Looking at the sizes for the HLS modes on programmes using --info, the sizes for the same HLS mode varies quite a lot. Using Eastenders from 10/7/15 (b061y4bt) as an example, the --info output shows that using hlsvhigh, the download could be anywhere between 329MB and 396MB depending on whether you use 1 2 3 or 4.
hlsvhigh1=329MB
hlsvhigh2=396MB
hlsvhigh3=385MB
hlsvhigh4=396MB
Why is there such a big difference in the same mode?
@user-30: Please can we have some sort of <code> tags back to put around commands?
We're working on additional formatting for the editor - give it a bit more time. Code formatting isn't as necessary as it was - the new system preserves line breaks and doesn't muck around with punctuation.
File sizes are only computed from metadata, which may vary a bit between CDNs and may be off by several minutes or more anyway. Don't take those numbers as gospel. You can only be absolutely sure you have a complete download by actually checking the file.
I neglected to mention the biggest factor: different CDNs provide different bit rates and video sizes within the broad categories defined for get_iplayer, e.g., vhigh.
Quote:I neglected to mention the biggest factor: different CDNs provide different bit rates and video sizes within the broad categories defined for get_iplayer, e.g., high.
Does that mean that 1 CDN's version will be better than another CDN's version or is it all the same download in the end just worked out differently?
(11-07-2015, 10:08 PM)@user-30: Please can we have some sort of <code> tags back to put around commands?
I haven't put the buttons into the post tools yet but you should be able to use
[c] for in line highlighting.
Close it with [/c].
For code blocks use
[code]. Close it with the same scheme
Hopefully the above displays and doesn't eat the tags...
(11-07-2015, 11:43 PM)Does that mean that 1 CDN's version will be better than another CDN's version or is it all the same download in the end just worked out differently?
Some are listed as having slightly larger video size or bit rate, but it's academic because I don't think get_iplayer can access those additional streams (the ones with CDN identifiers beginning with "mf_"). They aren't compatible with the method get_iplayer uses to parse HLS stream data. Those streams appeared after the last release and so aren't excluded automatically since they were unknown at the time.