Brucer Loc: Bedminster, NJ

Hello. I have an issue I hope you find interesting. I create 4-image series I synchronize while developing in Lightroom 6, then I export all 4 at 300 DPI into a photo folder, before uploading them into Affinity Photo & selecting Focus Merge. Then I export the single-image 300 DPI product into a photo folder. Recently, I checked on the weight of an image in a series of 4. The original RAW file was shot from the Small size option of my D850, so the RAW weight is only 22.2 MB. But one of the series weighed only 6.66 MB. To check on Lightroom's functioning, I exported a jpg I had developed from a RAW image weighing 53.4 MB; it weighed 26.2, so I began to wonder if something's wrong.

For a few days, on occasion I exported at 300 DPI, the "resize" box unchecked, and the linear dimensions were right--a Small image about 4000 on one side, about 2600 on the other, a Large image about 8000 on one side, 5500 on the other--a few images to see if Lightroom corrected itself. It's still doing the same, but I wondered if it always has, so I did the Focus Merge process on a series of images for which I set the development parameters a couple of months ago. I have about 40 "Final" stacked images in total, but all of the series at 300 DPI that produced them, I've deleted; I've kept only the RAW files, so the idea was to reproduce what I had done & check on weights. Result--each of the series weighs between 28.6 and 31.2 MB, but the Final image, after all 4 images were merged in Affinity Photo, is 51.7 MB.

So how did Affinity Photo increase what I assume is pixel density (though I might be wrong in my assumption that more weight is more pixel density, it did make sense that maybe a merging of 4 photo does increase something like that)? I went back and checked on the weights of my earliest Final merged images from last winter, and I found weights pretty much between 220 MB and 230 MB, as if to confirm my notion that merging the images increases weight so that the Final image is a kind of conglomerate of pixels from those images merged, whether or not this has anything to do with resolution or detail, notions I'm not clear about.

Then I looked at more recent Finals: 50.3 MB, 55.6 MB, etc. All of them around that much. So I wonder if, since last winter, my Lightroom has somehow crapped out. Normally, I export with the resize box checked and one edge selected at 1000, the other at 667, 72 DPI. The images are good enough to display digitally. If I were to get prints done, as I hope to, I would export without resizing, but wouldn't I want the images' weights to be pretty close to what the RAW file's weight is, or should I really lose so much as I've noted when exporting as jpg? But then, why the change from like 230 MB to 53.1 MB after merging images? Some of those early series were 8 images, but even so, half of 230 is 115, not 50.

CHG_CANON Loc: the Windy City

So to begin: bytes have no 'weight'. Even a million of them (aka a megabyte) still has no weight.

Next, we ask on nearly a weekly basis: how many pixels fit inside a dot? I feel like Diogenes with his lamp wandering around ancient Athens searching for someone, just one honest man, who can just once say exactly how many pixels fit into just 1 dot? In the same vein, if megabytes had a weight, I could begin to ask: I have a 1-pound bag of bytes, how many do I have?

When you check the resize box in the LR Export, you only have options for the PIXEL RESOLUTION of the resulting image. All your above analysis and observation of the file sizes are absent the more important aspect of the export JPEG: What are the pixel dimensions of the resulting file(s)? Also missing is the JPEG 'quality' setting that also can impact how large / small is the resulting file size in bytes / megabytes.

This older post speaks to resizing images as well as the fallacy of physical 'dot per inch (DPI)' in pixel-based digital photography:

Recommended resizing parameters for digital images

JD750 Loc: SoCal

I wouldn't put too much stock in the "weights" as you say. (Weights! I like it!) If the image "weight" changed when checking "300 DPI" then it would seem the software must have an internal default for physical images size. Maybe you can change that.

JPEGs will be smaller (weigh less) than RAW files because the JPEG is compressed. The level of JPEG compression as well as the size in pixels both affect the "weight".

Physical image size on-screen depends on the screen resolution and screen size. Say you have a 1000 x 1000 pix image, and you display it at 1:1 on an old VGA display 640x480 pixels, it won't fit. You will only see a portion of the image. However on a 15" Retina display 2880 x 1440 pix, it will easily fit and won't even take up 1/2 of the width.

It's further complicated because software will scale scale up and down for you. Open a 4k x 5k pixel image on a 2880 x 1440 screen, it will fit because the software scales it down to fit. However it's no longer 1:1.

So you really want to use DPI for prints, not for screens. Consider pixels in the context of how big you want the 1:1 image to appear on a screen.

CHG_CANON Loc: the Windy City

For pixel-based digital technology, i.e., cameras, monitors, TVs, phones, image files, and so forth, just focus on pixels.

When you display your pixel-based images, consider your target pixel-based device. Try to size your images to your target electronic display.

Consider your 4K TV that is really only 3840 pixels 'wide'. Imagine if you had a TV at the same pixel density as your 43-inch Samsung Crystal UHD, $357 at Walmart.com. This "4K" TV is just 38-inches side. What if you wanted to see your 24MP image from your Nikon D7200? That would need a TV about 60-inches wide and 35-inches tall (roughly 5-feet by 3-feet).

More importantly: there is not a single 'dot' in that entire 4K TV, only pixels, just like there are only pixels in your image file and your digital camera.

When you print your digital image, here too, there are only pixels in your image file. There is not a single dot anywhere in your pixel-based image file.

When you consider the quality of the print, consider the target print size and use your spreadsheet, or your pencil and paper, or just your fingers and toes, and calculate the pixels to inches (or cm) of the target print size. Determine your target pixel resolution, in inches, and confirm you have 150 to 300 ppi, pixels per inch, based on the pixel size of the image, the closer to 300ppi the better. Example, your 6000-pixel wide image file, printed to 20-inches wide, yields a 200 pixel per inch print resolution.

Longshadow Loc: Audubon, PA, United States

Pixels and <mega>bytes are not directly related.
It depends on the arrangement of the different colors in the image.
All white or all black will yield smaller file sizes that an image that has very colorful fall trees throughout the image. It takes more bytes to store a "complicated" image than a simple image.
The file size is what it is depending on the image contents and compression rate (high quality/low quality).

You can fill your car trunk with popcorn or potatoes. The final car weight for each will be different.
Just like the final file size will be different for different images.

CHG_CANON Loc: the Windy City

You can fill your trunk with gigabytes and you'll still never know the difference ....

Longshadow Loc: Audubon, PA, United States

JD750 Loc: SoCal

Well a byte is digital data and as such it must occupy a physical location. That location is not 0 mass. Be that a hard drive or computer memory, there is a physical location where that byte is stored. On computer memory is silicon, it has mass, on spinning rust it is the platter and IronOxide, which also has mass. HD's are TB sized now but a few years ago they were GB and before that MB sized. A GB would be a lot of those! So a GB of data definitely has mass, i.e., weight.

Longshadow Loc: Audubon, PA, United States

No, the media that it is stored on has mass.
A byte itself has none.

A thought has no mass, but when written on paper, does that give it mass?
The mass of the combined ink and paper?
What is the mass of a byte written on paper?

You can store one byte or a bazillion, the mass of the media will not change.

Brucer Loc: Bedminster, NJ

Yeah, I don't where I got the notion of weight, but I do understand that a byte is a datum. I did forget to let you guys know I have had jpg quality set to 100 all along. The only reason I export jpgs at full size (the resize box unchecked) is because I anticipate printing some of these relative few "Final" merged images.

When I print some of the other images, I'll simply export without resizing the image, then, or do whatever the printing service requires, but not if my Lightroom 6 is not exporting images at the best resolution I can get. I certainly think that since a byte is a datum, the number of an image's bytes have to do with its resolution. That's always been obvious to me and I would hope I'm not mistaken. Maybe there's an element of absurdity above a certain range, as if a million gigabyte image would be no sharper or detailed.

CHG_CANON Loc: the Windy City

Images have lots of things, although dots are not one of those many things. The relevant issue for pixel-based images is their pixel resolution.

"Best Resolution" is a some what moving target. The gold standard is a print that achieves 300 pixels of the image to each printed inch of the physical print. But, consider too the print size, as well as the viewing distance. When you look at a billboard, that is not a print 300 ppi (pixels per inch). That billboard is measured in feet (yards), where pixels per inch isn't really relevant.

If we return to the original post / question: what are the pixel dimensions of the files being discussed? If you just pick a new image and export right now, what are your Export parameters and what are the resulting pixel dimensions of the resulting file, as well as the file size?

Brucer Loc: Bedminster, NJ

Almost all of what I export is just for use on my BenQ and Laptop (& I post some online). That will change eventually, but for now, I have time & money only for these 72 DPI images, sized at 1000 x 677 and taking up about 650 KB. I could resize 72 DPI at 1920 x 1280, but those take up 1.2 MB of space or so, and at least to my eye, seem no sharper and more detailed on the screen. These 1000 X 667 images really do look GREAT on the BenQ; usually I'm shooting large, not small, D850 RAW files that come in around 60 MB, sometimes 72, but eventually, I will want to print.

That's why I develop the series of those I stack at 300 DPI, the resize box left unchecked, so that after I might make straight the horizon, cropping ever so little, the dimensions remain at about 8000 X 5500. You would think one of these images of the series I stack in Affinity Photo would be about the same 60 MB after being developed and exported. Not quite as much as a RAW file with all its information, but not like half as much.

CHG_CANON Loc: the Windy City

Brucer, it seems that all my comments about dots are not reaching you. There are no dots in pixel-based images. DPI -- as in dots per inch -- are as relevant to pixel-based images as the appendix is to the human, a forgotten holdover from another time and age. Keep in mind we're not there to look over your shoulder, rather we must work through you for the relevant details.

An image around 8000 x 5500 is a 46MP image, so we're thinking you have an EOS R5 or similar camera. When I search the EOS R5 manual, it seems the expected file size is 45MB. Maybe you have a Sony and you're looking / comparing the Sony compressed RAW file sizes, something that may change in the comparison when looking at the JPEG? Also, different cameras produce different bit-depth image files, that will impact the byte-size of the resulting files.

So, a bit more details about what you're talking about, at the specific technical details, will assist / accomplish a more enlightened review of the technical issues and your original questions and observations.

JD750 Loc: SoCal

That ridiculous. We are talking about the bytes in an image not the word byte.

A word written requires the mass of the ink and occupies the underlying paper. And that has mass. Take the ink and underlying paper out of a book what do you have?

if you fill a disk drive with information is it heavier? Yes because memory is charge storage. You have added electrons to the system, and they have mass. 9.10938356 × 10-31 kilograms per electron.

But I am arguing that the underlying structure is part of the mass required to create and store the information. Without that structure there can be no digital information.

Digital information must be stored or moving in a system or else it’s not information. If it’s stored or running in a system it is occupying the mass of the storage element.

Yes a bazillion bites takes more storage space than one thus it requires more mass.

tcthome Loc: NJ

Either way, I don't use pixels when exporting from LR6. I click in the little box that says pixels (in the image sizing section of the export dialog) & choose inches in the drop down list when I'm concerned with resizing like for my computer monitor.