I gots a couple questions regarding this RADAR thing.

1: When people say it sounds better than ProTools what are they comparing it too? Once the thing is digital it shouldn't 'sound' like anything. Are you saying that the converters are better? Better than what, the stock stuff from Digi or some other option from somebody like Lucid or Prism or something. Can RADAR be used with third party converters?

2. Is the file format RADAR data is stored in openly documented? By this I mean if I have a bunch of stuff that was recorded on one of these things do I need a RADAR system to retrieve the data?

projectMalamute wrote:I gots a couple questions regarding this RADAR thing.

1: When people say it sounds better than ProTools what are they comparing it too? Once the thing is digital it shouldn't 'sound' like anything. Are you saying that the converters are better? Better than what, the stock stuff from Digi or some other option from somebody like Lucid or Prism or something. Can RADAR be used with third party converters?

I believe they're referring both the converters and the "internal" sound of the software. Audio softwares have "sound", this sound is the algorithm used for the internal summing and routing I believe. or maybe how the software accesses the data. I'm not a software engineer, so these are just guesses. Of course, different software, different algorithm, hence you get a different sound.

From what I gather, Radar is going only up to 48kHz. 48kHz sample rate really doesn't "translate" the analog signal well. You might not notice it until you get to high frequency signals, I don't know how high, maybe 15k?

I don't think you can really load any plug-ins on radar tracks, I mean, I don't think the system supports that, and I believe that that's really helping the sound. I was making some comparisons with an Audio Precision unit once, I was comparing a certain analog EQ with it's equivalent digital plug-in. When doing it on 44 kHz, it was just crazy, the graph of the processed sound had ramps all over it, you could see the how bad the resolution is. It looked much better when I did the tests on(or should I use "in") 96kHz, the digital graph was smooth and it much pretty much perfectly the analog graph.

So eliminating the plug-ins would help, I think.
Sorry for my descriptions and explanations not being that scientific, I'll try to dig more into that.

projectMalamute wrote:2. Is the file format RADAR data is stored in openly documented? By this I mean if I have a bunch of stuff that was recorded on one of these things do I need a RADAR system to retrieve the data?

I believe that you can access the data, never checked that though.

eliya wrote:I believe they're referring both the converters and the "internal" sound of the software. Audio softwares have "sound", this sound is the algorithm used for the internal summing and routing I believe. or maybe how the software accesses the data. I'm not a software engineer, so these are just guesses. Of course, different software, different algorithm, hence you get a different sound.

That's right. The summing algorithms (and other audio processing algorithms) are different. That's partly why people use those summing amps with their direct outs rather than summing in the box. I remember a few years back discussing in a studio about how Pro-Tools wasn't too bad if you mixed through a desk and just used it as a recorder rather than anything else.

I've not heard a Radar system, but some of the guys at work who do the outside broadcasts use this which they record with in DSD format - one bit delta-sigma like what's used for SACD. I listened to a surround mix they were doing which certainly wasn't bad at all, but there was no comparison for reference.

eliya wrote:48kHz sample rate really doesn't "translate" the analog signal well. You might not notice it until you get to high frequency signals, I don't know how high, maybe 15k?

A 48KHz sample rate will give you an audio bandwidth of just under 24KHz.
The limit of human hearing is 20KHz, less if you're older and your hearing has deteriorated.

Type "Nyquist Theorem" into wikipedia and you'll get a bunch of stuff on this!

Rodabod wrote:

eliya wrote:I believe they're referring both the converters and the "internal" sound of the software. Audio softwares have "sound", this sound is the algorithm used for the internal summing and routing I believe. or maybe how the software accesses the data. I'm not a software engineer, so these are just guesses. Of course, different software, different algorithm, hence you get a different sound.

That's right. The summing algorithms (and other audio processing algorithms) are different. That's partly why people use those summing amps with their direct outs rather than summing in the box. I remember a few years back discussing in a studio about how Pro-Tools wasn't too bad if you mixed through a desk and just used it as a recorder rather than anything else.

I've not heard a Radar system, but some of the guys at work who do the outside broadcasts use this which they record with in DSD format - one bit delta-sigma like what's used for SACD. I listened to a surround mix they were doing which certainly wasn't bad at all, but there was no comparison for reference.

But I thought we were talking about a system that replaces a tape machine: 24 ins, 24 outs, no plugins. If you ask the computer to do something to the sound of course it's going to sound different, my question is how and why RADAR sounds better than anything else if it's just being used as a tape machine. It doesn't look like anything I couldn't put together myself with enough channels of high quality converters and a linux box.

Mister_Tog wrote:You do have a point, but you can easily make several exact copies of whatever it is you are storing the files on and store them in multiple locations with minimal climate control concerns.

Sure you can. There's all kinds of stuff you can do.

Mister_Tog wrote:With tape, you can only have one truly first generation copy.

Well, that's not true for mixdown, but even for tracking - you can have a second generation that sounds pretty good compared to the first.

news story on the incident wrote:Universal Music Group, an unrelated company, leased space in one of the video library vaults for master copies of reel-to-reel audiotapes of music from the 1940s and 1950s, but all of the archive had been copied, much of it digitally, as the site was being phased out, a spokesman said Monday, "so in a sense nothing was lost."

But your point about the original story - these were audio masters from the 40s and 50s. Some of that stuff was 60 years old, ready to play and needed to be killed by fire to be made useless.

You're saying that the digital copies of that stuff are going to be as good to go in 2060? Maybe you have a job as a spokesman for Universal Music Group.

= Justin

projectMalamute wrote:But I thought we were talking about a system that replaces a tape machine: 24 ins, 24 outs, no plugins.

If you are not summing, then the sound is purely down to the conversion and codec/bitrate of choice, so you are correct - a soundcard and Linux would do this. I'm not sure if Radar offers summing, but if it does then this would also affect the sound.

projectMalamute wrote:

Rodabod wrote:

eliya wrote:I believe they're referring both the converters and the "internal" sound of the software. Audio softwares have "sound", this sound is the algorithm used for the internal summing and routing I believe. or maybe how the software accesses the data. I'm not a software engineer, so these are just guesses. Of course, different software, different algorithm, hence you get a different sound.

That's right. The summing algorithms (and other audio processing algorithms) are different. That's partly why people use those summing amps with their direct outs rather than summing in the box. I remember a few years back discussing in a studio about how Pro-Tools wasn't too bad if you mixed through a desk and just used it as a recorder rather than anything else.

I've not heard a Radar system, but some of the guys at work who do the outside broadcasts use this which they record with in DSD format - one bit delta-sigma like what's used for SACD. I listened to a surround mix they were doing which certainly wasn't bad at all, but there was no comparison for reference.

But I thought we were talking about a system that replaces a tape machine: 24 ins, 24 outs, no plugins. If you ask the computer to do something to the sound of course it's going to sound different, my question is how and why RADAR sounds better than anything else if it's just being used as a tape machine. It doesn't look like anything I couldn't put together myself with enough channels of high quality converters and a linux box.

That's exactly the point. Radar only has the algorithm to access the data submit it to it's path(track) and then convert it. Other audio programs are summing it to a master track(if you're not mixing OTB), or internal routing to virtual busses, etc'. This stuff degrading your audio.

And also, because Radar doesn't "let" you use plug-ins, you're forced to do everything outside, while when using pro tools or such, you might use a plug-in and not an outboard processor which you might want to save for other tasks. You're most likely to do some processing ITB, especially if it's for small/easy tasks. The use of these plug-ins will detract your audio quality.

I believe that Radar is also like a tape machine because of the workflow. I believe it's pretty straight forward and there are no gazillion menus and you don't have to look at the screen(If I remember correctly, you can use radar without a monitor, at least in some versions). Because Radar isn't involved with any 3rd plug-ins, it's probably much more stable than any other audio program with some plug-ins. Radar's OS is dedicated to Radar and is developed by the company and fully tested by the company, where with other formats(i.e. audio programs), the OS isn't developed by the company making the software. It's heavily tested by the QA department of the company, but it's never enough. Not to mention the updates that the OS maker is putting out every week or so. This stuff is fucking up the system.

matt_stevens wrote:

eliya wrote:48kHz sample rate really doesn't "translate" the analog signal well. You might not notice it until you get to high frequency signals, I don't know how high, maybe 15k?

A 48KHz sample rate will give you an audio bandwidth of just under 24KHz.
The limit of human hearing is 20KHz, less if you're older and your hearing has deteriorated.

Type "Nyquist Theorem" into wikipedia and you'll get a bunch of stuff on this!

It's been a long time since I read about nyquist and not sure if I remember it right.
The higher we get in frequency, the less "sampling points" we have, no? so 15k signals will be less transparent, no?

I know we don't hear above 20kHz, but I believe we "feel" these frequencies in a way. Things sound differently when they're not there.

anyhow, reading about nyquist now.

Rodabod wrote:

projectMalamute wrote:But I thought we were talking about a system that replaces a tape machine: 24 ins, 24 outs, no plugins.

If you are not summing, then the sound is purely down to the conversion and codec/bitrate of choice

And the quality of the analogue line stages that come before the AD and after the DA of your chosen digital machine/program.

God, Pyramix. How do your colleagues find it? I can't wait til we scrap it at work, but that's mainly because we've had no training on it, so providing 'support' is hit and miss at best. The computers Merging provided are fucking rubbish though, and all sound like turbines, which is not particularly great for a really expensive computer, built buy a DAW manufacturer specifically for audio use...