Is the difference in size a difference in compression or the size of the embedded preview and/or a difference in the version of the DNG spec supported, or are all those things identical? Have you extracted the previews from each conversion’s DNG, if there are any, to see if they are identical or not?
Are the same number of threads being used by the Adobe RAW engine in the ACR and DNGC conversions?
Have you looked at the GPU load during each conversion?
I set the same thumbnail size for each ("Medium"), but I don't see any other setting that would govern the preview.
I watched the thread count for Camera Raw in Task Manager and saw it jump by as much as 8 at some points, so I'm assuming it's giving jobs to each of my cores. Certainly the overall utilization was high enough (35%-40% and higher occasionally) to prove it wasn't just running on one core. I'm not sure how to gauge threading at a finer level without employing much more powerful monitoring/debugging tools.
It occurred to me that Camera Raw may be using my Camera Raw defaults to generate the color in the embedded preview, where the DNG Converter probably isn't. Good idea to look at the previews... However, they don't appear to have different colors when extracted by either IrfanView or Windows Photo Viewer, and they're 1024 x 683 in every case.
I re-tested. The GPU doesn't seem to be tasked with any of the processing. Its temperature remains stable in every case. Another observation: The count remaining seems to decrement in close pairs with the 64 bit ACR, whereas it just counts down steadily with the 32 bit ACR and DNG Converter.
You can see how many CPU cores are being used by watching the performance tab with core performance separated, as long as nothing else much is going on. Do the file sizes of the extracted previews match each other or the difference in the RAW files.
If you go into the DNGC’s DNG Compatibility / Custom settings you can see that DNGs are compressed by default but you can turn this off, so that could be a difference if you can get one that matches ACR’s conversion by varying the settings.
Noel Carboni wrote:
It occurred to me that Camera Raw may be using my Camera Raw defaults to generate the color in the embedded preview, where the DNG Converter probably isn't.
Actually, I'm pretty sure DNG Converter WILL respect ACR "Defaults"...the only other thing I can think of that might make a difference is the DNG Preferences...you have to dig down in the DNG Conter prefs to get to them but make sure the backwards compatibility is the same for both DNG Converter and ACR.
As long as both DNG Converter and ACR are in lockstep as to version, there should be no funtional differences between them in terms of the DNG at the end. Also not that when you open a raw image into Camera Raw, the fisrt step is to essentially convert the raw data to in effect, a DNG data set. That can make ACR faster cause it's already read th file. As to why ACR in 64 bit is faster than in 32 bit, it's 64 bit processing as well as the amount of ram ACR can access from Photoshop running as a 64 bit app vs a 32 bit app or hosted by Bridge (also 32 bit).
As to what is different in the head, don't know...it may be talked about in the DNG Spec, otherwise that's an Eric question...
You can see how many CPU cores are being used by watching the performance tab with core performance separated, as long as nothing else much is going on.
Not really. Particular threads do not normally maintain a long-term affinity for particular cores, so the usage gets spread around as Windows does its multitasking stuff and what you normally see is just activity across the board, as seen in the screen grab below (that was DNG Converter running). Only inasmuch as the overall percentage is higher than 1/n (where n is the number of cores) can we infer a certain amount of multithreading.
And yes to both of you - every setting that I could find is perfectly consistent across the board, including compression. And as I mentioned before, the embedded previews are coming out exactly identical, both in size and color.
Noel, see my note in the other thread. When ACR converts raw to DNG it adds the default camera raw data to the EXIF. The DNG converter does not add anything to EXIF. I think this additional EXIF data accounts for the small difference in file sizes.