Need to Confirm Process of Converting JPEG to 16 Bit TIFF...

[Rhondak72]

When I'm doing photography while traveling, many times I will shoot in JPEG with the highest quality setting on my Canon 50D. I understand that RAW is the best choice. However, I sometimes shoot in JPEG instead of RAW so that I can capture more photos and not run the risk of running out of disk space before my traveling is done.

I understand that if I was to apply adjustments and then save the JPEG file that data loss will occur due to the compression that takes place. So, usually what I’ll do is pick out the JPEG photos that I want to keep, convert those files to TIFF (uncompressed) and then apply Photoshop edits from there. I do this so that the photos don’t suffer data loss as I apply adjustments and then save those adjustments.

Now, I'm learning about 8 bit dept vs. 16 bit depth. Someone once told me that it is better to work in 16 bit mode instead of 8 bit mode when making adjustments to photos in Photoshop. My understanding is that 16 bit depth allows for so much more tonal ranges/colors to be considered than 8 bit depth when applying adjustments. (Because 16 bit allows for trillions of colors vs. the 16 or so million colors that are considered in 8 bit mode) To clarify, I understand that I’m truly working on an image that originally was 8 bit depth and that I can’t convert that original image to become a complete 16 bit image. I already know that JPEG is not as good of quality as RAW because the JPEG image was never 16 bit in the first place... With that said, I’m told that by converting to 16 bit mode during the editing process that Photoshop will not limit the ADJUSTMENTS to 8 bit depth; however, it will allow for 16 bit depth color to be included within the applied adjustments. Therefore, the adjustments will be of better quality than if I applied adjustments in 8 bit mode where adjustment colors and tonal ranges are more restricted… I was just wondering if anyone has used this process or confirm if it is correct?

In addition to that, I have found that if I save my JPEG file as a TIFF file and then re-open that TIFF file, it will show that it as an 8 bit TIFF file in Photoshop. But, if go to Image > Mode > 16 Bits/Channel and then Save As a TIFF file again, the next time I open that TIFF file, it will then display as a 16 bit TIFF file instead of an 8 bit TIFF file.
So, if I do the steps mentioned above, does that mean my photo will now be treated as a 16 bit lossless TIFF file when additional adjustments are applied in Photoshop (without having to change the mode to 16 bit depth)? Is it correct to say that I won't encounter further banding or data loss as changes are saved (since it is an uncompressed TIFF file) AND I will be editing in true 16 bit mode which will automatically allow my adjustments to be 16 bit depth in terms of the optional color and tonal schemes within those adjustments. Is that correct?

The reason I am asking is that if I’m thinking correctly in all of this that I may automate a process that will batch my photos to be saved from JPEG to 16 bit TIFF files. This way, I won’t have to worry about forgetting to go to Image > Mode > 16 Bits/Channel during the editing process or forget to convert the JPEG to TIFF before applying adjustments.

I realize the file becomes very large but my reasoning is that this process would at least ensure that the file is no longer a lossy file and also ensures that I’m editing in 16 bit mode. Once I’m finished with all adjustments, I can then save the final version as JPEG to CD and then get rid of the large TIFF file. (In the end, I'm told that if a JPEG is saved only a couple of times at the highest quality setting that the data loss usually is very minimal and pretty much not noticable to most people...)

Anyway, if anyone can confirm this proces or provide other suggestions, that would be great. Thanks!

[jiro]

I don't see any issue on your idea and I think it sounds OK to me. My only comment is the use of the word "lossless" since primarily jpeg files are lossy files. You are merely stretching the color and tonal values that you can work on with the image due to the 8-bit to 16-bit conversion but still the original file already lost some detail due to the in-camera jpeg conversion. 16 million shades of colors is a lot even at 8-bits, more so 281 trillion colors with 16-bit files!

If you are OK dealing directly with jpeg files then it's fine. If you want to further extend the "flexibility" of your edit I think the best route would still be to deal with RAW right from the start. Making the wrong decision with regards to exposure and then using jpeg as your preferred file would result with a lot of clipping either in the shadow or highlight areas. This is where RAW file excels. Hope this helps from a practical perspective.

[jeeperman]

I am thinking that this seems like extra work to have less available quality to work with when extra cards a relatively cheap. Buy a few new cards and shoot in RAW, would be I think an option that in the end would serve you well.

[Rhondak72]

Thanks for your input, Willie!

[Rhondak72]

Thanks Jeeperman! I may just go ahead an invest in a couple more cards.

Rhonda

[revi]

Just a couple of thought:
I agree with the others that suggested getting another few memory cards: easiest solution, will allow you to shoot in Raw, and they are quite sturdy.
Another option would be (if you have a computer handy) to get an external HDD and transfer your photos every evening (ideally 2 HDD, but well...).

As for the conversion to TIFF, can't you set PS to import images into 16-bit by default? in that case, I'd save the work to PS native format: you are sure you won't lose information between sessions, and the files might well be a lot smaller than 16-bit TIFFs (I found that for me, a 1-layer TIFF was double the size of a PNG, both using maximum lossless compression).

On the other hand, I'm not sure you win a lot by working in 16-bit, when you start from an 8-bit JPG file. Also, those JPG have already had a lot of treatments performed on the original data, including sharpening. And the sharpening could cause you grief in the later treatments (as it always introduces artifacts, which are not noticable in the original JPG, but can show up in later treatments).

[Colin Southern]

Hi Rhonda,

Your theory is fine - but the big question is perhaps "how well does the practice align with the theory"?

The thing to understand about JPEG files is that they're designed from the ground up to be "efficient" size wise, and they accomplish this in a number of cunning ways ...

1. Most cameras capture a LOT more information than we can ever use (assuming that the photo is exposed correctly). So one of the first things to happen is a LOT of information is thrown away (basically "if it's information we can't see, then throw it away") (allowing for a moderate safety margin though; "just in case").

2. Next, the image is "normalised" ... the difference between some levels is so small that our eyes can't tell the difference ... so these values are normalised into one.

3. Last, the image is compressed (thus realising the benefits of #2 above)

An 8 bit image can contain up to 256 levels for each of the 3 channels ... and yet the human eye struggles to differentiate even 200 levels for each channel ... so it's possibly for an image to contain even a moderate amount of degradation in some circumstances without it being visually obvious.

So - to relate this back to your situation ...

If your capture is reasonably accurate (exposure & white balance wise) then modest 8 bit edits will probably be fine. If you're making large corrections to the captures (especially to levels) then you're probably going to see degradation no matter which bit depth you process in (you can't reveal information that isn't there). Lastly, if you're making complex edits (adjusting many aspects of the image) then you'll get fewer rounding errors by processing in a 16 bit workflow.

In summary ...

1. A RAW capture is ALWAYS going to win in terms of image manipulation potential - and storage is cheap - so that would be my first option

2. There's no downside to processing in a 16 bit workflow as you describe - it's just that there probably isn't going to be much advantage if you're capturing the images correctly in the first place.

Hope this helps

[Sonic4Spuds]

Hi all,

Looks like you covered most everything already, but I just have to comment that I am really loving working in 32 bit non distructive editing.

Have been using the development version of gimp for the past two weeks and am really loving it.

-Sonic

[Colin Southern]

Hi all,

Looks like you covered most everything already, but I just have to comment that I am really loving working in 32 bit non distructive editing.

All we need to do now is wait 50 years for the cameras to catch up

[Jo Brown]

Lots of great suggestions on this thread. I'll just add my two cents.

JPEGs are processed by the camera. Outside of a few adjustments you have no control over the results. We have one little camera that really nails the jpeg processing - especially if we leave everything set to auto - kind of uncanny. However, every other camera we have doesn't do a terribly good job of it in our opinion. So we shoot in RAW, which lets us determine the outcome. This means we have to 'own' any mistakes ourselves unless the lens lets us down.

As for storage issues, when we are travelling we take along a laptop and two external hard drives. Every night we download to the laptop, scrap the mistakes and then move the files to the two external hard drives. The hard drives are mirrors of each other so we have a back up at all times.

It's time consuming and a bit of work - I guess that is why we call our vacations 'working holidays'.

[Colin Southern]

As for storage issues, when we are travelling we take along a laptop and two external hard drives.

Hi Jo,

Sounds like a great system -- my only other thought was "I hope nobody breaks in and steals the camera, laptop, and both drives"! (I suspect it's happened quite often).

[Sonic4Spuds]

All we need to do now is wait 50 years for the cameras to catch up

Thanks Colin,

As I use several applications that can creat up to 32 bit images natively (3D rendering apps) it is really useful to me.

-Sonic

[Jo Brown]

"I hope nobody breaks in and steals the camera, laptop, and both drives"!

Well that would certainly be a downer. So far so good.

[xpatUSA]

Welcome to the forum!

May I ask why you provided a link to an unintelligible web page to do with the use of Visual Basic for image manipulation? - Not all of us are expert in Object Oriented Programming . . .

Didn't really understand the post either, sorry.