More instructions for VIC-20 carts on Zimmers

[eslapion]

Hello everyone! Happy holiday season!

I just sent about half a dozen VIC carts instructions for posting on Zimmers. They should appearhere quite soon.

If anyone wishes to help me in my efforts, I am looking for pretty much all Atarisoft carts manuals except Pole Position which is already scanned and been sent to Zimmer's too.

Enjoy!

[nbla000]

I think is a good idea to post original manual scans of games on Zimmers btw we may update the Denial wiki too...

[Mayhem]

Well if they ended up being converted to PDF, then possible inclusion into GB20 awaits. Something to plan for v0.3 perhaps...

[eslapion]

Well if they ended up being converted to PDF, then possible inclusion into GB20 awaits. Something to plan for v0.3 perhaps...

As a person having experience in the printing and publishing industry, I have to tell you that the PDF format is a very problematic format. PDF is a derivative of postscript, a page description language that can describe images in a level of quality high enough for color magazines and ultra large posters. Its not meant to be size efficient. Its meant to provide the highest level of quality.

It is completely resolution independent while all scans are pixels based and therefore completely resolution dependent.

The consequence is that no matter what scale you look a page at, you never have any idea if what you see on your screen is a one to one pixel match of the actual scans causing scaling artefacting no matter what you do.

Also, the way it deals with grayscale or B/W monochrome pictures is absolutely miserable.

Example:
I have the FD-2000 manual in PDF format and it takes about 80Mbytes with 128 pages. I have the CMD HD instructions with 171 pages in monochrome GIF format at 300 dpi and it takes 5MBytes... same general level of quality. BTW, GIF uses a non lossy compression algorythm that's nowhere near as efficient as JPEG.

It is very easy to convert a bunch of GIFs into a PDF, assuming you want to waste the space but it is virtually impossible to turn back a PDF of scanned images back into an image format without introducing massive amounts of scaling artefacting.

PDF is a format that tries to be a jack of all trades by supporting both bitmap images of all color depths as well as structured graphics and fonts. Unfortunately, there is a high price to pay to achieve that... digital bloat and complete loss of future convertibility.

[formater]

I have the same experiences as Slapion with the PDF format, and for documents with fewer colors or Black and Withe I prefer a PNG format (JPG are problematic with letters)

[Mayhem]

I'd need a one file format with all the scans included to link. Don't want to link ZIPs inside GB20 for the manuals now. Hence PDF is about the only way to go, or TXT (which is how all the manuals are currently I use, except if they could be fully captured in a single JPG).

[eslapion]

I'd need a one file format with all the scans included to link. Don't want to link ZIPs inside GB20 for the manuals now. Hence PDF is about the only way to go, or TXT (which is how all the manuals are currently I use, except if they could be fully captured in a single JPG).

As much as possible, I try to use JPEG for scans with color drawings or images and monochrome GIFs for text or monochrome line art drawings.

The built-in previewer of Windows XP/Vista does an excellent job with both to view at any size because it generates anti-alias rather than do a nearest neighboor type of display scaling.

I will not convert text page to TXT format as retyping all the text is truly a tiring and excessively long job. Especially when text is used to construct graphic images. Then keeping the right layout becomes nightmarish.

I place these files in a ZIP files strictly for convenience because JPGs and GIFs are already highly compressed. ZIP does shrink it any further. There is no need to keep them in a ZIP file.

I think the best solution for GB20 would be to have a good software viewer that does smart scaling (convert monochrome GIFs to grayscale then perfom bicubic scaling and just decompress JPEGs and also do bicubic scaling as needed) for proper screen size display and use a small script file that tells the software viewer what files matches what page of the instruction related to a specific software.

It would make instructions more than one file but it would alleviate the debilitating technical implications of using an unoptimized and rigid format like PDF.

PDFs are good for turning something you typed with nice fonts in Word or a page layout software into something you can distribute without having to print but it completely sucks when there are color pictures or even scanned text.

Concerning your suggestion of having a single JPEG:

Generally, having scanned text in a JPEG file is also something I try to avoid as much as possible. When I feel it is absolutely necessary to do it (because of dirt on the scanned pages or other technical reasons), I make the JPEG in a completely different way than a color picture.

Poorly made JPEG files that has text in it can make the text completely unreadable because of compression artefacting. JPEGs are made to compress photographic images, not scanned text (as formater pointed out).

For these reasons, having both a color cover art and the monochrome text instruction in a single JPEG files is a serious technical blunder I refuse to make. Color is one thing, grayscale is another and complete monochrome is yet another. Mixing these types of images in a single picture file results in needlessly huge files or unreadable documents. I will do neither.

[Mayhem]

Except with the Gamebase frontend, you can't link to external software when placing files in the "EXTRAS" slots.

When you wish to view a file with the suffix .XXX with Gamebase, Windows launches the appropriate piece of software to use said file as defined by its internal database as if you'd double clicked on it yourself in say Windows Explorer. So a text file comes up in Notepad and a PDF comes up in Acrobat (if you have it installed).

You'd have to manually install a viewer and then link all picture formats to it if you wish to use that solution for Gamebase.

[carlsson]

Personally, I prefer a bulky PDF with all text, illustrations and pure images kept together than an array of loose JPEG, text and other files. The only optional file format I can think of is HTML or some other markup language, where you'd OCR the text portions to hypertext, inline the small images at the right location and use some magic marker to determine end of page. It is however near impossible to make a printable HTML document with fixed page breaks, at least I don't know a good way.

Lately me and some other friends have begun to archive old computer magazines in the size of 25-60 pages each. We simply scan each page as raw images in some nice resolution (I aim at about 900x1200 pixels) and then cram it all down to a PDF. That is exactly the kind of thing Eslapion dislikes, but it makes the process very simple. To run OCR on those magazines is currently out of the question, much less re-create the layout in a professional DTP program. If those 25-40 MB downloads are too big for you, the easy option is to avoid downloading them.

[Mayhem]

Given you'd be downloading GB20 in one big ZIP file anyhows (comprised of its individual parts) of say 200-300MB total, then the PDF route isn't really an issue. I'd prefer it for that project, if nothing else.

[PaulQ]

Personally, I prefer a bulky PDF with all text, illustrations and pure images kept together than an array of loose JPEG, text and other files. The only optional file format I can think of is HTML or some other markup language, where you'd OCR the text portions to hypertext, inline the small images at the right location and use some magic marker to determine end of page. It is however near impossible to make a printable HTML document with fixed page breaks, at least I don't know a good way.

Lately me and some other friends have begun to archive old computer magazines in the size of 25-60 pages each. We simply scan each page as raw images in some nice resolution (I aim at about 900x1200 pixels) and then cram it all down to a PDF. That is exactly the kind of thing Eslapion dislikes, but it makes the process very simple. To run OCR on those magazines is currently out of the question, much less re-create the layout in a professional DTP program. If those 25-40 MB downloads are too big for you, the easy option is to avoid downloading them.

I tend to agree. In this day and age of cheap and easily accessible high speed internet connections, along with cheap $10 storage devices starting in the 4 gigabyte range, megabytes have become the kilobytes of yesteryear. Even dialup has gone high speed.

I remember when I thought 1200 baud was high speed, and 300 baud was the norm. Today's slowest dial up connections are many times faster and much more reliable, delivering a 25 megabyte file in the same time today as a 300 baud modem used to give us a 130k file. A better comparison is to Rogers Hi-speed ultra-lite, which costs roughly the same as a 2nd phone line would alone and delivers 500kbps. That gives us a 40MB file in the same time as a 96k file at 1200 baud. Not bad at all.

[eslapion]

@ Carlsson and DQ:
The main reson I dislike PDF is not its size. Bandwidth and storage is not a premium. The main reason is its a one way dead end.

As I said, converting a bunch of monochrome GIFs into a PDF is very easy and results in no loss of quality.

However, because a PDF is resolution independent, converting a PDF back into a bitmap (for correcting errors or removing dirt in one image among many possible reasons) is nearly impossible to do without massive loss of image quality.

Then again, its practicality of use in the short term vs technical flexibility for future alterations or improvement. PDF is definitely the easiest format for a person to use but its a locked safe when it comes to future modifications.

Adding a premium on size is just the icing on the cake.

Anyways, generally PDF files compress relatively well because unlike JPEGs or GIFs, their content isn't already compressed.

You'd have to manually install a viewer and then link all picture formats to it if you wish to use that solution for Gamebase.

Considering the way I generate the instructions I have published on Zimmers, all you have to do to view them in Windows XP or Vista is expand the files in a folder and then right click on the first file and select "Preview".

The rest is completely intuitive.

Personally, I prefer a bulky PDF with all text, illustrations and pure images kept together than an array of loose JPEG, text and other files. The only optional file format I can think of is HTML or some other markup language, where you'd OCR the text portions to hypertext, inline the small images at the right location and use some magic marker to determine end of page. It is however near impossible to make a printable HTML document with fixed page breaks, at least I don't know a good way.

So far, I have not been confronted to a case where actual text (not scanned text) illustrations and color images needed to be bundled together. In such a case, I would agree that PDF is probably the best way to go. Otherwise that would become unmanageable.

In most cases I have a color illustrated cover with monochrome scans for text pages. In these cases, I simply use filenames that would put these in the right order and use the proper type of image format for the type of scan needed.

Lately me and some other friends have begun to archive old computer magazines in the size of 25-60 pages each. We simply scan each page as raw images in some nice resolution (I aim at about 900x1200 pixels) and then cram it all down to a PDF. That is exactly the kind of thing Eslapion dislikes, but it makes the process very simple.

I may dislike it but I would have to agree that in this specific case this is probably the best way to go. That's because in a single page (in a single scan) you can have color images, monochrome text and color text.

Using JPEG compression would introduce artefacting to the text and using monochrome scans would ruin everything that's in color. Because everything is already a pot pourrit of types then all you can do is scan with maximum resolution and color depth and use non-lossy types of compressions... such as zipping a PDF document.

[PaulQ]

However, because a PDF is resolution independent, converting a PDF back into a bitmap (for correcting errors or removing dirt in one image among many possible reasons) is nearly impossible to do without massive loss of image quality.

I don't understand why you would convert a PDF into a bitmap to remove errors or remove dirt in an image. A PDF is meant to be an end product, like a printed page. One would not print out a document and then scan it back in on a scanner to edit it; but that is effectively what you are suggesting. One would expect you would work on a source image (preferably in a lossless format), then export to a PDF for mass consumption once all error corrections and dirt removal has been completed.

[eslapion]

I don't understand why you would convert a PDF into a bitmap to remove errors or remove dirt in an image. A PDF is meant to be an end product, like a printed page. One would not print out a document and then scan it back in on a scanner to edit it; but that is effectively what you are suggesting. One would expect you would work on a source image (preferably in a lossless format), then export to a PDF for mass consumption once all error corrections and dirt removal has been completed.

My experience tells me that digital archaeology is not something where its wise to have an "end product". In fact, very often its impossible.

For example if I have the instructions for a specific game but not the color cover. Would I choose not to include it because I can't make my PDF an "end product"?

Then 5 years later somebody comes along he's got the cover but I died in an accident and I am gone. How is he going to add the cover to the archive I created if he doesn't have the rest of the instructions? Or he might have them but in a much degraded condition.

This leaves you with two pityful choices; a set of instructions with the color cover but in barely readable condition and another which are of good quality but to which you can't add the nice color cover... you have to have both. That sucks.

There are many instructions I archived that have dots or dirt in some visible areas. Somebody in the future may have the specific problematic pages in much better condition but other areas in poorer condition. How do you combine the good pages already published with the newly found ones if what was published is a PDF document? Good luck!

Nothing is an "end product" when it comes to preservation. Its always an ongoing work.

[Mayhem]

Ah... well it isn't really an issue with GB20 now as the packaging scans are held as different files to the manual within the EXTRAS section anyhow.