<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>Please do not take my word on this as gospel before one of us
      tests it. Don't send it to Spotify yet until we know for sure.  <br>
    </p>
    <p><br>
    </p>
    <p>Bob</p>
    <p><br>
    </p>
    <p><br>
    </p>
    <br>
    <div class="moz-cite-prefix">On 10/8/18 12:09 PM, Grimm | Eelco
      wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:F76AB116-FF43-4A88-B4A3-2CCA0AB01EA9@grimmaudio.com">
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      <div>Hi Kevin,</div>
      <div><br class="">
      </div>
      <div>Interesting point. The question then becomes how many fixed
        point based subband decoders are still around (in percentage of
        the total). Another question is whether the path from the
        decoder to the volume block (that likely will perform the
        normalization) is floating point. </div>
      <div><br class="">
      </div>
      <div>All the best,</div>
      <div>Eelco</div>
      <div><br class="">
        <blockquote type="cite" class="">
          <div class="">
            <div dir="ltr" class="">I'm pretty sure integer decoders are
              still in use. Not on PCs or even smartphones but in
              smaller audio appliances (e.g. Bluetooth speakers). In any
              case, you've got to get back to integer at some point to
              feed the DAC. So you either give the encoder the extra
              headroom as Spotify is suggesting (see Ian's blog) or you
              attenuate after the decoder as you've suggested. Maybe you
              end up doing both conservatively in which case you've lost
              perhaps 6 dB compared to 0 dBFS PCM which is obviously not
              a small discrepancy.
              <div class=""><br class="">
                <div class="">
                  <div class="">
                    <div dir="ltr" class="gmail_signature"
                      data-smartmail="gmail_signature">
                      <div dir="ltr" class="">Kevin Gross - AVA Networks</div>
                    </div>
                  </div>
                  <br class="">
                </div>
              </div>
            </div>
            <br class="">
            <div class="gmail_quote">
              <div dir="ltr" class="">On Mon, Oct 8, 2018 at 7:45 AM Bob
                Katz <<a href="mailto:bobkatz@digido.com" class=""
                  moz-do-not-send="true">bobkatz@digido.com</a>>
                wrote:<br class="">
              </div>
              <blockquote class="gmail_quote" style="margin:0 0 0
                .8ex;border-left:1px #ccc solid;padding-left:1ex">
                <div text="#000000" bgcolor="#FFFFFF" class="">
                  <p class="">Ian, it's been a while since I performed
                    this test. I would have to retest to be sure, but to
                    the best of my recollection, <b class="">the
                      clipping is ONLY on the repro side and only when
                      converted to fixed point.</b> Remember that all
                    the codecs (both encode and decode side) are 32 bit
                    floating point. <br class="">
                  </p>
                  <p class="">If you run an attenuator in the floating
                    point domain after the decoder, you can fix any over
                    issues which would manifest as clipping when
                    converted to fixed. Test it yourself. Run Apple's AU
                    lab with the round trip codec and Bitter placed in
                    critical places. <br class="">
                  </p>
                  <p class="">Then put a 24 bit dithering plugin and run
                    a distortion test on a sine wave to make sure it's
                    not an illusion. If you guys don't perform this test
                    before the weekend, I'll try to do it myself again.
                    <br class="">
                  </p>
                  <p class=""><br class="">
                  </p>
                  <p class="">Best wishes,<br class="">
                  </p>
                  <p class=""><br class="">
                  </p>
                  <p class="">Bob</p>
                  <br class="">
                  <div class="m_-238104789122760974moz-cite-prefix">On
                    10/7/18 7:39 PM, Ian Shepherd wrote:<br class="">
                  </div>
                  <blockquote type="cite" class="">
                    <pre class="">Hi All !

My recent blog post about lossy encode clipping has had quiet a bit of attention:

<a class="m_-238104789122760974moz-txt-link-freetext" href="http://productionadvice.co.uk/spotify-upload-true-peak/" target="_blank" moz-do-not-send="true">http://productionadvice.co.uk/spotify-upload-true-peak/</a>

Including an email from Robert Bleidt @ Fraunhofer. He’s asking if I have specific examples of encodes that have “baked in” the clipping, because his position is that it shouldn’t happen, these days.

I’d like to give him a helpful reply and I’ve experienced problems myself in the past with encoding from Logic and Audacity, for example, but the truth is I don’t have a collection of examples on hand to offer him.

Do any of you have recent examples that I can point him at, or know for sure of current encoders that consistently have issues with extremely loud content ?

Thanks for any suggestions,

Ian
_______________________________________________
MLA mailing list
<a class="m_-238104789122760974moz-txt-link-abbreviated" href="mailto:MLA@grimmaudio.nl" target="_blank" moz-do-not-send="true">MLA@grimmaudio.nl</a>
<a class="m_-238104789122760974moz-txt-link-freetext" href="http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl" target="_blank" moz-do-not-send="true">http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl</a>
</pre>
                  </blockquote>
                  <br class="">
                  <div class="m_-238104789122760974moz-signature">-- <br
                      class="">
                    <div class="m_-238104789122760974moz-signature">
                      <div class="m_-238104789122760974moz-signature">
                        <div class="m_-238104789122760974moz-signature">
                          <div
                            class="m_-238104789122760974moz-signature">
                            <div
                              class="m_-238104789122760974moz-signature">
                              <div
                                class="m_-238104789122760974moz-signature">
                                <pre class="m_-238104789122760974moz-signature" cols="80"><font class="" face="Courier">
If you want good sound on your album, come to
Bob Katz 407-831-0233 </font><font class="" face="Courier">DIGITAL DOMAIN MASTERING STUDIO
Author: <b class="">Mastering Audio              </b></font><font class="" face="Courier">
<a href="http://www.digido.com/" target="_blank" class="" moz-do-not-send="true">Digital Domain Website</a>

No trees were killed in the sending of this message. However a large number
of electrons were terribly inconvenienced.<font class="" color="#ff0000">
</font></font></pre>
                              </div>
                            </div>
                          </div>
                        </div>
                      </div>
                    </div>
                  </div>
                </div>
                _______________________________________________<br
                  class="">
                MLA mailing list<br class="">
                <a href="mailto:MLA@grimmaudio.nl" target="_blank"
                  class="" moz-do-not-send="true">MLA@grimmaudio.nl</a><br
                  class="">
                <a
                  href="http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl"
                  rel="noreferrer" target="_blank" class=""
                  moz-do-not-send="true">http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl</a><br
                  class="">
              </blockquote>
            </div>
            _______________________________________________<br class="">
            MLA mailing list<br class="">
            <a href="mailto:MLA@grimmaudio.nl" class=""
              moz-do-not-send="true">MLA@grimmaudio.nl</a><br class="">
            <a class="moz-txt-link-freetext" href="http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl">http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl</a><br
              class="">
          </div>
        </blockquote>
      </div>
      <br class="">
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
MLA mailing list
<a class="moz-txt-link-abbreviated" href="mailto:MLA@grimmaudio.nl">MLA@grimmaudio.nl</a>
<a class="moz-txt-link-freetext" href="http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl">http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl</a>
</pre>
    </blockquote>
    <br>
    <div class="moz-signature">-- <br>
      <meta http-equiv="content-type" content="text/html; charset=utf-8">
      <div class="moz-signature">
        <div class="moz-signature">
          <div class="moz-signature">
            <title></title>
            <div class="moz-signature">
              <div class="moz-signature">
                <div class="moz-signature">
                  <pre class="moz-signature" cols="80"><font face="Courier">
If you want good sound on your album, come to
Bob Katz 407-831-0233 </font><font face="Courier">DIGITAL DOMAIN MASTERING STUDIO
Author: <b>Mastering Audio              </b></font><font face="Courier">
<a href="http://www.digido.com/">Digital Domain Website</a>

No trees were killed in the sending of this message. However a large number
of electrons were terribly inconvenienced.<font color="#ff0000">
</font></font></pre>
                </div>
              </div>
            </div>
          </div>
        </div>
      </div>
    </div>
  </body>
</html>