This content has been marked as final. Show 6 replies
I don't understand why you would need the intermediate ACR step(?) That is, calibrate Vuescan the usual way, and save it as a TIF. Then, run Rags' script on the TIF.
Actually, if Vuescan's own calibration process worked properly, there would (ideally) be no need to calibrate again.
> Then, run Rags' script on the TIF ...
Oops! ... that wouldn't do much good. I also confused the color targets reccommended for each ... IT8 for Vuescan and the Macbeth CC for the ACR script. Now your question makes a bit more sense ... but it does combine 2 softwares, only one of which needs to be calibrated.
You can however, obtain an IT8 slide and use that for calibrating Vuescan.
My first reaction is that I would recommend traditional profiling for scanners and skipping ACR completely. That said, I enabled TIF scanner support as a result of a request from another user. He indicated he was satisfied with the results. The ACR calibration script does calibrate. With the proper target datasheet, it can be used with negatives. The read colors script merely reports the target/image differences. The DNG step should be unnecessary.
Since this is somewhat new territory with CS3, I would be willing to work with you offline to see if we can accomplish something that fits your workflow. Contact me via my web site.
As a TIF image, I would expect ACR to honor whatever image profile was already tagged on input. Thus, ACR could be used simply for aesthetic tweaking and things like cropping. I must admit I have not tested this thoroughly. It might be treating it as untagged raw values? If so, I have a lot of questions about what ACR does with JPG images.
If ACR is ignoring the TIF profile, it could explain why Vuescan is not giving you the results you expect. Have you run a target through Vuescan profiling, saved it as TIF, then just run read colors to evaluate the results? Then do the same after opening in ACR with everything defaulted.
Cheers, Rags :-)
>I would expect ACR to honor whatever image profile was already tagged on input.
Pretty sure that's what it does with all supported rendered files with profiles...if the file does NOT have a profile (no embedded profile), I'm pretty sure it assumes sRGB...
Rags and Jeff,
Thanks much for taking the time to reply. Indeed what I have described is a bit of a new approach made possible by the ability of the current ACR to read TIFF and/or the ability of the current Vuescan software to output a DNG.
I shall poke at this a bit and report back. Sadly, my investigations are on hold pending arrival of a motherboard replacement.
Thanks for the workflow and The links. I will let you know my results...