The No.1 Website for Pro Audio
Source COnnect - 192k vs 384
Old 22nd October 2019
  #1
Lives for gear
 
Led's Avatar
Source COnnect - 192k vs 384

Hi, I have a client who needs to add source connect to use here and there with a new VO artist overseas. They don't need to use video, just want to listen in and talkback etc, then record on the Pro Tools in their edit suite. The artist already has source connect so they want to go that way.
It's just recording mono cold voice reads, 24bit 48K, so can they get away with the 192kb standard version, or should I suggest they get the Pro with 384kb?
I've never used Source Connect so can't really tell them from my own experience.
Thanks for any advice.
Cheers
Old 22nd October 2019
  #2
Lives for gear
 

I have tested this a lot. 192 sounds really good but there is a very slight loss of low low end. 134 is just about 100% identical to PCM.

You should compair for yourself. You can get a demo and try it.
Old 22nd October 2019
  #3
Here for the gear
 
nixt's Avatar
192 is actually for stereo. When recording a mono voice you're limited to 96. It works but the difference is very noticeable compared to the pro version. When I work with people who are limited to 96, I ask them if they can record on their side and send me the files after the session.
Recently we have been using another system, more similar to Source Connect Now but which works better in my opinion, and without the latency that Source Connect sometimes has. It's 200 $ per year (or 20 per month) and when you have a subscription you can invite a remote talent for free. Check it out : https://www.sessionlinkpro.com/
The quality is pretty good (256 kb mp3) and you can use MTC for video (or just send LTC on the 2nd channel). I use Source Nexus to route the audio between Pro Tools and Chrome. Works like a charm.
Old 22nd October 2019
  #4
Lives for gear
 
Led's Avatar
Thanks
Old 22nd October 2019
  #5
Lives for gear
 
minister's Avatar
Quote:
Originally Posted by nixt View Post
192 is actually for stereo. When recording a mono voice you're limited to 96. It works but the difference is very noticeable compared to the pro version. When I work with people who are limited to 96, I ask them if they can record on their side and send me the files after the session.
To be clear, you are talking the Standard version. Pro goes up to 384 kbps.
Pro X goes to 1344.
Old 22nd October 2019
  #6
Here for the gear
 
nixt's Avatar
Quote:
Originally Posted by minister View Post
To be clear, you are talking the Standard version. Pro goes up to 384 kbps.
Pro X goes to 1344.
Yes, but still, these are the figures for stereo. With the standard version, you can record in mono at 96 kbps. With the pro version, up to 192 kbps.
Old 22nd October 2019
  #7
Lives for gear
Just a sidenote:

There's very little in spoken word above 12k, male or female. US Analog FM stereo has a hard cutoff at 15 kHz. JBL's large cinema arrays (5732) are spec'd -3 dB to 19k, and -10 dB to 20k. All of which can be handled quite nicely in a well designed 48 k s/r system.

For me, distortion is a lot more critical in a voice recording than extended highs...

Yes, Source Connect is a compressed system. But I can't find any spec as to whether its bit distribution is frequency-adaptive based on program material, or if it just assumes a faster connection wants more HF... anybody know for sure?
Old 22nd October 2019
  #8
Lives for gear
 

Quote:
Originally Posted by Jay Rose View Post
Just a sidenote:

There's very little in spoken word above 12k, male or female.
Hate to disagree here but I've been mixing dialogue for 20 + years. Vocal harmonics reach above 12k, for sure.
Old 22nd October 2019
  #9
Lives for gear
Quote:
Originally Posted by brandoncross View Post
...I've been mixing dialogue for 20 + years. Vocal harmonics reach above 12k, for sure.
No worries. There’s surely something up there. I said “very little”... certainly enough for good ears to hear, but not having >12k doesn’t affect intelligibility or performances. Or it least it didn’t for most Hollywood films pre-1990 or so.

My other examples were 15k for most pre-HD broadcast, and 20k even for new good theaters... all of which can be handled by standard 48k* sampling. The big question is what does medium bitrate Source Connect do within that band.

Anybody care to run some objective tests? It’ll be hard, I know, for psychoacoustic compression (which should do much better on narrow band test signals than on practical audio) but can give us some insights.

*Yes, there are benefits to 96k recording if you’re slowing down the voice. But there, the final result is still 48.
Old 22nd October 2019
  #10
Here for the gear
 
nixt's Avatar
Jay Rose, you might be mistaking Kbps and KHz... An mp3 file at 96 kbps doesn't sound near as good as a 256 kbps one. This has nothing to do with the bandwidth.
Old 22nd October 2019
  #11
Lives for gear
 
Led's Avatar
So, without derailing my own thread.. I recently had an issue where a delivery was bounced by a tv network here and was sent the master to check. Straight off the bat it was waaayyy too loud, it turns out the editor had gotten the specs wrong and thought it was for online, not tv. The other thing I noticed is the voice sounded different to my mix file, lacking air for want of a better term. I did some null tests and then some spectrum analysis and the restripe had somehow ended up with compressed audio (data not dynamic range), and had a pretty sharp drop off starting at around 16Khz. So for my money at least there is important stuff up there. Whether it matters to the viewing public I don't know..

Thanks for the replies re SC, the stuff in question will be recorded for HD TV here, so I'm going to suggest they go Pro instead of standard, but I'll get them to check what version the talent in the States has, because I think we'll be limited to that level regardless of what we have on our end?
Old 23rd October 2019
  #12
Lives for gear
Quote:
Originally Posted by nixt View Post
Jay Rose, you might be mistaking Kbps and KHz... An mp3 file at 96 kbps doesn't sound near as good as a 256 kbps one. This has nothing to do with the bandwidth.

No, these are two different things. The OP was about data speed AND compression algorithm, which are different things. No question that given the same algorithm (and designer choices), more speed will result in better sound... up to a point of diminishing returns.

The OP was also about spoken word. I added a sidenote about bandwidth, because in a lot of cases the software designer's choice between medium and high speed data is to extend the bandwidth. This may be an important consideration in a remote music session, but not relevant to a voice-over.

Note also Brandon's contribution saying he hears a low-end loss at 192 that he doesn't hear at 384. This is obviously that kind of software designer's choice. As you know, rolling off the lows doesn't save any data at all with standard PCM linear codecs (.wav and AIFF). What Brandon is hearing could be a decision to re-distribute SC's frequency bins at the higher rate, with fewer available for low frequencies.
Old 25th October 2019
  #13
Lives for gear
 
Mundox's Avatar
We recently tried Source Connect Now, which is a free version if one of the participants have a Source Connect account.
It sounded very good with minimal delay. And it goes up to 512kbps.
I'd give that a shot before investing in the pro, since the VO already has access to SC Now.
Old 25th October 2019
  #14
Gear Addict
 

Source Connect Now users, be aware there is a bug in version 77 of Chrome if you are using it.
Old 25th October 2019
  #15
Lives for gear
 
Led's Avatar
Quote:
Originally Posted by Mundox View Post
We recently tried Source Connect Now, which is a free version if one of the participants have a Source Connect account.
It sounded very good with minimal delay. And it goes up to 512kbps.
I'd give that a shot before investing in the pro, since the VO already has access to SC Now.
I made them aware of this as an option, but they want to go with something that has full tech support rather than something still officially in beta.
Reliability and support is more important than cost to them. I might try SC now at my place and see how it runs though.
Old 31st October 2019
  #16
Gear Head
 

The thing about all the Chrome solutions including Source-Connect Now is that they are more like Skype in many ways. Specifically Google designed the features these systems use in Chrome for general communications and not true fidelity. The goal is a to support a conversation and provide intelligibility where if better fidelity is possible then it streams at the desired quality level. Chromes goes for much lower latency and then if there is a bandwidth issue for a moment it will "conceal" the data loss with various techniques including time stretching (making some words looooonnnnger and others fasttocatchbackup) and variable bitrate fluctuations causing changes in the frequency response. One problem is that sometimes if you are not paying attention during the live voice record session, these artifacts can be missed and go unnoticed until the connection is over causing more of a headache and delay to resolve. The time stretching causes problems with ADR work, picture sync and even podcasting. In general the connection Chrome provides is more "voip" like and much more on a yo-yo as it lacks true error correction and instead just makes stuff up if there are any problems.

In comparison Source-Connect Standard/Pro provides a locked and clocked connection that can have higher latency (especially when connecting with Standard on older versions or on Windows). With Source-Connect errors a fixed or you hear a dropout. There is no "concealment" There is no time stretching. There is no bitrate change. Its binary... good or bad. Actually the AutoRestore/Replace system guarantee the recording is good even if errors manifest out as dropouts.

Regarding codecs, Chrome uses the Opus codec which is very good while Source-Connect Standard/Pro use the AAC-LD / AAC-ELD codec brewed by Fraunhofer. Some would argue that AAC is superior to Opus.

One other thing, while Source-Connect Standard provides 192kbps in stereo and only 96kbps in mono, its possible that due to the joint stereo nature of the AAC compression the higher bitrate provides more than 96kbps to the mono component of the stereo stream and so thus its possible that a mono voice at 192kbps AAC stereo sound better than a mono voice at 96kbps AAC mono.
Old 1st November 2019
  #17
Lives for gear
 
Mundox's Avatar
Just a note for whatever its worth, Source Connect Now works as a standalone app as well, so it doesn't rely solely on Chrome.
Old 1st November 2019
  #18
Gear Addict
 

Quote:
Originally Posted by Mundox View Post
Just a note for whatever its worth, Source Connect Now works as a standalone app as well, so it doesn't rely solely on Chrome.
Yes it does. We had to use it for one of our remote talent last week.
There was some metering issues, but no dropouts.
Old 1st November 2019
  #19
Lives for gear
 
minister's Avatar
Quote:
Originally Posted by Mundox View Post
Just a note for whatever its worth, Source Connect Now works as a standalone app as well, so it doesn't rely solely on Chrome.
No, that's just a helper app, it still flows through Chrome.
Old 1st November 2019
  #20
Gear Addict
 

Chrome has been updated to version 78
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump