Exporting from lightroom to meet a requirement

Hi everyone

I've got a technical question about how I get my images into a particular format for a website

Lightroom 5 is where I do all my processing, key wording & image management but I do have Photoshop which I use to stitch images together when creating panoramas.

I shoot raw format with a full frame canon 6d. File sizes typically 20-30mb.

My workflow is to process in Lightroom & export as a jpeg in srgb colour at the largest size.

A new agency I'm joining has the following requirement

Upload as a "JPEG 12 with a minimum uncompressed size of between 50-75mb

300dpi 8-bit JPEG with adobe RGB colour profile"

Given that some of my images are cropped, the exported jpeg file sizes vary so I know some will never make the uncompressed window of 50-75mb

So how do I ensure mine fit that requirement?

If I export from lightroom as a RGB 240dpi jpeg, & then change the dpi to 300dpi in Photoshop it seems to do the trick but it seems long winded to have to use Photoshop to do this.

Does anyone have any other suggestions on how I might do this using just lightroom?

I'll be honest & hold my hands up saying I simply don't get the 8bit stuff, nor the colour profile or uncompressed size, since my existing agencies have been happy with me using the lightroom default.

Thanks in advance guys!

"Upload as a "JPEG 12 with a minimum uncompressed size of between 50-75mb"

WOW, that is a big jpg. My RAW files from the Mk IV are in the 25mb range. Are you sure on that one?
My understanding there is no '12' in jpg quality. 10 is as high as it goes and 12 was retained for future development.

But at any rate, can't you do all you need with the Export function in LR?

File size requirements are such an odd request; I don't know why some companies still do this. They perpetuate the misunderstanding that file size is a meaningful indicator of, well, anything. It's not.

I'd say export at 100% JPEG quality (which is almost always unnecessary), and in the Resize box, make sure it says "Don't resize" which will retain all the original pixels.

I guess if you wanted to bloat the file size, you COULD specify a pixel dimension that's even larger than your original, and LR will scale it up. It does a pretty good job scaling up on export - up to a certain point, anyway.

I'd see if anyone at the agency could clarify their requirements, from megabytes to pixel dimensions.

Thanks both

Ebiggs - I've taken that from their site & it's massive isn't it. Having opened a handful of my files in Photoshop & gone into resize (to view the uncompressed size) only a tiny percentage will quality

Nathan - that's partly my thinking too, provided nothing is flagged to reduce the size it'll be at its Max. Lightroom is surely used by a large percentage of toggers these days to mean it can be done in there

Ha! Just for kicks, I converted a 5DII raw image in ACR and saved a full-sized JPEG at JPEG 12 in PS - 10MB. A full sized 5DIII raw image converted in ACR and saved as JPEG 12 in PS - 13MB. In each case, I converted to sRGB and embedded the profile.

Good luck. Shoot really really noisy images and you might almost start to double these values (noise affects compression efficiency). Upscaling an image to artificially create pixels to make a bigger image and a larger file size will do nothing but make your image look like crap, which will probably get down sampled anyway, but with all of the up sampling artifact in it.

TIFF files with/out compression might give a 50-70MB file size - a full sized 5DIII raw file conversion yields a 126MB 16 bit TIFF and a 30MB 8bit TIFF.

kirk

Photoshop highest quality jpeg setting is in fact "12". (at least on all the versions I have used a lot)

That said, i never understood any printer that asks for a size in MB, but they've been doing it for many years.

I'd be hard pressed to come up with a jpeg of 50MB without up-scaling a lot. I also wonder if they think they want a Tiff?

Adobe RGB is also an interesting requirement.

CyberDyneSystems wrote in post #17868755
That said, i never understood any printer that asks for a size in MB, but they've been doing it for many years.

After reading the OP's post, I get the impression that it is a stock agency that has issued these submission guidelines, not a printer. But I might be wrong on that point.

The requirement doesn't make sense to me, either, but it is (or has been) common practice for several of the places I submit to. Alamy, one of the world's largest and most successful stock houses, has had the file size minimum for quite some time (or maybe they used to have it and have recently abandoned the MB minimum requirement).

CyberDyneSystems wrote in post #17868755
I also wonder if they think they want a Tiff?

Alamy has always required jPegs, not TIFFs. I am not inferring that the OP is working with Alamy, I merely mention them because they were the stock house best known for specific file size minimums (as opposed to resolution minimums). In fact, I have never worked with a stock agency that wanted TIFFs, but I have worked with a couple of publishers that wanted TIFFs. Two, to be precise.

To get a 50 MB uncompressed file size for an 8 bit image you need 17476266 pixels, or approx 17.5 MP. For a 75 MB uncompressed file you will need 26214400 pixels or 26.2 MP. It is very simple really a single pixel 9 bit RGB image requires three Bytes, no more no less. To be useful you might want to start with four pixels, it allows for a small amount of information and produces a square image. That would require 12 Bytes. Since an 8 bit pixel uses 8 bit for each of the Red, Green and Blue channels, and as 8 bits is 1 Byte it becomes simple. Three Bytes is also 24 bits, and sometimes the total number of bits per pixel, not per channel are used.

So 8 Bit and 24 bit are effectively the same, 8 bit per colour for 24 total.

and 16 bit and 48 bit are the same 16 bits per colour for 48 bit total. A 16/48 bit pixel uses 6 Bytes per pixel.

So if you are quoted an uncompressed file size in 8 bit colour (24 bit is usually only used in scanners and for output from graphics cards) you divide the file size by 3 to get the number of pixels, but remember that 1 MegaByte (MB with a capital B) has a total of 1048576 bytes not 1 Million.

if you are being asked for 16 bit you will have to provide a TIFF file, as JPEG is 8 bit only, then you divide the file size by six for the number of pixels. If using uncompressed TIFF the file size will be a little bigger to allow for the header information, the EXIF data, and the .icc profile. The .icc colour profile can take up quite a bit of space, I have seen some printer profiles that occupy a whole MB of file space. TIFF files also have compression options, so may or may not be different in size when open.

it is really pointless in asking for a data size for an opened image, when photographers are used to dealing with pixels, and there is a fixed relationship between data size and pixel count. Back in the early days of digital imaging knowing the data size the image was going to occupy when it was opened was important. When the image data size might exceed the maximum amount of memory available to hold the image. In a 32 bit operating system the largest memory size that can be allocated allows for an 8 bit image that contains 1431655765 pixels, that's 14.32 Giga Pixels!

Just remember to completely ignore the size of a JPEG file on the disk, as reported by the operating system.

Alan

kirkt wrote in post #17868752
TIFF files with/out compression might give a 50-70MB file size - a full sized 5DIII raw file conversion yields a 126MB 16 bit TIFF and a 30MB 8bit TIFF.

kirk

?
Don't you mean 63 MB? Snow in Philly must be getting to you.

Tom is right. Alamy and others want jpgs but they specify the UNCOMPRESSED size, i.e. the size after the compressed jpg has been decompressed (opened) for use. And the pixel content of an image (in MP) decompressed from a jpg is always MB/2.86. So a 50 MB ex-jpg will be 17.5 MP and a 75 MB will be around 26 MP.

Ppi or dpi are irrelevant. Noise is irrelevant. The degree of compression previously applied to the file is irrelevant. Color space is irrelevant. (BTW, stock agencies ask for Adobe RGB because a significant segment of their market wants images for lithographic reproduction.)

An RGB jpg is always 8 bit and always compressed. Even a Q12 PS compression (= 94-100 in LR) will be from 3:1 to 5:1. But in the pre-compression or post-decompression versions each pixel is 24 bits, which is 3 bytes. But there are 1,024 bytes in a KB and 1,024 KB in a MB, so when we are talking about millions of pixels, the multiplier is not 3, but rather 2.86. But to anybody in the trade, the x3 conversion is instinctive and easy.

I used ZIP compression on the 8bit file to get 30MB. Uncompressed it is 66.3MB, compressed it is actually 33.3MB but I rounded down. To conserve energy for the blizzard. The range was for the extremes (16bit uncompressed and 8bit compressed).

If I take the same image and save it as a JPEG, I get a 13.6MB image (level 12). If I add noise using Filter > Noise > Add Noise at a slider value of 20, the file size becomes 45.9MB for a JPEG with quality 12.

if you are talking about the post-decompression file size, well that's a different story.

We have 2inches of snow so far. It is snowing heavily but the wind has not started yet.

Time for more beer.

kirk

Just add 30Mb worth of keywords to the file

Dan Marchant wrote in post #17869054
Just add 30Mb worth of keywords to the file

best advice yet!

thanks for the follow-up responses

to clear a few things up, yes the firm is a stock/image agency so trying to fit into their requirements & they won't accept tiffs, just jpegs

im also with alamy & am aware of their requirements, especially in terms of file size. they make it very easy though because when i upload, they've got validation built into their upload functionality to look at the image size & to decline anything i upload which isn't large enough - essentially it means if i deem my content to be good enough, i upload it without considering the size & i just acknowledge any they decline for size without having to worry in advance. in other words, im reactive to it, whereas this new firm need me to be proactive to it

clearly its in everyones interests for me not to step up my sizes (the buyers/agency won't like the reduced IQ & i don't want to bolt additional things to do in other applications in my workflow)

so back to my original question which was how i process these in lightroom (ideally just lightroom & not a combo of that & photoshop) it looks like exporting like this:-

* no restrictions on size (will give me the largest possible file)
* 300dpi
* jpeg
* open file in photoshop & view uncompressed size (can't see any other way of checking the uncompressed size)

daft question - is there a certain rule which can be applied (even if i use a certain website/calculator) where i can roughly obtain the uncompressed size when i know the default jpeg size? ie anything above say 10,000kb (im using that as an example only!!) = 50,000kb when uncompressed

As I noted above, the amount of compression applied to the jpg and its resulting size have no affect on the uncompressed size; all you need to know is the total number of pixels (height x width). Multiply that by 3 for a rough estimate. 2.86 to be more accurate.

For instance: 4000 x 6000 pixels = 24MP x 3 = 72MB