Converting from 96kHz/24bit to 44.1kHz/16bit

[777N]

Hi,
I recorded and mixed at 96kHz/24bit; now I have to convert to 44.1kHz/16bit. I used POW-R Dither (16bit) in the master fader and I bounced it (44.1kHz/16bit) but some strange noises appeared. Suggestions?
Thank you very much!

Here my specs:

Pro Tools LE 8.0.4 with a 003 Rack

Computer
Model : Sony VPCF12S1E
Workgroup : WORKGROUP
Host Name : pro_tools
User : x

Processor
Model : Intel(R) Core(TM) i7 CPU Q 740 @ 1.73GHz
Speed : 1.86GHz
Cores per Processor : 4 Unit(s)
Threads per Core : 2 Unit(s)
Type : Mobile, Quad-Core
Integrated Data Cache : 4x 32kB, Synchronous, Write-Thru, 8-way, 64 byte line size, 2 threads sharing
L2 On-board Cache : 4x 256kB, ECC, Synchronous, ATC, 8-way, 64 byte line size, 2 threads sharing
L3 On-board Cache : 6MB, ECC, Synchronous, ATC, 12-way, Exclusive, 64 byte line size, 8 threads sharing

Computer
Platform Compliance : Intel Centrino Mobile Technology
Mainboard : Sony VAIO
BIOS : AMI R1110Y6 05/31/2010
Bus(es) : PCI PCIe IMB USB FireWire/1394 i2c/SMBus
Multi-Processor (MP) Support : No
Multi-Processor Advanced PIC (APIC) : Yes
Total Memory : 6GB DDR3 SO-DIMM

Chipset
Model : Sony Clarksfield/Lynnfield DMI
Front Side Bus Speed : 2x 2.4GHz (4.79GHz)

Chipset
Model : Sony Core Mobile (Clarksfield) UnCore
Front Side Bus Speed : 2x 2.13GHz (4.26GHz)
Total Memory : 6GB DDR3 SO-DIMM
Channels : 2
Memory Bus Speed : 2x 665MHz (1.33GHz)

Memory Module(s)
Memory Module : Hynix (Hyundai) HMT351S6BFR8C-H9 4GB DDR3 SO-DIMM PC3-10700S DDR3-1334 (9-9-9-25 4-34-10-5)
Memory Module : Elpida EBJ21UE8BDS0-DJ-F 2GB DDR3 SO-DIMM PC3-10700S DDR3-1334 (9-9-9-25 4-34-10-5)

Video System
Video Adapter : NVIDIA GeForce GT 330M (32 SM4.1 1.26GHz, 1GB DDR3, PCIe 2.00 x16)

Graphics Processor
OpenCL Processor : NVIDIA GeForce GT 330M (48SP 6C 1.26GHz, 986.6MB)
CUDA Processor : NVIDIA GeForce GT 330M (48SP 6C 1.26GHz, 986.6MB)
Compute Shader Processor : NVIDIA GeForce GT 330M (32SP 1.26GHz, 986.6MB)

Storage Devices
Seagate ST9500420AS (500.1GB, SATA150, 2.5", 7200rpm, NCQ, 16MB Cache) : 466GB (C:) (E:)
Optiarc BD ROM BC-5500H (SATA150, BD-R, DVD+-RW, CD-RW, 8MB Cache) : N/A (D:)

Logical Storage Devices
Pro_Tools (E:) : 106GB (NTFS) @ Seagate ST9500420AS (500.1GB, SATA150, 2.5", 7200rpm, NCQ, 16MB Cache)
Win_7 (C:) : 360GB (NTFS) @ Seagate ST9500420AS (500.1GB, SATA150, 2.5", 7200rpm, NCQ, 16MB Cache)
Optical Drive (D:) : N/A @ Optiarc BD ROM BC-5500H (SATA150, BD-R, DVD+-RW, CD-RW, 8MB Cache)

Peripherals
LPC Hub Controller 1 : Sony PM55 LPC Interface Controller
Audio Device : Sony P55/PM55/3400 High Definition Audio
Audio Codec : Realtek Semiconductor High Definition Audio
Audio Device : Sony High Definition Audio Controller
Audio Codec : nVidia 000Ah
Audio Codec : nVidia 000Ah
Audio Codec : nVidia 000Ah
Disk Controller : Sony P55/PM55/3400 6 port SATA AHCI Controller
Disk Controller : Sony PCIe SD/MMC Host Controller
Disk Controller : Sony PCIe SD/MMC Host Controller
USB Controller 1 : Sony P55/PM55/3400 USB2 Enhanced Host Controller
USB Controller 2 : Sony P55/PM55/3400 USB2 Enhanced Host Controller
FireWire/1394 Controller 1 : Sony E832h
SMBus/i2c Controller 1 : Intel 801xx/63xx SMBus

Printers and Faxes
Printer : Microsoft XPS Document Writer (600x600, Colour)
Fax : Microsoft Shared Fax Driver (200x200)

Network Services
Network Adapter : Controller Gigabit Ethernet PCI-E generico Marvell Yukon 88E8057
Wireless Adapter 1 : Intel(R) Centrino(R) Advanced-N 6200 AGN (802.11g (ERP), TKIP, 54Mbps)

Power Management
Mains (AC) Line Status : On-Line
Battery No 1 : 100%

Operating System
Windows System : Microsoft Windows 7 Ultimate 6.01.7600
Platform Compliance : x64

Windows Experience Index
Current System : 5.9

[Top Jimmy]

If it were me, I would do both these steps separately instead of together at the same time, and I would not do it to a track that had the final limiting applied.

I would print the mix without L2007, Precision Limiter, L1, L2 or L3 squeezing the last few dB out. I would then SRC export the mix as 44.1/24. I would bring this file into a new 44.1/24 session where final limiting would be applied with a ceiling of (-.2) with dithering and then export as 44.1/16. Beware though that exporting from 24 to 16 in PT automatically applies dithering so exporting with Pow-R at 16 bits in a 24-bit file and truncating with another program or bringing this directly into your burning software would be appropriate.

[777N]

Originally Posted by Top Jimmy

If it were me, I would do both these steps separately instead of together at the same time, and I would not do it to a track that had the final limiting applied.

I would print the mix without L2007, Precision Limiter, L1, L2 or L3 squeezing the last few dB out. I would then SRC export the mix as 44.1/24. I would bring this file into a new 44.1/24 session where final limiting would be applied with a ceiling of (-.2) with dithering and then export as 44.1/16. Beware though that exporting from 24 to 16 in PT automatically applies dithering so exporting with Pow-R at 16 bits in a 24-bit file and truncating with another program or bringing this directly into your burning software would be appropriate.

Could you be a little more specific? Sorry, but I'm not so skilled with this process. In the master fader I use a compressor/limiter, an EQ and Maxim; should I export the mix without the compressor/limiter or Maxim (or both)?
Thank you very much!

[kbruff]

Originally Posted by 777N

Could you be a little more specific? Sorry, but I'm not so skilled with this process. In the master fader I use a compressor/limiter, an EQ and Maxim; should I export the mix without the compressor/limiter or Maxim (or both)?
Thank you very much!

Remove both...

[Top Jimmy]

When sample-rate converting, peaks above those in the source can be created. It's best not to SRC a file that's been strongly limited to 0 dBFS, and it's also not advisable to add dither to a file that has been limited to 0 dBFS.

The problem may be as simple as the ceiling you have set on Maxim. If you have the ceiling set at 0 dBFS, try reducing it a little. -.2 dBFS is popularly recommended by Bob Katz.

[777N]

Originally Posted by Top Jimmy

When sample-rate converting, peaks above those in the source can be created. It's best not to SRC a file that's been strongly limited to 0 dBFS, and it's also not advisable to add dither to a file that has been limited to 0 dBFS.

The problem may be as simple as the ceiling you have set on Maxim. If you have the ceiling set at 0 dBFS, try reducing it a little. -.2 dBFS is popularly recommended by Bob Katz.

Ok, got it! Thank you. Last question, Maxim has a preset for CD quality with own dither. If I use it should I not use POW-R? Which one is better?

[Top Jimmy]

Maxim can noise-shape the dither, but POW-R is well known to have some of the best noise-shaping curves around. There's no reason you can't use the dither function on Maxim, but if you're obsessive/compulsive, then by all means go with the POW-R.

But, if you're O/C about using POW-R, then you really should be applying it after the sample-rate conversion not before. Bounce the mix out at 44.1/24 without limiting. Bring into new 44.1/24 session. Apply limiting, dither and bounce out at 16.

[777N]

Originally Posted by Top Jimmy

Maxim can noise-shape the dither, but POW-R is well known to have some of the best noise-shaping curves around. There's no reason you can't use the dither function on Maxim, but if you're obsessive/compulsive, then by all means go with the POW-R.

But, if you're O/C about using POW-R, then you really should be applying it after the sample-rate conversion not before. Bounce the mix out at 44.1/24 without limiting. Bring into new 44.1/24 session. Apply limiting, dither and bounce out at 16.

Ok, done! But...
I've just discovered a funny thing about noises: you can hear them only with VLC, the mix works perfectly with Quick Time or Windows Media Player.
Ah I did not mention that I created a multiband compressor with 4 aux tracks. Should I apply it at the final stage too?
Thank you for your precious suggestions!