V-DAT Quality

voidar · Post by **voidar** » Sun Sep 24, 2006 11:33 am

The http://www.cockos.com/reaper (http://www.reaper.fm/) sequencer is actually supperior to Traction 2 as it has a wholy 64-bit signal path. I just tried the summing test wich it passed and T2 failed, even in 64-bit mode.

Cute little thing.

Updated link.

[ This Message was edited by: voidar on 2006-09-25 23:04 ]

Post by **hubird** » Sun Sep 24, 2006 12:57 pm

thanks man, it's on the list for my little nephew

astroman · Post by **astroman** » Mon Sep 25, 2006 3:56 pm

well, with such a tacky company name, an ingeneous logo and that favourite icon from the early Mac days it had to be a winner

I'm impressed - it might even beat S.A.W Studio for usability...

big thanks for the link, Tom

Post by **hubird** » Mon Sep 25, 2006 9:18 pm

what icon Astro?

voidar · Post by **voidar** » Mon Sep 25, 2006 10:18 pm

There are some interesting threads on the Reaper forum on why SAW allegedly sounds better than floating point sequencers.

http://www.cockos.com/forum/showthread.php?t=2473

http://www.cockos.com/forum/showthread.php?t=2076

While summing-tests seem to be equal, people report hearing a difference during playback. It might be due to the way SAW communicates with the sound-hardware.
While other applications have a integer/float/integer-conversion which might introduce jitter, SAW has a purer signal path.

This might be what some us experience when using VDAT, or even bussing out tracks to Scope mixers.
Does this matter? Sure, if you actually hear what is going on you are doomed to mix better.

voidar · Post by **voidar** » Sun Oct 08, 2006 2:09 pm

Final verdict.

As of today I would say Adobe Audition 2 is the winner in regards to compatibility with Scope HW.
It edits 32-bit PCM, and seems to record them too.

arela · Post by **arela** » Sun Oct 08, 2006 2:13 pm

oh no!
Suddenly a reason for updating my 1.5
....think i will sleep on it

voidar · Post by **voidar** » Sun Oct 08, 2006 2:37 pm

I think 1.5 can record 32-bit PCM through the 24-bit wave-source drivers. At least 1.5 reports the Creamware Play/Rec drivers as 32-bit PCM compliant.

Try to record a reverb-tail into 1.5.

1.5 has a lot of controll of how you can treat in-/out-comming audio. But in 2.0 I am not too certain whether it record to float at first, and to what format.

In the editor I see the last bits below the -160dB area, so it's definitely 2^32 precision.

There's not much of midi-editing here though which means a different application for automation.
Unless you feel like using ramps.

[ This Message was edited by: voidar on 2006-10-08 15:50 ]

arela · Post by **arela** » Mon Oct 09, 2006 1:02 am

I did a test , rec 2 vdat sessions
1 is 24 bit int and 2 is 32 bit int
......and i can open them in Audition 1.5.

But saving the files from Audition gives me this options:
1- 32-bit normalized float (type 3)
2- 4-byte pcm (type-1 32 bit)
3- 24 bit packed int (type-1 24 bit)
4- 24 bit packed int (type-1 20 bit)
5- 32 bit 24.0 float (type-1 24 bit non standard)
6- 16.8 float - obsolete/compatibility

What is the 4-byte type?

This is whith scope wave source

Then
I switched to scope 24 bit wave source
When i import the same files, the 32 bit file i edited (cut the last 20s) and saved as 16.8 bit, is destroyed

I cut 20s off one of the 24 bit files and saved it, same format.

Then,
back to wave source
The destroyed 32 bit file is still destroyed, but when i drag it to Cubase, it's ok. The other files is ok.

Then
I recorded 20s stereo with wave dest
and saved to the types mentioned. (new files)
Type 1, 2, 5 & 6 are 6891Kb while type 3 & 4 is 5169Kb.
All files could be opened again, both with wave source and 24 bit wave source

This is over my head

[ This Message was edited by: arela on 2006-10-09 02:04 ]

voidar · Post by **voidar** » Mon Oct 09, 2006 6:00 am

4-byte PCM is the proper 32-bit format that VDAT and STS loads/saves.

I tried recording 32-bit PCM through the Wave-dest 24 drivers, but AA1.5 only seems to save them at 16-bit resolution.
You can check in the Edit-window. Just zoom in at the end of the file. If you recorded a reverb-tail or something, you'll be able to see the last values as steps. We're talking below -90dB range here.

In 1.5 the last bit always defined the -96dB area.
In 2.0 it was below -160dB. Technically it should be -192dB.

From 2.0 manual:

The Microsoft Windows PCM format supports both mono and stereo files at a variety of resolutions and sample rates. It follows the RIFF (Resource Information File Format) specification and allows for extra user-information to be embedded and saved with the file. The WAV format reproduces digital audio by using PCM (Pulse Code Modulation)—PCM doesn’t require compression and is considered a lossless format.
You can include Broadcast Wave metadata in Windows PCM files. (See Broadcast Wave options.)
The following options are available for 32-bit files; no options are available for 8- or 16-bit files:

32-bit Normalized Float (type 3) – Default The internal format for Adobe Audition and the standard floating point format for type 3 .wav files. Values are normalized to the range of +/–1.0, and although values beyond this range are saved, clipping may occur in some programs that read them back in. (Adobe Audition won’t clip audio but will instead read the same value back if it’s beyond this range.)

4-byte PCM (type 1, 32-bit) Saves 32-bit audio as 32-bit integers, retaining 192 dB of dynamic range. However, signal-to-noise ratio in saved files is 144.5 dB, reflecting the 32-bit float format that Adobe Audition uses internally.

24-bit Packed Int (type 1, 24-bit) Saves straight 24-bit integers so any data beyond the bounds is clipped. The .wav BitsPerSample is set to 24 and BlockAlign is set to 3 bytes per channel.

24-bit Packed Int (type 1, 20-bit) Saves straight 24-bit integers so any data beyond the bounds is clipped. The .wav BitsPerSample is set to 20 and BlockAlign is set to 3 bytes per channel. The extra 4 bits are actually the remaining valid bits when saving, and they are used when reading (thus still giving 24-bit accuracy if those bits were actually present when writing). Applications either fill those last 4 bits with zeros or with actual data; analog/digital converters that generate 20 bits of valid data automatically set the remaining 4 bits to zero. Any type 1 format with BlockAlign set to 3 bytes per channel is assumed to be packed integers, and a BitsPerSample value between 17 and 24 will read in all 24 bits and assume the remaining bits are either accurate or set to zero.

32-bit 24.0 Float (type 1, 24-bit) – Non-Standard Saves full 32-bit floats (ranging from +/–8million), but the .wav BitsPerSample is set to 24 while BlockAlign is still set to 4 bytes per channel.

16.8 float – Obsolete/Compatibility The internal format used by Adobe Audition 1.0. Floating point values range from +/–32768.0, but larger and smaller values are valid and aren’t clipped since the floating point exponent is saved as well. The .wav BitsPerSample is set to 32 and BlockAlign is set to 4 bytes per channel.

Enable Dithering Dithers 32-bit files when they are saved to a PCM format (20-bit, 24-bit, or 32-bit). This option is available only for a 32-bit file that you select to save to a nonfloating-point type format. It applies a Triangular dither with a depth of 1.0 and no noise shaping. If you wish to apply a noise-shaped dither, use the Edit > Convert Sample Type command to dither the audio first, and then save the file without dithering enabled in the file format options.

Anyway, you need to tick off the "treat 32-bit PCM files as 16.8 float" box in setings, or so.

I am not totally convinced about AA yet. I think I should petition the REAPER creator to add 32-bit PCM recording support.
Because hardly any editing software reads the 32-bit .AIFF's it writes, except T2, VDAT and STS.

What I want to avoid is the ASIO/float/integer conversion that is so common in these applications. I want an app that reads our ASIO-drivers correctly!

arela · Post by **arela** » Mon Oct 09, 2006 9:37 am

Tail - i looked under my chair

The file i got up now, ends with what i thought was total silence, but zooming in shows it is not flat at all, but got peaks up to -62dB - from -160dB i think
Is that what you talk about voidar?
File Format: 24bit packed int (type 1)
File Type: 44.1kHz 32 bit float mono

Am i on the right track?

voidar · Post by **voidar** » Mon Oct 09, 2006 11:44 am

Audition automatically saves to 32-bit float which I don't think is a good idea.
It is 4 byte 32-bit PCM, type 1 that would be ideal.

Here is a STS recorded 32-bit file, in Edit view of AA.

It seems that AA views the last bit as -174dB after conversion.

"Big Meter" in SFP reports upon recording the signal as being -180.6dB.
Playing back the file on STS3000 via pre-view outputs, Big Meter reports -172.6dB.

[ This Message was edited by: voidar on 2006-10-09 12:48 ]

voidar · Post by **voidar** » Mon Oct 09, 2006 12:26 pm

I get better S/N-ratio when recording to AA via ASIO FLT than directly to STS.
Last bit of STS seems to define about -174dB to -inf while AA stops at -186dB.

"Big Meter" also reports some dB lower with the AA-file played back with STS, than the native STS-file.

So, recording to AA and converting it to 32-bit int actually gives better S/N.

This has been a technical "deaf"-test though.

-- I just did a test with VDAT. VDAT saves as low as -186dB, just like ASIO. So VDAT gives better results than STS (!). That's 4 steps on the "ladder" in Edit view.
Can someone confirm this?

[ This Message was edited by: voidar on 2006-10-09 14:05 ]

arela · Post by **arela** » Mon Oct 09, 2006 1:05 pm

the picture/link is broken

I also got the lastest Cool Edit,
so will do some comparding
I suspect Adobe used this engine intheir first issues of Audition.

voidar · Post by **voidar** » Mon Oct 09, 2006 1:17 pm

Seems like home.no is a bit unstable at the moment.

wayne · Post by **wayne** » Wed Oct 18, 2006 3:07 am

Better late than never....

Just had my first day properly using VDAT - and struth, it's a beauty

Low brass has never had this much depth & dynamic on this DAW. I'm a happy bloke

As S4L would say, this dingo is chokablok fulla rabbits

rock on regardless, W

Jonathan T · Post by **Jonathan T** » Wed Oct 18, 2006 9:29 am

Just for the record, I use SAWStudio and SFP and they work beautifully together.

The midi side of SAW is certainly quite basic but it is possible to carry out all the midi recording and editing functions I need one way or another. The rock solid sync with the audio tracks, plus the audio quality, makes up for the lack of midi bells and whistles. I've seen the 'problem syncing audio and midi data' message too many times in Logic (old PC version to be fair) for me to take it too seriously.

SAW does have a reasonably steep learning curve I suppose, as does any DAW, but it is entirely logical and has a very fast workflow once you get used to it. The virtual mixer is laid out very much like a traditional analogue board.

Regards

Jonathan

voidar · Post by **voidar** » Thu Oct 19, 2006 7:09 am

How do you use Scope with SAW?

I remember SAW sounding beautifull, but I got some glitches when using XTC-effects when automating the SAW mixer.

Jonathan T · Post by **Jonathan T** » Fri Oct 20, 2006 2:21 am

I don't use Scope in XTC mode with SAW, I run it in standard mode, usually with an STM2448 and the 24 bit ASIO drivers with about four or five stereo channels for each. Select Options/Audio Driver Model/Asio Protocol/Asio Scope in SAW, the Asio channels then appear in the Audio Device Setup menu. SAW's outputs then route to these four or five stereo inputs of the STM2448 which are then effectively subgroups. If I want to use a Creamware effect I route the relevant SAW channel(s) to another STM2448 channel and use the effect as an insert or Aux.
Creaware's midi sources appear as virtual midi ports within SAW Midi Workshop so I can route to Creamware synths/B2003 etc via these.
For recording guitars via Dynatube etc, or bass, I usually use STM2448 bus outputs routed via the ASIO 24 bit Dest back to SAW's record inputs.

Best regards

Jonathan