Thumbnails in the database

[Reeka]

I’m hoping someone can help me understand how and why thumbnails are saved to the database. It’s a subject that I have been neglecting during my more than 10 years using ACDSee. But it was when I recently ran a Database Maintenance procedure (ACDSee Pro 9) that I noticed several (approx 20%) of my folders were marked: “Folder that contains Thumbnails in the Database.” I would like to understand what is considered good practice. Is there some value in saving thumbnails in the database? I would like to understand what may have caused only some of my thumbnails to be saved and, more importantly, what control I have over this process going forward.
Thanks, Reeka

[Glen Barrington]

I suspect that what you are seeing might be related to the option to display processed thumbnails or not in raw images.As you may know, few if any, raw developer programs will actually write directly to raw files, this includes DNG files (with the exception of Lightroom for DNG). To my knowledge they all take a non destructive approach towards raw, and that means that the source data must remain untouched by the program at all times. That embedded thumbnail is a part of the raw file's source data.

The net result of this decision is that the thumbnail image embedded within the raw file can never be updated by the host program. To get around this, the raw developer programs save the thumbnails in the database for use by the program. In a raw developer that is built around a relational database, you never see this background operation since it is hidden by the normal database operations (BTW, I'm a retired Teradata DBA, and as such, I have some knowledge of database operations).

However, in those raw development programs that use a hierarchical database (the ACDSee programs, CaptureOne, and AfterShot Pro, to name just 3) hidden, but very real, folders are created to store the thumbnails. The actual names of, and the total contents of what they store are different based on the hierarchical DB engine the various programs use, but they tend to share certain similarities, they are normally hidden from the user, and are normally subfolders of the folder in which the raw photos themselves are stored.

Off hand, I can't think of any raw development programs that have this sort of sub folder stored in one location that would serve the entire catalog. This dispersion would be for access speed issues and for data integrity (if the HD of the pc on which the single thumbnail folder resides, fails at the exact location of the thumbnail folder, a consolidated thumbnail subfolder would be disastrous for the entire catalog, while a more dispersed set of catalog folders reduces risk, overall, and actually enhances access speed.). This is one of the reasons I am against dumping ALL of your photos into a single large "Photos" folder, you are effectively placing all your eggs into a single basket when the host raw development program can offer you multiple baskets for different sorts of eggs.

And once your database gets large enough, the inherent speed advantage of hierarchical databases over relational databases becomes really important. So I believe optimization for speed from the very beginning of your database means you won't have to redesign your database when your response time starts to get unacceptable. Which it will eventually, when you just dump everything into one big folder. But that's an explanation for a different question I guess..

If you tell your operating system to display hidden folders, you can usually see these normally hidden folders. However I would caution you against doing anything with those folders. Anything that needs to be done with those folders is built into ACDSee already (and more importantly, tested for accuracy and reliability). There is nothing you can do in them that ACDSee can't do better, more safely, and more efficiently.

Hope this helps.

[Emil]

I’m hoping someone can help me understand how and why thumbnails are saved to the database.

How: Read this.

Why: To show thumbnails for files that currently are not on line, like CD, DVD, memory stick, SD-cards or detached external harddisk. Also to speed up the displaying of the thumbnails.

For drives and folders that are excluded from the db AC will not store thumbnails in the db. For all other drives and folders AC will automatically collect meta data (incl. thumbnails) from the files and store them in the db. Database maintenance is used to get rid of obsolete meta data. Database optimization is used to compress the database.

Well, that's the theory - on almost every step of these I've seen queer thinks happen is the last decade.

[Reeka]

Many thanks to Glen & Emil for their input. Good information that helped me learn more about how the database works. Unfortunately, I still can't figure out why only a few of my many folders are marked: “Folder that contains Thumbnails in the Database.” Those folders do not contain any raw images so, if I understand what Glen wrote, the raw image concept doesn't provide an answer here. Emil wrote: "For drives and folders that are excluded from the db AC will not store thumbnails in the db. For all other drives and folders AC will automatically collect meta data (incl. thumbnails) from the files and store them in the db." Emil, does this mean that ALL the folders in my database should have thumbnails ...... not just the few that are identified? If there's another explanation that can be offered, I would certainly appreciate seeing it, but this is certainly no big deal for me. The Database Maintenance procedure allows for the removal of these thumbnails, if they're not wanted, and as things stand right now keeping them seems to have no effect on the performance of the software.

[Emil]

. . ., does this mean that ALL the folders in my database should have thumbnails

Hmm, I'd say not necessarily. AC can catalogue (and store the meta data in the db) of any type of file, this includes files for which AC can not produce thumbnails. In this case AC will display icons. AC can even store meta data for folders, not only their content.

Imho the database maintenance should list folder names for every folder with catalogued items. I almost never use it because I always delete or move catalogued files with AC if needed. So my db never contains orphans. If I need to replace hardware I also prefer to create a new catalogue.

The search functions in AC indeed needs time for every catalogued item, no matter if the catalogued item still exists or not. Eg. if your catalogued items are stored on several external drives, the search function's will not be any quicker if one of the drives is detached. From my experience AC's search functions handle 20,000 files a second. So, if your database contains meta data for 200,000 items a single quick search will need ten seconds. Combining search phrases like "cat +dog" will double the time!