SOUND ON SOUND

[Hugh Robjohns]

... it's best to give that to people who are used to dealing with the new generation of jobsworths that seem to have infested many broadcasters.

There aren't any jobsworths! It's just a computer saying NO!

That's the whole point of R28 and its equivalents... all technical parameters are now completely objective and measured automatically when the file is ingested. Your programme either provides the right numbers or it doesn't.

No human intervention or evaluation occurs any more. There's just a computer print-out.

[Hugh Robjohns]

Hugh, would what you're referring to by any chance be something like what NUGEN LM Correct 2 does?

Hugh will certainly reply, but he most likely simply meant "lower the fader" .

LM Correct is an automatic audio file processor that measures an audio file much faster than real time, and then automatically adjusts it's 'gain' to comply with a specified target loudness before re-saving the file. Widely used in post-production and programme ingestion applications.

But the same process can be achieved manually by running the programme through a loudness meter in real time, noting the numbers, and then applying a 'gain' change -- either manually with the output fader or by altering the level of audio clips en-masse, to correct any offset before exporting the file.

[Hugh Robjohns]

Updated Test numbers

2 minute song, Rendered LM Correct 2 file: -23.8 Integrated, -12.3 Max True Peak, 12.3 LRA

and then tested separately after

12 minute Dialogue, Rendered LM Correct2 file: -24.0 Integrated, - 2 Max True Peak, 12.1 LRA

:-) Those numbers suggest there is little or no compression on the dialogue -- which is not unusual...

Of course... measuring each section of the programme can be interesting, but it's the entire programme's Integrated Loudness figure that matters as far as the broadcaster acceptance is concerned.

[NewtoPT]

Thank you guys for the responses and clarifications, I am mixing/mastering one of the performances today, will keep posted.

[CS70]

Hugh, would what you're referring to by any chance be something like what NUGEN LM Correct 2 does?

Hugh will certainly reply, but he most likely simply meant "lower the fader" .

LM Correct is an automatic audio file processor that measures an audio file much faster than real time, and then automatically adjusts it's 'gain' to comply with a specified target loudness before re-saving the file. Widely used in post-production and programme ingestion applications.

But the same process can be achieved manually by running the programme through a loudness meter in real time, noting the numbers, and then applying a 'gain' change -- either manually with the output fader or by altering the level of audio clips en-masse, to correct any offset before exporting the file.

:thumbup:

[NewtoPT]

Having a problem regarding video playback in PT 12 (Windows 10) when I'm stopping/starting keep getting these errors below.

"The Video Engine reported an error. More information is in the Video Engine log file"

"The Video Engine reported a synchronization error. Try increasing the video Play Start Latency preference.

"The Avid Video Engine was unable to stop"

Now the video will start playing again like half the time once I hit play, sometimes it won't right away. But those errors come up nonstop. When I put the video offline, no errors happen.

Imported a DNxHD-36 MXF .. Tested a AppleProRes .mov, same errors. Even tried the original MP4 I used to convert from, Same errors. I set the Disk Cache to 8GB (I have 16GB of RAM) changed from 'frames' to 'blocks'., HW buffer size 1024. No luck. Confirmed fps is matching. Running my sessions off a T7 portable Samsung SSD ext. and the Video off that same Drive. Also tried importing the video off my other Samsung T5 external SSD to see if that would help but it didn't.

Anyone know what the problem could be?

[Eddy Deegan]

Running my sessions off a T7 portable Samsung SSD ext. and the Video off that same Drive. Also tried importing the video off my other Samsung T5 external SSD to see if that would help but it didn't.

Anyone know what the problem could be?

I have a Samsung T7 and I think there is a possibility that it's the problem. Although their transfer rate is very fast on average I've noticed that it regularly pauses I/O briefly (and sometimes not so briefly - up to 10 seconds on large bulk transfers) during which time the blue light swirls quickly, presumably while it's doing something internal.

I've noticed this both on reading (usually when accessing a directory for the first time) and writing (usually when doing large transfers). I've not seen much about this online but it's mentioned here.

I'm not saying it's guaranteed to be the culprit but I would definitely try a different brand of drive before ruling it out as a cause.

[NewtoPT]

Thanks Eddy. But that means it should be fine to still be running the actual PT sessions off the T7 right? Just not importing the video from it? The T5 had the same issues as well when importing from that when playing back. But I will try a different external brand altogether and see if that helps at all. Worth a shot, thanks for the suggestion.

[Eddy Deegan]

Thanks Eddy. But that means it should be fine to still be running the actual PT sessions off the T7 right? Just not importing the video from it?

I'm not so sure; it would rather depend on how much data PT 'bursts' from the drive each time it reads it. If PT has run out of buffered data, asks the drive for more and the Samsung is in 'hang on a minute, I'm busy' mode, then you'd certainly notice.

Of course none of this may apply in your specific scenario; I'm just mentioning that (at least some of) the T series Samsung drives have got issues when it comes to on-demand access and this could possibly be upsetting things!

[NewtoPT]

So I tested with different externals, unfortunately still having the same problem. Feel like I've tried everything at this point lol.

[James Perrett]

If you are just using the video for reference could you reduce the video resolution to reduce the load on the system?

[NewtoPT]

Just tested with a 360p video. Same errors. Looks like I may have to switch to Mac.

[The Red Bladder]

... it's best to give that to people who are used to dealing with the new generation of jobsworths that seem to have infested many broadcasters.

There aren't any jobsworths! It's just a computer saying NO!

That's the whole point of R28 and its equivalents... all technical parameters are now completely objective and measured automatically when the file is ingested. Your programme either provides the right numbers or it doesn't.

No human intervention or evaluation occurs any more. There's just a computer print-out.

1. That's just audio and has meant that all audio is crushed and reduced to the grot level of TV speakers - OK for rubbish like 'Strickly' or the news, but deadly for films or even music with a wide dynamic range.

2. Humans still seem to be looking at the image and rejecting for pointless things like not protecting for 14:9. (Though we did get something rejected by BBC2 a month ago for having the wrong count-in which was justified I suppose - it wasn't me Gov!)

3. Given the rubbish prices broadcasters are paying nowadays (the BBC being the worst by a considerable margin) I am shocked that they have the audacity to reject anything! I've seen one-hour TV documentaries in HD being sold for £500 to the BBC. That's not a joke!

[The Red Bladder]

Thanks Eddy. But that means it should be fine to still be running the actual PT sessions off the T7 right? Just not importing the video from it? The T5 had the same issues as well when importing from that when playing back. But I will try a different external brand altogether and see if that helps at all. Worth a shot, thanks for the suggestion.

It needs to be on an internal SSD.

I get the same thing but with different glitch reports in Premier or Vegas Pro or Resolve if I try to use an external drive via USB, inc. USB3.

[NewtoPT]

Hey guys, had a question regarding Loudness Meter .csv logs. Specifically Nugen VisLM2's.

Is it right that the excel file only generates in tandem when you analyze and render with LM Correct 2?

I had the first log exported, for some reason it only showed 2 minutes out of 4 minute audio
Then I just tested with LM Correct again and the VisLM2 log was written over with 4 minutes of data.

The Integrated Loudness measure in the Log is out of LKFS spec for stretches of time, but then goes back into Spec. As I'm sure should be the case. I know all the matters is that the Average of the enitre program is within LKFS range. But since (from my understanding) the Broadcaster will measure a chunk of the Audio couldn't this pose an issue?

Edit: I see the last number at the very bottom of the log is the LKFS of the whole audio. But what if I want to monitor it manually without LM Correct 2?, Let's say for instance my LKFS is 23.1 and I want to have a log in the VisLM folder where the data is computed off of an 23.1 LKFS reading that's been tested.