Your 5D3 M-RAW experience

Owning a 30D, I'm trying to mitigate my impatience while waiting for a deal on the 5D3 to come through.

I downloaded a set of RAW files to get an idea of today's high ISO noise and discovered how hard these files are on my old faithful PC when I play with them in DPP. On the other hand, my standard print size is A4, so I'm wondering whether M-RAW could be the right option for me, especially for portraits. After all, the 1D3 was a great pro camera with its 10MP, and even the current Canon flagship has considerably less resolution than the 5D3.

So, has anyone compared a TIFF from M-RAW to a TIFF from full RAW, subsequently downsized to 10MP ? Does the "downsizing" that the camera performs at raw level yield higher final image quality ?

How does a 5D3 M-RAW compare to a 1D3 full RAW ? Is there any gain from the 5D3's higher resolution on capture ?


Many thanks for your insight!

The M raws are great. But you need to always consider that when you do an upgrade like this the computer will need an upgrade as well. Especially if it is an older and less powerful one. I run a six core i7 980x and I need more power.

Hey thanks David!

You are right, I know you are. But the chore of setting up a new computer from scratch is so unappealing. Since I am a low-volume shooter I might find this slow motion bearable for the next ten years or so ...

With smaller RAW you'd loose dynamic range, how much is depends but hightlight/shadow recovery has more room with full RAW.

LR4.1 is slow on my PC so I did some football shots with std jpegs and I like them. So did some pictures of the kids yesterday and they came out real good. I know I know I should be shooting RAW but I need a new PC.

sonnyc wrote in post #14922498
With smaller RAW you'd loose dynamic range, how much is depends but hightlight/shadow recovery has more room with full RAW.

Hmm, the higher dynamic range in the 5D3 was something I was looking forward to. Wouldn't like losing any.

I use m-raw with my 7d all the time and i like it alot
I am sure you will have a blast with the 5diii

sonnyc wrote in post #14922498
With smaller RAW you'd loose dynamic range, how much is depends but hightlight/shadow recovery has more room with full RAW.

I don't believe that to be the case at all, but I could be mistaken. I though M and S Raw only aggregate data from multiple photosites into a single pixel. If the cluster of photosites is at max brightness, the aggregate is also equal to that same max brightness level, same with shadow, thus the file has the same capability for minimm and maximum tonal range. True, you do lose detail and resolution since you are aggregating data. But, Are you saying that at other than RAW, you have less bit depth to map tone levels?

<Edit>, I believe I am correct, look at this paper starting at the bottom of page 15 with the summary of sRAW and then to the ehd: dougkerr.net/Pumpkin/a​rticles/sRaw.pdf. The available bit depth is still the same. The CPS whitepaper also specifically mentions that sRAW and mRAW benefits are maximum data for post processing at the expense of resolution.

Bsmtril, I am quite sure you are correct, the raw data is simply resized down much like a jpg would be, using whatever algorithm Canon employed.

There's a guy on FM board did a test with the 5D2 RAW and sRAW and the samples clearly showed highlight being blown pretty bad when trying to recover from sRAW while at Full RAW details were still retained.

Maybe it changed now with the 5D3 but if it's important enough, I'd shoot full RAW.

sonnyc wrote in post #14925039
There's a guy on FM board did a test with the 5D2 RAW and sRAW and the samples clearly showed highlight being blown pretty bad when trying to recover from sRAW while at Full RAW details were still retained.

Maybe it changed now with the 5D3 but if it's important enough, I'd shoot full RAW.

Depending on the size of those details and the amount of crop, that could well have been a result of reduced resolution, and not less dynamic range. I'm not saying your wrong here, just that there could be other factors that were perhaps explained in his original test.

I almost never shoot raw, but when I do, it's usually Medium size on my 7D.


I shoot for newsprint, so I shoot JPG 99 percent of the time. I usually only shoot raw when I'm shooting portrait sessions or something for my personal clients, and I don't need 18mp raw files for that stuff.