Covermakers Chat Thread

[quartertone]

Hello old friends!
I'm slowly working my way back into LibriWorld after a long hiatus.
And the first thing I'm going to do is make/ride waves

A couple thoughts:

1) I feel like whenever there's conversation about the technical aspects of this project there's a lot of guess work and reverse engineering (looking at source code, webpage behavior, etc). Do we not have a contact on the IA side that can directly and definitely answer these questions? It would save a lot of time and headache to be able to communicate with the team. Sometimes it seems that Anne is our liaison in these matters but not always(?) It's not fully clear or transparent. Anne - do you have the ears of the IA technical team? If so, do comments/suggestions/questions posed here get passed on to them?

2) I tend to agree with Anne regarding the 300x300 vs 600x600 issue. (sorry.) An average image of 300x300 px is about 270kB. 600x600 quadruples that to just over 1MB. Yes, storage these days is cheap and getting cheaper by the day, but multiply this by the thousands of images we would be storing and it adds up. The other issue to consider is the energy cost of the added bandwidth. Search for energy cost of a byte and you can find plenty of reading material. The tl;dr is that according to some estimates, Internet use constitutes about 3% of carbon emissions globally, which is about as much as the airline industry. Admittedly LibriVox probably represents less than a grain of sand in the sand dunes of internet use, but I would prefer us not to be that one grain of sand that causes the avalanche.
On the other hand I alone would probably negate any benefits gained by this potential saving with just one night of binge-streaming Netflix...

It's hard to know where to draw the line. The cover images are works of art and it would be great to display them at a higher resolution somewhere though. Maybe we could put together a physical photo album with community-selected favorites and sell it as a fundraiser for LV/IA. It could be another medium through which to spread Librivox awareness. (Imagine: LibriVox cover album coffee table book; guests remark on it, and conversation ensues.) But this would take effort and coordination, and someone passionate about this side-project to take point. Maybe Michele?


ps - are the monthly Zoom chats not happening any more? It seems there hasn't been much activity on the Zoom threads for a few months.

[msfry]

To Quartertone:
As part of your post seems directed at me, I will answer:
Zoom sessions seem to be on hiatus for various reasons, and will hopefully resume at some point. I can't wrap my mind around hosting now due to house moving, but that doesn't prevent someone else from stepping up. My advice is to stay subscribed to Zoom Announcements and the Zoom Discussion Thread, and when someone offers to host, support them. I do miss visiting with everybody, enjoy meeting new faces, and learning stuff.

I would not be interested in curating any collection of "favorite" covers, which would surely hurt more feelings than it rewards and is not, as they say, "Librivoxy". I do, however, create collections of full color solo covers for any reader who wants them printed and framed for themselves, either 6 up or 12 up in 11x17 frames. They look great hanging on the wall. I have also experimented with sliding 12X, folded in half, into sheet protectors to create a three-ring binder book.

As to file size, the LV YouTube Channel has enormous viewership and they apparently don't object to displaying larger covers, which they store on their own servers (along with all our book files). The purpose of this discussion is how to get them over there.

And no, I will NOT cease binge-streaming Netflix!

[LCaulkins]

Hi, Quartertone, nice to see you again

Yeah, there are a lot of difficulties, both social and technical, involved in creating one, but what a cool idea, a conversation book of LV covers! My imagination can totally see that!

[BengtW]

Regarding YouTube I have update the uploading to search for itemimage file and use it if available. So the script now basically finds the 300p cover link and will check if a file with "_itemimage" exists and use it. So it the cover is called example.jpg it will look for example_itemimage.jpg. If not found it will use the normal cover file. If the itemimage file exists this will look a lot better and will look as sharp as in the PDF I guess. So if used it will be picked up now but is based on this naming convention as there are no other references or links I can use to find this image source.

[redrun]

So if used it will be picked up now but is based on this naming convention as there are no other references or links I can use to find this image source.

I don't have experience with covers from our side or Archive's, but just in case: it looks like extracting images from PDFs could be automated if necessary.

Here's what I've found along those lines:
- 'pdfimages' (on Ubuntu Linux, this is in the package 'poppler-utils') can extract all images from a PDF into a folder, given a format and a naming convention.
- 'imageinfo' (package 'imageinfo') can tell you the pixel size of each image. If you already use ImageMagic, apparently one of the many tools in that suite can do this too.

Some PDFs will have multiple images in them, so that will mean some code to compare, and pick which one to use. I suspect this is the kind of thing that one could get "working" in an afternoon, but "reliable" only after it's seen some real-world use and tweaking.

[BengtW]

I have tired these tools already and it does not work. Many extracts includes parts outside the cover or splits into multiple images. Sometime its rotated. As the PDF files are not following the strict format it cannot be use reliably. I use them to find the cover artist but this is again a best effort using OCR methods. Mistakes in cover artist can be manually updated in the description if needed. For the cover I need something more reliable as a mistake will encode the video incorrectly which cannot be corrected manually without a complete re encode.

[annise]

I think Michele is the only one using multiple templates now and of course that would make automating it a nightmare. And our Mac users seem to use a different template. What you have done seems fine - many of my images for example are made from book covers and are quite low resolution and look fuzzy any higher.

Anne

[carthinius]

Regarding YouTube I have update the uploading to search for itemimage file and use it if available. So the script now basically finds the 300p cover link and will check if a file with "_itemimage" exists and use it. So it the cover is called example.jpg it will look for example_itemimage.jpg. If not found it will use the normal cover file. If the itemimage file exists this will look a lot better and will look as sharp as in the PDF I guess. So if used it will be picked up now but is based on this naming convention as there are no other references or links I can use to find this image source.

So ... if we don't need _itemimage for Internet Archive anymore and it would just be used for the Youtube upload, what about making _itemimage a 1920 x 1080 px widescreen image?

[TriciaG]

So ... if we don't need _itemimage for Internet Archive anymore and it would just be used for the Youtube upload, what about making _itemimage a 1920 x 1080 px widescreen image?

Erm...

In case you're being serious... That would require a total redesign of the cover, since it's a different aspect ratio.

BengtW or Michele - could you please provide an example or two of a YouTube video that the picture is especially fuzzy on? Part of the issue on this is that some covers simply won't look better because the original image is fuzzy anyway. And there's at least one cover maker that doesn't make the image in 600x600 but works directly in Word, and it extracts from there to some other size.

[carthinius]

So ... if we don't need _itemimage for Internet Archive anymore and it would just be used for the Youtube upload, what about making _itemimage a 1920 x 1080 px widescreen image?

Erm...

In case you're being serious... That would require a total redesign of the cover, since it's a different aspect ratio.

I am serious. Of course, it wouldn‘t be a must to do this ratio, but if it is technically possible, why not support it?

[TriciaG]

Not gonna happen.

[msfry]

I sent BengtW this image in Youtube's required rectangular ratio to see if that might work as a template. He could slip our 300 x 300 images into the slot, and upscale it to the Youtube rectangles dimensions. It would not solve the "fuzziness" issue where it exists, but would display the entire cover instead of cropping it. Haven't heard back from him on that yet.

A lot of our images are fuzzy on the Youtube platform, but the one I remember offhand is one of my own: Comparing cd case to youtube:
https://ia601206.us.archive.org/12/items/librivoxcdcoverart34/kingofelflandsdaughter_2001.pdf
https://www.youtube.com/results?search_query=the+king+of+elfland%27s+daughter

[TriciaG]

A lot of our images are fuzzy on the Youtube platform, but the one I remember offhand is one of my own: Comparing archive to youtube:

That one on YouTube, the image is zoomed in. The title is cut off, and the girl is even barely all in the image. If it were regular size, it would probably look fine.

[annise]

Many Youtube makers just turn out rectangular images into squares - and if they care so little they are unlikely to do anything different, it will not increase their earnings. I always try to make our ad hard to remove just so we at least get some credit
Anne
We are discussing BengtW's channel in this thread aren't we ?

[redrun]

I am serious. Of course, it wouldn‘t be a must to do this ratio, but if it is technically possible, why not support it?

Image file sizes grow quadratically as you increase their dimensions, so while quadrupling an already small 300x300 image to 600x600 still gives a reasonable file size, even a square 1080x1080 is just bananas in comparison.
If cover-makers are making art that needs that resolution, they ought to start their own online galleries or something! That kind of work should be seen, but LV's catalog site isn't the place for showing it off. Neither, I would guess, are most YouTube channels.

Technical limitations are one thing, but we have generously been given this space for the purpose of public domain audio, and there ought to be no doubt that's what we're using it for, whether we also donate money to Archive's broader mission or not.