SRC for remastering

[lukejs]

when remastering a song which has already been dithered and SRC´d to 16 bit 44 khz, I heard it´s best to upsample to 24 bit 96khz. What´s the best way to do this ? I use Sonar, and work at 24 96, but when I import an audio file, I´m pretty sure it doesn´t resample it up to 96 khz. Do I have to buy something like R8 Brain Pro for best results in SRC or is there something out there for free which is just as good ? I´ve used Adobe Audition, but it seems to sound worse than the original file. Any tips on this process of SRC to 96 khz, and then mastering processing , and then reconverting to 44.1 and 16 bit ? With all the SRC and re dithering, are the gains from SRC and re dithering worth the trouble ? Thanks for any insight on this matter , Luke

[Cellotron]

Originally Posted by lukejs

when remastering a song which has already been dithered and SRC´d to 16 bit 44 khz, I heard it´s best to upsample to 24 bit 96khz.

Actually, what would be "best" is to get the original mixes at their original sample rate and bit rate (if digital) or get the original tapes (if it was first mixed to analog)! Obviously this isn't always available for projects like this.

Anyway - my own usual workflow when presented with 16bit/44.1kHz mixes is to simply play them as received via an excellent DAC to my analog process chain, capture back at 24bit/44.1kHz, and then if I am doing further processing, use digital processors which internally up and down sample (i.e. Sonoris LPEQ in "HQ" mode, Voxengo Elephant 2.5 in oversampling mode - usually with me at 4x) prior to adding dither and requantizing.

I think you'll find that most ME's will also play out to their analog process chains at the received sample rate but then will upsample prior to their digital process chain - i.e. most users of the hardware Waves L2 believe it sounds better processing at the 2xFs rates than 1xFs. Sample rate converters used for this task are usually hardware like the Weiss SFC-1.

With all the SRC and re dithering, are the gains from SRC and re dithering worth the trouble ?

Best thing is to ignore the internet blabber and instead do blind a/b/x's between the same track processed at received rate vs. one up and then processed and then down sampled in order to make your own judgements!!

Thanks for any insight on this matter

As far as SRC algorithms - a couple of ones for PC that spec out the best currently and worthy of investigating with a/b's to hear how which one might work for you are
r8brain (Pro version) from Voxengo - http://www.voxengo.com
and Saracon from Weiss - http://www.weiss.ch

Best regards,
Steve Berson

[lukejs]

thanks for the info ! I´ll have to try going out to the analog chain and do some comparisons.. I did notice that when I did a loopback test , I could hear the difference that an extra round of ADDA conversion makes. I figured that that src would be more transparent. I have the uad-1 plugins like the pultec pro and precision eq which do upsampling, so maybe I´m ok with just working on the original file that I import and say the heck with SRC. The only external EQ´s I have are the one´s in my Allen and heath GS3000 console, so maybe it´s best for me to just stay in the box for now. thanks for your advice !! Luke

[Cellotron]

Originally Posted by lukejs

thanks for the info ! I´ll have to try going out to the analog chain and do some comparisons.. I did notice that when I did a loopback test , I could hear the difference that an extra round of ADDA conversion makes. I figured that that src would be more transparent.

Yes - in general a good current SRC algorithm will be more transparent than a round from DA back to AD, and if your intent is only to do an src conversion then I'd stick to the digital realm for this. The only reason I would do a loop out into the analog realm is if I wanted to use my analog processors which have characteristics that I usually find preferable to my digital tools - plus I am using Lavry and Mytek converters which makes the round trip a pretty harmless thing.

I have the uad-1 plugins like the pultec pro and precision eq which do upsampling, so maybe I´m ok with just working on the original file that I import and say the heck with SRC.

Again - I'd do a test first on one track comparing the same processing changes done to the 44.1kHz file, vs. the same processing changes on the file upsampled, processed and then downsampled - and then a/b them to hear whether one or the other sounds better to you. To me which of these things will sound best is potentially dependent on the DAW app, the plugins, and the SRC algorithm used so there is no one set answer.

The only external EQ´s I have are the one´s in my Allen and heath GS3000 console, so maybe it´s best for me to just stay in the box for now.

Yup - you definitely you don't want to put things through a mid budget console for processing the 2 track!!!!!! - you'll be much better off staying in the box in that case.

Best regards,
Steve Berson

[smallbutfine]

Hi,
I was thinking for a while about this topic...
I'm really wondering if the non-pro version of r8brain would be fully suitable for all conversion tasks...for free.
I guess the SRC is one of the bests and has been found just as good as barbabatch (in apple-world).
But i'm confused about dithering. As a user of Sonar, i have pow-r 3 for free - but r8brain can do dithering as well...does anyone with maybe better ears than mine know which of both is better - i am not really sure about this...

Kind regards,
Martin

[UnderTow]

Originally Posted by lukejs

I use Sonar, and work at 24 96, but when I import an audio file, I´m pretty sure it doesn´t resample it up to 96 khz.

It does actually or you would hear your 44.1Khz file playing really fast in your 96Khz project.

Do I have to buy something like R8 Brain Pro for best results in SRC or is there something out there for free which is just as good ?

The SRC in Sonar 6 has been improved compared Sonar 5. It might not be quite as good as R8Brain but it is no slouch. Try and it and see if it is good enough for you or not.

Any tips on this process of SRC to 96 khz, and then mastering processing , and then reconverting to 44.1 and 16 bit ? With all the SRC and re dithering, are the gains from SRC and re dithering worth the trouble ?

It depends of the material and the quality of any plugins you use. Listening is the only way to find out if it is worth it for you and the material or not.

Alistair