Two approaches:

-- Master in Csound itself: all compression, level setting, equalization, etc., is performed by the Csound orchestra (this is what I usually do, as it is easy to repeat and documents itself).

-- Master in an audio editor such as Pro Tools (I use Adobe Audition if I don't use Csound itself).

In either case, always render at 88.2 KHz to float samples. This way you have a creamier, high-resolution soundfile for concert diffusion or audiophile listening. You also leave yourself a high-resolution file for any further tweaking you may require.

Just before creating your CD, convert the high-resolution soundfile to 44.1 KHz/16 bits with dither (Adobe Audition will do this for you automatically from a track list of high-resolution soundfiles).

ALWAYS USE THE HIGHEST QUALITY YOU REASONABLY CAN EVERY STEP ALONG THE WAY, all right? Like an ARTIST? 

I was looking at Jan van Eyck oil paintings in the Metropolitan Museum the other day. Looked like they were painted last year instead of 1430 something. Purer colors, and less faded by far, than the Courbets of 1860 something in the same museum. Jan perfected the methods of oil painting, used only lasting pigments, mixed them into the linseed oil himself or had his apprentices do it under his watchful eye. This is the way to go.

Hope this helps,
Mike

-----Original Message-----
>From: Dave Seidel 
>Sent: Apr 21, 2008 10:08 AM
>To: csound@lists.bath.ac.uk
>Subject: [Csnd] request for advice: mastering for CD
>
>Hi all,
>
>I'm going to be putting together a small CD-R release in the next couple 
>of months, and I need some advice on mastering.  It's a very low-budget 
>project on a tiny local label, so I need to do the mastering myself. 
>We'll be burning the CDs ourselves as well.
>
>The music itself is all purely generated from Csound opcodes (poscil3 is 
>usually the generator) and I use reverbsc.  None of the pieces use samples.
>
>My first though is to start by rendering the pieces to 44.1K/16, since 
>that's the "native" format does CDs.  (Normally I render to 96/24.)  But 
>I am open to any and all advice from people who have already been down 
>this road.
>
>- Dave
>
>
>
>Send bugs reports to this list.
>To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Thanks, Mike.

As I mentioned, I normally render to 96/24, and that's what I play in a 
concert situation.  I'm guessing that you recommend 88.2K rather than 
96K because dithering it down to 441.1K is simpler (just a divide by 2), 
correct?

Is it possible to use Csound to perform this step?  I don't own 
Audition.  Audacity can resample, but I don't see any option for 
dithering.  I may be able to do it with Reaper, though, will have to 
look into that.

- Dave

Michael Gogins wrote:
> Two approaches:
> 
> -- Master in Csound itself: all compression, level setting, equalization, etc., is performed by the Csound orchestra (this is what I usually do, as it is easy to repeat and documents itself).
> 
> -- Master in an audio editor such as Pro Tools (I use Adobe Audition if I don't use Csound itself).
> 
> In either case, always render at 88.2 KHz to float samples. This way you have a creamier, high-resolution soundfile for concert diffusion or audiophile listening. You also leave yourself a high-resolution file for any further tweaking you may require.
> 
> Just before creating your CD, convert the high-resolution soundfile to 44.1 KHz/16 bits with dither (Adobe Audition will do this for you automatically from a track list of high-resolution soundfiles).
> 
> ALWAYS USE THE HIGHEST QUALITY YOU REASONABLY CAN EVERY STEP ALONG THE WAY, all right? Like an ARTIST? 
> 
> I was looking at Jan van Eyck oil paintings in the Metropolitan Museum the other day. Looked like they were painted last year instead of 1430 something. Purer colors, and less faded by far, than the Courbets of 1860 something in the same museum. Jan perfected the methods of oil painting, used only lasting pigments, mixed them into the linseed oil himself or had his apprentices do it under his watchful eye. This is the way to go.
> 
> Hope this helps,
> Mike

Another option for resampling is ecasound.  It's quite easy to use and 
harnesses libsamplerate for the conversion.  I've been very pleased with 
its quality.

jp


Dave Seidel wrote:
> Thanks, Mike.
>
> As I mentioned, I normally render to 96/24, and that's what I play in 
> a concert situation.  I'm guessing that you recommend 88.2K rather 
> than 96K because dithering it down to 441.1K is simpler (just a divide 
> by 2), correct?
>
> Is it possible to use Csound to perform this step?  I don't own 
> Audition.  Audacity can resample, but I don't see any option for 
> dithering.  I may be able to do it with Reaper, though, will have to 
> look into that.
>
> - Dave
>
> Michael Gogins wrote:
>> Two approaches:
>>
>> -- Master in Csound itself: all compression, level setting, 
>> equalization, etc., is performed by the Csound orchestra (this is 
>> what I usually do, as it is easy to repeat and documents itself).
>>
>> -- Master in an audio editor such as Pro Tools (I use Adobe Audition 
>> if I don't use Csound itself).
>>
>> In either case, always render at 88.2 KHz to float samples. This way 
>> you have a creamier, high-resolution soundfile for concert diffusion 
>> or audiophile listening. You also leave yourself a high-resolution 
>> file for any further tweaking you may require.
>>
>> Just before creating your CD, convert the high-resolution soundfile 
>> to 44.1 KHz/16 bits with dither (Adobe Audition will do this for you 
>> automatically from a track list of high-resolution soundfiles).
>>
>> ALWAYS USE THE HIGHEST QUALITY YOU REASONABLY CAN EVERY STEP ALONG 
>> THE WAY, all right? Like an ARTIST?
>> I was looking at Jan van Eyck oil paintings in the Metropolitan 
>> Museum the other day. Looked like they were painted last year instead 
>> of 1430 something. Purer colors, and less faded by far, than the 
>> Courbets of 1860 something in the same museum. Jan perfected the 
>> methods of oil painting, used only lasting pigments, mixed them into 
>> the linseed oil himself or had his apprentices do it under his 
>> watchful eye. This is the way to go.
>>
>> Hope this helps,
>> Mike
>
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body 
> "unsubscribe csound"

Thanks, Julian, not sure if I can run it on Windows though, and I don't 
have a decent Linux box at the moment.

- Dave

Julian Peterson wrote:
> Another option for resampling is ecasound.  It's quite easy to use and 
> harnesses libsamplerate for the conversion.  I've been very pleased with 
> its quality.
> 
> jp
> 
> 
> Dave Seidel wrote:
>> Thanks, Mike.
>>
>> As I mentioned, I normally render to 96/24, and that's what I play in 
>> a concert situation.  I'm guessing that you recommend 88.2K rather 
>> than 96K because dithering it down to 441.1K is simpler (just a divide 
>> by 2), correct?
>>
>> Is it possible to use Csound to perform this step?  I don't own 
>> Audition.  Audacity can resample, but I don't see any option for 
>> dithering.  I may be able to do it with Reaper, though, will have to 
>> look into that.
>>
>> - Dave
>>
>> Michael Gogins wrote:
>>> Two approaches:
>>>
>>> -- Master in Csound itself: all compression, level setting, 
>>> equalization, etc., is performed by the Csound orchestra (this is 
>>> what I usually do, as it is easy to repeat and documents itself).
>>>
>>> -- Master in an audio editor such as Pro Tools (I use Adobe Audition 
>>> if I don't use Csound itself).
>>>
>>> In either case, always render at 88.2 KHz to float samples. This way 
>>> you have a creamier, high-resolution soundfile for concert diffusion 
>>> or audiophile listening. You also leave yourself a high-resolution 
>>> file for any further tweaking you may require.
>>>
>>> Just before creating your CD, convert the high-resolution soundfile 
>>> to 44.1 KHz/16 bits with dither (Adobe Audition will do this for you 
>>> automatically from a track list of high-resolution soundfiles).
>>>
>>> ALWAYS USE THE HIGHEST QUALITY YOU REASONABLY CAN EVERY STEP ALONG 
>>> THE WAY, all right? Like an ARTIST?
>>> I was looking at Jan van Eyck oil paintings in the Metropolitan 
>>> Museum the other day. Looked like they were painted last year instead 
>>> of 1430 something. Purer colors, and less faded by far, than the 
>>> Courbets of 1860 something in the same museum. Jan perfected the 
>>> methods of oil painting, used only lasting pigments, mixed them into 
>>> the linseed oil himself or had his apprentices do it under his 
>>> watchful eye. This is the way to go.
>>>
>>> Hope this helps,
>>> Mike
>>
>>
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body 
>> "unsubscribe csound"
> 
> 
> 
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe 
> csound"
> 
> 

Denis Crowdy wrote:

> On Thu, Apr 24, 2008 at 11:22 PM, Dave Seidel  wrote:
> > Thanks, Mike.
> >
> >  As I mentioned, I normally render to 96/24, and that's what I play in a
> > concert situation.  I'm guessing that you recommend 88.2K rather than 96K
> > because dithering it down to 441.1K is simpler (just a divide by 2),
> > correct?
> 
> I'm pretty sure the answer to this is no - it is by no means a simple
> divide by 2.  The author of libsamplerate (Erik de Castro Lopo) gives
> a series of lectures to my audio engineering students

I really enjoyed that Denis. Thanks.

> and demonstrates
> how important the filtering and conversion process is so that aliasing
> artefacts don't appear after the conversion,

To expand on this a bit, there are two basic categories of doing sample
rate conversion

  - algorithmically naive and computationally cheap ways
  - good algorithms

When using the first method, yes, downsampling by 2 is better/easier
than downsampling by 960/441.

However, when using good algorithms (ie the SRC_SINC_* in libsamplerate
and similar converters in other libraires and applications) the
difference in quality between downsampling by 2 over downsampling by
960/441 is unlikely to be measurable.

> (similar to those at http://src.infinitewave.ca/) to demonstrate this.

Excellent web site. For a little more on what these spectral plots
mean and how they are obtained, see my blog post:

   http://www.mega-nerd.com/erikd/Blog/CodeHacking/SecretRabbitCode/src_compare.html

A later post

   http://www.mega-nerd.com/erikd/Blog/CodeHacking/SecretRabbitCode/progress.html
   http://www.mega-nerd.com/erikd/Blog/CodeHacking/SecretRabbitCode/rel_0_1_3.html

describes a reworking of the converters I did just recently which put
the best of the libsamplerate converters (with a SNR of 145dB)
extremely close to the best of the commercial converters (160dB SNR or
better). Personally, I think 145dB SNR should be more than sufficient
for any audio application.

Cheers,
Erik
-- 
-----------------------------------------------------------------
Erik de Castro Lopo
-----------------------------------------------------------------
With 22,100,000 legitimate businesses in the US alone,
allowing each to send only one UCE per *year* gets every
mailbox 60,547 emails per day. There will either be email
without UCE or there will be no email.

Not at all Denis, I welcome the information, and really appreciate it.

Erik, you have whetted my appetite to try libsamplerate, so I am trying 
to get it built, but I've run into a problem building libsndfile (using 
MinGW in WinXP).  I successfully got through initial make step, but in 
"make check", everything is failing with errors like the following:

c:/music/SecretRabbitCode/libsndfile-1_0_17/libsndfile-1.0.17/src/test_file_io.c:351: 
undefined reference to `psf_fopen'

There are similar messages for other functions with the "psf_" prefix. 
Am I missing a library?  Or can I just use the pre-compiled version?

- Dave

Denis Crowdy wrote:
> Thanks Erik - interesting stuff.  Apologies Dave if I led things
> slightly awry there.
> 
> Certainly shows that that I'm better off using csound and
> libsamplerate than our expensive proprietary DAW when it comes to some
> operations anyway - sample rate conversion being one of them...
> 
> Denis

Never mind, I was able to use the precompiled version of libsndfile to 
build libsamplerate.  I'm now going to play with sndfile-resample and 
see how it goes.  I had previously found a free (as in beer) program 
called r8brain[1] that seems to do a very decent job, so I'll use the 
output from that as comparison.  Of course, my only basis of comparison 
is my ears. :-)

Does libsamplerate perform dithering when it downsamples?

- Dave

[1] http://www.voxengo.com/downloads/?highlight=r8brain

Dave Seidel wrote:
> Not at all Denis, I welcome the information, and really appreciate it.
> 
> Erik, you have whetted my appetite to try libsamplerate, so I am trying 
> to get it built, but I've run into a problem building libsndfile (using 
> MinGW in WinXP).  I successfully got through initial make step, but in 
> "make check", everything is failing with errors like the following:
> 
> c:/music/SecretRabbitCode/libsndfile-1_0_17/libsndfile-1.0.17/src/test_file_io.c:351: 
> undefined reference to `psf_fopen'
> 
> There are similar messages for other functions with the "psf_" prefix. 
> Am I missing a library?  Or can I just use the pre-compiled version?
> 
> - Dave
> 
> Denis Crowdy wrote:
>> Thanks Erik - interesting stuff.  Apologies Dave if I led things
>> slightly awry there.
>>
>> Certainly shows that that I'm better off using csound and
>> libsamplerate than our expensive proprietary DAW when it comes to some
>> operations anyway - sample rate conversion being one of them...
>>
>> Denis
> 
> 
> 
> 
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe 
> csound"
> 
> 

Dave Seidel wrote:

> Never mind, I was able to use the precompiled version of libsndfile to 
> build libsamplerate.  I'm now going to play with sndfile-resample and 
> see how it goes.  I had previously found a free (as in beer) program 
> called r8brain[1] that seems to do a very decent job, so I'll use the 
> output from that as comparison.  Of course, my only basis of comparison 
> is my ears. :-)

r8brain is actually very very good, at least according to this
comparison:

     http://src.infinitewave.ca/

> Does libsamplerate perform dithering when it downsamples?

Dithering actually doesn't have anything to do with sample rate
conversion. Rather, dither is something you need to add when 
reducing bit depth, ie float or 24 bit PCM to 16 bit PCM.

However, the answer is no, the sndfile-resample program does
not do dithering. I would suggest that you use sndfile-resample
on a floating point input and output files and do the bit depth
reduction separately.

Unfortunately I don't have a recommendation for bit depth reduction.
Its something I have been meaning to add to libsndfile for some
time.

Cheers,
Erik
-- 
-----------------------------------------------------------------
Erik de Castro Lopo
-----------------------------------------------------------------
"The growing and dangerous intrusion of this new technology,
threatens an entire industry's economic vitality and future
security." -- Jack Valenti (MPAA president) on the video
cassette recorder, 1982.

Thanks, Erik.  I will probably continue to use r8brain for now, since it 
performs the bit depth reduction in conjunction with the SRC.  But I 
will certainly keep my eye on libsamplerate.

- Dave

Erik de Castro Lopo wrote:
> r8brain is actually very very good, at least according to this
> comparison:
> 
>      http://src.infinitewave.ca/
> 
>> Does libsamplerate perform dithering when it downsamples?
> 
> Dithering actually doesn't have anything to do with sample rate
> conversion. Rather, dither is something you need to add when 
> reducing bit depth, ie float or 24 bit PCM to 16 bit PCM.
> 
> However, the answer is no, the sndfile-resample program does
> not do dithering. I would suggest that you use sndfile-resample
> on a floating point input and output files and do the bit depth
> reduction separately.
> 
> Unfortunately I don't have a recommendation for bit depth reduction.
> Its something I have been meaning to add to libsndfile for some
> time.
> 
> Cheers,
> Erik