mattbasta · June 16, 2017 20:24
diff --git a/rob_walch_response.txt b/rob_walch_response.txt
 Thanks for your quick response, Rob. I'd like to bring up a few things about it, if you don't mind indulging me. The goal of my post is to challenge claims and assumptions about VBR, and I hope to either reinforce or disprove these. Essentially, I hope to put citations on the arguments for and against VBR.

 The post you linked to states that some devices do not support VBR, and substantiates that with quotes from the ISO/IEC standard for MPEG. The quotes that it cites--to show that VBR is not a required feature of MP3--actually seem to confirm that VBR is in fact a standard feature of MP3. The statement in question:

 Layer III supports variable bitrate by switching the bitrate index.

 Layer III, in the context of this quote, refers to MPEG-2 Audio Layer 3, which is itself MP3. Layer 1 and Layer 2 refer to MPEG-1/2 Audio Layer 1 and Audio Layer 2 (MP1 and MP2 respectively), both of which are separate audio codecs that are unrelated to MP3 as it's used today. ISO/IEC 11172-3:1993 and ISO/IEC 13818-3:1995 both seem to state that Layer 3 (any by definition, MP3) does not make variable bitrate support optional, and in fact VBR is a strict superset of CBR (all CBR files are valid VBR-encoded files). And since Fraunhofer's expensive MP3 patent licensing program covered decoders as well as encoders, would it not have make sense for patent licensees to pay for a decoder that maximizes compatibility (like Fraunhofer's own decoder)?

 I did some research to try to identify devices without VBR support that would reasonably be able to play podcasts (i.e., more than 128MB of flash storage, not a DVD player, not a car CD player, etc.) and could not find any electronics produced after 2006 that fit this profile. I was hoping that you could share any experiences of your own, even anecdotal, that could help back up this claim.

 The post also says:

 Today there is no need for [VBR] - available bandwidth and storage today is much different than 15 and 20 years ago.

 Is it not also the case that the audience for podcasts in the United States has more than doubled in the last ten years, and that the number of podcasts on Apple Podcasts alone has increased by 10x? Surely the increase in podcast production has made distribution more costly because of sheer volume alone. According to Josh Morgan's (@pluralofyou) research, the median podcast duration has nearly doubled since 2007, from 25 minutes to over 40 in 2015. On Amazon S3, a podcast with 10,000 subscribers distributing 100MB MP3 files would spend $360 in hosting costs to put out weekly episodes, while VBR encoding resulting in (a commonly cited, conservative decrease in file size of) 20% would cost only about $285 at roughly the same audio quality.

 In countries like India, China, and even Australia, where bandwidth can be very expensive for consumers, minimizing the file size of content would conceivably help encourage new listership. Especially in developing international markets, where consumer electronics are almost exclusively modern Android and iOS devices that support VBR playback, is there not an argument to be made for supporting globalization of podcasting by optimizing content encoding in all ways that we can?


 I really do appreciate you taking the time to respond. Thanks again,

 Matt Basta
 https://pinecast.com
	Thanks for your quick response, Rob. I'd like to bring up a few things about it, if you don't mind indulging me. The goal of my post is to challenge claims and assumptions about VBR, and I hope to either reinforce or disprove these. Essentially, I hope to put citations on the arguments for and against VBR.

	The post you linked to states that some devices do not support VBR, and substantiates that with quotes from the ISO/IEC standard for MPEG. The quotes that it cites--to show that VBR is not a required feature of MP3--actually seem to confirm that VBR is in fact a standard feature of MP3. The statement in question:

	Layer III supports variable bitrate by switching the bitrate index.

	Layer III, in the context of this quote, refers to MPEG-2 Audio Layer 3, which is itself MP3. Layer 1 and Layer 2 refer to MPEG-1/2 Audio Layer 1 and Audio Layer 2 (MP1 and MP2 respectively), both of which are separate audio codecs that are unrelated to MP3 as it's used today. ISO/IEC 11172-3:1993 and ISO/IEC 13818-3:1995 both seem to state that Layer 3 (any by definition, MP3) does not make variable bitrate support optional, and in fact VBR is a strict superset of CBR (all CBR files are valid VBR-encoded files). And since Fraunhofer's expensive MP3 patent licensing program covered decoders as well as encoders, would it not have make sense for patent licensees to pay for a decoder that maximizes compatibility (like Fraunhofer's own decoder)?

	I did some research to try to identify devices without VBR support that would reasonably be able to play podcasts (i.e., more than 128MB of flash storage, not a DVD player, not a car CD player, etc.) and could not find any electronics produced after 2006 that fit this profile. I was hoping that you could share any experiences of your own, even anecdotal, that could help back up this claim.

	The post also says:

	Today there is no need for [VBR] - available bandwidth and storage today is much different than 15 and 20 years ago.

	Is it not also the case that the audience for podcasts in the United States has more than doubled in the last ten years, and that the number of podcasts on Apple Podcasts alone has increased by 10x? Surely the increase in podcast production has made distribution more costly because of sheer volume alone. According to Josh Morgan's (@pluralofyou) research, the median podcast duration has nearly doubled since 2007, from 25 minutes to over 40 in 2015. On Amazon S3, a podcast with 10,000 subscribers distributing 100MB MP3 files would spend $360 in hosting costs to put out weekly episodes, while VBR encoding resulting in (a commonly cited, conservative decrease in file size of) 20% would cost only about $285 at roughly the same audio quality.

	In countries like India, China, and even Australia, where bandwidth can be very expensive for consumers, minimizing the file size of content would conceivably help encourage new listership. Especially in developing international markets, where consumer electronics are almost exclusively modern Android and iOS devices that support VBR playback, is there not an argument to be made for supporting globalization of podcasting by optimizing content encoding in all ways that we can?


	I really do appreciate you taking the time to respond. Thanks again,

	Matt Basta
	https://pinecast.com