is more critical).

If you use JRSS with 5.1 output and check "Move Center Channel to L/R" then a quadraphonic recording will have the mains identical to the original. Two channel content will just mix the surrounds for ambiance. With this setting, it doesn't even matrix a center channel when the original source has no center.

I'm not sure how JRiver sees the quadrophonic recording regarding channel mapping, but if they aren't right, you can remap the channels correctly using Parametric EQ > Mix Channels.

From 16.0.86

4. Changed: When using DSP Studio > Output Format with the option 'Move center to front L/R' enabled, simulated center sound (like from JRSS) will not be folded into the front speakers.

[GreggP]

If you use JRSS with 5.1 output and check "Move Center Channel to L/R" then a quadraphonic recording will have the mains identical to the original. Two channel content will just mix the surrounds for ambiance. With this setting, it doesn't even matrix a center channel when the original source has no center.

I'm not sure how JRiver sees the quadrophonic recording regarding channel mapping, but if they aren't right, you can remap the channels correctly using Parametric EQ > Mix Channels.

Sorry, I'm a little confused...

Using JRSS with 5.1 output, and enabling 'Move Center Channel to L/R' for a quadraphonic recording (L,R,SL,SR), there will be no changes to the front-left, front-right, surround-left and surround-right channels. The center channel will be silent and the LFE or subwoofer channel will have bass frequencies generated by the JRSS algorithm. Is that correct?

Same setup with a 5 channel recording (L,R,C,SL,SR) the center channel will be copied equally to the front-left and front-right. The mix on the front speakers could be louder with the addition of the signal from the center. The center channel will be silent and JRSS will generate bass frequencies for my sub. Is this correct? If so, not good.

Same setup with a 5.1 recording (L,R,C,LFE,SL,SR) the center channel will be copied equally to the front-left and front-right. Again the mix on the front speakers could be louder after adding the signal from the center channel. The center channel will be silent, etc. Is this correct? If so, not good.

Also, why would I want ambiance from my surrounds with a 2 channel mix? Not to sound like an absolute purist, but I attempt to replay the music pretty much the way it was produced and engineered. I've tried numerous pseudo surround upmixing methods, both real-time and recorded and I don't really like them. I do like music that's been engineered as surround or quadraphonic, which I go out of my way to purchase, but I tend to avoid upmixes.

Thanks for your help.

[GreggP]

This is not a common problem.

Can you reproduce it?  Could you provide a sample file?

Thanks.

I can reproduce it by playing any of my multichannel recordings. For example, when playing a 4-channel/24bit/96KHz track from the quadraphonic version of Miles Davis - Bitches Brew, the sound goes from the cool jazz with vibraphone, etc., to very industrial electronica. An interesting effect, but not really what I wanted here.

I'm not sure what you mean by providing a sample. I can't easily route the output from by multichannel DAC to the inputs and make a sample recording. Nor can I just use a bunch of microphones.

If you'd like a copy of one of my multichannel tracks, I could possibly send a snippet if you provide an FTP site.  The whole track would be too large to send via email.

[Matt]

I can reproduce it by playing any of my multichannel recordings. For example, when playing a 4-channel/24bit/96KHz track from the quadraphonic version of Miles Davis - Bitches Brew, the sound goes from the cool jazz with vibraphone, etc., to very industrial electronica. An interesting effect, but not really what I wanted here.

4.0 to 5.1 channel conversion should not change any of the 4.0 channels.  So I don't understand this.

If you mute the center, does it still sound different?

[mojave]

Sorry, I'm a little confused...

Using JRSS with 5.1 output, and enabling 'Move Center Channel to L/R' for a quadraphonic recording (L,R,SL,SR), there will be no changes to the front-left, front-right, surround-left and surround-right channels. The center channel will be silent and the LFE or subwoofer channel will have bass frequencies generated by the JRSS algorithm. Is that correct?

Yes, that is correct. The only issue, as I mentioned, might be the channel mapping for the rears (they might go to the center and LFE/sub). I don't have a quadraphonic recording so I don't really know.

Same setup with a 5 channel recording (L,R,C,SL,SR) the center channel will be copied equally to the front-left and front-right. The mix on the front speakers could be louder with the addition of the signal from the center. The center channel will be silent and JRSS will generate bass frequencies for my sub. Is this correct? If so, not good.

From my understanding, whenever it mixes channels JRSS adjusts the gain and the the mixed channels are still in proper balance with the unmixed channels. In this case, the mains are still the correct volume for the sub and surrounds. JRSS will generate the bass for the sub.

Same setup with a 5.1 recording (L,R,C,LFE,SL,SR) the center channel will be copied equally to the front-left and front-right. Again the mix on the front speakers could be louder after adding the signal from the center channel. The center channel will be silent, etc. Is this correct? If so, not good.

Yes, the center will be silent. This setting would be used with a 5.1 recording for those using a phantom center. This is what I currently do. The mix on the mains is still correctly balanced for volume.

Also, why would I want ambiance from my surrounds with a 2 channel mix? Not to sound like an absolute purist, but I attempt to replay the music pretty much the way it was produced and engineered. I've tried numerous pseudo surround upmixing methods, both real-time and recorded and I don't really like them. I do like music that's been engineered as surround or quadraphonic, which I go out of my way to purchase, but I tend to avoid upmixes.

I was just stating how it works if you have it setup that way and use a 2 channel mix. Just check "For stereo sources only mix to 2.1" and then you won't get the surround channels. Some setup different zones with all settings the same except for how the zone handles the different channels. You could have a 2 channel zone, a quadraphonic zone, and a 5.1 zone. Maybe someday JRiver will have a zone tag and we can just tag our content and it will play with the correct zone settings.

According to sound engineers on some other forums, many of the 7.1 Blu-ray releases just use upmixing in the studio just like JRSS or ProLogicIIx and then record the created channels on the disc. I think Toy Story 3 was the first movie with a 7.1 soundtrack that had the extra tracks specially mixed.

 

[GreggP]

4.0 to 5.1 channel conversion should not change any of the 4.0 channels.  So I don't understand this.

If you mute the center, does it still sound different?

After muting the center channel, with my software mixer app., I still get lots of distortion.

BTW, I never used JRSS v2.0 with MC15 because of the other issues with upmixing/downmixing 4 and 5 channel sources. So, I decided to try to enable JRSS v2.0 with these multichannel recordings to see how it worked. As it turns out, the same distortion that I'm experiencing with MC17 happens with MC15.

Another interesting discovery:

With MC17, if for the Output Format settings in DSP Studio, I set
Channels: 5.1 channels
Mixing: JRSS v2.0
Subwoofer: Silent (use Room Correction to redirect bass to subwoofer)

result - distortion

then, when I change the 'Mixing' setting to 'No upmixing or downmixing' and don't make any other changes...

result - everything works the way I wanted. There's no longer any distortion and the bass frequencies below 80 Hz are hopefully copied to the subwoofer channel. I say 'hopefully' because there's no way to tell if the bass frequencies are copied or moved. I have the speakers in Room Correction all set to copy, so hopefully this is the way it is working.

So, by enabling JRSS and setting subwoofer to silent, then disabling JRSS, it works.

Does this make sense?

Thanks again for your help!

[Matt]

result - distortion

Is it possible you're just clipping?  What is the peak level at the lower left of DSP Studio?  Make sure you have 'Clip Protection' enabled.

I just don't understand how 4.0 -> 5.1 mixing could cause any distortion on the 4.0 channels.

So, by enabling JRSS and setting subwoofer to silent, then disabling JRSS, it works.

This is same bug we discussed above.  This is a workaround, but won't be necessary in a coming build.

[GreggP]

Is it possible you're just clipping?  What is the peak level at the lower left of DSP Studio?  Make sure you have 'Clip Protection' enabled.

I just don't understand how 4.0 -> 5.1 mixing could cause any distortion on the 4.0 channels.

No, it's not clipping. And clip protection is enabled. I've also tried disabling it and it makes no difference.

[Matt]

No, it's not clipping. And clip protection is enabled. I've also tried disabling it and it makes no difference.

Could you send me a file that shows distortion?  I'd like to see if I can reproduce it.

I'm matt at jriver dot com.

We don't have an FTP, but maybe you could use MediaFire, RapidShare, or similar?

Thanks.

[mojave]

4.0 to 5.1 channel conversion should not change any of the 4.0 channels.  So I don't understand this.

If you mute the center, does it still sound different?

If there are only four channels in the source, does JRSS view them as L, R, C, LFE or L, R, LS, RS? If it is the first method then the wrong channels would get mixed/routed and it would sound bad.

[Matt]

If there are only four channels in the source, does JRSS view them as L, R, C, LFE or L, R, LS, RS? If it is the first method then the wrong channels would get mixed/routed and it would sound bad.

It should view them as L, R, LS, RS. 

I don't have any real quad material for testing, so I'm hoping for a test clip.

[GreggP]

It should view them as L, R, LS, RS. 

I don't have any real quad material for testing, so I'm hoping for a test clip.

I'd be happy to send you something, but will need someplace to send it, like an FTP server. I don't know of any other way unless there was someway to setup a private torrent or something.

[KH]

Has this issue been solved? I have the same problem, and many example files. By some unknown reason, the only player I have found that can actually play these files (quadraphonic and some 5.1) are PowerDVD. But I'd like to use JRiver since it's better for playing audio files (when it works). Perhaps I could share i Dropbox folder with the address Matt provided above?

[Matt]

Has this issue been solved? I have the same problem, and many example files. By some unknown reason, the only player I have found that can actually play these files (quadraphonic and some 5.1) are PowerDVD. But I'd like to use JRiver since it's better for playing audio files (when it works). Perhaps I could share i Dropbox folder with the address Matt provided above?

Please start a new thread, and describe your settings in DSP Studio > Output Format carefully.

Thanks.

[KH]

Not sure what you did in the last update of JRiver a few days ago, but now all my Quad-albums can be played! I was btw. wrong when I said it worked with PowerDVD (which makes 5.1 of every 4.0 by default), it was only Winamp that worked with default setup. But now also JRiver plays 4.0, which is great! I first thought the crossover didn't work, but the issue was probably that these old recordings don't have as much bass information as modern recordings, so I had to turn up the volume quite a bit to hear the bass. So now the files can be played like they're supposed to, and I can set the crossover to redirect the bass to the sub.

My dream for the future would be to able to play streamed music (like Spotify) through JRiver, e.g. by navigating the streaming solution on iPad, and redirect it through the magic of JRiver (crossovers etc.), using something like the Airserver third-party application for Mac (only as a JRiver services on Windows)...

[Matt]

My dream for the future would be to able to play streamed music (like Spotify) through JRiver

This is possible, although still a little rough around the edges:
http://yabb.jriver.com/interact/index.php?topic=70242.0