PDA

View Full Version : Big Party



pcourtney
2011-09-12, 11:12
I have been asked to help provide the music at a wedding this weekend, and this will involve about ten receivers (SB Duets) , luckily I can use ethernet that is already in place in every main room, and I can run some more Cat5 where I need it, I have my own 24 port switch that will act as the hub for the computer, so networking wise I am all set.

but I did a smaller party not so long ago and we had a pair of active stereo speakers in every room, my feeling was that this was definitely overkill because nobody could really listen to the music in the proper stereo imaging sense whilst dancing and chatting away, after all it's just a party :-)

So this time I want to put just one active speaker in every room, and this will also cost 50% less in speaker hire fees for the night, but I need to ask if I put a stereo to mono adapter on every duet receiver will it be OK, I have tested it just now and it seems OK to my ears, but is this simple fix the best way, or am I missing something ??

SuperQ
2011-09-12, 13:29
Yea, it should be fine to combine the output this way.. but it might be easier to mix to mono using a custom file type in SoX. It would also eliminate your worry about the plugs.

JJZolx
2011-09-13, 08:02
Yea, it should be fine to combine the output this way.. but it might be easier to mix to mono using a custom file type in SoX. It would also eliminate your worry about the plugs.

Ok, I had to try this. I tested it using the Sheffield/XLO Test & Burn-in CD, which has channel and phase tests. Worked like a charm. What was interesting is the second track is a relative phase test. The in-phase part worked normally, but the out-of-phase test was completely silent.

Create a file named custom-convert.conf in the same folder as convert.conf and restart Squeezebox Server. I only tested the first rule, with is for standard 16/44.1 Flac streaming, which normally streams natively. The second rule is used for downsampling high bitrate Flac files for players that can't play them natively. Of course, this assumes all of your files are Flac. You'd need additional conversion rules for other types of files.



flc flc * *
# FT:{START=--skip=%t}U:{END=--until=%v}
[flac] -dcs $START$ $END$ -- $FILE$ | [sox] -q -t wav - -t flac -C 0 - remix 1,2

flc flc transcode *
# FT:{START=--skip=%t}U:{END=--until=%v}D:{RESAMPLE=-r %d}
[flac] -dcs $START$ $END$ -- $FILE$ | [sox] -q -t wav - -t flac -C 0 $RESAMPLE$ - remix 1,2


I'm not sure how much overhead this adds to the server, but with 10 streams you probably want to test it first.

pcourtney
2011-09-13, 08:08
oh that is very cool JJ, thank you very much.

JJZolx
2011-09-13, 08:21
Just be aware:

With 10 players, even if they're all sync'd together, SBS is sending out 10 individual audio streams. This means that your server will be running 10 of the above conversion processes. You'll have 10 instances of Flac (decoding Flac) and 10 instances of SoX (mixing and encoding back to Flac) all running at the same time.

Make sure you test this first to be sure you server can handle it. If not, your cabling option may be the better solution.

pcourtney
2011-09-13, 08:28
will test tonight and report back, if it fails I may have to install SBS 7.6.1 on a more powerful 2U Intel Server we have at work, lucky I have one spare, thanks for the heads up.

andyg
2011-09-13, 09:22
Just be aware:

With 10 players, even if they're all sync'd together, SBS is sending out 10 individual audio streams. This means that your server will be running 10 of the above conversion processes. You'll have 10 instances of Flac (decoding Flac) and 10 instances of SoX (mixing and encoding back to Flac) all running at the same time.

Make sure you test this first to be sure you server can handle it. If not, your cabling option may be the better solution.

Actually if they are synced, you should only have 1 flac+sox conversion running.

rayman1701
2011-09-13, 09:24
Only 1 thing to mention about your original idea with the plugs, since you only get 1 channel, if you were to play for instance something from an early stereo Beatles album, you might end up with just music or just vocals depending on which channel you are using for your output. Which is kinda in the same area of thought as JJZolx's post about the out of phase test tracks. Just something to keep in mind.

JJZolx
2011-09-13, 09:33
Actually if they are synced, you should only have 1 flac+sox conversion running.

Good news. Thanks.

pcourtney
2011-09-13, 15:12
Only 1 thing to mention about your original idea with the plugs, since you only get 1 channel, if you were to play for instance something from an early stereo Beatles album, you might end up with just music or just vocals depending on which channel you are using for your output

I tried this out with the cable supplied in the Duet box, and tested this out with a plug from my local Maplins store that I bought today

http://www.maplin.co.uk/gold-plated-phono-socket-splitter-218491

and it successfully combines both stereo channels into one combined phono signal/channel, so very cheap to do this, and the quality was very good too !

But to my surprise the JBL Eon 15 G2 speakers that we are going to use have two 1/4" input jacks and two-band equalization. This would normally allow a live double act to reinforce vocals and instruments without an external mixer, but going from the SB Receiver to the JBL I no longer have to worry, as the mixer inside the JBL now takes care of it :-)

I always wondered why people like JBL pro gear !

pcourtney
2011-09-13, 15:15
Actually if they are synced, you should only have 1 flac+sox conversion running.

good to know, just a quickie, is that also the case if I am streaming WAV to the ten SB Receivers ( no flac or sox ) , and the receiver just receives the wav and decodes to PCM natively ?

JJZolx
2011-09-13, 16:53
good to know, just a quickie, is that also the case if I am streaming WAV to the ten SB Receivers ( no flac or sox ) , and the receiver just receives the wav and decodes to PCM natively ?

Is what the case? If there's no conversion process to run then there would be no transcoding applications running. You do still have 10 individual streams being sent by the Squeezebox Server, though, whether or not the 10 players are synched.

I tried figuring out a conversion rule for streaming WAV as PCM mixed to mono, but had no luck at all. Kept getting chipmunks, as it seems to want to play at double speed. Probably the same problems that I encountered in this thread (http://forums.slimdevices.com/showthread.php?t=87260) when trying to downsample Flac and stream as PCM.

But you can stream WAV files mixed to mono if you stream them in Flac, using the following rule in custom-convert.conf:



wav flc * *
# FD:{RESAMPLE=-r %d}
[sox] -q -t wav $FILE$ -t flac -C 0 $RESAMPLE$ - remix 1,2

Phil Leigh
2011-09-15, 00:35
good to know, just a quickie, is that also the case if I am streaming WAV to the ten SB Receivers ( no flac or sox ) , and the receiver just receives the wav and decodes to PCM natively ?

Yes it is true, but why on earth would you want to stream wav in this party setup?

pcourtney
2011-09-15, 13:40
Is what the case?

I think Phil is saying the opposite to you, that if there is no transcoding going on, eg streaming native wav files to the receiver, then only 1 synced stream is sent to all ten receivers, which is how it should be IMHO

Quote: Originally Posted by andyg
Actually if they are synced, you should only have 1 flac+sox conversion running.

Quote: Originally Posted by pcourtney
good to know, just a quickie, is that also the case if I am streaming WAV to the ten SB Receivers (no flac or sox transcoding required) , and the receiver just receives the wav and decodes to PCM natively ?

JJZolx
2011-09-15, 13:52
No, there are definitely 10 streams going out over the network. Each player must be sent its own stream. There's only one conversion process, though. All streams send data from that one process. If the network bandwidth used to send one 16/44.1 PCM stream is 1.411 Mbps, then streaming to 10 players requires 14.11 Mbps.

pcourtney
2011-09-15, 13:58
Yes it is true, but why on earth would you want to stream wav in this party setup?

2 reasons, one because we can, a fast ethernet network should be able to handle this, no other users or computers, just my iPad, and 10 receivers, if SBS cannot handle it, then I reckon SBS has much bigger problems with the whole architecture of how it's been coded in the first place,

and two because the 1gb USB drive he has given me to look at is chock full of wav files, a lot these are old vinyl LP's that have been ripped to a hard drive using the clever little USB PhonoPlus

http://solodeals.com/USB-PhonoPlus-V2-ARTcessory/M/B000BBGCCI.htm

and all the cracks, pops and noise removed, along with replaygain etc apparently is was a labour of love that took him over a year to do 500 albums.

http://gnormalize.sourceforge.net/

I have now bought one of these USB PhonoPlus jobbies to rip some of my old vinyl, but I don't think I will be taking as much care, just nice to have them on the NAS for a rainy day :-)

maggior
2011-09-15, 14:09
a lot these are old vinyl LP's that have been ripped to a hard drive using the clever little USB PhonoPlus

http://solodeals.com/USB-PhonoPlus-V2-ARTcessory/M/B000BBGCCI.htm



I have one of these too that I got a few years ago. It's a great sounding strudy little device.

aubuti
2011-09-15, 14:45
and two because the 1gb USB drive he has given me to look at is chock full of wav files,
Did you mean 1TB? Personally I'd be more risk averse and convert them to mp3 or stream as mp3. No one will notice the lossless/lossy difference in a setting like that, but they will notice if you push the network over its limit. But I have no doubt you'll choose the hard way :-)

Phil Leigh
2011-09-16, 00:47
No, there are definitely 10 streams going out over the network. Each player must be sent its own stream. There's only one conversion process, though. All streams send data from that one process. If the network bandwidth used to send one 16/44.1 PCM stream is 1.411 Mbps, then streaming to 10 players requires 14.11 Mbps.

Yes that is what I meant.

Phil Leigh
2011-09-16, 01:01
2 reasons, one because we can, a fast ethernet network should be able to handle this, no other users or computers, just my iPad, and 10 receivers, if SBS cannot handle it, then I reckon SBS has much bigger problems with the whole architecture of how it's been coded in the first place,

and two because the 1gb USB drive he has given me to look at is chock full of wav files, a lot these are old vinyl LP's that have been ripped to a hard drive using the clever little USB PhonoPlus

http://solodeals.com/USB-PhonoPlus-V2-ARTcessory/M/B000BBGCCI.htm

and all the cracks, pops and noise removed, along with replaygain etc apparently is was a labour of love that took him over a year to do 500 albums.

http://gnormalize.sourceforge.net/

I have now bought one of these USB PhonoPlus jobbies to rip some of my old vinyl, but I don't think I will be taking as much care, just nice to have them on the NAS for a rainy day :-)



The default behaviour of SBS is to stream WAV files on disk as FLAC over the network (with a single conversion at the server end if all players are synched)... unless you manually change this. I was simply asking why you would change this to stream 10 lots of WAV over the network instead of 10 lots of FLAC - you seem to be going to some trouble for zero benefit (and potentially network bandwidth issues, although that may not affect you).
The format of the files on disk is irrelevant.
SBS can handle it fine, but many peoples networks (maybe not yours) could have an issue.

Blisteringblue
2011-09-16, 04:33
Did you mean 1TB? Personally I'd be more risk averse and convert them to mp3 or stream as mp3. No one will notice the lossless/lossy difference in a setting like that, but they will notice if you push the network over its limit. But I have no doubt you'll choose the hard way :-)

I'm no audiophile in any shape or form but I am an IT man and have to agree with this one.

Surely in a party environment the music is a secondary event anyway? People are there to celebrate a wedding, eat, drink and socialise? They are not going to notice (or I suspect care) if the music is streamed as 320k MP3 or fully loaded WAV files?

While I can see it is a big bash & sounds like the host wants to impress his guests, but surely nothing wrong with MP3 in that environment?

Phil Leigh
2011-09-16, 04:45
I'm no audiophile in any shape or form but I am an IT man and have to agree with this one.

Surely in a party environment the music is a secondary event anyway? People are there to celebrate a wedding, eat, drink and socialise? They are not going to notice (or I suspect care) if the music is streamed as 320k MP3 or fully loaded WAV files?

While I can see it is a big bash & sounds like the host wants to impress his guests, but surely nothing wrong with MP3 in that environment?

From the available data, very few people on the planet can reliably tell the difference between streamed 320kbs mp3 and flac. Of those, only a very tiny percentage even claim to be able to hear any difference between an SB playing streamed flac or streamed WAV... and their case is not borne out by any available measurement techniques.

You are playing these files in mono. That's going to have a much bigger impact on the sound quality than the streaming format in some cases. Just monoing (via cables or SOX) stereo files into mono will seriously mess up some of them (i.e. the ones with out-of-phase l/r components - many tracks from 1967-75 can surprise you).


Anyway, enjoy the party!

As others have said, I'd stream as FLAC or MP3 for a stress-free party...

pcourtney
2011-09-16, 05:04
thanks everyone for all the good advice, have explained this to him, and we are converting about 7 hours worth of wav to flac, when I did a test for him this morning, he was very happy with the overall sound quality, yes this was in his office at work, but all the same it sounded pretty good, I think it's going to be fine, and I will now set this all up tomorrow morning and make sure all ten rooms are sounding OK before the guests arrive.

have a good weekend everyone !

aubuti
2011-09-16, 05:21
One more bit of advice: make sure those MP3s have replaygain tags. And, assuming that you'll be playing a mix of tracks from various albums, set the SBs to use either Track Gain or Smart Gain.

SuperQ
2011-09-17, 10:52
2 reasons, one because we can, a fast ethernet network should be able to handle this, no other users or computers, just my iPad, and 10 receivers, if SBS cannot handle it, then I reckon SBS has much bigger problems with the whole architecture of how it's been coded in the first place,

and two because the 1gb USB drive he has given me to look at is chock full of wav files, a lot these are old vinyl LP's that have been ripped to a hard drive using the clever little USB PhonoPlus

http://solodeals.com/USB-PhonoPlus-V2-ARTcessory/M/B000BBGCCI.htm

and all the cracks, pops and noise removed, along with replaygain etc apparently is was a labour of love that took him over a year to do 500 albums.

http://gnormalize.sourceforge.net/

I have now bought one of these USB PhonoPlus jobbies to rip some of my old vinyl, but I don't think I will be taking as much care, just nice to have them on the NAS for a rainy day :-)

Be careful with backups of those WAV files. You might want to generate a checksum database since WAV doesn't have good built-in error detection for corrupt files. FLAC does have error detection.

pcourtney
2011-09-18, 07:49
just to thank everyone for the advice, the party went very well, without a hitch, no drama, everyone happy :-


just to say a big hearty thanks to you all for helping me out on this one!

JJZolx
2011-09-18, 07:51
What did you end up doing?

pcourtney
2011-09-18, 08:13
saved his wav collection onto my USB drive, then converted his wav's to FLAC, and used Foobar ReplayGain, then copied all the files to a VortexBox that I setup specially for the occasion, simples :-)

http://wiki.slimdevices.com/index.php/RGFoobar

http://wiki.slimdevices.com/index.php/AutomaticVolumeAdjustment

this was also a helpful tutorial for me that I followed courtesy on Sonos

https://sonos.custhelp.com/cgi-bin/sonos.cfg/php/enduser/std_adp.php?p_faqid=438

if you use VortexBox ver 1.09, this needs to be observed
http://info.vortexbox.org/tiki-index.php?page=ReplayGain

http://vortexbox.org/forum/general/add-uniform-replay-gain-to-entire-collection/

but I think if you use the latest ver 1.10 it is now doing RG properly according to Ron Olsen, and I quote "The changes to the replay gain code came after the VB 1.9 ISO was released" nb Ver 1.10 was released on the 7th Sept

http://vortexbox.org/2011/09/vortexbox-1-10-released/





This was an interesting read too

http://forums.slimdevices.com/showthread.php?t=87061

pcourtney
2011-09-18, 09:40
One more bit of advice: make sure those MP3s have replaygain tags. And, assuming that you'll be playing a mix of tracks from various albums, set the SBs to use either Track Gain or Smart Gain.

thanks, that was good advice, I used "Smart Gain", even though some advocate to use "Track Gain"

You have to set your SqueezeBox players to use the RG tags.
I did this by simply going to the SBS gui
SBS > Settings > Player > Audio page
Looked for the "Volume Adjustment/Replay Gain" line
Changed this value to "Smart Gain", and clicked "Apply".

nb I had to do this for all ten players, but it sounded pretty much OK all through the party, so it was worth doing this I think.

JJZolx
2011-09-18, 10:25
About the mono?

pcourtney
2011-09-18, 10:57
About the mono?

I did mention in message 10, that I did not have to worry about this in the end, as to my big surprise the JBL Eon 15 G2 speakers that we hired had two 1/4" input jacks and two-band equalization. This would normally allow a live double act to reinforce vocals and instruments without an external mixer, but going from the SB Receiver to the JBL the onboard mixer inside the JBL took care of it !

I did do some tests with a cable that combined the stereo into one mono (both left and right channels combined) and it also worked fine, a lot of worry about nothing I guess, but it needed testing to be sure :-)

aubuti
2011-09-18, 15:05
thanks, that was good advice, I used "Smart Gain", even though some advocate to use "Track Gain"
If you were playing a mix of tracks, without consecutive tracks from the same album in the _same_order_ as the album, then SmartGain==TrackGain. And even if you had some of those, for the rest of your mix it was effectively doing TrackGain. Which in this case, is Smart.

pcourtney
2011-09-18, 16:46
If you were playing a mix of tracks, without consecutive tracks from the same album in the _same_order_ as the album, then SmartGain==TrackGain. And even if you had some of those, for the rest of your mix it was effectively doing TrackGain. Which in this case, is Smart.

when I go into Player Settings/Audio
Volume Adjustment / Replay Gain

I am offered 3 possible settings for RG
1) Track
2) Album
3) Smart

If I am setting up a carefully constructed playlist (a mix of tracks) and I don't think I would hardly ever have a consecutive track playing from the same album, which of the above 3 settings should I choose then ?

Are you saying it does not matter, Track and Smart are the same ?

MrSinatra
2011-09-18, 18:29
did u use smart crossfading?

i'd like to see a beat mixing plugin.

aubuti
2011-09-20, 13:54
when I go into Player Settings/Audio
Volume Adjustment / Replay Gain

I am offered 3 possible settings for RG
1) Track
2) Album
3) Smart

If I am setting up a carefully constructed playlist (a mix of tracks) and I don't think I would hardly ever have a consecutive track playing from the same album, which of the above 3 settings should I choose then ?

Are you saying it does not matter, Track and Smart are the same ?
Please see the documentation for the replay gain settings. Even the basic info that shows up in the "i" icon on SBS settings (if I remember correctly). Under certain conditions (eg, random or carefully constructed mixes), Track Gain and Smart Gain give the same results. Under other conditions (eg, playing an entire album), Album Gain and Smart Gain give the same results.

Basically, track gain sets the levels based on the track gain RG value. So all tracks play at about the same level, whether it's an acoustic ballad or some death metal track. Much like broadcast radio.

Album gain uses the album gain RG value. So it preserves the dynamic range of songs on the same album. Think "Blackbird" and "Helter Skelter" from the White Album: track gain would play them at the same volume, whereas album gain would play Helter Skelter much louder.

Smart gain uses album gain if the tracks are consecutive tracks from the same album, and track gain for anything else. So if your carefully constructed playlist never has consecutive tracks from the same album, then Smart Gain will give you the same results as Track Gain, which is probably what you would want.

MrSinatra
2011-09-20, 14:52
when you do RG, there's only two gain tag frames/fields, track and album. thats it. (and two peak files that go with them to avoid clipping).

all smart gain does is pick which one to use, based on upcoming playlist. thats it.

so all its designed to do, is keep the user from manually having to switch between track or album [gain].

Phil Leigh
2011-09-21, 06:49
when you do RG, there's only two gain tag frames/fields, track and album. thats it. (and two peak files that go with them to avoid clipping).

all smart gain does is pick which one to use, based on upcoming playlist. thats it.

so all its designed to do, is keep the user from manually having to switch between track or album [gain].

And it does it perfectly. Just set SMARTGAIN and be happy :-)

pcourtney
2011-09-24, 14:08
thanks everyone, will be using Smart Gain whenever I do another party, cheers :-:

jephens
2011-09-24, 20:13
Great thread, thanks!!